Received: 14 April 2019 Revised: 26 August 2019 Accepted: 27 August 2019

DOI: 10.1002/er.4883

REVIEW PAPER

Energy management system for smart grid: An overview

and key issues

Sumit K. Rathor | D. Saxena

Department of Electrical Engineering,

Summary
Malaviya National Institute Technology
Jaipur, Rajasthan, India Energy crisis and the global impetus to “go green” have encouraged the inte-
gration of renewable energy resources, plug‐in electric vehicles, and energy
Correspondence
Sumit K Rathor, Department of Electrical
storage systems to the grid. The presence of more than one energy source in
Engineering, Malaviya National Institute the grid necessitates the need for an efficient energy management system to
Technology Jaipur, Rajasthan 302017, guide the flow of energy. Moreover, the variability and volatile nature of
India.
Email: [email protected] renewable energy sources, uncertainties associated with plug‐in electric
vehicles, the electricity price, and the time‐varying load bring new challenges
to the power engineers to achieve demand‐supply balance for stable operation
of the power system. The energy management system can effectively
coordinate the energy sharing/trading among all available energy resources,
and supply loads economically in all the conditions for the reliable, secure,
and efficient operation of the power system. This paper reviews the framework,
objectives, architecture, benefits, and challenges of the energy management
system with a comprehensive analysis of different stakeholders and
participants involved in it. The review paper gives a critical analysis of the
distributed energy resources behavior and different programs such as demand
response, demand‐side management, and power quality management
implemented in the energy management system. Different uncertainty quanti-
fication methods are also summarized. This review paper also presents a
comparative and critical analysis of the main optimization techniques used to
achieve different energy management system objectives while satisfying multi-
ple constraints. Thus, the review offers numerous recommendations for
research and development of the cutting‐edge optimized energy management
system applicable for homes, buildings, industries, electric vehicles, and the
whole community.

KEYWORDS
distributed energy resources (DER), energy management system (EMS), energy storage systems
(ESS), plug‐in electric vehicle (PEV), renewable energy sources (RES)

Int J Energy Res. 2020;1–43. wileyonlinelibrary.com/journal/er © 2020 John Wiley & Sons, Ltd. 1
2 RATHOR AND SAXENA

1 | INTRODUCTION parameters, intermittent generation of renewable energy,

and high charging demands of PEVs, which would even-
1.1 | Motivation tually lead into catastrophic issues such as energy imbal-
ance, power quality issues, loss of reliability, resiliency,
According to a report published by the Organization for security of the system, and regulatory issues such as
Economic Co‐operation and Development (OECD) uneven profit distribution to consumers. Moreover, the
economic outlook and interim economic outlook May penetration of the new type of sources, ie, renewable
2018,1 China, India, and the United States secure the energy along with the PEV in the distribution network,
top three positions concerning gross domestic product has transformed the traditionally passive distribution sys-
(GDP) growth rate. However, according to British Petro- tems into the active distribution systems. Thus, with
leum (BP) statistical review of world energy in June more than one energy sources and storage system, there
2017,2 these countries rank worst concerning the air qual- is a requirement of controlling the energy flow among
ity index as compared with other countries. The major the various sources. To maximize the utilization of the
sources of pollution include generating stations, transpor- potential of new types of sources and new types of loads
tation fleet, and industries owing to the use of oil in one in the power grid and to minimize their adverse effects,
form or another, which is one of the main contributors to guarantee the continuity of load supply in all condi-
to greenhouse gas (GHG) emission. The electricity and tions and to increase the power system stability, an
power sector has a mere 2% share of oil consumption as energy management system is essential. International
compared with the transportation sector that has a stag- Electrotechnical Commission standard IEC 61970 defines
gering 56% oil consumption owing to increasing flight energy management system as “a computer system com-
and freight services3 and the rest being consumed by prising a software platform providing essential support
industrial, commercial, and residential users. services and a set of applications providing the function-
Meeting the emission goals pledged by the countries ality needed for the effective operation of electrical gener-
under the United Nations Framework Convention on Cli- ation and transmission facilities to assure adequate
mate Change (UNFCCC) will still leave the world with security of energy supply at minimum cost”. Energy man-
13.7 billion tonnes of CO2 (or 60%) above the recom- agement system for the smart grid can help maintain
mended level if the predefined temperature increase of supply‐demand balance, satisfying all the system con-
2°C is to be met by 2035.4 Growing power demand, pollu- straints to achieve economic, reliable, and secure opera-
tion due to GHGs emissions and a limited amount of fos- tion of the power system. Optimization is an integral
sil fuels are significant challenges in the emerging power part of energy management, which also ensures the
industry. Around 10% of the total power generated is lost decreased cost of energy production. Thus, the energy
in transmission and distribution, with 40% of this total management system controls and minimizes the quantity
loss occurring at the distribution level only.5 A possible and the cost of energy required for a specific application
and promising solution of energy crisis and pollution is by combining all systematic procedures.
to reduce overall dependence on fossil fuels by replacing
them with renewable energy sources (RES) such as pho-
tovoltaic (PV), wind power, and fuel cells. Moreover, 1.2 | Review methodology
replacing conventional gasoline transportation by electric
transportation, ie, plug‐in hybrid electric vehicle (PHEV) This paper utilizes the methodology of conducting critical
and plug‐in electric vehicles (PEVs), in addition to inte- reviews proposed by Baumeister and Leary, 1997,7 to
grating energy storage systems (ESSs) or battery energy provide interpretive analysis of existing literature in a
storage system (BESS) into an existing grid are some of particular topic of interest to reveal strengths, weak-
the alternative possible solutions to combat the exponen- nesses, contradictions, controversies, inconsistencies,
tially increasing (GHGs) emissions. Renewable energy is and/or other relevant issues with respect to theories,
characterized by variability and intermittency in their hypotheses, research methods, or results. The critical
output. Intermittency of RES can be overcome by inte- review follows a systemic review in the first step, where
grating more than one RES along with storage systems relevant keywords are defined and searched in peer‐
and backup sources. reviewed journals for different groups. Figure 1 illustrates
Similarly, charging of a large number of the PEV may the different keywords used for searching in relevant
affect power grid operation adversely.6 Hence, the afore- peer‐reviewed journals in the online articles database.
mentioned technologies in addition to mitigating GHGs The keywords in group 1 were selected based on critical
emission also pose numerous challenges, such as elements of study, namely, “energy management system”
increased complexity of the system, uncoordinated grid of an “energy sources.” Also, the environment of the
RATHOR AND SAXENA 3

FIGURE 1 Review methodology for energy management system

energy management system and the programs imple- (SHEMS), building energy management system (BEMS),
mented within were included in the keyword generation and energy management system aggregator (EMSA),
scheme. Based on the result outcomes from the first data- but very few have discussed uncertainties modeling,
base search using the keywords of group 1, the keyword optimization/programming techniques, and solution
list is further expanded to form group 2, which includes approaches used in EMS. In literature, one can find several
synonyms or the frequently used terms for group 1 cate- comprehensive review papers on EMSs. Cheng et al8
gory keywords. In the next step, results from the online presented a comprehensive review on microgrid energy
database were checked and evaluated for relevance to management system. The review topics include vital
the topic by screening the abstract and categorizing the components of a microgrid, comparison of centralized and
work identified as accepted for further comprehensive decentralized architecture, and its control methods. Shareef
study and analysis or rejected. At last, the interpretive et al9 reviewed the HEMS considering demand responses
analysis and evaluation of the identified work were done (DRs), smart technologies, and intelligent controller. The
to reveal merits, demerits, strengths, limitations, contro- residential DR programs in HEMS are compared, and
versies, inconsistencies, and other issues concerning applications of artificial intelligence for load scheduling in
components, stakeholders, performance, implementation, HEMS are also discussed. Rafique and Jianhua10 presented
and several other aspects of the energy management review on the EMS, generation, and demand predictors,
system. Thus, this critical review article provides with critical reviews on primary and secondary control of
constructive information to the scholars working in the microgrid including forecasting techniques for wind, solar,
direction of the energy management system about and load. Poompavai and Kowsalya11 presented a review
limitations of prior research, its strengths, and provides study on energy management strategies applicable for PV
direction to study and explore the area for further and wind‐fed water pumping system, but uncertainties
improvement. associated with PV and wind generation are not addressed
in the paper. Azuatalam et al12 provide a technoeconomic
assessment of seven different HEMS strategies including
1.3 | Contributions
their forecast, and it was found that practical implementa-
Research on energy management system (EMS) has tion of these strategies is limited due to the assumption
received more and more attention in recent years. Several made in modeling of different EMS. Zia et al13 provided a
review papers have addressed different types of EMS critical review on microgrids EMS with insights on various
strategies such as home energy management system solution approaches and communication requirement
(HEMS), smart home energy management system for microgrid. Vivas et al14 reviewed different energy
 4

TABLE 1 Existing literature review on the energy management system

Solution Year of
Sr. No. Ref. Types Architecture Solar Wind ESS PEV Stakeholders Uncertainty DR/DSM Approaches Publication
1 Poompavai and Kowsalya11 EMS ✓ ✓ 2019
12
2 Azuatalam et al EMS ✓ ✓ ✓ ✓ 2019
22
3 Yousefi et al EMS ✓ ✓ 2019
18
4 Zou et al EMS ✓ ✓ 2019
5 Cheng et al8 EMS ✓ ✓ ✓ ✓ 2018
6 Shareef et al9 HEMS ✓ ✓ ✓ 2018
15
7 Hannan et al BEMS ✓ 2018
13
8 Zia et al EMS ✓ ✓ 2018
14
9 Vivas et al EMS ✓ ✓ ✓ 2018
23
10 Zakaria et al BEMS ✓ 2019
17
11 Khan et al EMS ✓ ✓ 2018
12 Salimi and BEMS ✓ 2018
Hammad16
13 Hannan et al25 EMS ✓ 2018
10
14 Rafique and Jianhua EMS ✓ ✓ ✓ ✓ 2017
26
15 Ahmad et al EMS ✓ ✓ ✓ 2017
27
16 Carreiro et al EMS ✓ ✓ ✓ ✓ ✓ 2017
28
17 Meng et al EMS ✓ ✓ 2016
24
18 Zhou et al SHEMS ✓ ✓ ✓ ✓ ✓ 2016
19
19 Martinez et al EMS ✓ ✓ 2016
20 Lee and Cheng20 EMS ✓ 2016
21 Olatomiwa et al21 EMS ✓ ✓ ✓ ✓ ✓ 2016
22 Beaudin and HEMS ✓ ✓ ✓ 2015
Zareipour29
23 Proposed review EMS ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ‐

Abbreviations: BEMS, building energy management system; DR, demand response; DSM, demand‐side management; EMS, energy management system; ESS, energy storage system; HEMS, home energy management
system; PEV, plug‐in electric vehicle; SHEMS, smart home energy management system.
RATHOR AND SAXENA
RATHOR AND SAXENA 5

management approaches for hybrid renewable energy the overall stakeholder involvement and the distributed
systems based on hydrogen backup. It was concluded that energy resources (DER) behavior with their roles and
strategies that considers technical and economic criteria responsibilities in the EMS for smart grid environment along
are the most efficient and secure, though more complex with uncertainty quantification and different programs such
algorithms are required for their implementation. Hannan as DR, DSM, and power quality management (PQM) imple-
et al15 reviewed the prospects of internet of energy (IoE)‐ mented in an EMS. The main objective of this review paper is
based building energy management system (BEMS) for to provide a comprehensive critical survey of EMS concept
enhancing the energy utilization in a building. The review and framework with all stakeholders, DER behavior, and
suggests the requirement of advanced controllers integrated solution approaches to recognize trends, opportunities,
with IoE for sustainable energy usage in a building. The challenges, and potential barriers regarding the implementa-
paper also highlights the issues and challenges related to tion of EMSs to improve overall system performance. The
reliability, security for data access, scalability, cost, and findings provide new insights into the prospects of EMS for
weather, in the implementation of IoE‐based BEMS. Salimi researchers, policy makers, and manufacturers.
and Hammad16 presented a critical review of the methods The main contributions of the paper are summarized
for the collection and application of parameters related to as follows:
occupancy that effects the total energy consumption of the
building. Different control strategies were examined espe- 1. Overview of the EMS in the smart grid with its
cially considering heating, ventilation, and air conditioning different objectives and comparison of different
(HVAC) and lighting systems to understand the merits and architectures of the EMS.
limitations of existing methods and to identify future 2. Role and responsibilities of various stakeholders
research gaps. Role of a multiagent systems for distributed involved with the EMS and detailed analysis of DER
microgrid EMS is presented in Khan et al,17 and a compre- behavior.
hensive review of different optimization techniques is also 3. Critical reviews on uncertainty management, DR,
provided. Zou et al18 surveyed EMS for interconnected and PQM.
multi‐microgrids (MMGs). Various topologies of MMG 4. Critical and comparative analysis of solution
such as radial, daisy chain, and mesh topology are pre- approaches to the EMS.
sented. The paper reviewed the existing studies of EMS in 5. Key issues and challenges in the implementation of
microgrids and the existing decentralized optimization the EMS.
approaches. More review articles such as IoE‐based BEMS
by Hannan et al,15 energy management in PEV by Martinez Section 2 provides the objectives, architecture, and dif-
et al,19 energy savings by EMS by Lee and Cheng,20 and ferent types of EMS at the distribution level. Section 3
energy management strategies in hybrid renewable energy gives significant stakeholders and participants in EMS
systems by Oltamiwa et al21 do not cover all the major including system operator (SO), information communica-
aspects of EMS. Yousefi et al22 give a comprehensive analy- tion technologies, advanced metering infrastructure
sis and comparison of different forecasting models PV, PEV, (AMI), and aggregators. Section 4 discusses different
heat pump, and load demand. Zakaria et al presented a DER behavior in an EMS. Section 5 throws light on vari-
review of different uncertainty models for stochastic ous uncertainty quantification methods and other pro-
optimization approaches used for renewable energy grams implemented in an EMS such as DR, DSM, and
applications. It was concluded that for renewable energy PQM. Different solution approaches adopted for EMS in
systems, stochastic optimization methods give better the existing literature are reviewed in Section 6. Section
performance as compared with the deterministic 7 provides critical reviews addressing the challenges and
optimization methods.23 Different HEMS architectures issues in real‐time implementation of an EMS. Finally,
and their functional modules are reviewed in Zhou et al.24 Section 8 provides the concluding remarks. Figure 2 illus-
The paper also investigates different strategies to schedule trates the structure of the review paper excluding Section
home appliances with the objective to reduce electricity cost 1 “Introduction” and Section 8 “Conclusion.”
and improve energy efficiency. Table 1 shows existing
literature review articles on different types of energy
management strategy and their contribution concerning 2 | ENERGY MANAGEMENT
its types, architecture, components, stakeholders, solution SYSTEM
approach, uncertainty handling techniques, and programs
like DR and demand‐side management (DSM). The evolution of the EMS began in the 1960s, which was
Although several review papers on EMS have been termed as a control center, and later came to be known as
published recently, this is the first review that focuses on energy control center (ECC) during 1970s. It was further
6 RATHOR AND SAXENA

FIGURE 2 Structure of a review paper

renamed as supervisory control and data acquisition‐EMS system, better power quality, improve life expectancy,
(SCADA‐EMS) when SCADA system based on advanced and reduce maintenance and downtime. Integration of
computer came into existence during the early 1990s PEVs, ESS, and RES without proper coordination may
and finally evolved into real time known as EMS, which cause the system to deviate from its intended perfor-
includes various control techniques like load control mance, for example, PEV charging and discharging in
(LC), DSM, and distribution management system uncoordinated manger may lead to thermal stress on
(DMS).30 The function of EMS is to optimally allocate the distribution network and transformer,37 and uncoor-
different energy sources to the customers along with inte- dinated integration of RES may create reactive power
grating sustainable power sources without compromising problem that causes deviation in voltage and other
the reliability, security, and safety of the system.31 An issues.38 While setting up the EMS goal, technical objec-
EMS can monitor, supervise, optimize, and control the tive and related constraints should be considered. Eco-
consumers, distribution, transmission, and generation nomic objectives refer to the total operating cost of
facility.32 Thus, the prime duty of EMS is to create a bal- energy,39 cost to the customers,40 profit maximization of
ance between supply and demand in an efficient manner, aggregators and parking lot owners,41 etc. While setting
which should be cost‐effective within operating con- up economic objective technical constraints are also con-
straints and uncertainties (uncertainties in the EMS sidered, without considering technical constraints EMS
architecture include variation in generation from renew- optimization may provide the best economic outcome,
able energy resources, electricity price, and customers' but the system may be subjected to brownout, blackout,
behavior). It can work for real‐time SCADA applications, or appliances/equipment operating under a distribution
dispatch, control, energy scheduling, and accounting as network may be damaged. Environmental objectives
well as transmission security management. The EMS is cover GHGs emission minimization,42,43 where different
becoming more and more complicated, as the grid fossil fuel‐based generation (main utility or distributed
is evolving with the integration of PEV and RES, ESS, generation [DG]) and RES are scheduled, to achieve low
buildings high power requirement, and many more fac- carbon footprint generation. Social‐economic‐based
tors. However, the internet of things (IoT) and machine objectives are considered where economic activity is
learning are gaining popularity simultaneously, and both shaped by the social processes, which include the various
are very helpful for the efficient functioning of the EMS program in which end users, aggregators, agents, and net-
in the network. work operators (NOs) take active participation to achieve
the best economic outcome.44,45 The benefit obtained
from social welfare‐based programs is shared among all
2.1 | Objectives of EMS
the players, including the end users.
An EMS may have a single objective or multiple objec-
tives that may be technical, economic, technoeconomic,
2.2 | Architecture of EMS
environmental, and social‐economic. Numerous studies
have been done on EMS, but most of the research contri- Efficacy of EMS depends on the control architecture of
butions are focused on economic objectives. Technical the EMS, and the adopted solution approach, commonly
objectives of EMS include power quality,33 equipment used architectures are centralized, decentralized (distrib-
performance,34 and degradation of transformers,35,36 and uted) and hierarchical EMS.46 The three control architec-
its consideration provides better performance of the tures of EMS are discussed below.
RATHOR AND SAXENA 7

2.2.1 | Centralized EMS communication network available, the decentralized

architecture can be categorized into three modes of oper-
The centralized EMS architecture is characterized by one ation, ie, fully independent, partially independent, and
central controller with a high‐performance computing entirely dependent.46 In the fully dependent mode of
unit and a dedicated, secure communication network to operation, the local/distributed controllers communicate
manage the utilization of energy. The central controller with each other via a central entity. Though the commu-
may be a utility or an aggregator, which collects informa- nication among local controllers is dependent on a cen-
tion such as energy generation of DER, energy consump- tral entity, the decision is taken by themselves and not
tion pattern of each load/consumer, meteorological data, the central entity. In fully independent structure, commu-
other required information from market operators nication between local controllers is independent of a
(MOs), etc from all nodes to execute the optimization pro- central entity. While in the partially independent mode
grams to meets its objectives, for efficient operation. This of operation, the local controllers communicate with each
centralized control structure provides optimal global per- other and also with the central entity.49 The redundancy
formance, but it has its fair share of demerits as well. of controllers and communication contribute higher reli-
Since all information is gathered and handled at one ability in a decentralized EMS architecture as compared
place, the computational burden increases, especially if with centralized EMS architecture. Centralized EMS pro-
a large number of assets are to be controlled, thereby vides global optimization hence provides a reduction in
making this control structure less effective for real‐time total operating cost with consideration of all constraints,
communication requirements.8 The centralized EMS while local optimization in decentralized EMS will not
architecture is difficult to expand and may get interrupted be able to provide a solution for minimization of total
on integrating a new source/component, which has dif- operating cost as a whole. Whereas the decentralized
ferent operating cost and constraints. Moreover, it entails architecture provides the benefit of low computational
a single point of failure. In Olivares et al,47 a centralized burden and low response time as well and is more reli-
EMS for the isolated grid is proposed using model predic- able as compared with centralized EMS for the smart
tive control (MPC) technique where the problem is split grid. Ahmad et al26 proved that decentralized EMS has
into unit commitment and optimal power flow, to reduce better reliability as compared with centralized EMS. A
the computational burden. The problem is solved itera- reliability model based on Monte Carlo state sampling is
tively by multistage mixed‐integer linear programming implemented and tested on five bus residential microgrid
(MILP) and nonlinear programming (NLP), which makes where reliability index is compared for centralized and
it reliable for real‐time implementation. In Patnam and decentralized EMS in a microgrid.
Pindoriya,48 centralized stochastic energy management
approach using the sequential procedure for short‐term
operation of the aggregator is proposed to minimize the 2.2.3 | Hierarchical EMS
cost of consumers and risk of the aggregators using a
day ahead and real‐time operation. The approach is tested Based on decentralized EMS, the architecture of hierar-
on a 15‐node radial active distribution network consisting chical EMS is proposed and tested by few researchers
of 420 residential customers with the objective of minimi- especially for an MMG or microgrid community (MGC)
zation of consumer cost considering electricity price, PV system where the individual microgrids are intercon-
output, uncertainty, phase unbalance, and peak to nected to form a MGC. In hierarchical architecture, the
average ratio and solved by mixed‐integer nonlinear pro- system is divided into multiple levels of control with
gramming (MINLP). different control objectives at each level.50 Generally,
two‐level or three‐level structures are proposed. The flow
of information is only between adjacent levels, and no
2.2.2 | Decentralized EMS information is exchanged between units at the same level.
Most of hierarchical EMS includes supervisory control
The decentralized EMS architecture is characterized by level, optimization control level, and execution control
distributed processing system with each node having level. Furthermore, there is the possibility of sublevels
autonomous control capability and peer‐to‐peer commu- at each of the levels depending on the modeling of the
nication with other nodes. Thus, the decentralized EMS system. At optimization control layer, EMS is imple-
architecture overcomes the limitations of the centralized mented, which provide cost‐effective and reliable opera-
architecture by enhancing the expandability, allowing tion of the system if modeled with proper uncertainty
greater flexibility of operation, and avoiding single‐point handling techniques. Since coordination between adja-
failure.8 Depending on the level of decentralization and cent levels is required for this architecture, there will be
 8

TABLE 2 Merits and demerits of centralized, decentralized, and hierarchical EMS

Objective
Architecture Optimization Technical Economic Environmental Merits Demerits Ref.

Centralized Model predictive control Operating cost ‐ ● Centralized EMS ● Customer privacy is Olivares et al47
minimization provides global less
optimization ● A dedicated strong
● Reduction in overall communication link is
operating cost needed
● Uniform for the whole ● Computational
network so easy to burden is high
implement ● Involves single‐point
failure
● Poor flexibility and
expandability
Mixed‐integer linear Minimization of Minimization of the total ‐ Zaree and
programming (MILP) customers' cost Vahidinasab40
dissatisfaction level
Particle swarm Total operating cost ‐ Papari et al39
optimization (PSO) minimization
Decentralized Market‐oriented ‐ ‐ CO2 ● Provide higher privacy ● The separate objective Sugimoto et al42
programming (MOP) minimization to customers needs to consider
● Computational ● Exact optimization is
burden distributed on not possible
each EMS ● Increased total
● Higher flexibility operating cost
compared with central ● Effective
EMS communication
system and effective
synchronization are
required
MOP ‐ Operating cost CO2 Yakire et al43
maximization minimization
Monte Carlo state Reliability index ‐ Cheng et al46
sampling
Distributed model ‐ Total operating cost ‐ Mendes et al52
predictive control minimization
Convex optimization ‐ PEV user's convenience ‐ Wen et al53
maximization

(Continues)
RATHOR AND SAXENA
 RATHOR AND SAXENA

TABLE 2 (Continued)

Objective
Architecture Optimization Technical Economic Environmental Merits Demerits Ref.
Hierarchical Model predictive control Minimize unscheduled Microgrid cost ● Layerwise control can ● Complicated control Bazmohammadi
power exchange minimization enhance reliability and architecture and et al50
accuracy implementation
● Best suited for multi‐ ● High level of
microgrid or microgrid communication
community requirement
● Operating cost and
power quality both
can handle well with a
hierarchical control
architecture
Multi‐agent system Improved transient Minimization of operating GHG emission Ahmad et al54
performance cost considered
Markov decision process Minimize line losses and ‐ ‐ Zhuang and
maintain voltage Liang55
MILP Charging discharging of Minimize the operating ‐ Tian et al56
ESS, power exchange cost of MG
minimization

Abbreviations: EMS, energy management system; ESS, energy storage system; GHG, greenhouse gas; MG, microgrid; PEV, plug‐in electric vehicle.
9
10 RATHOR AND SAXENA

no transfer of information and energy if there is a com- generation and transmission level and end user level.
munication fault in the upper level. Bazmohammadi Since the customer is empowered and plays a key role
et al50 proposed a two‐level hierarchical stochastic EMS in smart grid, EMS at the distribution level can monitor
for operation management of interconnected four control and optimize the local system performance. At
microgrids. The upper layer performs energy scheduling. the customer end, some of the commonly implemented
It is at this level the power is exchanged between EMS are SHEMS,24,57 HEMS,9,24,29,57-61 BEMS,15,62-66
microgrids and the main grid, and the power exchanged and plant energy management system (PEMS).67 At the
within the microgrids network is calculated and commu- end users level, the ultimate role of EMS is to minimize
nicated. At a lower level, decisions are taken based on energy usage by properly scheduling the devices in speci-
chance‐constrained MPC considering various uncer- fied time horizons. The aim of BEMS, HEMS, PEMS, and
tainties. The proposed strategy not only minimizes oper- SHEMS is almost the same, but the functioning and
ating cost of considered four microgrids but also modeling of such EMS programs are different. HEMS
minimizes total average real‐time power imbalances in controls devices and manages power consumption of a
100 random scenarios. Han et al51 presented hierarchical house using the electricity tariff information received
energy management for PV/hydrogen/battery in an from different utilities. PEMS is mainly applicable to
islanded DC microgrid. Here, two control layers are con- industries where devices like heating, cooling, lighting,
sidered, local control layer and system control layer. or other curtailable loads can be scheduled to operate
Operating characteristics of sources in DC microgrid are during low tariff rates. Other than HEMS, BEMS, and
controlled in the local control layer, and in system control SHEMS, many another type of EMS are also cited in liter-
layer, equivalent consumption minimization strategy is ature such as factory energy management system
applied. Simulation results are validated using the RT‐ (FEMS)68 and electric vehicle EMS (EV‐EMS)69; in this
LAB real‐time simulation. Table 2 summarizes the merits paper, HEMS and BEMS are discussed in detail.
and demerits of the three control architectures.
Thus, the centralized architecture allows economic
implementation and simple maintenance of EMS, which 2.3.1 | Home energy management system
gives global optimal results; however, high computational
burden, high cost of communication, the privacy of data, “A home energy management system is a technology
and reliability are few issues with it. For microgrids, platform comprising of both hardware and software that
decentralized architecture with distributed processing allows the user to monitor energy usage and generation
offers higher reliability, low computational burden, but and to manually control and automate the use of energy
it does not guarantee global optimal results. For within a household.” Huang et al58 proposed a DR appli-
MGC/MMG, hierarchical architecture is a suitable option cation to optimize usage of appliances using chance
that provides different control levels. However, the com- constraint optimization to accommodate the uncer-
plexity of implementation is high. In the framework of tainties of electricity price and the load, and to reduce
EMS for smart grid, the robustness of EMS can be defined the computational burden, both particle swarm optimi-
as the ability to handle uncertainties in the presence of zation (PSO) and two‐point estimate method are used
uncertain loads, sources, and uncertain electricity prices while maintaining a high level of accuracy. Huang
to provide optimal results. Accurate uncertainty modeling
is the key criterion to achieve robustness in EMS, TABLE 3 Key features of HEMS and BEMS
whether it is centralized, decentralized, or hierarchical
HEMS BEMS
architecture of EMS.
• Optimize home energy usage • Efficient demand control
and response system
2.3 | Types of EMS at the distribution • Decision‐based on pricing policy • Minimize energy
level and tariff consumption within the
building
Environment preservation and energy saving are the two • Effective DR participation • Short‐term forecast of
key concerns in the emerging power system. The emerg- building load
ing grid aims to increase the responsibility and respon- • Consumer interaction and • Minimize waste of
siveness of each player in the electricity market towards involvement in energy thermal/heat energy
the clean environment and to the energy‐efficient, eco- management and conservation
nomical, optimal, and reliable operation of the system. Abbreviations: BEMS, building energy management system; DR, demand
The goal can be taken care of by the EMS at the response; HEMS, home energy management system.
RATHOR AND SAXENA 11

et al58 and Ahmad et al70 propose an optimized home 2.3.2 | Building energy management
energy management system (OHEMS) to minimize the system
electricity bill by scheduling the household appliances
and ESS in response to the dynamic price of electricity. BEMS can control and monitor the building energy needs
Qureshi et al59 implemented HEMS using Raspberry pi to minimize the overall energy consumption, considering
3 Model‐B to minimize electricity bill by considering the comfort and convenience of the customers. Signifi-
user‐defined priority‐based scheduling of loads in cant loads considered in BEMS modeling are HVAC,
response to real‐time electricity pricing. Kikusato et al60 lightning, and charging loads. BEMS can be applied for
presented a study for optimized utilization of PV output commercial and residential buildings. A building automa-
and mitigation of PV power forecast error using PEVs tion and control system (BACS) is proposed in Ożadowicz
charge‐discharge management through data exchange and Grela62 to minimize power and energy consumption,
between HEMS and grid energy management system using time‐driven and the event‐driven mechanism for
(GEMS) and hence minimizes the operation cost of the calculating demand. Ożadowicz et al63 suggested BACS
residential customers without disturbing the PEV usage. with demand elasticity model. The active DSM is done
Zhai et al61 presented a case study of HEMS for a resi- by extracting accurate and reliable information about
dential customer to take part in DR using smart plugs. the current and expected energy use. Manusilp and
DR is achieved efficiently using an appliance load moni- Banjerdpongchai64 investigated the optimal dispatch
toring method, taking into account the comfort level of strategy of cogeneration to minimize total operating cost
customers. Correa‐Florez et al71 presented a stochastic (TOC) and total CO2 emission (TCOE). BEMS proposed
framework base HEMS. HEMS modeling considered by the authors consists of combined heat and power
here include PV, water heater, and batteries, where (CHP), which utilized waste heat energy. The results
energy procurement cost is minimized in the wholesale show that the thermal energy storage (TES) was able to
electricity market. reduce 4.15% TOC and 5.25% of TCOE under optimal eco-
There is one another control used in HEMS known as nomic operation, while the reduction in TOC was 1.85%
homeostatic control (HC), first introduced back in early and in TCOE was 6.25% under environmental optimal
1980 by Schweppe et al, which is inspired from homeosta- operation as compared with cogeneration without TES.
sis regulation in a living organism where the organisms Mbungu et al74 presented a study on building EMS using
constantly adapt their intake and expenditure of energy single‐input single‐output MPC (SISO‐MPC) for smart
by continuously interacting with their environment to grid framework. The author has demonstrated that the
maintain their consistent internal equilibrium. Similarly, energy demand can be managed in real‐time electricity
HC tries to keep an equilibrium between demand and price environment. MPC strategy with some realistic con-
supply by interacting with the customers. The residential straint and numerous input parameter controls electricity
customers are kept informed of electricity time‐varying usage to achieve 50% to 60% saving in total electricity
prices and have the liberty to decide their independent cost. Mbungu et al65 also presented an MPC‐based
consumption rather than following the imposed electric- approach, to minimize the cost of energy flow under
ity consumption. HC finds its major application in the real‐time electricity price scenario. The power demand
microgrid, which is connected to the utility grid to supply for commercial buildings of Tshwane (in South Africa)
power to a group of residential customers. The microgrids is investigated, and the cost paid to the utility is reduced
can continuously interact with both the utility and the to around 46%. Mbungu et al75 presented a study on
customers and adapt to changing circumstances to main- building EMS system modeled using MPC technique for
tain the balance between supply and demand in a man- the time of use (ToU) electricity tariffs (off‐peak, pea,
ner to benefit both customer and utility. Yanine and and standard) to minimize the cost of electricity for com-
Córdova72 modeled HC and achieved energy efficiency mercial buildings. It was concluded that the MPC‐based
and sustainability in a grid‐connected wind PV system is robust enough to handle the system with
microgeneration system for two cases, one with ESS and numerous constraints, and the optimal cost of electricity
other without ESS. Yanine et al73 presented another case to be paid is highly affected by the reference turning
study on a small rural community in Chile using real parameter. Table 3 summarizes the key features of HEMS
electricity consumption data for a grid‐connected and BEMS.
microgrid with HC for reconciling power supply and To maintain the operating efficiency of home appli-
energy demand response management (EDRM). The use ances and to reduce the electricity cost, HEMS/SHEMS
of an ESS is proposed for improved performance in terms and BEMS are gaining more popularity. Also, with the
of thriftiness, energy efficiency, and energy sustainability increase in DG such as PV and wind, and implemented
of the system as a whole. programs like DSM and DR, the task of HEMS becomes
12 RATHOR AND SAXENA

more complicated. The following points need to be with communication links to pass on the necessary infor-
considered for the practical implementation of HEMS mation. Vital monitoring, measurement, and control,
and BEMS. information is passed upstream, and downstream stake-
holders of EMS through information and communication
• Computational time plays an important role and technology (ICT) interface. Figure 3 illustrates different
needs to be minimized by robust controllers. stakeholders and participants of the EMS with the pool
• The privacy of the user's data is an important concern information processing system. Most of the information
that needs to be addressed. flow is bidirectional. The two‐way information flow pro-
• Local implementation reduces the cost of communica- vides decision support to DER, consumers, PEV owners,
tion and links. aggregators, and agents, especially for programs like DR
• Comfort level of the customer should be a priority so and DSM, which enables consumers to actively partici-
that every end user keen to take part in such pate thereby providing services for the flexible and
programs. efficient operation of the grid. Big data analysis,76 IoT,77
and IoE (use of IoT with DER)78 enable to analyze, share,
and utilize information gathered from different levels of
an EMS. Implementation of such technologies has been
possible with the help of ICT infrastructure and commu-
3 | STAKEHOLDERS AND
nication protocols.
PARTICIPAN T S OF E MS

Significant participants of EMS are the main utility,

small‐scale prosumers (consumers and producers), distri- 3.1 | System operator
bution system operator (DSO), transmission system
operator (TSO), etc. Each participant has their function- Every grid has its standard and hierarchy of operators.
ing boundaries predefined in EMS algorithms. Stake- The responsibility of these operators is to ensure contin-
holders and participants of the smart grid are connected ual matching of supply and load demand economically

FIGURE 3 Stakeholders and participants of the energy management system [Colour figure can be viewed at wileyonlinelibrary.com]
RATHOR AND SAXENA 13

with a certain level of predefined reliability, quality, and Sophisticated, reliable, and fast communication infra-
at minimum environmental impacts. The TSOs and inde- structure is, in fact, necessary for communication among
pendent system operators (ISOs) have their role and the large number of distributed elements, such as genera-
responsibilities at the transmission level, ie, between gen- tors, substations, ESSs, and end users, to enable real‐time
erating plants and distribution substation, whereas the exchange of data and other necessary information for the
power delivery at the low voltage level is the responsibil- management of the system. An effective communication
ity of DSOs and utilities. Figure 4 illustrates the role and system is essential for ensuring improvement in terms
responsibility of each player. The role and responsibilities of efficiency, reliability, flexibility, and investment return
of these operators vary with the countries depending on for all those involved in an EMS as producers, operators,
the organization of the electricity sector. Many country and customers. Communication technologies used for
grid codes use the term SO (or network operator in the effective EMS implementation can be broadly classified
UK) for TSOs and DSOs, and there is a MO at higher level into wired and wireless technologies. Wired and wireless
as an economy handling authority. The role of SO is to technologies along with the respective range and trans-
tackle the conditions of congestion (maximum loading mission rate are given in Figure 5. Table 4 compares
online), overloading (beyond congestion), etc.79 The rapid different wired and wireless technologies concerning the
growth of DER in the energy sector necessitates the need standard, concerning protocols, bandwidth, data transfer
for a mediator, where TSOs/DSOs play an essential role. rate, applications, advantages, and disadvantages. Com-
The ultimate role of TSO/DSO is to optimize investments, munication requirement of EMS depends on the type
capacity management, and the behaviors of system par- and scale of EMS implementation. Centralized EMS
ticipants. TSO is mainly responsible for the operation of requires dedicated and robust communication links.
ancillary services and direct integration of resources at SCADA, the programmable logic controller (PLC), and
the transmission grid. Primary conventional sources and optical fiber are well‐known, old technology but still the
abundant RES can provide their services directly to TSO most preferred technology for wired communication in
and receive operation signals from them, whereas DER centralized EMS implementation. Decentralized EMS
and consumers provide their services to and receive oper- can have distributed communication links and are lim-
ational signals from the DSO. DSO facilitates open access ited to a shorter range, such as WLAN, RF, and ZigBee.
to the competitive electricity market, enables customers BEMS, EV‐EMS, FEMS, and HEMS require short‐
to be prosumers, and optimizes the uses of DER to pro- distance information flow that can be done with cost‐
vide secure, sustainable, and affordable supply. With the effective technologies such as Bluetooth, ZigBee, WLAN,
transition of the distribution network from passive to Wifi, and WiMax.
active network with the increasing penetration and utili- Ding et al80 presented a simulation case study on a
zation of DER, a high level of coordination is required novel communication scheme employed for a distributed
between TSOs and DSOs to keep the grid stable. EMS for the smart grid to maximize total social welfare
cost that balances generation side expenses transmission
line cost including losses and consumer payment. A
3.2 | Information and communication new event trigged communication for each generator,
technologies and each load is proposed instead of continuous or peri-
odic sampling data transfer and was validated on IEEE
ICTs represent a fundamental element in the practical 9‐bus system. The proposed method significantly reduces
implementation, growth, and performance of EMS. the information flow while maintaining the same

FIGURE 4 Role and responsibilities of transmission/distribution system operator (TSO/DSO)

14 RATHOR AND SAXENA

3.3 | Advance metering and pool

information system

AMI is a prerequisite for the management of energy flow

between the consumers/prosumers and the utilities. The
services provided by AMI can be categorized into five ser-
vices, namely, consumption awareness, market services,
scheduling and control, network services, and diagnos-
tics.82 Figure 6 gives the application of services and the
data input required for each of these services. The basic
building blocks of AMI include SMs, communication net-
work, meter data acquisition system, and meter data
management system, which can communicate with the
HAN. An SM measures the physical quantities, records
different events like outages or any changes in contrac-
tual rated power, etc, and converts the information into
digital form before transmitting it to a central acquisition
system. In addition to collecting data on energy flow, SM
FIGURE 5 Wired and wireless communication technologies and
also measures voltage level and power at customer's pre-
transmission rate
mises to be sent to DSO. The acquired information is used
by DSO for network management (eg, minimizing net-
control performance as with continuous data transfer. work losses and for customer management, ie, it provides
Aftab et al81 presented the IEC 61850 and XMPP validated data to retailers to be used mainly for billing
communication‐based EMS for microgrids with a high reasons). Energy management, at any level, relies on the
penetration of EV‐EMS. Since in microgrid, there is a information sent or received from AMI.
bidirectional power flow from PEVs and home base PV The effective execution and functioning of AMI and
installation, the whole approach is data‐driven to main- SMs mainly depend on the choice of communication pro-
tain resiliency, reliability, and security of the system. tocol. The bidirectional capability of the SMs between end
IEC 61850 modeling of EV components was extended to users and the utility allows the customer to take active
included vehicle‐to‐grid (V2G) support of PEV, which in participation in various programs run by the utility such
turn provides effective communication. The communica- as DR and real‐time pricing. One‐way and two‐way com-
tion is proposed through the XMPP protocol, which munications are two modes of data transfer; one‐way
enhances network security and scalability. The additional communication is preferred for low cost and where no
components required for communication in SHEMS, EV‐ control action is required, while the two‐way communi-
EMS, BEMS, and HEMS are home area network (HAN), cation is preferred where control action is required for
smart meters (SMs), human‐machine interface (HMI), programs like home automation, PEV control, and DR.
smart plugs, sensors, and control device (actuators), In such programs, wireless communication is the most
smartphones, and smart appliances. Wired communica- preferred mode of communication. A large variety and
tion such as DSL, PLC, and optical cable is still the most quantity of data from AMI is required to perform the task
preferred communication for EMS due to its higher range of optimal energy allocation. Various types of
of data transmission; still, cost is an issue. While wireless data/measurements from hundreds of SMs is collected
technologies such as ZigBee, Bluetooth, WiMAX, Wifi, at data concentrator and is processed and repackaged
and GSM are preferable for remote locations with a low before sending it further to centrally at the pool informa-
rate of data transmission and lesser cost, but interference tion processing system. The raw data provided as input to
is an issue with wireless technologies. With recent EMS are forecasted generation data, weather data, data
advancement in EMS deployment, more sensors, local from geographical information system (GIS), consump-
controllers, and meters are needed to be integrated for tion forecast, and road traffic data, etc. Additionally,
monitoring and controlling purposes. Wireless sensor PEV owners can provide input to PEV agents for schedul-
and IoT controller devices are becoming very important ing charging and discharging of PEV. PEV agent can pass
elements in monitoring and managing energy use. How- the information to EMS for further processing. Data are
ever, the interoperability and network security are a typically obtained from various wired and wireless links,
major concern for the communication technology to as discussed in the previous section. Missing data or
maintain user's privacy. corrupted data can be repaired or predicted with the
RATHOR AND SAXENA 15

use of state estimation. Processing of data involves the to avoid economic losses, user's privacy, and reliability.
categorization, clustering, prediction, correction, etc, Both the customers and utility are benefitted from AMI
which is used for the decision‐making process using concerning reliability, security, operational, and financial
optimization/algorithm programs. Based on the concep- benefits. Still, there are many challenges in the imple-
tual framework of the standard AMI, Mbungu et al83 mentation of AMI, such as high capital cost, integration
proposed the design of an optimal single‐phase energy issues, and standardization.
meter to manage the electricity demand. The energy
meter enables the customer to set a minimum electricity
3.4 | Aggregators
tariff, thereby optimizing the overall energy cost in a
real‐time environment. The quadratic MPC with realistic Deregulation, decarbonization policy, and the recent
constraints is used to achieve the desired objective. advances in ICT have led to increased penetration of
Sampling time of metering data and the latency in the DERs. With increased penetration and utilization of
communication between the metering point and the final DERs, consumers have become more active participants
node are the two crucial parameters that are to be taken of the electric grid either acting as a prosumer or by
care of in AMI. Data authenticity, confidentiality, and engaging in DR programs and can provide many system
integrity are an essential part of any EMS implementa- services, which may lead to a more efficient and flexible
tion. Protecting EMS from malicious attack is necessary power system. The complexity of the grid increases with

TABLE 4 Wired and wireless communication technologies performance comparison

Transfer
Protocols Rate Range Standard Application Advantages Disadvantages

Wired SCADA 51.8 Thousands IEEE STD. Preferred for Can handle extensive data, Require skilled
Mbps mile 1815‐2012, centralized allows sensor connections, operator, higher
to 2.48 IEEE 60870 EMS RTU connection possible, cost, high
Gbps fast response complexity
PLC 200 200 or more ISO/IEC Useful for Easy interfacing, easy in The high complexity
Mbps 14908‐3 centralized programming of setup, difficulty
EMS in error debug
Optical 100 Approximately IEEE 802.3ah For distributed Great bandwidth, fast rate, Limited application,
fiber Mbps 100 km and cheap, less signal fragility, repeater
to 2.5 (without a centralized degradation, long life needed for long
Gbps repeater) EMS distance
application
Wireless WLAN 54‐900 100 m IEEE 802.11 Preferred for Speed, accuracy, and flexibility Higher connection
Mbps home reduced data
automation transfer rate,
lower bandwidth
Bluetooth 721 Kbps 100 m IEEE 802.15.1 Best suited for Cost‐effective, less interference Less security,
2.4 GHz HEMS interference from
RF, power
consumption
ZigBee 20‐250 10‐100 m IEEE 802.15.4 Home Less power consumption, cost‐ Less privacy, low
Kbps automation effective, reliable and self‐ transmission rate,
915 MHz application healing short
transmission
range
WiMax 70 Mbps 50 km IEEE 802.16j/ Suitable for Can serve hundreds of users, Weather can affect,
2‐6 GHz m AMI long distance interference
problem, high
cost
Cellular 14.4‐100 5‐100 km LTE (2009, For automation Handy in nature, easy upgrade, Vulnerable in
Mbps 3G), LTE applications wireless connectivity nature, the higher
Advanced initial cost
(2011, 4G)

Abbreviations: EMS, energy management system; HEMS, home energy management system; PLC, programmable logic controller; SCADA, supervisory control
and data acquisition.
16 RATHOR AND SAXENA

FIGURE 6 Services provided by advanced metering infrastructure (AMI)

the participation of prosumers. To tackle the challenges and aggregators serve as master control. If multiple EMS
such as price volatility and unbalance in supply and devices are incorporated and work in a decentralized
demand, EMSA plays an important role, which is a new mode in the network, all can try for achieving benefits
type of electricity service provided to optimize the con- for the end user and utility that may lead to a violation
sumption of energy for a group of consumers by produc- of power quality limits or may lead to instability in the
ing energy and by selling excessive energy. According to grid. That is why EMSA is needed to coordinate the
the smart grid task force report “Regulatory Recommen- whole process. Each small generation facility (FC, PV,
dations for the Deployment of Flexibility”37 published in wind, or any DG) and the consumer cannot take direct
2015, aggregator is considered as a legal entity that pro- participation in the electricity market for selling or buy-
vides an opportunity to harness the flexibility potential ing electricity, and aggregator plays the role of mediators
of small end users by promoting their access to the retail for such consumer or producer to participate in the power
electricity market to sell load flexibility and gain benefits exchange market. Figure 7 illustrates the coordination
or lower energy bill. Aggregator involves with EMS task between aggregators at the supply and demand side with
known as EMSA provides ancillary services, DR, and load market and SOs.
flexibility and reduces the burden to the SO with direct
interaction with end users, and numerous benefits can
be achieved. EMSA can allow much better management
of power system requested by SO to match supply and
demand by controlling load or utilizing locally available
RES. Various types of aggregator for EMS programs have
been proposed by researchers such as EMSA, PEV
aggregator, DER aggregator, load aggregator, and
microgrid aggregators. An aggregator is a legal agent
and can take part in financial electricity market (intraday,
a day ahead, etc.) participating in regulating power mar-
ket. EMS may have device management system (DMS),
battery management system (BMS), and power conver-
sion system (PCS), where the user interfaces act as a slave FIGURE 7 Aggregator at demand and supply‐side
RATHOR AND SAXENA 17

Aggregator role is to sum up such small prosumers in electric car, accounting for more than 40% of the electric
sufficient volume to take part in the market, and the rev- cars sold in the world and more than double the amount
enue generated is shared among producer, consumer, and sold in the United States. With the global impetus to “go
aggregator. An aggregator is not a generator or marketer, green,” promotional policies, and incentives given by the
it is identified as a retailer, which purchase electricity and government, sale of PEVs is increasing every year all over
resell to end users. Sardou and Azad‐Farsani presented a the world. PEV can prove to be the key solution to reduce
microgrid aggregator under an uncertain system with the the emission of GHG. Smog rating indicates vehicle tail-
objective of expansion cost minimization and line loading pipe emissions that contribute to local and regional air
index minimization using gravitational search algorithm pollution, creating problems such as health issue, smog,
(GSA).84 Graditi et al85 presented a study on optimal bid- and haze. A score of 10 indicates the cleanest vehicle.94
ding strategy for DER aggregator where sources like DG, Table 6 gives the US government fuel economy informa-
storage devices, and DR are considered in the day‐ahead tion, which indicates PEV to be the cleanest solution for
market taking into account customers power demand transportation. PEV includes both PHEV (series, parallel,
flexibility. The uncertainty of the system is modeled using and power‐split powertrain) and PEV (BEV/all‐electric
a roulette wheelbase scenario generation and scenario vehicle). They differ in construction, fuel economy, and
reduction. The profit of an aggregator increased by proper efficiency. Energy management algorithms for PHEVs
interaction (communication) among all local energy con- are crucial for vehicle performance. PHEV can operate
trollers under the optimal bidding strategy, which is in charge sustaining and charge‐depleting mode. During
solved by the MILP framework. The concept of virtual startup, it works in charge depleting mode and later on
power plant (VPP) is proposed to enhance power genera- switches to the charge sustaining mode once the battery
tion and trading of power in the electricity market, which reaches to its threshold state of charge. These different
is a cloud‐based distributed power plant that aggregates energy management modes are automatically controlled
the small‐scale DER. Liang et al67 proposed an EMS to depending on the state of charge, vehicle speed, engine
minimize the TOC of the VPP along with maintaining torque, environment, temperature, battery temperature,
power quality of the network. VPP structure consists of and air conditioning.96 The two basic control modes of
a combination of aggregation of consumers and RES. PEV are charge depleting and charge sustaining, depend-
Meng et al86 demonstrated dispatch of a virtual ESS in ing on the state of charge. The very first level is at home
the smart distribution network for load management where PEVs can contribute to HEMS. Second are the
where significant air conditioning loads are considered, parking lots (commercial buildings, malls, offices, and
and active power curtailment is done among aggregators municipal parking lots) subjected to availability of
while maintaining the room temperature at a comfort parking space, type of charger (fast/slow charger), and
level. Table 5 summaries various types of aggregator total parking duration, which are provided with big
involved in EMS programs. To interact with end users, charging stations with fast charging. Wireless communi-
EMSA required ICT infrastructure for IoT, net‐zero cation is a prime requirement due to the mobile nature
energy building (NZEB) programs to exchange data for of PEV.
further processing or to send control signals to manage PEVs that can operate in grid‐to‐vehicle (G2V) (electric
the load. To deal with individual end users, the cost of power flows into PEV) and vehicle‐to‐grid (V2G) (PEV
IoT should be low, and wireless communication should acts as a DG injecting power into the grid) mode of opera-
be preferred with utmost reliability and security. The tion in an EMS are associated with both opportunities and
practical deployment of EMSA still faces challenges and challenges. They add value to the power system by
barriers such as regulatory framework, technical stan- providing many ancillary services such as (a) frequency
dard, and protocols, which need to be addressed. There and voltage regulation,97,98 (b) peak load shaving and val-
is a need for cooperation between academia, industry, ley filling with bulk power support,99,100 (c) reactive power
and policy makers to develop low cost, reliable, and support,101,102 (d) spinning reserve, and (e) none spinning
secure EMS. reserves103; benefit of such services can be achieve with
the help of EV‐EMS illustrated in Figure 8. PEVs act as
the distributed mobile energy storage device, facilitating
4 | DER BEHAVIOR better integration of renewable energy and participate in
active energy management to ensure DSM for reliable
4.1 | PEV behavior and EV‐EMS and stable operation of the emerging grid.104,105 However,
uncontrolled charging of a large number of PEVs can lead
According to Global PEV outlook93 report, in 2016, China to an added undesirable peak in the distribution system,
was by far the largest manufacturer and seller of the reduced reserve margins, and increased power losses,
18 RATHOR AND SAXENA

TABLE 5 Role of aggregators

Ref. Type Algorithm Role Objective

Rahmani‐Andebili and Demand‐side resources Dynamic game Loads aggregator and Enhance profit,
Venayagamoorthy87 aggregator theory PEV aggregator customers' income,
and market efficiency
Mohiti et al88 EV aggregator ADMM As a mediator between The total cost of the
Distribution network system minimized
(DN) and PEV
Shafie‐Khah et al89 Demand response Stochastic Increase customer Improves stability,
aggregator optimization participation in the reliability, and
electricity market market efficiency
Pei et al90 Microgrid aggregator Stochastic Real‐time balancing Minimize economy and
optimization market bidding improved bidding
strategy
Sauba et al91 RES aggregator Genetic algorithm Domestic prosumers Increase profit for the
RES aggregation and prosumers and
act as a mediator minimize the penalty
Somma et al92 DER aggregator Stochastic MILP Acting toward grid as Better for finding
one entity for DER bidding curve and
resources improve the
interaction of DER
with local energy
sources (LES)

Abbreviations: ADMM, alternating direction method of multipliers; DER, distributed energy resources; EV, electric vehicle; MILP, mixed‐integer linear pro-
gramming; PEV, plug‐in electric vehicle; RES, renewable energy sources.

voltage deviations, and deterioration of power quality.106 battery (charge), battery‐to‐home loads (discharge), and
Uncontrolled charging of PEVs may also have adverse battery to the grid (discharge). The main considered oper-
effects on voltage control, power quality (harmonics and ating modes are G2V, V2G, vehicle‐to‐home (V2H), and
subharmonics), and supply‐demand balance and can also vehicle‐to‐building (V2B) to reduce operating costs for a
cause the protection relay settings to be changed.106 As consumer having single and dual EVs. Table 7 gives differ-
PEVs support the bidirectional transfer of energy between ent modes of energy flow from PEV, along with their
PEV and distribution system, the PEV battery is interfaced advantages and limitations. Numerous studies have been
with utility and house load point through power done on the charging of PEVs for loss minimization, opti-
electronics‐based bidirectional AC/DC converter. The mal power utilization, fast charging, price‐based charging,
converter controls power flow direction between grid‐to‐ and incentive‐based charging. In the market, level 1, level

TABLE 6 US government fuel economy data95

Fuel
Economy Annual
Sr. Vehicle Cityhwy, Fuel Saving/Year, GHG Smog
No. Type MPGe Cost, $ $ Rating Rating

1 Gasoline vehicle 26 2150 1850 7 6

2 Gasoline vehicle (Gas Guzzler Tax) 11 5400 14 400 1 5
3 Diesel vehicle 35 1650 4350 9 6
4 Electric vehicle 99 600 9600 10 10
5 Flexible fuel vehicle (E85) 24 2300 1100 7 6
6 CNG 29 1050 7350 10 8
7 Hydrogen 56 1450 5350 10 10a

Abbreviations: CNG, compressed natural gas; GHG, greenhouse gas; MPGe, mile per imperial gallon.
a
Score 10 indicates the cleanest vehicle from smog exhaust.
RATHOR AND SAXENA 19

FIGURE 8 The typical structure of electric vehicle energy management system EV‐EMS [Colour figure can be viewed at
wileyonlinelibrary.com]

2, and level 3 chargers are available with charging time discharging limits of PEVs in the parking lot. Effective
around 18 hours, 8 hours, and 30 minutes, respectively.114 exploitation of RES can also be taken into account, along
Based on the locations of the charging facility, PHEV with charge and discharge management.
charging can be categorized into two groups: residential PEV aggregator plays a significant role in managing
charging/home charging and nonresidential PEVs charging/discharging power in the parking lot to
charging/public charging/parking lots/charging stations. maximize their effective utilization along with customer
Home charging has individual charging points directly satisfaction and to minimize the cost without increasing
connected to the low‐voltage distribution network, with the adverse impact on the grid. From PEV owners' point
slow or average charging rate with the moderate power of view, they wish to opt fast charging of the vehicle or
requirement of 1.6 to 9.6 kW. However, charging stations maximize their revenue while discharging and minimize
have several connecting points for fast charging connected charging cost using optimal charging strategies.109 G2V
to MV distribution networks.107,115 PEVs charge/discharge mode of operation might slightly affect the performance
management plays a crucial role in EMS. State of charge of of main or microgrid, but V2G mode affects system
the battery, tentative travel distance, load conditions, and parameters due to in‐feed, for example, protection
owner's willingness to participate in V2G energy transfer schemes need to be changed because of reverse power
are few of the factors to decide the charging and flow. The solution can be achieved with the help of
20 RATHOR AND SAXENA

TABLE 7 Energy flow from PEV batteries

Mode Benefits Limitations Ref.

G2V Reduction in GHG High complexity in integration

Economic operation Lack of user participation and awareness Hadley and Tsvetkova106
Load flattening possible Dedicated communication infrastructure required Fernandez et al107
Require distribution system upgradation
Smart charging required
V2G Increase user participation Dedicated communication infrastructure Sexauer et al34
Reactive power support May result in high battery degradation Johnston et al,108 Tabari and Yazdani,109
and Amini and Islam110
Frequency regulation Fair pricing policy required Graham111
Improved reliability
Peak shaving111
V2H Reduced electricity bill Market interaction and exchange reduces Graham111
Support in outages Subject to availability of PEV An et al112
Power balancing
More benefit during high tariff
Load shedding mitigation112
Peak shaving111
V2B Benefit for HVAC loads User contract needs to be signed Johnston et al108
Peak load reduction Complicated operation Bessa and Matos103
Reduced electricity bill Pricing issues for discharging/charging Tomić and Kempton104
Uncertainty balancing Market interaction reduces Clement‐Nyns et al113
Less investment for backup power
Power quality improvement

Abbreviations: GHG, greenhouse gas; G2V, grid‐to‐vehicle; HVAC, heating, ventilation, and air conditioning; PEV, plug‐in electric vehicle; V2B, vehicle‐to‐
building; V2G, vehicle‐to‐grid; V2H, vehicle‐to‐home.

aggregator that acts as middlemen for the proper opera- the battery, number of batteries to be a charged, and the
tion of the grid without violating the constraints of the owner behavior. The uncontrolled charging of PEVs
system and decide the mode of operation of PEV to be may occur coincidentally during peak hours because of
V2G or G2V depending on the actual or predicted price the consuming habit, which will adversely affect the
of electricity and power profiles. It also helps to make a operational performance of power systems dramati-
reliable offer for the day ahead of ancillary services. cally.37,107 Abdelsamad et al115 stated that uncoordinated
Aggregator in most of the cases has all the information charging leads to an overload of transformer and cable in
of PEVs related to the owner behavior, charging‐ the secondary distribution system (SDS). To mitigate
discharging time, battery capacity, weather conditions, these adverse effects of PEVs, smart charging is the solu-
traffic data, etc that can be helpful for prediction of charg- tion. Smart charging refers to the controlled charging,
ing or discharging power from PEVs. In Sortomme and which aims to reduce the operational cost of utilities
El‐Sharkawi100 and Amini and Islam,110 allocation of and cost of energy to PEV owner. PEV charging stations
parking, lots for minimization of electricity cost, and in the market with control capabilities provide opt‐out
maximization of profit for the aggregators is discussed. options to drivers in the event of smart charging control
Special charging lots are designed with bidirectional char- or DR. Kamankesh et al117 demonstrated that smart
ger, which coordinates with the aggregator and the grid, charging is best among the three charging strategies,
for optimal parking lot allocation to minimize distribu- namely, uncontrolled, controlled, and smart charging
tion system losses. Another work in El‐Zonkoly et al116 schemes, to overcome issues of grid constraints violation
presented a study on IEEE 33‐bus radial network with and to achieve economical operation. Energy manage-
the multiobjective algorithm to determine the number ment of PEV for V2G and G2V mode is discussed in
of parking lots, location and size of parking lots, energy Nunna et al.118 In the case where accurate forecast data
scheduling, and minimizing overall energy cost using to the microgrid controller is not available, the
artificial bee colony (ABC). charging‐discharging schedule is decided by real‐time
Charging power requirement from the grid is highly generation and demand (5 to 15 minutes prior) known
uncertain and depends on a variety of parameter such as decentralized control. Whereas when accurate forecast
as the type of chargers, level of charging, the health of data are available on a day ahead or half‐day ahead basis,
RATHOR AND SAXENA 21

then charging‐discharging scheduling of PEVs is centrally flywheel energy storage system (FESS) setpoint using
decided by aggregators. Another attractive feature of V2G linear MPC, and the second controller is a real‐time adap-
mode of PEVs is reactive power compensation. This prop- tive hysteresis controller for minimizing wind power pre-
erty allows the PEV battery to discharge without the diction error. Foley et al126 presented a comparative study
exchange of active power, thus avoiding any excess load on wind power uncertainty and prediction. Kundur
on the grid.119 Reactive power compensation studies for et al123 presented a case study on the Dutch electricity
single PEV and for charging stations are presented in pool. Short‐term probabilistic prediction of wind power
Buja et al119 and Bo et al,120 respectively. Thus, PEVs is made to minimize the penalty of regulation cost and
have both benefits and limitations when integrated to to maximize the revenue. For the secure operation of
the grid, and it is required to validate the practicability the smart grid, an accurate prediction or forecast of RES
of PEV role in DR, DSM, and as mobile energy generation is highly demanded. Zhao et al127 presented
storage/load with the integration of RES. The uncertainty case study on MMG with system of systems (SoS) archi-
is still an issue that needs to deal with PEV. The comfort tecture with hierarchical decentralized EMS using bilevel
level of the customer at the same time should be consid- optimization to tackles issues of RES uncertainty (wind
ered with all technical, economic, and social factors. and solar) for reliable, economical, and power balance
problem. RES power can be either a grid‐connected or
standalone. Table 8 gives different EMS in a system inte-
4.2 | RES behavior grated with RES with their objectives and solution
approaches.
Extensive RES technologies exist, however, most com- Mbungu et al135 presented a study on the use of renew-
monly employed are wind and solar in most of the coun- able energy, eg, PV, wind with battery storage to mini-
tries. Many researchers proved the efficacy of EMS mize the utilization of utility grid for electricity usage in
algorithms with consideration of RES into the system the industry using MPC technique. Based on the context
model.121,122 The intermittent nature of RES and an accu- and background of the study, it is found that RES play
rate prediction of power mainly for wind and solar to an important role in improving security, climate changes,
avoid mismatch in supply and demand are key challenges and economical benefits to all the stakeholders involved
with RES participation in EMS. The output power from with RES. To meet the energy requirement of rural and
RES may not be adequate to meet demand at some time remote areas, especially where no distribution network
while at other times, the generated power may be more is available, the use of hybrid renewable energy system
than the demand. Any imbalance between generation (HRES) working in a microgrid environment is increas-
and supply would affect the system frequency, which ing. In addition to the reliability problem due to intermit-
may result in loss of the stability of the system.123 The tent nature of RES, the cost of HRES is also high. Thus,
unpredictable and variable nature of RES makes it less there is a need to develop efficient and economically prof-
reliable with regard to generation planning and schedul- itable hybrid microgeneration of small kilowatt range
ing of power from RES. Wind power is dependent on that can be used for residential purposes or for providing
wind speed, and solar radiation received at the surface small backup power applications. Yanine and Sauma136
of the earth is also variable and intermittent. It depends proposed a theoretical model for implementing supervi-
on cloud cover variability, atmosphere aerosol level, geo- sory control strategies based on HC for grid‐interactive
graphical location, and other atmospheric parameters. A HRES; the HRES is analyzed as a complex adaptive sys-
possible solution to handling uncertainty in wind power tem that can learn and change to sustainable systems.
is to consume the whole power locally. The consumption The simulation study on hybrid energy system without
can be in the form of heating load, PEV charging, etc. energy storage connected to the grid to demonstrate the
Caballero et al124 presented a study on the business above concept is presented by Yanine et al,137 presented
optimal design of system comprising of the small‐scale a simulation study on hybrid energy system without
grid‐connected hybrid energy system of PV and wind tur- energy storage connected to the grid. Homeostatic regula-
bine. The author has applied the concept of net billing tor for the homeostatic controller has been modeled and
and net metering for bidirectional power flow from con- simulated for five distinct criteria and tested for different
sumers end and proved that such a scheme is beneficial scenarios for energy efficiency, thriftiness, and energy
for customers. It is also shown that there is a significant sustainability. Many authors proved that utilization of
reduction in the life‐cycle cost of installation, subjected RES provides flexibility in power system and potential
to space and investment. Abdeltawab and Mohamed125 benefits. However, they did not discuss variability and
minimize the power losses associated with wind flywheel uncertainties involved while integrating RES. Presence
using two ESS controllers. The first controller defines a of RES requires more flexibility in power system for
22 RATHOR AND SAXENA

TABLE 8 Energy management with consideration of RES

Solution
Ref. Configuration Approach Objective of EMS

Liao and Ruan128 Standalone/PV/ESS Linear programming Improved system, dynamic performance, and
efficiency. The solar cell array powers the
steady‐state energy, and the battery
compensates the dynamic energy
Ismail et al129 Stand‐alone PV/ESS/ Linear programming To minimize cost using PV and ESS, during
microturbine (MT) excess load demand MT is utilized
Dash and Bajpai130 Stand‐alone PV/ESS/FC Linear programming EMS aims to divert excess energy of PV to the
electrolyzer
Finn and Fitzpatrick131 Grid‐connected WT Linear programming WT is used to supply load during high demand
periods
Ozbilen et al132 WT/Hydro/FC Linear programming To manage intermittent renewable power,
hydrogen energy storage is proposed to be
utilized during peak load periods
Abedi et al133 PV/WT/FC Differential algorithm + Minimize cost, power balancing, and fuel
fuzzy consumption by utilizing RES energy
Dahmane et al134 Standalone WT/PV/ESS/ Linear programming Cost‐saving by utilizing PV as the main source of
diesel energy, WT used as a supplement of PV, and
diesel generator is as an additional source to
charge ESS batteries

Abbreviations: EMS, energy management system; ESS, energy storage system; PV, photovoltaic; RES, renewable energy sources; WT, wind turbine.

supply and demand compared with traditional power sys- grid with a variety of energy sources, sometimes, the gen-
tem without RES. Hence, there is a need for modeling eration will be more than the load demand, and then the
uncertainties associated with RES effectively and to intro- excess energy from the renewables can be stored in bat-
duce standards and policies for the complex interdepen- tery stacks. These batteries storage can act as battery
dencies of the renewable‐based energy system. exchange stations for PEV where fully charged batteries
can replace low SoC batteries of the vehicle. These batte-
ries can also act as a voltage support device or power
4.3 | ESSs behavior quality compensator. Ultimately, it can provide ancillary
services like peak power shaving, load shifting, flattening
ESSs are yet another essential component of any EMS. load profile, and spinning reserves. For ESS, two most
Various energy storage technologies such as mechanical common configurations are distributed (several units
(pumped hydroelectric and compressed air), electrochem- placed in different locations) and aggregated (all ESS
ical (flow battery, secondary battery, and supercapacitor), located at one place). Most of the ESS generates DC
chemical (hydrogen and fuel cell), and thermal (molten power, so a power electronic interface is required to inte-
salt) are gaining more popularity due to long lifetime, less grate them to AC power systems or AC MG. Semiconduc-
operating cost, environmental safety, and the better sys- tor switches of power electronic interface, ie, converters,
tem performance. With the improvement in battery tech- offer more controllability for ESS. Mbungu et al138 pro-
nology and PEVs in the market, it is now possible to store posed a dynamic distributed energy storage strategy
energy that can be utilized to fulfill demand during peak (DDESS) for smart energy coordination of autonomous
load hours. PEV batteries act as mobile energy storage residential home. DER considered by the authors are
system (MESS) or distributed energy storage system ESS, PV, and wind power connected to the main utility,
(DESS) that can be connected to any location and acts and the optimal energy flows of DERs are coordinated
as a source to local loads or supply power to grid. Simi- to minimize the electricity cost. For MG, different control
larly, solar tied batteries can be utilized to supply power strategies such as V/f control, PQ control (for aggregated
when needed. Technoeconomic factors have to be taken ESS), or droop control (for distributed) can be applied for
into account while using ESS, such as installing energy ESS. ESSs provide the best solution to maintain grid reli-
storage capacity of about 2% to 5% of the total installed ability in a scenario where a large percentage of genera-
capacity will require a large no of cells. In the emerging tion will come from RES.
RATHOR AND SAXENA 23

5 | F A C T O R S TO BE CO N S I D ER E D to apply. The stochastic model‐based analysis is the cor-

FOR EMS IMPLEMENTATION rect way to analyze and handle uncertainty, which
predicts output based on a random set of values using
5.1 | Uncertainty management probability distribution. Most commonly used methods
techniques used to handle uncertainty in the EMS are Markov chain
model, Monte Carlo (MC), and Markov chain Monte
The emerging power system is characterized by many Carlo (MCMC) method. The word Monte Carlo came
new technologies such as the integration of RES, PEVs, from casinos in Monte Carlo, a concept based on random
DSM, DG sources, and energy storage facilities. New number generation in games at the casino. MC model has
technologies have helped the grid to become smarter been used in previous studies115,117,140-142; to handle
but also bring in challenges in the form of various uncer- uncertainty in microgrid, PEV charging profile, charging
tainties such as uncertainty in generation due to intermit- power demand, storage capacity, reliability, and power
tent nature of RES, uncertainty in forecasted load due to quality are determined. MC used by a researcher where
random nature of the load, or unanticipated breakdown demand management and load modeling are required,
of transmission equipment or generation unit. Another especially where stochastic human interaction is
uncertainty occurs due to PEV owner's behavior of charg- involved.143 Markov chain model is a memoryless ran-
ing or discharging into the grid. These uncertainties make dom process that can predict the next state based on the
the exact prediction or forecast of generation and con- current state.
sumption difficult, thus affecting the reliability and secu- Zhou et al proposed a solution to handle uncertainty
rity of the power system. It has been analyzed that there due to PEV charge/discharge pattern using capacity fore-
are two types of uncertain parameter, namely, technical casting model. Here, eight different uncertainties are con-
parameters and economic parameters.139 Technical sidered, and the precise charge/discharge prediction is
parameters are further classified as operational parame- achieved.144 Robustness of EMS refers to the ability of a
ters and topological parameters. Operational parameters system to resist changes without changing its previous
include the information of demand and renewable gener- configuration. However, there is no clear‐cut definition
ation, and topological parameters include the information of robust EMS (REMS). Still, we can define REMS as
about the disruption of any line or generating unit. More- the one that is able to handle uncertainties effectively,
over, the economic parameter can be further classified as ie, able to model all the uncertainties with good accuracy,
macroeconomic, which include generation cost and thereby adequately integrating them into the system.
microeconomic parameters, which include economic pol- Valencia et al145 proposed a REMS using an MPC frame-
icies and regulation. Uncertainty modeling has been done work. The available energy of wind energy sources is
by various methods while implementing EMS. The differ- predicted using the fuzzy prediction interval model. The
ent methods of handling uncertainties are given in results indicate that the model was robust against
Figure 9. Probabilistic nature of power system requires the uncertain wind power variations though there was a
stochastic methods to evaluate system reliability and per- slight increase in operating cost due to the addition of
formance. Due to different uncertainty or randomness in additional reserves. In Rezaei et al,146 robust energy man-
the power system, deterministic models are not suitable agement is proposed for the islanded microgrid to

FIGURE 9 Uncertainty handling methods classification

24 RATHOR AND SAXENA

manage frequency excursion arising from the RES gener- Advancement in power electronics technology and the
ation and load mismatch. Information gap decision increasing number of nonlinear loads are the major cause
theory (IGDT) is applied to handle the uncertainties of of poor power quality. Moreover, recently added PEV
microgrid and solved by MILP framework. Many other load, with a bidirectional charger, also injects harmonics
type of uncertainty handling method such as robust,147 in the system thereby deteriorating the power quality.
fuzzy,148 scenario based,149 linearization based,150 The traditional task of EMS was to deal with
unscented transform,151 probabilistic,152,153 possibilistic scheduling/dispatching, and the power quality was
Monte Carlo,154 Gaussian mixture model,155 estimation looked after at another level of control. The task of EMS
distribution,156 and stochastic inventory theory157 are also was limited to cost and optimal utilization of energy,
proposed by researchers to model uncertainties in an while the power quality remained untracked at the same
EMS. Selection of appropriate method to handle uncer- time. With the deregulation of the electricity market,
tainty is required in the power system to avoid a large maintaining good power quality is of utmost importance
gap between supply and demand. Monte Carlo simula- for the utility to retain/attract customers. Therefore, an
tion (MCS) and scenario‐based methods are usually EMS must consider such power quality events and con-
applied in many types of research works due to their sim- straints while scheduling significant loads/sources.
plicity. However, the computational time is always an Figure 10 shows the functioning of modern EMS incorpo-
issue with such a method. There is a need to develop rated with PQM. Few researchers suggested the use of
faster and accurate methods to handle uncertainties effec- unified power quality conditioner (UPQC), static var
tively in an EMS. compensator (SVCs) to control the quality of power in a
microgrid, rather than using costly and complicated
devices.158 If the energy allocation is done in such a
5.2 | Power quality management way as to reduce power quality issues, it would be a better
way to accomplish two tasks (energy allocation and
PQM is the process of providing solutions to minimize power quality) at the same time.
the impact of external and internal disturbances/events Luo et al45 proposed a DC microgrid based on sustain-
that can impact the uptime or performance of a plant or able buildings where PEVs batteries are used as a storage
a specific process. Power quality issues vary from high‐ device to balance voltage fluctuation, to improve power
speed events such as voltage impulses, transients, and quality, and to achieve power balance in the DC
high‐frequency noise to wave shape faults, voltage swells microgrid in the grid‐connected or islanded mode of oper-
and sags, overvoltage, under voltage, and harmonics. ation. Ovalle et al38 presented an optimal charging sched-
Quality of power delivered to consumers depends on ule of PEV providing voltage regulation and maintaining
the source and on the load at the consumer's end. the benefits of the customers in a residential building

FIGURE 10 Energy management system integrated with power quality management [Colour figure can be viewed at wileyonlinelibrary.
com]
RATHOR AND SAXENA 25

with the uncoordinated charging of PEV fleet. the reliable operation of the power system, better effi-
Safavizadeh et al159 proposed reactive power manage- ciency, minimization of voltage regulation problem, and
ment of inverter‐based DERs to control the voltage profile mitigation of carbon emission. With the liberalization of
and distribution system operation when threatening with the electricity market and technological advances in
short‐term voltage magnitude variations. The event of smart grid, DSM is becoming more popular, which bene-
short circuit and unbalancing in the distribution system fits both the customer and the utility. DSM programs can
may lead to many power qualities issues. Detecting the be categorized into two categories, one is energy response
symmetrical components is still a challenge for power (ER) or energy efficiency conservation (EEC) programs,
engineering, especially when there is a change in and the other is DR programs.161-163 EEC programs
amplitude‐phase, and there is the influence of harmonic encourage customers to gain some economic benefits by
and noise. Naidoo et al160 proposed a novel approach of giving up some of their energy usage.164-167 The energy
estimating symmetrical components under harmonic usage can be reduced through activities such as reducing
and noise signal conditions using nonlinear adaptive the settings of thermostat,168,169 modifying projects,170-172
tracking of amplitude, phase, and frequency of nonsta- DR, on the other hand, is a highly flexible program that
tionary sinusoidal waveform in real time using hardware can be customized to the energy consumption and finan-
experiment and simulation study. EMS is itself a compli- cial objectives of the participants. Six load shaping objec-
cated task, and the inclusion of a power quality character- tives are used in DSM to make the load follow the
ization module will make it more tedious task. Though, generation as close as possible. These six load shaping
the combination of EMS and PQM would give better grid objectives, illustrated in Figure 11 are, namely, peak clip-
performance and cost‐effective solution. ping, valley filling, load shifting, strategic conservation,
strategic load growth, and flexible load shaping.143,173
Peak clipping is the reduction of system load peak dur-
5.3 | DSM, DR, and pricing policy ing high power demands of electricity and provides the
benefit of installation of the additional generation unit
DSM and DR are two essential components of any EMS can be delayed. It can be achieved by direct load control
that contribute to improving system load profile, efficient (DLC) method or through customer support with proper
use of system asset, and reduce demand peaks. It leads to awareness. It can also include an energy conservation

FIGURE 11 Load shaping concept [Colour figure can be viewed at wileyonlinelibrary.com]

26 RATHOR AND SAXENA

method where the customer should be encouraged to use metering, appliance control, DLC, demand bidding,
energy‐efficient devices. In Valencia et al,145 PEVs time‐based programs, and frequency regulations. The
are used as smart appliances for DSM, which can be exact matching between supply and demand is a chal-
scheduled to participate in the load shaping strategy. lenging task for a power engineer. Fail to do so may cre-
The peak shaving strategy is achieved by coordination ate reliability issues in system or power quality‐related
of multiple domestic PEVs charging and discharging. issues such as voltage variation, frequency variation,
The owners of PEVs have the choice to select the pre- brownout, and blackout. Nowadays, awareness of the
ferred time zone from the three time zones given to customer, AMI, communication facility, IoT, etc provide
charge their PEV based on their battery status. The out- satisfactory performance in DR programs. DR can be a
come is the reduction in peak demand and a reduction very effective program from a reliability point of view
in active power losses. when the integration of RES is done with EMS program.
Valley filling is the process of utilization of energy pro- It can create balance and smoothen the demand curve
duction and efficient delivery to end users during periods using the time‐differing concept for scheduling the appli-
of low demand specifically during night time. In Jian ances.27 DR can be incentive‐based (LC), price‐based (an
et al174 and Khatiri‐Doost and Amirahmadi,175 valley fill- effective pricing policy), and demand reduction bids
ing strategy is presented for large centralized scale PEV's based (demand bidding/buyback programs) to the con-
charging with the less computational burden as com- sumer to connect/disconnect their load depending on
pared with other existing valley filling strategies. Zhang the signal. Shafie‐Khah et al179 compared DR program
and Zheng176 demonstrated the capability of PEV to work for a price‐based and incentive‐based with different sub-
in V2G and G2V mode for dynamic economic/emission programs such as ToU, real‐time pricing (RTP), critical
dispatch for valley filling and peak shaving, respectively, peak pricing (CPP), and emergency demand response
with consideration of constraints like power balance, program (EDRP). Controlling method for the load is also
power flow, ramp rate, and BDC and solved by the proposed, such as Interruptible/Curtailable (I/C services),
multiobjective optimization algorithm. DLC. Setlhaolo and Xia163 presented residential demand
Load shifting involves shifting load from on‐peak to response (RDR) and implemented at the household level
off‐peak period. In Xing et al,177 load shifting is demon- to manage peak load by the use of differentiated prices
strated for PEV with bidirectional capability in smart grid and an incentive payment to control the demand.
environment to flatten the total demand curve. Hu et al Figure 12 represents the DR process, pricing scheme,
presented a mathematical programming formulation for and benefits.
HEMS to manage demand for the smart grid using the Retailing concept, which is usually related to commer-
concept of optimal load shifting.178 In this study, devices cial activities like selling electricity, billing, metering,
like PEV, battery storage, and automatic window are con- procuring, and pricing nowadays, is accepted and imple-
sidered, and the DR model is evaluated for cost saving for mented by many electricity governances. Electricity
consumers. Overall, efficient demand‐side management producer sells power to the resellers (electricity utility
functioning can be achieved by load limiters, smart companies, competitive power providers and electricity

FIGURE 12 Pricing policy and demand response

RATHOR AND SAXENA 27

marketers) in the wholesale market and finally purchase such as PV and wind, which require operators to balance
and sale of electricity to the consumers is done in the the forecasting error production.182 After direct deregula-
retail market.79 Retail electricity product falls into two tion marketing of RES is allowed in power exchange.
category, one is guaranteed price product, and another Usually, TSO has to trade electricity generated by RES a
known the spot price product. Guaranteed price product day ahead, it is done in time horizon up to 36 hour, so
has two distinct features of price allocation in advance it allows the correction in forecasted errors. In the day
or for a specific period of the contract. The second feature ahead, market trading is settled the day before deliv-
related to the end user's concern for the quantity of sup- ery,183 ie, on Saturday 4 PM, day‐ahead auction is to be
ply. Both price products have four building blocks, done for energy delivery of Sunday, it is based on fore-
namely, flat rate, ToU, seasonal, and fixed bill.180 How- casted information. The day‐ahead market is beneficial,
ever, flat rate tariff is most appealing to the customer especially in the case when generators are having a high
because of the simplicity of understanding. Other pricing startup cost and time. SO can decide for such a long pro-
schemes are distributed DR, usage‐based dynamic pric- cess. Another is intraday scheduling,183 the prediction for
ing, static pricing policy, and the quadratic cost function next hour ahead to schedule the current day or period
for RTP. In the electricity market, three common terms that has already started. Moreover, the time range of
are a price cap, price floor, and price collar. The price intraday can be 15 to 30 minutes before usage.
cap is the maximum price that can be charged. It is Both the program DR and DSM are introduced to max-
assigned by the utility, selling a call to the retailer. The imize the utilization of existing assets, improving effi-
end user has the right but not to obligate to buy electric- ciency, reliability, flattening the load profile, and other
ity, and this price is known as the exercise price. It pre- benefits. DSM being implemented since the 1980s, not a
vents the retailer from setting high spot price. Floor new technique but the decision variables involved in
price can limit electricity from a utility to lower the retail DSM program nowadays increased due to the presence
price. Price collar is a combination of floor and cap, of DER, and other power quality constraints. Traditional
where the customer buys electricity at the spot within methods and tool are not effective to deal with the behav-
the collar. If the spot price is outside the collar, then cus- ior of consumers, investors, end users, and other institu-
tomer buys power at electricity price.180 Misra et al181 tional factors for decision making step for local and
proposed dynamic pricing policy (D2P), where home individual level. Research should be a focus on the inno-
and roaming pricings are considered, and results show vation for coevolution of infrastructure and exist environ-
34% increment in the utility of PHEV compared with ment condition from the current state of the art of
existing optimal charging of PHEVs. Day‐ahead electric- multiscale energy modeling.
ity market can be treated as a financial market where in
case of the real‐time electricity market is a physical mar-
ket. Both works on the pool trading concept where vari- 6 | S OLUTION APPR OACHES FOR
ous curves are made based on the pool. ISO accepts bids EM S
from aggregator or marketer and determines the market
clearing price (MCP). Only it is found by plotting supply Several programming methods and optimization algo-
bid on the supply curve, and demand bids on the demand rithms are developed for energy management schemes
curve, the intersection of both are known as MCP. Intra- such as battery management, PEV charging/discharging
day electricity market concepts accepted by market management, renewable sources energy management,
players due to increasing penetration of renewable energy and LC management. The ultimate goal of EMS strategy

FIGURE 13 Process for solving an energy management problem

28 RATHOR AND SAXENA

is to minimize or maximize the objective function, which exact optimization methods, as the name indicates, guar-
can include cost, carbon footprint (GHG emission), effi- antee an optimal solution. However, computation com-
ciency, reliability, power quality, losses, energy distribu- plexity increases in this case. Therefore, they are not
tion, load profile, etc. Various solution approaches have preferred for real‐time application, especially if it takes
been used in EMS to achieve the optimal and efficient a longer time to find a solution. Whereas the approximate
operation of the emerging grid. Figure 13 illustrates the optimization methods do not guarantee an optimal solu-
process for solvation energy management strategy with tion but give a solution within a reasonable timeframe
all possible inputs, objectives, technical and economic with the optimum or nearest result. Heuristic or
objective with optimization framework, and desired out- metaheuristic optimization techniques belong to approxi-
come. A general classification of the solution approaches mate algorithms that provide the result at best value
can be done as exact optimization methods and approxi- within the reasonable time frame and can be applied to
mate optimization methods. The exact optimization online or real‐time problems. In the following paragraph,
methods, as the name indicates, guarantee an optimal these solution approaches to EMS are discussed.
solution. However, computation complexity increases in
this case. In this paper, solution approaches for EMS are
grouped into four categories, ie, mathematical program- 6.1 | Mathematical programming‐based
ming based, heuristic, metaheuristic, and another EMS (equation‐based)
solution approach, as illustrated in Figure 14. The math-
ematical programming‐based optimization methods Depending on the nature of the objective function and
belong to the exact solution approach methods.184 The constraints, mathematical programming‐based methods

FIGURE 14 Optimization/programming approaches applied to the energy management system

RATHOR AND SAXENA 29

can be further categorized as convex programming, qua- system is subjected to dynamics with discrete decisions
dratic programming (QP), geometric programming (GP), variables and nonlinearity, MINLP is helpful to find an
MILP, MINLP, and linear programming. The EMS optimal solution that combines the function of NLP and
problem solved by these optimization techniques share MILP (as subproblem).189 Table 9 provides a comparative
common properties, such as they can be expressed in analysis of different programming method used for
mathematical terms, can have continuous or discrete var- achieving an optimal solution in an EMS.
iables, are provably difficult (NP‐complete or worse), and
may include dynamical aspects. Mathematical optimiza-
tion requires careful thought at modeling time, and the
6.2 | Heuristic‐based solution approaches
model must be exploited for reliability and efficiency.
for EMS
Composite problems suffer from computational complex-
Heuristic approaches are the most basic and knowledge‐
ity and the curse of dimensionality. Bellman proposes the
based method to provide approximate solutions to the
basic concept of dynamic programming (DP) based on the
given problem based on prespecified rules. For EMS, heu-
principle of optimality DP for optimal energy manage-
ristic approaches are well designed to reduce the compu-
ment is getting popular among researchers because it
tational complexity for the assigned task for the optimal
can be applied to any nonlinear, nonconvex, mixed‐
solution. Rahim et al198 presented a review study on
integer problems. It splits the whole problem into differ-
energy management controller performance designed
ent stages and solves it in a discretized form based on
for the HEMS for minimization of electricity bill and peak
time, state, and decision variables. In Wang and Liang,157
in load profile using the heuristic method. Performance
DP is effectively applied to minimize daily energy cost for
of EMS is tested for the ToU tariff and inclining block
PHEV energy management problem and to reduce the
rate (IBR) tariff with maintaining the user's comfort level.
computational time. In Rotering and Ilic,185 DP is applied
Klein et al199 presented a study on multiagent comfort
to find the economic optimal solution for vehicle owner
and energy system (MACES)‐based model for BEMS to
for charge control of PHEV in the deregulated electricity
minimize energy consumption and improve the end user
market. QP and convex optimization‐based methods are
comfort level. A real‐world testbed data are utilized for
widely used by researchers with the knowledge of system
simulation and modeling, which includes temperature,
condition. QP optimizes a quadratic objective function of
user preference, behavior, and schedules incorporated
variables subjected to linear constraints on these vari-
into multiagent Markov decision problems (MDP).
ables. The computation time of QP is less as compared
Sianaki and Masoum200 presented a simulation base
with since DP requires the problem to be gridded in time
study for HEMS for smart grid and modeled multiagent
and states, thereby increasing the memory and computa-
intelligent decision‐making support system. The author
tional demand exponentially with the increase in the
applied fuzzy‐TOPSIS‐based robust tool for decision
number of gridded variables, an effect known as the curse
making for end users and other variables, which in turn
of dimensionality. In Chen et al,186 QP method is used to
minimize cost and energy consumption. Such models
formulate the nonlinear relationship between engine fuel
can help end users to maximize their participation in
rate and input parameter like battery power.
DR programs. Table 10 presents a comparison of
In Rajasekharan and Koivunen,187 GP is applied for
heuristic‐based approached for EMS. Megahed et al202
consumption balancing/leveling of a household with
presented a study on energy management in the zero‐
the energy storage devices. In convex programming, the
energy building using neural network predictive control.
objective is a convex function if it is minimization prob-
Neural network and model and predictive control tech-
lem or else concave function if it is maximizing the
niques are combined to enhance control and reducing
program. However, all constraints are convex functions.
disturbances. A neural network is mainly applied for fore-
Problems solved by linear programming can be solved
casting energy from intermittent energy resources such as
by convex programming because all linear functions are
wind and PV.
convex. In Cortés and Martínez6 and Wu et al,57 convex
programming is tested and found to be rapid and efficient
for control decision for optimum power allocation among 6.3 | Metaheuristic‐based solution
PEV battery, home battery, home power demand, and the approaches for EMS
utility grid. Most common tool available to solve the con-
vex problem is CVX incorporated in Matlab. MILP is a Most metaheuristic methods, such as genetic algorithms
programming method where some of the objective (GAs), PSO, ant colony optimization, simulated anneal-
variables are constrained to be integer while additional ing (SA), bee colony optimization, and many other
variables are permissible to be noninteger.188 When the methods, are stochastic and are motivated by nature,
30 RATHOR AND SAXENA

TABLE 9 Comparative analysis of mathematical optimization methods

Method Objectives Advantage Drawbacks Uncertainties Architecture

Dynamic Reduced fuel High Offline, problem Not considered Decentralized

programming157,185,190,191 consumption for computational formulation,
PHEV for civil and capability computational
military challenge
Geometric programming187 Minimize electricity Simple to work and Degree of difficulty Forecasted Centralized
consumption and implement is high
electricity bill
Quadratic Optimal utilization of Fast convergence, Real‐time use not Not considered Decentralized
programming177,186 battery energy and applicable for possible
engine status real‐time
Convex Maximize home Reliable, efficient, Higher complexity Not considered Centralized
programming57,192,193 economy, with moreover,
satisfying comfort suitable for real‐
time operation
Linear programming181 Charging cost Suitable for real‐ Work with linear Forecasted Centralized/
minimization time application variables only, Decentralized
not applicable to both
more than two
variables
MILP163,193 MG operating cost Accuracy higher Lower modeling Not considered Decentralized
minimized than MINLP capabilities as
compared with
MINLP
MINLP143,194-197 Smart charging/ Better in modeling High solution time Forecasted Centralized
discharging strategy capability and lower result
quality

Abbreviations: MG, microgrid; MILP, mixed‐integer linear programming; MINLP, mixed‐integer nonlinear programming.

physical, or biological principle and try to resemble them solution in low execution time, which is an important
making a balance between exploration and exploita- aspect for a smart grid environment, particularly when
tion.203 Most of the metaheuristic methods are a large number of distributed resources are taking part.
population‐based and are a nongreedy algorithm. For Table 11 provides a comparative analysis of different
EMS, some methods have been applied by researchers, metaheuristic techniques used in EMS.
namely, GA, fuzzy, SA, tabu search (TS), PSO, firefly,
and ABC. These metaheuristic techniques are computa-
tionally robust but do not require objective to be con- 6.4 | Other solution approaches for EMS
vex.204 Ruiz‐Cortés et al205 used a GA for optimal
charge/discharge scheduling of batteries in microgrid of Besides mathematical programming‐based, heuristics, and
prosumers to minimize energy exchange losses by mini- metaheuristics, few other solution approaches have been
mizing the power at the point of common coupling with used in EMS to enhance the modeling of the problem,
main grid and microgrid. Radosavljević et al206 proposed for achieving computational benefits and accuracy in
energy and operation management of microgrid using results. Methods such as fuzzy,212 MPC,213 sliding mode
PSO to minimize total energy and operating cost of a control,214 and stochastic programming215 are explored
microgrid by optimally adjusting the control variables of by many authors. Out of the above‐mentioned approaches,
energy and operation management. Sousa et al207 pro- stochastic and robust methods are best suited for uncer-
posed an SA approach for energy management in a tainty handling; in former approach, uncertainty is char-
VPP. A case study was done on a 33‐bus distribution net- acterized by the distribution of parameters, while in later
work with 66 generators, 32 loads, and 1000 gridable one, we assume these quantities belong to certain sets,
vehicles, and the results of SA are compared with MINLP and no distributional assumptions are made. Prodan and
approach. SA approach proved to be more effective for Zio216 proposed an approach for reliable MG EMS using
energy resource management in VPP, providing good receding horizon control to minimize the operating cost
RATHOR AND SAXENA 31

TABLE 10 Comparative analysis of heuristic‐based solution approaches

Method Objectives Advantage Drawbacks Uncertainties Architecture

Markov Minimization of energy Best suited for decision making Can solve linear Forecasted Decentralized
decision199 consumption and improve problems
the comfort level of users
TOPSIS200 Minimize consumption and Simple to implement, a limited Not much reliable for Forecasted Decentralized
cost, improve DR number of criteria and the optimal solution
participation alternatives provide result
Backtracking Reduced energy consumption Intuitive approach, simple to Inefficient for a huge Not considered Decentralized
search201 for HEMS and improve implement number of branches,
comfort large space required
ANN202 Improve and simple control Suitable for forecasting and The parallel processing Forecasted Decentralized
and reduce disturbances decision making power required,
unexplained behavior

Abbreviations: ANN, artificial neuron network; DR, demand response; HEMS, home energy management system.

by proper scheduling of battery set. MPC‐based concept is payment. Table 12 provides a comparative analysis of
used for the uncertain environment of generators and con- other approaches applied to EMS. Tungadio et al222 pre-
sumption profile with certain realistic constraints. Chen sented a study on the MPC approach for controlling power
et al217 proposed real‐time price‐based DR management flow in AC‐tie line and frequency deviations in a system
for residential appliances using stochastic optimization composed of two microgrids. The simulation results were
and robust optimization approach. The stochastic optimi- compared with those obtained using the open control
zation is used for minimizing the expected electricity pay- loop; it was found that that closed‐loop base MPC is much
ment for the entire day, keeping the financial risks faster and accurate as compared with open‐loop control
associated with the uncertainty real‐time electricity price. for a complex dynamical system.
Robust optimization is used for considering price uncer- Selection of solution approach needs to apply specific
tainty intervals for minimizing the worst‐case electricity criteria before applying it to the problem or even before

TABLE 11 Comparative analysis of metaheuristic approaches

Method Objectives Advantage Drawbacks Uncertainties Architecture

205,208
Genetic algorithm The operation, Easy to implement, derivation Computational Forecasted
emission cost free, and stochastic time is high
minimized and
increase trade profit

Centralized
Particle swarm MG operating cost Derivative‐free, easy to Computational time is high, real‐ Point Centralized
algorithm206 minimization implement, limited time implementation difficult estimation
input required
Tabu search209 Reduce operating cost and Flexible in nature, Computational complexity Scenario based Centralized
losses adaptive memory
Artificial bee Operating cost of MG Simplicity, robustness, The higher the complexity of the Forecasted Centralized
colony210 minimized and flexibility system for the sequential
process
Simulated VPP operating cost The less computational To check the optimality of result, Forecasted Centralized
annealing207 minimized time required other method required such as
branch and bound
Firefly211 MG operating cost Can deal with highly Probability of trapping in local Scenarios Centralized
minimized nonlinear, minima based
multimodal
optimization

Abbreviations: MG, microgrid; VPP, virtual power plant.

32 RATHOR AND SAXENA

TABLE 12 Comparative analysis of other approaches

Method Objectives Advantage Drawbacks Uncertainties Architecture

Fuzzy218 Charging/discharging Handle incomplete data, flexible, High computational ANN Centralized
management and and simple time and rules need to
operating cost be well designed
minimization
Model DR studied and Superior performance with Costly installation and ANN Centralized
predictive minimization of extensive data, predictive higher skill required
control219 operating and emission
cost
Sliding Improvement in power‐ Robust to uncertainty, applicable Quick switching Forecasted Decentralized
mode sharing and manage for nonlinear systems (chattering issue),
control220 active and reactive limited to a single
power input
Stochastic215 Investment and operating Can handle uncertainty well, Lot information and Scenario based Centralized
cost minimization excellent for decision making data required;
distribution of data
required
Robust221 Real‐time demand Ability to cancel measured and Implementation for real‐ Not considered Centralized
response unmeasured disturbances, prior time require higher
implementation and knowledge does not require skills
energy trading

Abbreviations: ANN, artificial neuron network; DR, demand response.

formulating the problems. In the problems mentioned increases and its performance deteriorates. The
above, some of these are computationally time‐ decentralized structure offers a solution, but it requires
consuming and required large memory space. Mathemat- synchronization among components, and continuous
ical programming requires more time compared with two‐way communication makes it less cost‐effective.
heuristic and metaheuristic. Most of the metaheuristic The hierarchical EMS in the area is needed to be explored
problems are population‐based, and it can find a near‐ further for efficient operation and control of the grid. In
optimal or optimal solution from search space with a less hierarchical EMS, layerwise decisions are taken and proc-
computational burden. On the other hand, heuristic essed with consideration of multiobjective such as power
approaches are knowledge‐based, which provide an quality, cost, and comfort to customers. Communication
approximate solution to the problem. For heuristic solu- is another important aspect to be considered for the
tion approaches, prior knowledge of energy management real‐time application of any EMS. Cost and data rate are
is required with certain assumptions and relaxation to the two factors that are considered for the deployment
obtain solution nearer to optimal with less computational of EMS in residential and rural areas. Bluetooth, Wifi,
time. Stochastic, robust, MPC, SMC, etc are difficult to and Zigbee are preferred in these areas.223 However, for
implement, but computational tractability and accuracy implementation of EMS in utilities and microgrids, cover-
in results are guaranteed. age rate along with data rate is essential, and optical net-
work, 3G, and 4G are better choices.224 As discussed
7 | CRITICAL REVIEWS AND ahead in the section, there are many issues to be resolved
for the practical and real‐world implementation of EMS.
R E A L‐ WORLD A PPLI CAT ION
Many simulations and case studies have been done, but
Implementation of EMS seems relatively easy in the mod- only a few presented the practical and real‐world imple-
ern era due to advancement in communication technol- mentation of EMS, whether related to building EMS,
ogy, modern infrastructure, and sensors, but there are home EMS, microgrid EMS, etc. However, in the field,
many challenges in the practical implementation of several implementations of EMS have been done by big
EMS as shown in Figure 15. Most of the EMS studied in players like General Electric, ABB, Schneider, and Sie-
literature have centralized control architecture as it mens. Still, some gaps remain between institutional
requires simple maintenance, but with an increase in developers and policy maker, and there is a strong need
penetration of RES, PEVs, ESSs, its computational time to bridge the gap.
RATHOR AND SAXENA 33

FIGURE 15 Challenges ahead in energy management system implementation [Colour figure can be viewed at wileyonlinelibrary.com]

7.1 | GHG emission infrastructure, strict government policies, and high mar-
gins of car manufacturing companies in selling
EMS is a multiobjective task that includes environmental diesel/petrol cars. United Nations environment program
issues in addition to technical and economic issues. Two GHG indicator helps any organization for accounting
major objectives of EMS are GHG minimization and and reporting GHG emission as a single figure.225 The
reduction in energy consumption. Proper scheduling of generation utilities and the industrial sector can use such
conventional generators (fossil fuel‐based power plants), guidelines for calculating their GHG emission, which can
industrial loads, and PEV is required to minimize overall be key input data for any EMS.
GHG emission rate. Different generating units have dif- Several research works have been published on envi-
ferent emission levels. To reduce GHG emission from ronmental factors, climate change, CO2 minimization,
conventional power plants, the plant‐specific characteris- and GHG mitigation. However, no work has been
tics such as fuel type, efficiency, emission rate, ramp‐up reported in the following directions.
time, and age of the plant should be an input data to
EMS for the optimal dispatch of these units and changing • Core collaboration of clean development mechanism
the dispatch of the unit at the emission margin. Peak load (CDM) and its economic viability.
reduction reduces emissions, as peaking plants tend to be • An EMS strategy that correlates its objective with
less efficient than other plants on the system. Moreover, CDM and GHG mitigation.
when electricity demand is high, transmission and distri- • Replacement of traditional energy resources with
bution equipment tend to be less efficient, resulting in alternate small‐scale low pollutant energy sources
overall greater system losses. Thus, effective DSM and and clean energy.
DR play an important role in reducing emissions as • Earth hour public participation and awareness.
energy efficiency is a direct measure to reduce GHG. • Energy demand analysis and CDM implementation
Nuclear power offers a promising alternative, but there studies in public and private transportation segment
are issues of safety, waste disposal, and proliferation that which can contribute to EMS.
need attention. The GHG emission from the conventional • Energy demand analysis for both local and urban
vehicle can be minimized by encouraging people to use area, taking into account the locally available energy
PEV and PHEV. However, PEVs have a long way to go resources is the need of the hour, which can contrib-
in developing countries owing to lack of charging ute to EMS.
34 RATHOR AND SAXENA

7.2 | DR implementation energy management and DR, but they also have unin-
tended outcomes with respect to customers privacy.
To enhance grid performance and usability of available Power usage information stored in meters can be manip-
energy resources, DR program is an effective tool to benefit ulated by interested groups to extract the personal infor-
to utility and customer both. Still, there are many obstacles mation of customers like their lifestyle and financial
in effective implementation of DR programs in the real sys- state, energy usage habits, their routine, activities likings,
tem. The success of the DR program is based on economic beliefs, and behavior. The data can be tempered for any
growth rates, investment in the energy sector, and imple- misuse like manipulating the billing, electricity prices,
mentation in an urban area, specifically with high densi- or even control commands that can threaten the whole
ties of population and awareness of energy and operation of EMS. In real‐time EMS, all collected data
environmental issues.226 Integration of RES, PEV, and are processed in a distributed or centralized manner via
ESS may provide benefits to utility, but the ultimate chal- internet and hardware connection. The storage concen-
lenge of energy balance and sustainable energy requires trators, the processing (cloud computing), and the service
the active participation of all players, including the end providers are all third parties and must ensure the integ-
users. Acceptance of the DR programs mainly depends on rity of data. The energy sector is vulnerable to face
its usefulness and ease of understanding to the end users. cyberattack, and private information of the users need
The following points need attention for effective to be protected by the utility/operators. Key steps for data
implementation of DR programs: protection and privacy are given below.

• Lack of active participation of end users is one of the • For EMS and DR programs, customers should have
major concerns and barrier in the implementation of the right to own their data and only used as per the
the DR program. Full control of loads (interruptible/ permission of the customer.
controllable) to utility/operator may make customers • Strong legislation and policies need to be imple-
uneasy to take participation in the DR program. There mented to enhance the security and privacy of the
is a need to create DR awareness and understanding user's data.
and to equip consumers with sufficient information • Program such as “Data protection directive 95/46/EC
regarding DR programs enabling them to make well‐ implemented by the European Commission”
informed decisions. needs to be implemented in all countries regarding,
• The reliability of response, once the DR request is initi- “Recommendation on data protection impact assess-
ated, has to be explored further. Most of the consumers ment template for smart grid and smart metering
continue with their everyday routine and habits systems.”228
despite receiving feedback on energy consumptions
from the in‐home display units. The unanticipated or
rather irrational behaviors, as well as preferences of 7.4 | Customer awareness and
consumers, are few issues that need to be considered participation
and accounted for in the evaluation analysis of DR pro-
grams. It is crucial to understand the motivations of the Without customer participation, the energy‐efficient oper-
consumer for engaging in or rejecting a DR program. ation cannot be achieved. Other than consumers, the
• The financial compensation offered against the incon- concept of prosumers should be introduced to customers
venience of users for DR remains insufficient to pro- where installation of PV and utilization of PEV can be
mote the development of the DR sector.203 So there suggested. In such cases, energy sharing and balancing
is a need for attractive incentive policies for increasing might become more precise and economic; still, the chal-
the participation of users. lenge of managing large prosumers needs to bear.
• To optimally regulate DER and provide DR, the con-
cept of VPP is proposed in several papers.227 However, • Energy sharing and stakeholder engagement need to
the response from VPP is slow due to the presence of be focused, where incentives should be offered to the
inertia, which needs to be addressed. prosumers to drive the market in a positive direction.
• Funding and investment for new infrastructure and
awareness of existing subsidy schemes offered by gov-
ernment agencies should be communicated to the
7.3 | Data security and privacy
customers.
Security and privacy of user data are severe challenges in • Fair and straightforward pricing policy that can be eas-
EMS implementation. Use of SMs is the first step towards ily understood by the customers’ needs to be explored.
RATHOR AND SAXENA 35

7.5 | Outdated infrastructure control architectures, types, benefits, and challenges of

the EMS with a comprehensive analysis of different
EMS includes monitoring, control, and optimization of stakeholders and participants of the EMS is presented.
energy, which requires a continuous exchange of infor- The three control architectures, centralized,
mation between SOs, utilities, and end users. Outdated decentralized, and hierarchical EMSs, are explored. The
communication protocols and transmission networks paper comprehensively investigates the behavior of differ-
may hold back the EMS implementation. Implementa- ent participants, such as RES, PEVs, and energy storage
tion of EMS and other programs require massive infra- in the EMS. The paper also addresses different uncer-
structural changes in a traditional power network, tainty from various sources and loads in smart grid and
including smart metering devices, communication, and the methods of handling them effectively, active DR,
data processing systems, central data management system DSM, and the optimization approach used in energy
communication devices, and computing mechanism, management to achieve the desired objectives consider-
which involve huge capital investment. Funding and ing all constraints. Various challenges and issues in the
investment from government and private firms for such implementation of an EMS with different stakeholder
a modern infrastructure is another challenge. Research under consideration are discussed, which would lead to
work done on EMS sometimes becomes limited to docu- future research and development towards a more
ments and prototyping only. Real‐time implementation advanced EMS. Research area to be considered for future
is hardly done due to the high cost incurred in the research work are the following (a) enhancement of cost‐
replacement of existing infrastructure. effective smart grid network with reliable and secure
communication is the prime requirement that needs to
• A public‐private partnership should be initiated to a be addressed; (b) fast and accurate modeling of uncer-
feasible level for green energy and modern infrastruc- tainties is an area that needs to be further explored; and
ture implementation. (c) real‐time, cost‐effective implementation of EMS
through hardware needs to be designed and developed.
Significant points are highlighted here.
7.6 | Highly uncertain systems

The emerging grid with new types of sources and loads 1. Advanced metering and communication are the fore-
bring in new challenges to power operators and planners most requirements for integration of RES into the
in the form of different uncertainties. RES generation is grid, controlled and monitored by the EMS.
highly variable and intermittent, quantifying PEV load 2. The REMS is essential for optimal allocation of
too is highly uncertain due to unplanned trips, traffic, energy sources in the grid.
driving distance, and driving behavior of PEV owners. 3. Computational complexity and burden need to be
The response of the customer to DR, electricity price, considered while designing and implementing EMS
and deviation of the load from the forecasted load are programs.
some other uncertainties faced by power engineers and 4. More focus is needed for handling uncertainty in the
operators. The uncertainties if not modeled perfectly will system, whether it is due to unpredictable human
affect the solution of the optimization programs imple- behavior for load connection/disconnection time,
mented in the EMS. Since it is difficult to model the exact PEVs driving pattern, weather conditions, intermit-
behavior of an uncertain parameter, accurate modeling of tency in sources, etc.
uncertainties is a challenging task in an EMS. 5. Security and privacy of customer data need to be
ensured with the implementation of new regulations
• Different uncertainties need to be modeled effectively. and policies.
• Mitigation of uncertainty from PEV and RES can be
taken care of by storage system, VPP, and DR
programs.
ORCID
Sumit K. Rathor https://orcid.org/0000-0003-3049-7022
8 | CONCLUSION

EMS ensures optimal utilization of distributed energy RE FER EN CES

sources to realize the economical, sustainable, and reli- 1. (2019, Feb). OECD Economic Outlook and Interim Economic
able operation of the smart grid. In this paper, a compre- Outlook. Available: http://www.oecd.org/eco/outlook/eco-
hensive and critical review on the concept, objectives, nomic‐outlook/
36 RATHOR AND SAXENA

2. (2017). BP Statistical Review of World Energy June 2017. Avail- 19. Martinez CM, Hu X, Cao D, Velenis E, Gao B, Wellers M.
able: https://www.bp.com/content/dam/bp/en/corporate/pdf/ Energy management in plug‐in hybrid electric vehicles: recent
energy‐economics/statistical‐review‐2017/bp‐statistical‐review‐ progress and a connected vehicles perspective. IEEE Trans Veh
of‐world‐energy‐2017‐full‐report.pdf Technol. 2016;66(6):4534‐4549.
3. International Energy Outlook 2017 [Online]. Available: 20. Lee D, Cheng C‐C. Energy savings by energy management sys-
https://www.eia.gov/outlooks/ieo/pdf/0484(2017).pdf tems: a review. Renew Sustain Energy Rev. 2016;56:760‐777.
4. Azzouz MA, Shaaban MF, El‐Saadany EF. Real‐time optimal 21. Olatomiwa L, Mekhilef S, Ismail MS, Moghavvemi M. Energy
voltage regulation for distribution networks incorporating high management strategies in hybrid renewable energy systems: a
penetration of PEVs. IEEE Trans Power Syst. review. Renew Sustain Energy Rev. 2016;62:821‐835.
2015;30(6):3234‐3245. 22. Yousefi M, Hajizadeh A, Soltani MN. A comparison study on
5. Berger LT, Iniewski K. Smart Grid Applications, Communica- stochastic modeling methods for home energy management
tions, and Security. Hoboken, New Jersey: Wiley; 2012. system. IEEE Trans Ind Inf. 2019;15(8):4799‐4808.
6. Cortés A, Martínez S. A hierarchical algorithm for optimal 23. Zakaria A, Ismail FB, Lipu MH, Hannan M. Uncertainty
plug‐in electric vehicle charging with usage constraints. models for stochastic optimization in renewable energy appli-
Automatica. 2016;68:119‐131. cations. Renew Energy. 2020;145:1543‐1571.
7. Baumeister RF, Leary MR. Writing narrative literature reviews. 24. Zhou B, Li W, Chan KW, et al. Smart home energy manage-
Review of General Psychology. 1997;1(3):311‐320. ment systems: concept, configurations, and scheduling
8. Cheng Z, Duan J, Chow M‐Y. To centralize or to distribute: strategies. Renew Sustain Energy Rev. 2016;61:30‐40.
that is the question: a comparison of advanced microgrid man- 25. Hannan MA, Hoque MM, Hussain A, Yusof Y, Ker PJ. State‐of‐
agement systems. IEEE Ind Electron Mag. 2018;12(1):6‐24. the‐art and energy management system of lithium‐ion batteries
9. Shareef H, Ahmed MS, Mohamed A, Al Hassan E. Review on in electric vehicle applications: issues and recommendations.
home energy management system considering demand Ieee Access. 2018;6:19362‐19378.
responses, smart technologies, and intelligent controllers. IEEE 26. Ahmad F, Alam MS, Asaad M. Developments in xEVs charging
Access. 2018;6:24498‐24509. infrastructure and energy management system for smart
10. Rafique SF, Jianhua Z. Energy management system, generation microgrids including xEVs. Sustain Cities Soc. 2017;35:552‐564.
and demand predictors: a review. IET Gener Transm Distrib. 27. Carreiro AM, Jorge HM, Antunes CH. Energy management
2017;12(3):519‐530. systems aggregators: a literature survey. Renew Sustain Energy
11. Poompavai T, Kowsalya M. Control and energy management Rev. 2017;73:1160‐1172.
strategies applied for solar photovoltaic and wind energy fed 28. Meng L, Sanseverino ER, Luna A, Dragicevic T, Vasquez JC,
water pumping system: a review. Renew Sustain Energy Rev. Guerrero JM. Microgrid supervisory controllers and energy
2019;107:108‐122. management systems: a literature review. Renew Sustain
12. Azuatalam D, Paridari K, Ma Y, Förstl M, Chapman AC, Energy Rev. 2016;60:1263‐1273.
Verbič G. Energy management of small‐scale PV‐battery sys- 29. Beaudin M, Zareipour H. Home energy management systems:
tems: a systematic review considering practical a review of modelling and complexity. Renew Sustain Energy
implementation, computational requirements, quality of input Rev. 2015;45:318‐335.
data and battery degradation. Renew Sustain Energy Rev. 30. Handschin E, Petroianu A. Energy Management Systems:
2019;112:555‐570. Operation and Control of Electric Energy Transmission Systems.
13. Zia MF, Elbouchikhi E, Benbouzid M. Microgrids energy man- Berlin, Germany: Springer‐Verlag; 1991.
agement systems: a critical review on methods, solutions, and 31. Hussain A, Bui VH, Kim HM. A resilient and privacy‐
prospects. Appl Energy. 2018;222:1033‐1055. preserving energy management strategy for networked
14. Vivas F, De las Heras A, Segura F, Andújar J. A review of microgrids. IEEE Trans Smart Grid. 2018;9(3):2127‐2139.
energy management strategies for renewable hybrid energy 32. Nge CL, Ranaweera IU, Midtgård O‐M, Norum L. A real‐time
systems with hydrogen backup. Renew Sustain Energy Rev. energy management system for smart grid integrated
2018;82:126‐155. photovoltaic generation with battery storage. Renew Energy.
15. Hannan MA, Faisal M, Ker PJ, et al. A review of internet of 2019;130:774‐785.
energy based building energy management systems: issues 33. Bialasiewicz JT. Renewable energy systems with photovoltaic
and recommendations. IEEE Access. 2018;6:38997‐39014. power generators: operation and modeling. IEEE Trans Ind
16. Salimi S, Hammad A. Critical review and research roadmap of Electron. 2008;55(7):2752‐2758.
office building energy management based on occupancy moni- 34. Sexauer JM, McBee KD, Bloch KA. Applications of probability
toring. Energ Buildings. 2019;182:214‐241. model to analyze the effects of electric vehicle chargers on distri-
17. Khan MW, Wang J, Ma M, Xiong L, Li P, Wu F. Optimal energy bution transformers. IEEE Trans Power Syst. 2013;28(2):847‐854.
management and control aspects of distributed microgrid using 35. Lunz B, Yan Z, Gerschler JB, Sauer DU. Influence of plug‐in
multi‐agent systems. Sustain Cities Soc. 2019;44:855‐870. hybrid electric vehicle charging strategies on charging and
18. Zou H, Mao S, Wang Y, Zhang F, Chen X, Cheng L. A survey battery degradation costs. Energy Policy. 2012;46:511‐519.
of energy management in interconnected multi‐microgrids. 36. Millner A. Modeling lithium ion battery degradation in electric
IEEE Access. 2019;7:72158‐72169. vehicle. In: Innovative Technologies for an Efficient and Reliable
RATHOR AND SAXENA 37

Electricity Supply (CITRES), 2010 IEEE Conference on. Wal- 52. Mendes PR, Normey‐Rico JE, Bordons C. Distributed energy
tham, MA,USA: IEEE; 2010:349‐356. management system for V2G networked microgrids. In: Inno-
37. Green RC II, Wang L, Alam M. The impact of plug‐in hybrid vative Smart Grid Technologies Conference‐Latin America
electric vehicles on distribution networks: a review and out- (ISGT Latin America), 2017 IEEE PES. Quito, Ecuador: IEEE;
look. Renew Sustain Energy Rev. 2011;15(1):544‐553. 2017:1‐6.
38. Ovalle A, Hably A, Bacha S, Ahmed M. Voltage support by opti- 53. Wen C‐K, Chen J‐C, Teng J‐H, Ting P. Decentralized energy
mal integration of plug‐in hybrid electric vehicles to a residential management system for charging and discharging of plug‐in
grid. In: Industrial Electronics Society, IECON 2014‐40th Annual electric vehicles. In: Wireless Communications & Signal Process-
Conference of the IEEE. Dallas, TX: IEEE; 2014:4430‐4436. ing (WCSP), 2012 International Conference on. Huangshan,
39. Papari B, Edrington CS, Vu T. Stochastic operation of intercon- China: IEEE; 2012:1‐6.
nected microgrids. In: 2017 IEEE Power & Energy Society 54. Ahmad J, Tahir M, Mazumder S. Improved dynamic perfor-
General Meeting. Chicago, IL, USA: 2017:1‐5. mance and hierarchical energy management of microgrids
40. Zaree N, Vahidinasab V. An MILP formulation for centralized with energy routing. IEEE Transactions on Industrial Informat-
energy management strategy of microgrids. In: Smart Grids ics. 2018.
Conference (SGC), 2016. Kerman, Iran: IEEE; 2016:1‐8. 55. Zhuang P, Liang H. Hierarchical and decentralized stochastic
41. Honarmand M, Zakariazadeh A, Jadid S. Optimal scheduling of energy management for smart distribution systems with high
electric vehicles in an intelligent parking lot considering vehicle‐ BESS penetration. IEEE Trans Smart Grid. 2019;1.
to‐grid concept and battery condition. Energy. 2014;65:572‐579. 56. Tian P, Xiao X, Wang K, Ding R. A hierarchical energy man-
42. Sugimoto Y, Miyamoto T, Kumagai S, Mori K, Kitamura S, agement system based on hierarchical optimization for
Yamamoto T. CO2 emission basic unit control mechanism in microgrid community economic operation. IEEE Trans Smart
a distributed energy management system using the market ori- Grid. 2015;7(5):2230‐2241.
ented programming. In: Sustainable Energy Technologies, 2008. 57. Wu X, Hu X, Teng Y, Qian S, Cheng R. Optimal integration of
ICSET 2008. IEEE International Conference on. Singapore: a hybrid solar‐battery power source into smart home nanogrid
IEEE; 2008:520‐525. with plug‐in electric vehicle. J Power Sources. 363:277‐283.
43. Yakire K, Miyamoto T, Kumagai S, Mori K, Kitamura S, Yama- 2017/09/30/2017
moto T. An energy distribution decision method in distributed 58. Huang Y, Wang L, Guo W, Kang Q, Wu Q. Chance constrained
energy management systems by the market‐oriented program- optimization in a home energy management system. IEEE
ming. In: 2006 SICE‐ICASE International Joint Conference. Trans Smart Grid. 2018;9(1):252‐260.
Busan, South Korea: IEEE; 2006:5983‐5988. 59. Qureshi MU, Girault A, Mauger M, Grijalva S. Implementation
44. He F, Wu D, Yin Y, Guan Y. Optimal deployment of public of home energy management system with optimal load sched-
charging stations for plug‐in hybrid electric vehicles. Transp uling based on real‐time electricity pricing models. In: 2017
Res B Methodol. 2013;47:87‐101. IEEE 7th International Conference on Consumer Electronics ‐
45. Luo Z, Hu Z, Song Y, Xu Z, Lu H. Optimal coordination of Berlin (ICCE‐Berlin). Berlin: IEEE; 2017:134‐139.
plug‐in electric vehicles in power grids with cost‐benefit analy- 60. Kikusato H, Mori K, Yoshizawa S, Fujimoto Y, Asano H,
sis—Part I: enabling techniques. IEEE Trans Power Syst. Hayashi Y, Kawashima A, Inagaki S, Suzuki T. “Electric vehicle
2013;28(4):3546‐3555. charge‐discharge management for utilization of photovoltaic by
46. Cheng Z, Duan J, Chow MY. Reliability assessment and compar- coordination between home and grid energy management sys-
ison between centralized and distributed energy management tems,” IEEE Trans Smart Grid, 2018;10(3):3186‐97.
system in islanding microgrid. In: 2017 North American Power 61. Zhai S, Wang Z, Yan X, He G. Appliance flexibility analysis
Symposium (NAPS). Morgantown, WV, USA: IEEE; 2017:1‐6. considering user's behavior in home energy management sys-
47. Olivares DE, Cañizares CA, Kazerani M. A centralized energy tem using smart plugs. IEEE Trans Ind Electron.
management system for isolated microgrids. IEEE Trans Smart 2018;66(2):1391‐1401.
Grid. 2014;5(4):1864‐1875. 62. Ożadowicz A, Grela J. An event‐driven building energy
48. Patnam BSK, Pindoriya NM. Centralized stochastic energy management system enabling active demand side management.
management framework of an aggregator in active distribution In: 2016 Second International Conference on Event‐based Control,
network. IEEE Trans Ind Inf. 2018;15(3):1350‐1360. Communication, and Signal Processing (EBCCSP). 2016:1‐8.

49. Celik B, Roche R, Suryanarayanan S, Bouquain D, Miraoui A. 63. Ożadowicz A, Grela J, Babar M. Implementation of a demand
Electric energy management in residential areas through coor- elasticity model in the building energy management system.
dination of multiple smart homes. Renew Sustain Energy Rev. In: 2016 Second International Conference on Event‐based Con-
2017;80:260‐275. trol, Communication, and Signal Processing (EBCCSP).
Krakow, Poland: IEEE Conference; 2016:1‐4.
50. Bazmohammadi N, Tahsiri A, Anvari‐Moghaddam A, Guer-
rero JM. A hierarchical energy management strategy for 64. Manusilp K, Banjerdpongchai D. Optimal dispatch of cogenera-
interconnected microgrids considering uncertainty. Int J Electr tion with thermal energy storage for Building Energy
Power Energy Syst. 2019;109:597‐608. Management System. In: 2016 13th International Conference on
Electrical Engineering/Electronics, Computer, Telecommunications
51. Han Y, Zhang G, Li Q, You Z, Chen W, Liu H. Hierarchical
and Information Technology (ECTI‐CON). Chiang Mai,Thailand:
energy management for PV/hydrogen/battery island DC
IEEE Conference; 2016:1‐6.
microgrid. Int J Hydrogen Energy. 2019;44(11):5507‐5516.
38 RATHOR AND SAXENA

65. Mbungu NT, Bansal RC, Naidoo R, Miranda V, Bipath M. An and Energy Systems (ICPES). Toronto, ON, Canada: IEEE Con-
optimal energy management system for a commercial building ference; 2017:106‐110.
with renewable energy generation under real‐time electricity 79. Hunt S. Making Competition Work in Electricity. New York:
prices. Sustain Cities Soc. 2018;41:392‐404. Wiley Publication; 2002.
66. Meng M, Guo M, Yuan Y, et al. Research of DC micro‐grid 80. Ding L, Wang LY, Yin G, Zheng WX, Han Q‐L. Distributed
based on sustainable buildings and electric vehicles. In: 2014 energy management for smart grids with an event‐triggered
17th International Conference on Electrical Machines and Sys- communication scheme. IEEE Trans Control Syst Technol.
tems (ICEMS). Hangzhou, China: IEEE Conference; 2018;99:1‐12.
2014:269‐272. 81. Aftab MA, Hussain SS, Ali I, Ustun TS. IEC 61850 and XMPP
67. Liang Z, Alsafasfeh Q, Jin T, Pourbabak H, Su W. Risk‐ communication based energy management in microgrids con-
constrained optimal energy management for virtual power sidering electric vehicles. IEEE Access. 2018;6:35657‐35668.
plants considering correlated demand response. IEEE Trans 82. Pitì A, Verticale G, Rottondi C, Capone A, Lo Schiavo L. The
Smart Grid. 2019;10(2):1577‐1587. role of smart meters in enabling real‐time energy services for
68. Endo M, Nakajima H, Hata Y. Simplified factory energy man- households: the Italian case. Energies. 2017;10(2):199.
agement system based on operational condition estimation by 83. Mbungu NT, Naidoo R, Bansal RC, Bipath M. Optimal single
sensor data. In: Automation Science and Engineering (CASE), phase smart meter design. J Eng. 2017;2017(13):1220‐1224.
2012 IEEE International Conference on. Seoul, South Korea:
84. Sardou IG, Azad‐Farsani E. Network expansion planning with
IEEE Conference; 2012:14‐19.
microgrid aggregators under uncertainty. IET Gener Transm
69. Masjosthusmann C, Köhler U, Decius N, Büker U. A vehicle Distrib. 2018;12(9):2105‐2114.
energy management system for a battery electric vehicle. In:
85. Graditi G, Di Somma M, Siano P. Optimal bidding strategy for a
Vehicle Power and Propulsion Conference (VPPC), 2012 IEEE.
DER aggregator in the day‐ahead market in the presence of
Seoul, South Korea: IEEE Conference; 2012:339‐344.
demand flexibility. IEEE Trans Ind Electron. 2018;66(2):1509‐1519.
70. Ahmad A, Khan A, Javaid N, et al. An optimized home energy
86. Meng K, Dong ZY, Xu Z, Zheng Y, Hill DJ. Coordinated dis-
management system with integrated renewable energy and
patch of virtual energy storage systems in smart distribution
storage resources. Energies. 2017;10(4):549.
networks for loading management. IEEE Trans Syst Man
71. Correa‐Florez CA, Gerossier A, Michiorri A, Kariniotakis G. Cybern Syst. 2017;49(4):776‐786.
Stochastic operation of home energy management systems
87. Rahmani‐Andebili M, Venayagamoorthy GK. Investigating
including battery cycling. Appl Energy. 2018;225:1205‐1218.
effects of changes in power market regulations on demand‐side
72. Yanine F, Córdova FM. Homeostatic control in grid‐connected resources aggregators. In: 2015 IEEE Power & Energy Society
micro‐generation power systems: a means to adapt to changing General Meeting. Denver, CO, USA: IEEE Conference; 2015:1‐5.
scenarios while preserving energy sustainability. In: 2013 Interna-
88. Mohiti M, Monsef H, Mazidi M, Anvari‐Moghaddam A, Guer-
tional Renewable and Sustainable Energy Conference (IRSEC).
rero JM. A decentralized model for coordinated operation of
Ouarzazate, Morocco: IEEE Conference; 2013:525‐530.
distribution network and EV aggregators. In: 2018 IEEE Inter-
73. Yanine FF, Caballero FI, Sauma EE, Córdova FM. Building national Conference on Environment and Electrical
sustainable energy systems: homeostatic control of grid‐ Engineering and 2018 IEEE Industrial and Commercial Power
connected microgrids, as a means to reconcile power supply Systems Europe (EEEIC/I&CPS Europe). Palermo, Italy: IEEE
and energy demand response management. Renew Sustain Conference; 2018:1‐6.
Energy Rev. 2014;40:1168‐1191.
89. Shafie‐Khah M, Fitiwi DZ, Catalão JPS, Heydarian‐Forushani
74. Mbungu T, Naidoo R, Bansal R, Bipath M. Smart SISO‐MPC E, Golshan MEH. Simultaneous participation of demand
based energy management system for commercial buildings: response aggregators in ancillary services and demand
technology trends. In: 2016 Future Technologies Conference response exchange markets. In: 2016 IEEE/PES Transmission
(FTC). San Francisco, CA, USA: IEEE Conference; and Distribution Conference and Exposition (T&D). Dallas, TX,
2016:750‐753. USA: IEEE Conference; 2016:1‐5.
75. Mbungu NT, Naidoo RM, Bansal RC. Real‐time electricity pric- 90. Pei W, Du Y, Deng W, Sheng K, Xiao H, Qu H. Optimal bid-
ing: TOU‐MPC based energy management for commercial ding strategy and intramarket mechanism of microgrid
buildings. Energy Procedia. 2017;105:3419‐3424. aggregator in real‐time balancing market. IEEE Trans Ind Inf.
76. Ku T, Park W, Choi H. Demand response operation method on 2016;12(2):587‐596.
energy big data platform. In: 2018 Tenth International Confer- 91. Sauba G, Burgt J v d, Varvarigos M, Makris P, Mamounakis I,
ence on Ubiquitous and Future Networks (ICUFN). Prague, Vergados DJ. VIMSEN—new energy management tool for RES
Czech Republic: IEEE Conference; 2018:823‐825. aggregators. In: 2016 IEEE International Conference on Emerging
77. Ozadowicz A, Grela J. Control application for Internet of Technologies and Innovative Business Practices for the Transfor-
things energy meter—a key part of integrated building energy mation of Societies (EmergiTech). Balaclava, Mauritius: IEEE
management system. In: 2015 IEEE 20th Conference on Emerg- Conference; 2016:124‐129.
ing Technologies & Factory Automation (ETFA). Luxembourg, 92. Somma MD, Graditi G, Siano P. Optimal bidding strategy for a
Luxembourg: IEEE Conference; 2015. DER aggregator in the day‐ahead market in the presence of
78. Ku T, Park W, Choi H. IoT energy management platform for demand flexibility. IEEE Trans Ind Electron. 2019;66(2):
microgrid. In: 2017 IEEE 7th International Conference on Power 1509‐1519.
RATHOR AND SAXENA 39

93. “Global EV Outlook 2017 Two million and counting,” 2017, 111. Graham R. Comparing the benefits and impacts of hybrid elec-
Available: www.iea.org. tric vehicle options. Electric Power Research Institute (EPRI),
94. “United States Environmental Protection Agency” 2017. Palo Alto, CA, Report. 2001;1000349.
95. “Fuel Economy, US Govt. source for fuel Economy Informa- 112. An F, Frank A, Ross M. Meeting both ZEV and PNGV goals
tion,” 2017. with a hybrid electric vehicle—an exploration. SAE Technical
96. Tie SF, Tan CW. A review of energy sources and energy man- Paper0148‐7191. 1996.
agement system in electric vehicles. Renew Sustain Energy Rev. 113. Clement‐Nyns K, Haesen E, Driesen J. The impact of charging
2013;20:82‐102. plug‐in hybrid electric vehicles on a residential distribution
97. Kempton W, Udo V, Huber K, Komara K, Letendre S, Baker S, grid. IEEE Trans Power Syst. 2010;25(1):371‐380.
Brunner D, Pearre N. A test of vehicle‐to‐grid (V2G) for energy 114. Luo C, Huang Y‐F, Gupta V. A consumer behavior based
storage and frequency regulation in the PJM system. Mid‐ approach to multi‐stage EV charging station placement. In:
Atlantic Grid Interactive Cars Consortium. 2009. 2015 IEEE 81st Vehicular Technology Conference (VTC Spring).
98. Quinn C, Zimmerle D, Bradley TH. The effect of communica- Glasgow, UK: IEEE Conference; 2015:1‐6.
tion architecture on the availability, reliability, and 115. Abdelsamad SF, Morsi WG, Sidhu TS. Optimal secondary dis-
economics of plug‐in hybrid electric vehicle‐to‐grid ancillary tribution system design considering plug‐in electric vehicles.
services. J Power Sources. 2010;195(5):1500‐1509. Electr Pow Syst Res. 130:266‐276. 2016/01/01/2016
99. Mohamed A, Salehi V, Ma T, Mohammed O. Real‐time energy 116. El‐Zonkoly A, Ashour H, Ahmed A. Optimal allocation, sizing
management algorithm for plug‐in hybrid electric vehicle and energy management of PHEV parking lots in distribution
charging parks involving sustainable energy. IEEE Trans Sus- system. In: 2014 5th International Renewable Energy Congress
tainable Energy. 2014;5(2):577‐586. (IREC). Hammamet, Tunisia: IEEE Conference; 2014:1‐5.
100. Sortomme E, El‐Sharkawi MA. Optimal scheduling of vehicle‐ 117. Kamankesh H, Agelidis VG, Kavousi‐Fard A. Optimal schedul-
to‐grid energy and ancillary services. IEEE Trans Smart Grid. ing of renewable micro‐grids considering plug‐in hybrid
2012;3(1):351‐359. electric vehicle charging demand. Energy. 2016;100:285‐297.
101. Kesler M, Kisacikoglu MC, Tolbert LM. Vehicle‐to‐grid reactive 118. Nunna HSVSK, Battula S, Doolla S, Srinivasan D. Energy man-
power operation using plug‐in electric vehicle bidirectional agement in smart distribution systems with vehicle‐to‐grid
offboard charger. IEEE Trans Ind Electron. 2014;61(12):6778‐6784. integrated microgrids. IEEE Trans Smart Grid. 2018;9(5):1‐1.
102. Manbachi M, Farhangi H, Palizban A, Arzanpour S. A novel 119. Buja G, Bertoluzzo M, Fontana C. Reactive power compensa-
Volt‐VAR Optimization engine for smart distribution networks tion capabilities of V2G‐enabled electric vehicles. IEEE Trans
utilizing Vehicle to Grid dispatch. Int J Electr Power Energy Power Electron. 2017;32(12):9447‐9459.
Syst. 2016;74:238‐251. 120. Bo S, Dragicevic T, Savaghebi M, Vasquez JC, Guerrero JM.
103. Bessa RJ, Matos MA. Economic and technical management of Reactive power support of electrical vehicle charging station
an aggregation agent for electric vehicles: a literature survey. upgraded with flywheel energy storage system. In: 2015 IEEE
Eur T Electr Power. 2012;22(3):334‐350. Eindhoven PowerTech. Eindhoven, Netherlands: IEEE Confer-
104. Tomić J, Kempton W. Using fleets of electric‐drive vehicles for ence; 2015:1‐6.
grid support. J Power Sources. 2007;168(2):459‐468. 121. Alamo DH, Medina RN, Ruano SD, et al. An advanced fore-
105. Mwasilu F, Justo JJ, Kim E‐K, Do TD, Jung J‐W. Electric vehi- casting system for the optimum energy management of island
cles and smart grid interaction: a review on vehicle to grid and microgrids. Energy Procedia. 2019;159:111‐116.
renewable energy sources integration. Renew Sustain Energy 122. Mehrjerdi H, Bornapour M, Hemmati R, Ghiasi SMS. Unified
Rev. 2014;34:501‐516. energy management and load control in building equipped
106. Hadley SW, Tsvetkova AA. Potential impacts of plug‐in hybrid with wind‐solar‐battery incorporating electric and hydrogen
electric vehicles on regional power generation. Electr J. vehicles under both connected to the grid and islanding modes.
2009;22(10):56‐68. Energy. 2019;168:919‐930.

107. Fernandez LP, San Román TG, Cossent R, Domingo CM, Frias 123. Kundur P, Paserba J, Ajjarapu V, et al. Definition and classifi-
P. Assessment of the impact of plug‐in electric vehicles on cation of power system stability. IEEE Trans Power Syst.
distribution networks. IEEE Trans Power Syst. 2011;26(1): 2004;19(2):1387‐1401.
206‐213. 124. Caballero F, Sauma E, Yanine F. Business optimal design of a
108. Johnston B, McGoldrick T, Funston D, et al. Continued design grid‐connected hybrid PV (photovoltaic)‐wind energy system
and development of the University of California, Davis without energy storage for an Easter Island's block. Energy.
FutureCar. SAE SPEC PUBL, SAE, WARRENDALE, PA, 2013;61:248‐261.
(USA), Feb 1998. 1998;1359:53‐66. 125. Abdeltawab HH, Mohamed YA‐RI. Robust energy manage-
109. Tabari M, Yazdani A. An energy management strategy for a ment of a hybrid wind and flywheel energy storage system
DC distribution system for power system integration of plug‐ considering flywheel power losses minimization and grid‐
in electric vehicles. IEEE Trans Smart Grid. 2016;7(2):659‐668. code constraints. IEEE Trans Ind Electron.
2016;63(7):4242‐4254.
110. Amini MH, Islam A. Allocation of electric vehicles' parking lots
in distribution network, in ISGT 2014. Washington, DC, USA: 126. Foley AM, Leahy PG, Marvuglia A, McKeogh EJ. Current
IEEE Conference; 2014, pp. 1‐5. methods and advances in forecasting of wind power genera-
tion. Renew Energy. 2012;37(1):1‐8.
40 RATHOR AND SAXENA

127. Zhao B, Wang X, Lin D, et al. Energy management of multiple 142. Abronzini U, Attaianese C, D'Arpino M, et al. Optimal energy
microgrids based on a system of systems architecture. IEEE control for smart charging infrastructures with ESS and REG.
Trans Power Syst. 2018;33(6):6410‐6421. In: Electrical Systems for Aircraft, Railway, Ship Propulsion
128. Liao Z, Ruan X. A novel power management control strategy and Road Vehicles & International Transportation Electrifica-
for stand‐alone photovoltaic power system. In: 2009 IEEE 6th tion Conference (ESARS‐ITEC), International Conference on.
International Power Electronics and Motion Control Conference. Toulouse, France: IEEE Conference; 2016:1‐6.
Wuhan, China: IEEE Conference; 2009:445‐449. 143. Akhavan‐Rezaei E, Shaaban MF, El‐Saadany EF, Karray F.
129. Ismail MS, Moghavvemi M, Mahlia TMI. Techno‐economic New ems to incorporate smart parking lots into demand
analysis of an optimized photovoltaic and diesel generator response. IEEE Trans Smart Grid. 2016;9(2):1376‐1386.
hybrid power system for remote houses in a tropical climate. 144. Zhou B, Littler T, Foley A. Electric vehicle capacity forecasting
Energ Conver Manage. 69:163‐173. 2013/05/01/2013 model with application to load levelling. In: Power & Energy
130. Dash V, Bajpai P. Power management control strategy for a Society General Meeting, 2015 IEEE. Denver, CO, USA: IEEE
stand‐alone solar photovoltaic‐fuel cell–battery hybrid system. Conference; 2015:1‐5.
Sustainable Energy Technol Assess. 9:68‐80. 2015/03/01/2015 145. Valencia F, Collado J, Sáez D, Marín LG. Robust energy man-
131. Finn P, Fitzpatrick C. Demand side management of industrial agement system for a microgrid based on a fuzzy prediction
electricity consumption: promoting the use of renewable interval model. IEEE Trans Smart Grid. 2015;7(3):1486‐1494.
energy through real‐time pricing. Appl Energy. 113:11‐21. 146. Rezaei N, Ahmadi A, Khazali AH, Guerrero JM. Energy and
2014/01/01/2014 frequency hierarchical management system using information
132. Ozbilen A, Dincer I, Naterer GF, Aydin M. Role of hydrogen gap decision theory for islanded microgrids. IEEE Trans Ind
storage in renewable energy management for Ontario. Int J Electron. 2018;65(10):7921‐7932.
Hydrogen Energy. 37(9):7343‐7354. 2012/05/01/2012 147. Zhai M, Liu Y, Zhang T, Zhang Y. Robust model predictive
133. Abedi S, Alimardani A, Gharehpetian GB, Riahy GH, control for energy management of isolated microgrids. In:
Hosseinian SH. A comprehensive method for optimal power 2017 IEEE International Conference on Industrial Engineering
management and design of hybrid RES‐based autonomous and Engineering Management (IEEM). Singapore, Singapore:
energy systems. Renew Sustain Energy Rev. 16(3):1577‐1587. IEEE Conference; 2017:2049‐2053.
2012/04/01/2012 148. Zaheer SA, Choi S, Jung C, Kim J. A modular implementation
134. Dahmane M, Bosche J, El‐Hajjaji A, Dafarivar M. Renewable scheme for nonsingleton type‐2 fuzzy logic systems with
energy management algorithm for stand‐alone system. In: input uncertainties. IEEE/ASME Trans Mechatron. 2015;20(6):
2013 International Conference on Renewable Energy Research 3182‐3193.
and Applications (ICRERA). Madrid, Spain: IEEE Conference; 149. Tahmasebi M, Pasupuleti J. Electricity demand uncertainty
2013:621‐626. modeling using enhanced path‐based scenario generation
135. Mbungu NT, Naidoo R, Bansal RC, Bipath M. Optimisation of method. In: 2017 6th International Youth Conference on Energy
grid connected hybrid photovoltaic–wind–battery system using (IYCE). Budapest, Hungary: IEEE Conference; 2017:1‐5.
model predictive control design. IET Renew Power Gener. 150. Zhang N, Kang C, Xia Q, et al. A convex model of risk‐based
2017;11(14):1760‐1768. unit commitment for day‐ahead market clearing considering wind
136. Yanine FF, Sauma EE. Review of grid‐tie micro‐generation sys- power uncertainty. IEEE Trans Power Syst. 2015;30(3):1582‐1592.
tems without energy storage: towards a new approach to 151. Savin AA, Guba VG, Maxson BD. Covariance based uncer-
sustainable hybrid energy systems linked to energy efficiency. tainty analysis with unscented transformation. In: 82nd
Renew Sustain Energy Rev. 2013;26:60‐95. ARFTG Microwave Measurement Conference. Columbus, OH,
137. Yanine FF, Caballero FI, Sauma EE, Córdova FM. Homeostatic USA: IEEE Conference; 2013:1‐4.
control, smart metering and efficient energy supply and con- 152. Shafie‐Khah M, Neyestani N, Damavandi M, Gil F, Catalão J.
sumption criteria: a means to building more sustainable Economic and technical aspects of plug‐in electric vehicles in
hybrid micro‐generation systems. Renew Sustain Energy Rev. electricity markets. Renew Sustain Energy Rev. 2016;53:1168‐1177.
2014;38:235‐258. 153. Žiha K. Statistical variability vs. probabilistic uncertainty. In:
138. Mbungu NT, Bansal RC, Naidoo RM. Smart energy coordination Proceedings of the ITI 2013 35th International Conference on
of autonomous residential home. IET Smart Grid. 2019. Information Technology Interfaces. Cavtat, Croatia: IEEE Con-
139. Verma PP, Srinivasan D, Swarup K, Mehta R. A review of ference; 2013:271‐274.
uncertainty handling techniques in smart grid. Int J Uncer- 154. Wang W, Maio FD, Baraldi P, Zio E. A hybrid Monte Carlo and
tainty Fuzziness Knowledge Based Syst. 2018;26(03):345‐378. possibilistic approach to estimate non‐suppression probability
140. Gao S, Chau KT, Liu C, Wu D, Chan CC. Integrated energy in fire probabilistic safety analysis. In: 2017 2nd International
management of plug‐in electric vehicles in power grid with Conference on System Reliability and Safety (ICSRS). Milan,
renewables. IEEE Trans Veh Technol. 2014;63(7):3019‐3027. Italy: IEEE Conference; 2017:552‐559.
141. Xiao H, Pei W, Yang Y, Qu H, Qi Z, Kong L. Multi‐objective 155. Sarabi S, Davigny A, Courtecuisse V, Riffonneau Y, Robyns B.
optimal operation of micro‐grid with plug‐in electric vehicles. Potential of vehicle‐to‐grid ancillary services considering the
In: Transportation Electrification Asia‐Pacific (ITEC Asia‐ uncertainties in plug‐in electric vehicle availability and
Pacific), 2014 IEEE Conference and Expo. Beijing, China: IEEE service/localization limitations in distribution grids. Appl
Conference; 2014:1‐5. Energy. 2016;171:523‐540.
RATHOR AND SAXENA 41

156. Qi X, Wu G, Boriboonsomsin K, Barth MJ. Development and 173. Gellings CW. The concept of demand‐side management for
evaluation of an evolutionary algorithm‐based online energy electric utilities. Proc IEEE. 1985;73(10):1468‐1470.
management system for plug‐in hybrid electric vehicles. IEEE 174. Jian L, Zheng Y, Shao Z. High efficient valley‐filling strategy
Trans Intell Transp Syst. 2017;18(8):2181‐2191. for centralized coordinated charging of large‐scale electric vehi-
157. Wang X, Liang Q. Energy management strategy for plug‐in cles. Appl Energy. 2017;186:46‐55.
hybrid electric vehicles via bidirectional vehicle‐to‐grid. IEEE 175. Khatiri‐Doost S, Amirahmadi M. Peak shaving and power
Syst J. 2017;11(3):1789‐1798. losses minimization by coordination of plug‐in electric vehi-
158. Khadem SK, Basu M, Conlon MF. Capacity enhancement and cles charging and discharging in smart grids. In:
flexible operation of unified power quality conditioner in smart Environment and Electrical Engineering and 2017 IEEE Indus-
and microgrid network. AIMS Energy. 2018;6(1):49‐69. trial and Commercial Power Systems Europe (EEEIC/I&CPS
159. Safavizadeh A, Yousefi G, Karshenas H. Voltage variation Europe), 2017 IEEE International Conference on. Milan, Italy:
mitigation using reactive power management of distributed IEEE Conference; 2017:1‐5.
energy resources in a smart distribution system. IEEE Trans 176. Zhang X, Zheng L. Coordinated dispatch of the wind‐thermal
Smart Grid. 2017;10(2):1907‐1915. power system by optimizing electric vehicle charging. Clust
160. Naidoo R, Pillay P, Visser J, Bansal RC, Mbungu NT. An adap- Comput. 1‐1.
tive method of symmetrical component estimation. Electr Pow 177. Xing H, Lin Z, Fu M. A new decentralized algorithm for opti-
Syst Res. 2018;158:45‐55. mal load shifting via electric vehicles. In: Control Conference
161. Behrangrad M. A review of demand side management business (CCC), 2017 36th Chinese. Dalian, China: IEEE Conference;
models in the electricity market. Renew Sustain Energy Rev. 2017:10708‐10713.
2015;47:270‐283. 178. Hu RL, Skorupski R, Entriken R, Ye Y. A mathematical pro-
162. Benetti G, Caprino D, Della Vedova ML, Facchinetti T. Electric gramming formulation for optimal load shifting of electricity
load management approaches for peak load reduction: a demand for the smart grid. IEEE Trans Big Data. 2016;0:0‐0.
systematic literature review and state of the art. Sustain Cities 179. Shafie‐Khah M, Heydarian‐Forushani E, Osório GJ, et al.
Soc. 2016;20:124‐141. Optimal behavior of electric vehicle parking lots as demand
163. Setlhaolo D, Xia X. Combined residential demand side response aggregation agents. IEEE Trans Smart Grid.
management strategies with coordination and economic 2016;7(6):2654‐2665.
analysis. Int J Electr Power Energy Syst. 79:150‐160. 2016/07/ 180. Faruqui A, Eakin K. Pricing in Competitive Electricity Markets.
01/2016 New York: Springer Science & Business Media; 2012.
164. Dixon RK, McGowan E, Onysko G, Scheer RM. US energy 181. Misra S, Bera S, Ojha T. D2P: distributed dynamic pricing
conservation and efficiency policies: challenges and opportuni- policyin smart grid for PHEVs management. IEEE Trans Paral-
ties. Energy Policy. 2010;38(11):6398‐6408. lel Distrib Syst. 2015;26(3):702‐712.
165. Xia X, Zhang J. Mathematical description for the measurement 182. Kiesel R, Paraschiv F. Econometric analysis of 15‐minute intra-
and verification of energy efficiency improvement. Appl day electricity prices. Energy Econ. 2017;64:77‐90.
Energy. 2013;111:247‐256. 183. Weitzel T, Glock CH. Energy management for stationary elec-
166. Zhang L, Xia X, Zhang J. Improving energy efficiency of tric energy storage systems: a systematic literature review.
cyclone circuits in coal beneficiation plants by pump‐storage Eur J Oper Res. 2018;264(2):582‐606.
systems. Appl Energy. 2014;119:306‐313. 184. Veerapen N, Ochoa G, Harman M, Burke EK. An integer linear
167. Malatji EM, Zhang J, Xia X. A multiple objective optimisation programming approach to the single and bi‐objective next
model for building energy efficiency investment decision. release problem. Inf Softw Technol. 65:1‐13. 2015/09/01/2015
Energ Buildings. 2013;61:81‐87. 185. Rotering N, Ilic M. Optimal charge control of plug‐in hybrid
168. Wang N, Zhang J, Xia X. Energy consumption of air condi- electric vehicles in deregulated electricity markets. IEEE Trans
tioners at different temperature set points. Energ Buildings. Power Syst. 2011;26(3):1021‐1029.
2013;65:412‐418. 186. Chen Z, Xia B, You C, Mi CC. A novel energy management
169. Saha A, Kuzlu M, Pipattanasomporn M. Demonstration of a method for series plug‐in hybrid electric vehicles. Appl Energy.
home energy management system with smart thermostat 2015;145:172‐179.
control. In: Innovative Smart Grid Technologies (ISGT), 2013 187. Rajasekharan J, Koivunen V. Optimal energy consumption
IEEE PES. Washington, DC, USA: IEEE Conference; 2013:1‐6. model for smart grid households with energy storage. IEEE J
170. Ye X, Xia X, Zhang J. Optimal sampling plan for clean Sel Top Sign Proces. 2014;8(6):1154‐1166.
development mechanism energy efficiency lighting projects. 188. Byrne RH, Nguyen TA, Copp DA, Chalamala BR, Gyuk I.
Appl Energy. 2013;112:1006‐1015. Energy management and optimization methods for grid energy
171. Wang B, Xia X, Zhang J. A multi‐objective optimization model storage systems. IEEE Access. 2018;6:13231‐13260.
for the life‐cycle cost analysis and retrofitting planning of 189. Belotti P, Kirches C, Leyffer S, Linderoth J, Luedtke J, Mahajan
buildings. Energ Buildings. 2014;77:227‐235. A. Mixed‐integer nonlinear optimization. Acta Numer.
172. Ye X, Xia X, Zhang L, Zhu B. Optimal maintenance planning 2013;22:1‐131.
for sustainable energy efficiency lighting retrofit projects by a 190. Bongermino E, Mastrorocco F, Tomaselli M, Monopoli V, Naso
control system approach. Control Eng Pract. 2015;37:1‐10. D. Model and energy management system for a parallel hybrid
42 RATHOR AND SAXENA

electric unmanned aerial vehicle. In: Industrial Electronics 207. Sousa T, Morais H, Vale Z, Faria P, Soares J. Intelligent energy
(ISIE), 2017 IEEE 26th International Symposium on. Edin- resource management considering vehicle‐to‐grid: a simulated
burgh, UK: IEEE Conference; 2017:1868‐1873. annealing approach. IEEE Trans Smart Grid. 2011;3(1):
191. Uebel S, Murgovski N, Tempelhahn C, Bäker B. Optimal 535‐542.
energy management and velocity control of hybrid electric 208. Elsied M, Oukaour A, Youssef T, Gualous H, Mohammed O.
vehicles. IEEE Trans Veh Technol. 2018;67(1):327‐337. An advanced real time energy management system for
192. Hu X, Martinez CM, Yang Y. Charging, power management, microgrids. Energy. 2016;114:742‐752.
and battery degradation mitigation in plug‐in hybrid electric 209. Arefifar SA, Ordonez M, Mohamed YA‐RI. Energy manage-
vehicles: a unified cost‐optimal approach. Mech Syst Signal Pro- ment in multi‐microgrid systems—development and
cess. 2017;87:4‐16. assessment. IEEE Trans Power Syst. 2016;32(2):910‐922.
193. Henao‐Muñoz AC, Saavedra‐Montes AJ, Ramos‐Paja CA. 210. Marzband M, Ghazimirsaeid SS, Uppal H, Fernando T. A real‐
Energy management system for an isolated microgrid with time evaluation of energy management systems for smart
photovoltaic generation. In: Synthesis, Modeling, Analysis and hybrid home microgrids. Electr Pow Syst Res. 2017;143:624‐633.
Simulation Methods and Applications to Circuit Design (SMACD), 211. Mohammadi S, Soleymani S, Mozafari B. Scenario‐based sto-
2017 14th International Conference on. IEEE; 2017:1‐4. chastic operation management of microgrid including wind,
194. Zhang H, Hu Z, Xu Z, Song Y. Evaluation of achievable photovoltaic, micro‐turbine, fuel cell and energy storage
vehicle‐to‐grid capacity using aggregate PEV model. IEEE devices. Int J Electr Power Energy Syst. 2014;54:525‐535.
Trans Power Syst. 2017;32(1):784‐794. 212. Wu Z, Zhang X, Brandt J, Zhou S, Jia‐Ning LI. Three control
195. Su W, Wang J, Roh J. Stochastic energy scheduling in approaches for optimized energy flow with home energy man-
microgrids with intermittent renewable energy resources. IEEE agement system. IEEE Power Energy Technol Syst J.
Trans Smart Grid. 2014;5(4):1876‐1883. 2015;2(1):21‐31.
196. Khodakarami A, Farahani HF, Aghaei J. Stochastic characteri- 213. Abdeltawab HH, Mohamed YAI. Market‐oriented energy man-
zation of electricity energy markets including plug‐in electric agement of a hybrid wind‐battery energy storage system via
vehicles. Renew Sustain Energy Rev. 2017;69:112‐122. model predictive control with constraint optimizer. IEEE Trans
197. Alipour M, Zare K, Abapour M. MINLP probabilistic schedul- Ind Electron. 2015;62(11):6658‐6670.
ing model for demand response programs integrated energy 214. Morstyn T, Savkin AV, Hredzak B, Tuan HD. Scalable energy
hubs. IEEE Trans Ind Inf. 2018;14(1):79‐88. management for low voltage microgrids using multi‐agent stor-
198. Rahim S, Javaid N, Ahmad A, et al. Exploiting heuristic algo- age system aggregation. IEEE Trans Power Syst.
rithms to efficiently utilize energy management controllers with 2018;33(2):1614‐1623.
renewable energy sources. Energ Buildings. 2016;129:452‐470. 215. Hu M‐C, Lu S‐Y, Chen Y‐H. Stochastic programming and mar-
199. Klein L, Kwak JY, Kavulya G, et al. Coordinating occupant ket equilibrium analysis of microgrids energy management
behavior for building energy and comfort management using systems. Energy. 2016;113:662‐670.
multi‐agent systems. Autom Constr. 2012;22:525‐536. 216. Prodan I, Zio E. A model predictive control framework for reli-
200. Sianaki OA, Masoum MA. A multi‐agent intelligent decision able microgrid energy management. Int J Electr Power Energy
making support system for home energy management in smart Syst. 2014;61:399‐409.
grid: a fuzzy TOPSIS approach. Multiagent Grid Syst. 217. Chen Z, Wu L, Fu Y. Real‐time price‐based demand response
2013;9(3):181‐195. management for residential appliances via stochastic optimiza-
201. Ahmed MS, Mohamed A, Khatib T, Shareef H, Homod RZ, Ali tion and robust optimization. IEEE Trans Smart Grid.
JA. Real time optimal schedule controller for home energy 2012;3(4):1822‐1831.
management system using new binary backtracking search 218. Chaouachi A, Kamel RM, Andoulsi R, Nagasaka K.
algorithm. Energ Buildings. 2017;138:215‐227. Multiobjective intelligent energy management for a microgrid.
202. Megahed TF, Abdelkader SM, Zakaria A. Energy management IEEE Trans Ind Electron. 2012;60(4):1688‐1699.
in zero‐energy building using neural network predictive con- 219. Solanki BV, Bhattacharya K, Canizares CA. A sustainable
trol. IEEE Internet Things J. 2019;6(3):5336‐5344. energy management system for isolated microgrids. IEEE
203. Trovão JP, Antunes CH. A comparative analysis of meta‐heuristic Trans Sustainable Energy. 2017;8(4):1507‐1517.
methods for power management of a dual energy storage system 220. Baghaee HR, Mirsalim M, Gharehpetian GB, Talebi HA. A
for electric vehicles. Energ Conver Manage. 2015;95:281‐296. decentralized power management and sliding mode control
204. Panwar M, Suryanarayanan S, Hovsapian R. A multi‐criteria strategy for hybrid AC/DC microgrids including renewable
decision analysis‐based approach for dispatch of electric energy resources. IEEE Trans Ind Inf. Singapore: Springer;
microgrids. Int J Electr Power Energy Syst. 2017;88:99‐107. 2017.
205. Ruiz‐Cortés M, González‐Romera E, Amaral‐Lopes R, 221. Guo Y, Zhao C. Islanding‐aware robust energy management
et al. Optimal charge/discharge scheduling of batteries in for microgrids. IEEE Trans Smart Grid. 2016;9(2):1301‐1309.
microgrids of prosumers. IEEE Trans Energy Convers. 2018;34 222. Tungadio D, Bansal R, Siti M, Mbungu N. Predictive active
(1):468‐477. power control of two interconnected microgrids. Technol Econ
206. Radosavljević J, Jevtić M, Klimenta D. Energy and operation Smart Grids Sustain Energy. 2018;3(1):3.
management of a microgrid using particle swarm optimization. 223. Han J, Choi C‐S, Lee I. More efficient home energy
Eng Optim. 2016;48(5):811‐830. management system based on ZigBee communication and
RATHOR AND SAXENA 43

infrared remote controls. IEEE Trans Consum Electron. plants considering correlated demand response. IEEE Trans
2011;57(1):85‐89. Smart Grid. 2017;10(2):1577‐1587.
224. Parikh PP, Kanabar MG, Sidhu TS. Opportunities and 228. (2019, March). Data protection Impact assessment template for
challenges of wireless communication technologies for smart smart grid and smart metering environment. Available:
grid applications. In: IEEE PES General Meeting. Providence, https://ec.europa.eu/energy/en/data‐protection‐impact‐assess-
RI, USA: IEEE Conference; 2010:1‐7. ment‐smart‐grid‐and‐smart‐metering‐environment
225. (March). Goal 13: take urgent action to combat climate change
and its impacts.
226. Gyamfi S, Krumdieck S. Price, environment and security: How to cite this article: Rathor SK, Saxena D.
exploring multi‐modal motivation in voluntary residential peak Energy management system for smart grid: An
demand response. Energy Policy. 2011;39(5):2993‐3004. overview and key issues. Int J Energy Res.
227. Liang Z, Alsafasfeh Q, Jin T, Pourbabak H, Su W. Risk‐ 2020;1–43. https://doi.org/10.1002/er.4883
constrained optimal energy management for virtual power