2424 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 62, NO.

4, APRIL 2015

A Review of Architectures and Concepts for

Intelligence in Future Electric Energy Systems
Thomas Strasser, Senior Member, IEEE, Filip Andrén, Member, IEEE, Johannes Kathan,
Carlo Cecati, Fellow, IEEE, Concettina Buccella, Senior Member, IEEE,
Pierluigi Siano, Senior Member, IEEE, Paulo Leitão, Senior Member, IEEE,
Gulnara Zhabelova, Valeriy Vyatkin, Senior Member, IEEE,
Pavel Vrba, Senior Member, IEEE, and Vladimír Mařík, Senior Member, IEEE

Abstract—Renewable energy sources are one key en- into the power grids is analyzed. Energy management
abler to decrease greenhouse gas emissions and to cope and demand response methods and important automation
with the anthropogenic climate change. Their intermit- paradigms and domain standards are also reviewed.
tent behavior and limited storage capabilities present a
new challenge to power system operators to maintain Index Terms—Ancillary services, automation architec-
power quality and reliability. Additional technical complex- tures, control concepts, demand response (DR), demand-
ity arises from the large number of small distributed gen- side management (DSM), distributed generation, energy
eration units and their allocation within the power system. storage, inverters, microgrid, power balancing, power
Market liberalization and changing regulatory framework networks, power system automation, renewable energy
lead to additional organizational complexity. As a result, sources, smart grid, standards.
the design and operation of the future electric energy sys-
tem have to be redefined. Sophisticated information and I. I NTRODUCTION AND M OTIVATION
communication architectures, automation concepts, and
control approaches are necessary in order to manage the
higher complexity of so-called smart grids. This paper pro-
vides an overview of the state of the art and recent devel-
T HE electric energy systems worldwide have to satisfy a
continuously growing demand for electricity and simulta-
neously provide a stable supply. Today, the worldwide power
opments enabling higher intelligence in future smart grids. generation is dominated by fossil fuels resulting in an increase
The integration of renewable sources and storage systems in CO2 emissions and global warming as indicated by the
“World Energy Outlook 2013” from the International Energy
Manuscript received March 31, 2014; revised June 29, 2014; ac- Agency [1]. In order to counteract, there is a clear trend toward
cepted August 14, 2014. Date of publication October 3, 2014; date of a sustainable electric energy system. Minimizing greenhouse
current version March 6, 2015. This work was supported in part by gas emissions caused by power generation will only be possible
the Austrian Climate and Energy Fund under Project DG-EV-HIL (FFG if renewable sources such as photovoltaic (PV) systems, wind
827987) and in part by the Czech Institute of Informatics, Robotics, and
Cybernetics, Czech Technical University in Prague. generators, biomass, and combined heat and power systems
T. Strasser, F. Andrén, and J. Kathan are with Electric Energy Sys- are being installed on a large scale [2]–[4]. They are typically
tems, Energy Department, AIT Austrian Institute of Technology, 1210 available in a decentralized way as distributed energy resources
Vienna, Austria (e-mail: [email protected]; filip.andren@ait.
ac.at; [email protected]). (DERs) [3]. Recent research results, technology developments,
C. Cecati is with the Department of Industrial and Information En- and regulatory alterations have been fundamentally changing
gineering and Economics, University of L’Aquila, 67100 L’Aquila, Italy, the framework conditions; the planning, management, and op-
with Harbin Institute of Technology, Harbin 150001, China, and also with eration of the power systems have to be redefined.
DigiPower Ltd., 67100 L’Aquila, Italy (e-mail: [email protected]).
C. Buccella is with the Department of Industrial and Information In addition, advanced metering, management, and optimiza-
Engineering and Economics, University of L’Aquila, 67100 L’Aquila, tion concepts on the consumer side are currently in the focus
Italy, and also with DigiPower Ltd., 67100 L’Aquila, Italy (e-mail: con- of research and demo projects [5]–[7]. The objective is to
[email protected]).
P. Siano is with the Department of Industrial Engineering, University effectively manage load peaks, by load shedding, peak-load
of Salerno, 84084 Salerno, Italy (e-mail: [email protected]). reduction, etc., to maintain or improve the security of supply.
P. Leitão is with the Polytechnic Institute of Bragança, 5301-857 The above-sketched developments are leading to complex ar-
Bragança, Portugal, and also with the Artificial Intelligence and Com-
puter Science Laboratory (LIACC), 4169-007 Porto, Portugal (e-mail:
chitectures with a tremendous amount of interconnected and in-
[email protected]). telligent components and subsystems, which have to exchange
G. Zhabelova is with the Department of Computer Science, Electri- both the information and the energy. Today, power utilities
cal, and Space Engineering, Luleå Tekniska Universitet, 97186 Luleâ, and system infrastructure operators are increasingly confronted
Sweden (e-mail: [email protected]).
V. Vyatkin is with the Department of Computer Science, Electrical, and with a highly dynamic and less predictable demand–supply
Space Engineering, Luleå Tekniska Universitet, 97187 Luleå, Sweden, balance. Moreover, consumers are evolving into so-called
and also with the Department of Electrical Engineering and Automation, prosumers—local energy consumers and producers.
Aalto University, 00076 Helsinki, Finland (e-mail: [email protected]).
P. Vrba and V. Mařík are with the Czech Institute of Informatics,
The availability of Information and Communication Technol-
Robotics, and Cybernetics, Czech Technical University in Prague, 166 ogy (ICT) and advanced automation concepts provides various
36 Prague, Czech Republic (e-mail: [email protected]; marik@ opportunities to operate highly interconnected power grids
labe.felk.cvut.cz). with corresponding components in a more effective way as
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org. today, known under the term smart grid [8]–[10]. According to
Digital Object Identifier 10.1109/TIE.2014.2361486 Yu et al. [8], it covers the intelligent integration of all users/

0278-0046 © 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
STRASSER et al.: REVIEW OF ARCHITECTURES AND CONCEPTS FOR INTELLIGENCE IN SYSTEMS 2425

stakeholders connected to electricity networks supporting a

sustainable, economical, and secure electricity supply.
This paper therefore discusses important automation archi-
tectures, smart devices, control concepts, and energy man-
agement principles enabling intelligence, decentralization, and
robustness in the field of future electric energy systems and
involved components. It provides a brief overview of the state
of the art, related work, important activities, and achievements
dedicated to smart grid systems and components. Necessary
functions and services to operate such an intelligent energy
infrastructure are discussed and summarized.
In this review paper, Section II gives an overview of the
future electric energy system, corresponding challenges, and
necessary functions, which have to be provided by smart com-
ponents, algorithms, and ICT/automation approaches exploring
concepts from the field of artificial intelligence (AI). The inte-
gration of DERs playing a major role in a sustainable future
system is discussed in Section III, followed by the topic of
energy storages in Section IV. The importance of managing the
Fig. 1. Smart grid vision: intelligent integration of all users/
demand side is pointed out in Section V, addressing demand stakeholders.
response (DR) and energy management principles. Necessary
automation and ICT concepts and paradigms, approaches, and users in a more intelligent manner exploring the AI principles
corresponding domain standards are reviewed in Section VI. [10], [13], [14].
Finally, Section VII deals with the key findings and with an
outlook about future research trends. B. Necessary Smart Grid Functions and Services
The management and operation of the future power sys-
tem and its components—particularly active power distribution
II. F UTURE E LECTRIC E NERGY S YSTEM
grids and microgrids [20]—require new and advanced control
The large-scale implementation of DER from renewable functions. They have to be integrated in the corresponding
sources during the last years fundamentally changed the design, automation solutions, intelligent electronic devices, and grid
planning, and operation of the power systems in various regions components. A brief overview of the most important functions
(U.S., Europe, Australia, etc.). It becomes already visible in and services is given by the following list [11], [21]–[23].
power transmission and distribution grids. System operators • Advanced monitoring and diagnostics: Monitoring and
and utilities have to manage the fluctuating power generation state estimation capability and real-time condition moni-
from DERs and uncoordinated responses to changing condi- toring of components not only in the medium-voltage but
tions of the power grids [10], [11]. also in the low-voltage distribution grids are important fu-
In a number of countries (Denmark, Germany, Italy, Spain, ture features. Self-diagnostic capabilities are helpful new
Australia, etc.), levels of renewables (PV, wind, hydro) have functions, which are valuable in a power grid with a large
already exceeded the local power grid’s hosting capacity, re- amount of DERs, flexible loads, and storage systems.
sulting in power quality disturbances. Smart grids are one of • Optimization/self-optimization capabilities: Fluctuating
the most promising solutions to use the existing power grid electricity generation from renewable sources requires the
infrastructure extended with proper ICT methods in a more effi- ability to (self-)optimize the operation in medium- and
cient way, allowing higher penetration levels of DER [12]–[19]. low-voltage distribution grids. In addition, the availability
of flexible loads and storage systems add additional de-
grees of freedom for operating future smart grids.
A. Toward Active Power Grids • Automatic grid (topology) reconfiguration: Support of au-
In the past, the power system operation has been done mainly tomatic or semiautomatic adjustment of the distribution
manually. The integration of DERs with smart power converters grid topology due to optimization purposes (e.g., maxi-
(PCs), the possibility to handle peak loads on the demand mum amount of distributed generators and flexible loads)
side, and technology developments in energy storage systems or due to fault management and power system restoration.
(ESS) together with advanced ICT solutions result in a higher • Adaptive protection: Automatic or semiautomatic adap-
automation degree, as outlined in Fig. 1. tion of protection devices (e.g., protection relays and
All components in the future smart grid—flexible loads, breakers) with respect to the actual power grid conditions
energy storages, smart substations with on-load tap changers (e.g., adaptation of the protection system settings due to
(OLTCs), DERs, metering systems, etc.—are interconnected in the bidirectional power flow caused by DERs).
addition to the power system with a corresponding commu- • Distributed power system management: Distributed con-
nication and automation infrastructure. Together with proper trol with automatic decision finding processes and
control approaches and strategies, a smart grid system can be proactive fault prevention have to be provided for the
implemented, which allows monitoring, managing, and opti- power system infrastructure operators in medium- and
mizing the future electric energy grids and its components and low-voltage grids.
2426 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 62, NO. 4, APRIL 2015

• Component level: For the integration of DER and dis-

tributed energy storages into the power system, advanced
component functions such as the provision of ancillary
services and adaptive protection possibilities are required.
Such intelligence is either used for local optimization
purposes (component behavior) or for supporting the op-
timization of subsystems and systems in a coordinated
manner on the higher levels. DER components, as shown
in Fig. 2, are typically part of this level.
• Subsystem level: Deals mainly with the optimization
and/or control of subsystems such as microgrids or home/
building energy automation systems. The objective is to
optimize (sub)systems with few components (DER, stor-
age system, etc.) in a coordinated manner. Typically,
distributed automation and control approaches are used.
A storage system together with a distributed generator
installed at the customer side can be considered as a
subsystem, as depicted in Fig. 2.
• System level: Power utility automation, demand-side man-
agement (DSM), or energy management is carried out
from a systems perspective in a coordinated way using
the provided functions and services of the underlying
subsystems and components. Central and distributed con-
Fig. 2. Necessary intelligence in a smart grid system on different
levels. trol approaches and corresponding algorithms are usually
applied, as represented in Fig. 2.
• Islanding possibilities/microgrids: Local operation of
Flexibility, adaptability, scalability, local intelligence/
islands/microgrids can improve the availability of the elec-
autonomy, and open interfaces are key requirements for corre-
tricity supply due to failures on higher voltage levels.
sponding ICT/automation systems and component controllers
• Distributed generation/distributed energy resources with
(DER, storage systems, OLTC, etc.) to support and to enable
ancillary services: Usage of ancillary services provided by
the above-described functions on the different levels [23]. The
DER (e.g., local voltage or frequency control and virtual
AI principles and techniques can be used for this purpose with
inertia) improves power grid optimization (e.g., enhancing
advantages on different levels.
the hosting capacity of DER in power distribution grids).
In the following sections, a review of major achievements,
• Demand response/energy management support: Electric
technology developments, and research results in the domain
loads and ESS provide additional flexibility in power
of DER (i.e., controlling distributed power generation), ESS
system operation. Energy management and DR principles
(i.e., managing storage capabilities), energy/demand-side man-
and smart functions have to be provided by the used ICT
agement (i.e., managing electricity consumption), and ICT/
infrastructure.
automation concepts is provided.
• Advanced forecasting support: Forecasting of (distributed)
generation and load profiles for optimized grid operation.
• Self-healing: Automatic or semiautomatic restoration of III. I NTEGRATION OF R ENEWABLE E NERGY R ESOURCES
grid operation in case of component/grid faults helps
power system and infrastructure operators. The integration of DER with the electrical power system re-
• Asset management/condition-dependent power system quires the usage of PCs, necessary for adapting fluctuating pro-
maintenance: Preventive maintenance according to duced energy with grid requirements; they are also necessary
component/device conditions and remaining lifetime. for efficient operation of ESS. The most appropriate topologies,
devices, control, and modulation techniques are fundamental in
order to fulfill grid codes/interconnection rules and for avoid-
C. Distributed Intelligence on Different Levels
ing the occurrence of instability or synchronization problems,
In order to realize the vision of smart grids and the corre- faults, and for obtaining high-efficiency operations. However,
sponding functions, distributed intelligence on different levels the nonlinear behavior of PC causes the injection of harmonic
in the power system infrastructure is necessary. The following distortions at the point of common coupling (PCC). More-
list and Fig. 2 provide a brief categorization of these levels and over, stable outputs could not be ensured in every condition,
indicate what kind of intelligence is needed. such as during deep voltage sags. The subsequent generated
• Subcomponent level: Typically, power electronics are used electromagnetic interferences may lead to undesired situations,
in PCs of DERs. Improvement of local component prop- loss of synchronization of DERs, grid instability, and faults.
erties such as harmonics and flicker due to advanced These situations can be mitigated by insertion of special con-
control algorithms and topologies is the main driver for verters, namely, power factor controllers, static compensators,
local intelligence on this level. The DER controller and and ESS, as well. In addition, the implementation of ICT tech-
hardware part in the system configuration shown in Fig. 2 nologies and concepts allows prevention of abnormal situations
can be considered as subcomponents. [24]–[26].
STRASSER et al.: REVIEW OF ARCHITECTURES AND CONCEPTS FOR INTELLIGENCE IN SYSTEMS 2427

To guarantee power quality at the input ac mains, harmonic Examples are local voltage control possibilities at the PCC,
standards and engineering recommendations must be also which can be also used in a coordinated way to maintain power
adopted to limit the level of distortion at the PCC [27], [28]. quality due to a high share of distributed generation.
PC topology plays a significant role in full integration of It is expected that, in the future, electromagnetic transformers
renewable energies with the electrical power grid. Nowadays, will be replaced by electronic counterparts, the so-called “solid-
there is a significant trend of considering multilevel converter state transformers (SSTs),” based on previous technologies [40]
(MLC) topologies for new smart-grid-related projects. They ba- (an application example for SST is provided in Section VI-A).
sically consist of arrays of power semiconductors in series and/ In fact, they can be easily integrated within the smart grid,
or in parallel, producing output voltage with discrete stair- interconnecting distinct electrical power grids operating with
stepped waveforms. Typical topologies of MLC are diode different characteristics (e.g., voltage and phase). They can
clamped or neutral point clamped (NPC), active NPC, capacitor simplify grid management during normal operating conditions
clamped or flying capacitor, and cascaded MLCs [29]–[31]. A and during faults. Moreover, they could simplify power factor
significant part of their behavior (namely, the quality of out- control and harmonic reduction as well. In general, they are
put waveforms) depends on control and modulation methods. more compact in size and are expected to become cheaper than
Popular methods are multilevel sinusoidal pulsewidth modula- traditional transformers, penalized by the high cost of copper
tion (PWM), phase disposition, multilevel selective harmonic and of magnetic materials [41]–[43].
elimination, harmonic mitigation, and space vector modulation Since power levels become higher, optimum modulation of
[32]–[38]. The higher the number of levels, the higher the PC is fundamental for satisfying power quality demand and for
quality of the output voltages and currents due to reduced improving design, performances, and efficiency. Elimination
electromagnetic emissions determined by chosen modulation or mitigation of one or more harmonics from output without
technique is. Due to their intrinsic characteristics, MLCs op- the use of passive filters is very important for achieving sig-
erate at low frequency and at medium voltage; moreover, they nificant improvement in terms of cost, reliability, and space.
eliminate common-mode currents. Additionally, they do not Genetic algorithms and particle swarm optimization [44]–[46],
require expensive passive filters, while achieving significant AI methods based on bacterial foraging algorithms, ant colony
improvements in terms of efficiency and cost. For such reasons, methods, and homotopy and continuation theory [47]–[49] are
their usage simplifies DER design and allows direct connection typical methods capable of finding the converter switching
of PC with power distribution lines. angles necessary for modulation. However, all these meth-
Another emerging topology in DER is the so-called modular ods require offline calculations, hence offer limited flexibility.
MLC (MMC) [39]. This can be considered a new and promising Recently, some analytical methods capable of identifying all
design approach consisting in connecting basic PC modules in possible switching angles eliminating a desired harmonic, for
series and/or in parallel. Typically, each module consists of a a fixed modulation index, in a five-level cascaded inverter have
half-bridge or a full H-bridge with a dc capacitor for energy been successfully developed, allowing real-time implementa-
storage and all the necessary sensors and control circuits. With tion without the use of lookup tables [38]. The transcendental
this approach design, realization and maintenance of high- equations characterizing the harmonic content can be converted
power and/or medium-voltage converters are simpler because in polynomial equations, which can be then solved using the
hardware can fully exploit modularity. Usually, the design of method of resultants from elimination theory [50], [51]. It has
the single modules is often very complex due to the numerous been shown that a pair of harmonics chosen among third, fifth,
constraints. Moreover, software and control become more com- and seventh can be eliminated in a five-level inverter output
plex, too, due to the additional intelligence and communication voltage using the method described in [52], based on graphical
capability needed by interactions among the modules. One analysis by using Chebyshev polynomials and the Waring’s
of the early applications of MMC was in high-voltage direct formulas. In order to generate switching PWM patterns with
current systems, but due to their significant advantages in terms low harmonic content, avoiding the elimination of some spe-
of flexibility, scalability, maintenance, and costs, their usage has cific harmonics, the total harmonic distortion can be treated
been extended to DER, e.g., megawatt-range wind turbines and as a global problem by using a general-purpose random-search
to other high-power applications as well. Of course, there are heuristic algorithm [53]–[55].
also some drawbacks, mainly caused by the intrinsic nature In summary, AI techniques and concepts play an important
of modular systems, namely, the use of connectors with po- role for controlling and optimizing elements on the subcompo-
tential interconnection problems, the potential instabilities due nent and component levels. It is expected by the authors that, in
to interaction among modules, the increased distances among the future, more intelligent algorithms and concepts are being
critical components, and the additional design complexity, due used on this level.
to the modularity.
Comparison among MLC and MMC is not simple; moreover, IV. ESS I NTEGRATION
MLC can be designed with a modular approach. However, in
the authors’ opinion, it should be expected a growth of both The increasing share of intermittent and distributed gener-
technologies and the diffusion of modular design approaches, ation units into the power system is the main driver for the
both in hardware and in software as well. integration of stationary storage systems into the power system.
In modern power systems, PCs are also providing ancillary Future demand for storage will include both long-term large-
services (e.g., local voltage or frequency control) using re- scale storage such as hydrogen or natural gas and small-scale
active and active power management functions and therefore short-term storage systems such as batteries or flywheels within
providing additional possibilities for the power grid operation. the distribution system [56], [57].
2428 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 62, NO. 4, APRIL 2015

A control architecture based on predefined droop character-

istics is discussed in [63]. It is shown that the simple adoption
of the droop characteristics for single systems decreases the
investment costs for the ESS by reducing the required capacity
significantly. One approach to achieve this functionality is
the implementation of a droop control function according to
IEC 61850 [17] or SunSpec [67]. Both the standard frameworks
not only provide a definition of standardized input and output
registers of inverter-based distributed generators and storage
systems but also define standardized functions and services.
Each function defined in IEC 61850 or SunSpec can be con-
figured based on a predefined set of parameters.
Marra et al. [68] discussed a threshold-based control al-
gorithm for increasing local consumption and simultaneously
providing voltage control. The defined threshold is based on the
nominal power of each PV array. When the PV penetration in a
feeder changes, the threshold for each storage system changes
as well. Once the threshold is set for each storage system, they
operate autonomously. While the strategy provides consistent
voltage control, the impact on local consumption is not dis-
cussed. The authors of the work showed that, with an increas-
Fig. 3. Exemplary simulation of an ESS providing voltage control for a ing PV penetration, the requirements for the storage system
local PV generator [65]. increase significantly. Only simple algorithms from the AI point
of view are being applied in the current solutions, but there is
The number of possible applications for grid-connected sta- space for more intelligent concepts and algorithms in future
tionary storage is large. Depending on the methodology and the developments.
granularity of the segmentation between 9 and 16, distinguished
services are reported in the literature [57]–[60].
Requirements to controls and ICT strongly depend on the B. Central Control Concepts
necessary service or combination of services and the storage
technology in use. A common picture showing the state of the In contrast to autonomous control, central approaches need
art for all services and control architectures is hence not possi- a functioning control unit and a communication network. At-
ble. This review focuses on the implementation and operation of tached components do not make decisions by their own. An
ESS providing services for power quality. The most important algorithm where storages provide voltage control with OLTC
service discussed is voltage control to maintain power quality simultaneously on the low- and medium-voltage levels is shown
and to allow for a higher hosting capacity of renewable sources in [62]. Single actions are rated based on their costs and
[61]–[64]. impact with offline load flow simulations. A central controller
Fig. 3 shows a simulation study of an ESS providing voltage optimizes component behavior based on measurements.
control for a local PV generator [65]. It absorbs all active The authors in [69] discussed a central control algorithm
power above a defined threshold (e.g., 1.02 per unit) to avoid for voltage and frequency control for microgrids based on
intolerable voltage levels. When they fall below the defined droop control. The battery storage system is the primary control
threshold, the ESS starts the discharging. unit for maintaining both voltage and frequency. It is only
The storage system is not the only component participating active when the microgrid changes to islanding mode. When
in voltage control schemes. Often, such functions are provided in islanding mode, the battery ESS (BESS) is coordinated with
by DER-based inverter systems and/or OLTC [62], [63]. In a local DERs over a central control unit.
number of cases, the analyzed storage systems also provide Sugihara et al. [70] proposed a central control approach
additional services such as local consumption of DER [66]. where the distribution system operator (DSO) has direct control
For the case of voltage control, different architectures, rang- access to distributed storage systems at customer sides. Access
ing from autonomous control with and without remote configu- is granted in exchange of a subsidy for the storage system.
ration ability to central control, are analyzed. The described approach takes advantage of the reactive power
compensation abilities of the storage converter. Voltage control
facilitated by active power control is not included in the pro-
A. Autonomous Control With Central posed approach. Like other control approaches, the decision
Configuration Ability foundation is based on characterization of the network via
This approach describes autonomous control of single com- load flow calculations. The authors showed that the described
ponents based on predefined function characteristics with the approach allows the DSO to maintain acceptable voltage level
opportunity to reconfigure functions from a central unit. The on a cheaper base than with other measures.
control algorithm can also be based on rules in combination The paradigm of the distributed versus centralized control
with threshold levels. The configuration of the single compo- approaches is in the focus of attention, but the trend toward
nents is often based on offline or online load flow calculations distribution is clearly identifiable. This is the only way out in
of the local power grid [62]. such complex systems.
STRASSER et al.: REVIEW OF ARCHITECTURES AND CONCEPTS FOR INTELLIGENCE IN SYSTEMS 2429

V. DR AND E NERGY M ANAGEMENT infrastructure. They state that models at multiple levels of
Future electric energy systems endowed with smart metering, analysis are essential in order to simulate a DR program using
intelligent electronic devices, and advanced ICT will be char- both a distributed and centralized control. Considering that DR
acterized by greater responsiveness and efficiency decisions by techniques are moving toward approaches based on a more and
the customers and the utility provider [5], [8], [9]. more interactive model of communication, active, interactive,
Recent research studies have confirmed, indeed, that DR and transactive electricity markets are proposed.
is essential for the operation of future smart grids [7], [71], In an active market, a price signal is sent from the central
[72]. DR denotes variations of the electric consumption by controller to the end-use customer so that he can respond to
customers when the price of electrical energy changes over time price variations. In an interactive market, the end-use customer
or in presence of financial incentives or reliability signals [71]. can also send information back to a central controller so that
Regulatory and policy frameworks, such as the Energy Policy the price signal can be changed. In a transactive market, inter-
Act of 2005 in the U.S., have been recently enacted, which active controllers are used, which are also able to automatically
promote DR and allow customers and load aggregators taking operate on behalf of the end-use electrical load.
part by means of DR resources in energy, capacity, and ancillary
services markets [73]–[75]. In addition, the FERC Order 719 B. DSEMS
contributed to remove obstacles to the participation of DR in In [81] and [82], the authors proposed a decision support
wholesale markets by allowing load aggregators bidding DR on and energy management system (DSEMS) that can be used for
behalf of retail customers into markets. In 2011, FERC Order residential, commercial, and industrial customers. This concept
745 determined that DR resources should be compensated at is modeled as a finite state machine and consists of some
the locational marginal price for their participation in wholesale scenarios that can be selected according to the customer’s
markets, thus establishing an equal treatment between demand- preferences. The inputs of the systems are the measured energy
side resources and generation [76], [77]. from available resources (i.e., from the distribution network
In Europe, even if it is recognized that DR could play a and local or other electrical or thermal energy sources) and
relevant role in favoring penetration of DER [78], policies are information related to hourly electrical and thermal energy tar-
more devoted to energy efficiency and DSM, rather than DR, iffs, status of the network in terms of components availability,
which is mainly considered in order to decrease the costs of users requirements, contract constraints, messages from the
load peaks. DSO, and environmental parameters. The system generates the
command signals for the management of thermal and electrical
A. DR Programs loads and the messages for the end user (i.e., information about
controlled devices, load state information, energy consumption
Nowadays, mainly due to both new regulatory and policy and energy saving suggestions for achieving energy saving, and
frameworks and new technologies, a great share of the end- messages from the DSO).
use electrical loads can be engaged in DR programs in a more The DSEMS is also endowed with a climate control, both
controlled manner. Automated control systems operating on a used during summer and winter seasons. The temperature set
continual basis are replacing traditional load control systems, point of the split is progressively modified in order to decrease
and these new systems are now available for commercial, in- the power absorption when the maximum power limit is ex-
dustrial, and individual residential customers. Often, residential ceeded or during periods when the cost of energy is high [81].
DR should be aggregated in order to compete in the wholesale In [82], the effectiveness of the DSEMS for residential
energy markets. applications is evaluated by varying the energy class of the
While the direct control of the end-use electrical loads ex- house from A to G and considering different scenarios (e.g.,
isted for decades, price-driven response programs, using a price comfort, economy, and energy) for two locations character-
signal as a means for demand control, are beginning to emerge ized by different external climatic conditions. The simulated
and to be the subject of study at the distribution level. Com- and experimental tests evidence that the use of DSEMS al-
bining the automation of the demand bids and the strategies to lows decreasing both energy consumptions and costs by not
respond to the growing user empowerment with respect to its deteriorating the user comfort. Higher benefits can be obtained
domestic consumption, DR programs are the ideal solution to for houses with high energy consumption and characterized by
reduce demand power and prevent grid congestions. a lower initial energy class with a maximum reduction of about
Diverse time-differentiated pricing models have been pro- 30% of the initial energy consumption achieved for class G
posed, such as real-time, day-ahead, time-of-use, or critical- houses [83]. This system is also useful for DSOs as it can help
peak pricing [71]. The basic ideas are the same for all these mitigating peak loads due to air conditioners usage.
models; first, it is required that retail prices reflect fluctuations
in wholesale prices so that users pay for what they consume at
C. DRP Implementation
the market price and that users themselves are free to decide
whether or when to buy the energy they need. In addition, Customers can take part to DR programs through an inter-
these models encourage users to shift high loads during off- mediary or directly with the utility. In the case when end-use
peak hours in order to reduce the cost of electricity and to help customers are gathered by intermediaries, generally called DR
reduce the peak-to-average ratio of demand [79]. providers (DRPs), the end-use customers’ aggregated capability
The authors in [80] aimed at evaluating the potential im- is presented to an organized market by DRPs [71]. Automated
pact that the implementation of DR methodologies and the response technologies, generally classified as control devices,
dynamic pricing could have on the operation of the electricity monitoring systems, and communication systems, are required
2430 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 62, NO. 4, APRIL 2015

this explicitly. The trend of the deeper exploration of AI will

continue.

VI. S MART G RID AUTOMATION AND C ONTROL

This section provides a survey of the applications of modern
AI and software technologies to perform control functions in
the field of smart grids. The requirements for such systems
impose the adoption of technologies based on principles of
decentralization of functions (e.g., monitoring and control) over
distributed, interoperable, loosely coupled, and reusable entities
used, for example, in DER, ESS, and DR, as covered above.
The AI principles and techniques do start to play the key roles
in the architectures, including the service-oriented architecture
(SoA), multiagent and holonic systems, and in standardization.

A. Architectures
Automation architectures of future smart grids are being
developed in a number of large-scale research programs world-
wide. One significant research initiative is conducted by the
center of Future Renewable Electric Energy Delivery and Man-
agement (FREEDM) funded by the National Science Founda-
tion, USA. The FREEDM center envisions the backbone of
the future grid to be a new power distribution infrastructure,
allowing for integration of smart grid components in a “plug-
and-play” manner and enabling bidirectional flow of electricity
[86]. This new energy delivery infrastructure is also denoted by
“Energy Internet” or “Internet of Energy” [86]. This vision is
akin to the (information) Internet.
Fig. 5 illustrates the FREEDM vision. The core enabling
Fig. 4. Energy management system in the smart grid infrastructure:
technology is the SST [87], which is a new generation power
(a) Hardware system setup. (b) Logical/automation architecture [71]. transformer using power electronics enabling bidirectional
electricity flow and supports ac and dc buses for direct con-
nection of components (e.g., DER and ESS). It is controlled by
in order to actuate the functionalities of a DRP and allow remote
distributed grid intelligence (DGI), which consists of two types
managing of peak load demand and energy consumption. An
of nodes: intelligent energy management (IEM) and intelligent
infrastructure to implement a DRP in a smart grid is presented
fault management (IFM).
in [71]. As shown in Fig. 4, it consists of energy management
The first key feature of FREEDM is a plug-and-play in-
systems (EMSs) and DRP, supervisory control and data acquisi-
terface, which enables components to seamlessly integrate
tion (SCADA), remote terminal units (RTUs), advanced meter-
with the rest of the grid. This interface includes an open and
ing infrastructure (AMI), state estimation algorithms (SEAs),
standard-based protocol, enabling interoperability and self-
and generation and load forecast system (GLFS). Measurement
awareness of the devices. The second feature is an energy router
data provided by AMI and RTUs are transferred to the EMS by
or an IEM device. The IEM device is based on the SST concept.
the SCADA system. The EMS performs monitoring, control,
It manages the energy flow, provides power management, and
and optimization tasks by using SEAs and GLFS.
has power-balancing capabilities. Furthermore, it manages con-
An AMI network, consisting of a smart meter, wide-area net-
nected DER, ESS, and flexible loads. The third feature is the
works, home area networks, meter data management systems,
DGI itself, formulated and implemented using open standards.
and neighborhood area networks, is used to measure, save, and
To address the fundamental challenges in controlling large
evaluate energy usage and to communicate information among
numbers of distributed devices and dynamic loads, new
the utility, the loads, and the consumers.
control theories and distributed agent-based algorithms are
In summary, the tasks of monitoring, diagnostics, and predic-
tive maintenance should play an important role in energy man- required and currently under development. The FREEDM
agement systems. The model-based solutions were recognized center envisions that the major challenge will be implementing
as very successful for this purpose, but it is extremely difficult DGI applications in a distributed manner across multiple
and expensive to obtain relevant models built from the first execution platforms (i.e., IEM and IFM). Experience so far
principles. Relevant approaches, including the model-based and has emphasized the need of open source open standard-based
fault-tolerant system approaches are reviewed in recent papers software and communication platforms. Toward this goal,
[84], [85]. In the DR energy management of today, the AI the center has proposed standard-based execution framework
principles and techniques (negotiations based on price models, for IFM and IEM applications [88]–[93]. This framework is
bidding and contracting) are strongly explored, despite that based on the IEC 61850 and IEC 61499 standards. The results
the authors from the energy management field do not mention have demonstrated successful deployment of agent-based
STRASSER et al.: REVIEW OF ARCHITECTURES AND CONCEPTS FOR INTELLIGENCE IN SYSTEMS 2431

Fig. 6. MAS applied to active power distribution grids [96].

Using MAS principles, each component, device, or actor is

monitored and controlled by an autonomous agent. Each agent
has a set of specific capabilities (such as controlling household
appliances) and local objectives (for example, minimization of
energy cost while preserving the end-user comfort). The agents
interact via messages to coordinate their behaviors in order
to achieve an equilibrium between the local and system-wide
goals (such as demand–supply balancing).
MAS technology has been employed for developing the
distributed ICT infrastructure within various projects such as
the FREEDM initiative and the CRISP project. A concept
of grid cells was developed to divide the network into inde-
pendently managed subnetworks. The agent assigned to each
grid cell is responsible for managing the current and future
production/consumption and handling the reconfiguration tasks
in case of fault detection [97]. A very similar approach con-
sidering microgrids as small-scale electricity grids operating in
low-voltage networks is presented in [98]. Each house in the
microgrid is equipped with the intelligent load controller (ILC),
which can communicate with house devices over the power line
in order to intelligently control their load. The agent embedded
in the ILC communicates with the microgrid central controller
agent, whose task is to balance production and consumption
Fig. 5. FREEDM architecture: (a) Energy Internet. (b) Key within the microgrid.
elements [86]. The MAS approach has been also exploited in designing the
power-balancing DGI application on a network of ARM concept of virtual power plants (VPPs). A VPP is comparable
controllers [94]. with a microgrid—it is an aggregation of generation units,
such as PV systems, combined heat and power units, or wind
turbines together with controllable loads. They are represented
B. Agent-Based Solutions by agents grouped into an MAS, which is represented by a
To master the high complexity of future electric energy sys- management agent as a single entity on the energy market [99].
tems, where the intelligence is distributed over autonomously To allow the smaller entities to trade the electricity locally
acting and interacting components, it is necessary to explore in order to achieve the demand–supply balancing, a concept of
architectures and programming techniques based on distributed electronic power markets was introduced. The PowerMatcher
AI. One of the most relevant technologies seems to be multia- tool is an MAS in which electricity generation and/or con-
gent systems (MAS). They have already proved its capabilities sumers are represented as agents trying to sell and buy energy.
for developing robust, flexible, and adaptive industrial automa- The process of price negotiation is managed by an auctioneer
tion systems, with application domains, including manufactur- agent that finds the equilibrium price considering the bids
ing control, dynamic product routing, production planning and collected from device agents [100]. The PowerMatcher tool has
scheduling, logistics, aerospace, air traffic control, and many been used also in other projects such as INTEGRAL, which is
others [95]. aimed at the distributed control and coordination of aggregated
In the effort to employ active elements in power grids with DERs [101], SmartHouse/Smart-Grid focused on integration of
the local intelligence and coordination capabilities, it is appar- smart houses into the smart grid [102], and EcoGrid targets the
ent that the MAS approach receives a significant popularity also large-scale demonstration of the smart grid solution with 2000
for developing smart grids [23], as shown in Fig. 6. residential consumers on the Danish island Bornholm [103].
2432 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 62, NO. 4, APRIL 2015

An agent-based architecture for smart grids is also presented interaction between house and smart meters and implementing
in [104]. The architecture is based on industrial standards DR/DSM functionalities [115]–[117].
IEC 61850 [105] and IEC 61499 [106] to facilitate the mi- According to Pagani and Aiello [113], SoA can help to
gration of agent technology into the industrial practice. The resolve such smart grid challenges as interoperability, scala-
agents are so-called intelligent logical nodes (iLNs), capturing bility, discovery, mobility, robustness, service integration and
data from IEC 61850 LN. The idea is that the power system composition, topology, advanced metering infrastructure, and
automation can be decomposed into LNs, capturing functional real-time constraints. However, one of the important issues with
requirements of the desired system. Then, LNs can be enhanced SoA is how to test and verify services and systems built on it.
with the desired algorithm. Each iLN is implemented with This also raises issues of trusted service providers and security
an IEC 61499 function block. The iLN architecture enables in SoA. These issues contribute into the bigger problem of
automatic generation of automation software and plant models smart grid transparency, security, and privacy.
given customer requirements captured with IEC 61850 means.
SysGrid is a tool supporting such automatic transformation, D. Holonic Control Principles
which is proposed in [107]. A number of projects were imple-
mented using iLN architecture [88], [104], [108]. Holonic principles [118] known from the AI field and gen-
The fault prevention and self-healing techniques for smart eral systems theory are being applied to the smart grids do-
grids based on MAS are also widely discussed. One technique main, trying to capitalize the intrinsic characteristics to model
presented in [109] is based on calculation of optimal network complex large distributed systems into potential benefits. An
configuration achieved by opening or closing the switching el- architectural model based on holonic MAS is proposed by [119]
ements to isolate the fault. Another similar technique considers to structure the software entities within the ICT infrastructure
partitioning the grid into multiple islands that disconnect from of smart grids. The whole network is assumed as a holon
the main grid in case of a failure. The agents responsible for that consists of some domain holons, namely, focusing on
those islands cooperatively prepare plans of action to isolate generation, transmission, distribution, control and operation,
the faulty parts from the power grid [109]. service provider, and market and customer domains.
Negeri et al. [120] applied the holonic approach to structure
the smart grid organized on autonomous prosumers that are re-
C. SoA Concepts cursively clustered at various aggregation layers. These holons,
named control holons, are organized in a bottom-up structure to
SoA is a principle based on discrete parts of software that form a complete control holarchy of the smart grid. Many func-
provides functionality as a service to other applications. SoA tions are included in this holon structure, namely, the environ-
is therefore designed to be independent of any vendor speci- ment state acquaintance, state analysis, database management,
fications, products, or underlying technology. Loose coupling forecasting, steering subholons, and scheduler. This holonic
of software parts provided by SoA provides flexibility and in- approach is extended by considering a SoA framework to sup-
teroperability to a system and enables dynamic reconfiguration port interoperability and reusability challenges [121], defining
possibilities. five major services: database service, state evaluation service,
As mentioned before, smart grids can be thought as the optimization service, transaction service, and stability service.
Energy Internet. From this point of view, SoA can become Frey et al. [122] proposed a holonic control architecture to
a core technology to enable easy interaction between het- address the current requirements of smart grids, namely, the
erogeneous devices and system integration. SoA can be also need to control a wide variety of heterogeneous producers and
used for integration with the legacy systems. Moreover, it can consumers to attain global objectives at the macro level, while
enable flexible integration of various actors such as distributed also taking into account local objectives and private interests
generators at industrial and residential levels, market players at the micro level. For this purpose, they identified three main
(auctioneers, buyers, sellers, regulators, etc.), consumers, gov- integration patterns for conflict resolution in self-managed sys-
ernment bodies, and power grid operators [110]. The ability of tems, namely, hierarchic, collaborative, and stigmergic. These
service discovery and its advertisement as the key intelligent patterns are complementary and could be successively used
features would facilitate dynamic nature of smart grids. according to the environmental conditions and the working
At the level of heterogeneous system components (DER, mode of the power grid.
EES, etc.), SoA can provide interoperability. Each node pro- A holonic MAS architecture is also introduced by [123] to be
vides a set of services with a defined interface. capable of adaptively controlling power distribution systems,
SoA requires protocols that describe how services are being e.g., allowing the operation of a distribution system as an island
sent, information are processed, services are described, and how mode in case of emergency, such as hurricanes/earthquakes or
to use them. To unify information exchange and representations, power grid failures.
many researchers propose to use the IEC 61850 and Common
Information Model (CIM) models [111]–[113]. Lehnhoff et al.
[112] adopted OPC Unified Architecture (OPC UA) for smart E. Important ICT and Automation Standards
grid applications, implementing SoA with OPC UA. Due to the increasing number of available ICT/automation
Systems built on SoA require enabling infrastructure. In solutions for smart grids and constantly changing requirements,
this context, Takagiwa et al. [114] proposed a service-oriented standardization is of crucial importance in order to handle inter-
network architecture for smart grids, including service-oriented operability and scalability requirements. These aspects have so
routers. In addition, the SoA principle is also seen to play an far been covered by several standardization organizations and
important role in Smart Home/Building applications, enabling international projects. Specht et al. [15] and Gungor et al. [14]
STRASSER et al.: REVIEW OF ARCHITECTURES AND CONCEPTS FOR INTELLIGENCE IN SYSTEMS 2433

provided a comprehensive overview of important smart grid IEC 61499 Compliance Profile (CP) for smart grid applications.
standards and roadmaps. The developed IEC 61499 CP provides a mapping between
As an international standardization organization, the Inter- the IEC 61850 structure with IEC 61499 artefacts, and it also
national Electrotechnical Commission (IEC) plays a very im- describes common rules for the usage of IEC 61499 together
portant role by providing common rules for the planning and with IEC 61850.
operation of smart grid systems. The “IEC Smart Grid Stan- Under the scope of the 4DIAC initiative [133], [134],
dardization Roadmap” suggests different core standards im- distributed reconfigurable control software for smart grids
portant for the implementation of smart grids, such as the and components was implemented based on IEC 61850 and
following: 1) IEC TR 62357 (service-oriented integration ar- IEC 61499. In particular, low-level control algorithms and
chitecture); 2) IEC 61970/IEC 61968 (CIM); and 3) IEC 61850 reconfigurable interfaces for the online update/adaptation of
(power utility automation) [17]. Other organizations also pro- control functions in DERs are covered by [135].
vide similar roadmaps for the implementation of intelligent Zhabelova and Vyatkin [104] introduced an automation
power grids. The “NIST Framework and Roadmap for Smart architecture that supports distributed multiagent intelligence,
Grid Interoperability Standards” [16], the “DKE German stan- combining IEC 61850 object-based modeling and interopera-
dardization roadmap for Smart Grids” [19], and the “IEEE ble communication with IEC 61499 function block executable
Guide for Smart Grid Interoperability of Energy Technology specification. The proposed architecture was applied to achieve
and Information Technology Operation with the Electric Power self-healing grid through collaborative fault location and power
System (EPS), End-Use Applications, and Loads” [18] are restoration.
important examples of such roadmaps and guidelines. Zhu et al. [136] proposed the function-block-based model for
The German DKE Smart Grids roadmap also goes one step flexible intelligent electronic devices, combining the IEC 61850
further by suggesting standards for the implementation of con- and IEC 61499 models. A prototype system was developed in
trol logic. For this, the usage of IEC 61131-3 for programmable MATLAB/Simulink.
logic controllers [124] or the IEC 61499 for distributed in-
telligent automation [106] is suggested. Both IEC 61131 and
IEC 61499 provide promising concepts for the development VII. C ONCLUSION AND O UTLOOK
of control and automation solutions in smart grids, which has The future electric energy system consists of a huge amount
already been demonstrated in previous works [19]. Recently, of interconnected components and supports bidirectional elec-
the OPC UA, which is defined in IEC 62541, has received tricity flow through the electrical network and an ICT infras-
an increased interest from the smart grid research community tructure. Important actors in such a smart grid system are
[125]. The previous OPC specification for data access (i.e., inverter-based DER, ESS, and flexible/controllable loads. Com-
OPC DA) was particularly developed for Windows-based plat- pared with today’s power system, the future infrastructure is
forms using mainly COM/DCOM technology, whereas the new characterized due to a higher amount of distributed components
OPC UA specification was particularly developed to provide a (in hardware and software), whereas the hierarchical structure
platform-independent information and communication model, of the power grids will still exist.
by using service-oriented principles for the data and informa- In this paper, the importance of power electronics and ad-
tion exchange. Due to its object-oriented and generic modeling vanced ICT/automation approaches has been highlighted as a
approach, there have been several studies conducted, where basis to design and operate the future power system, which
existing data models and protocols have been successfully has to deal with fluctuating distributed generation and limited
mapped to OPC UA, e.g., CIM and IEC 61850 [126], [127]. storage capacities provided by ESS. Power converters with
The IEC 61499 standard [128] was particularly developed ancillary services provide additional possibilities to optimize
as a reference architecture for the modeling and design of dis- the future power grids. Moreover, with DSM/DR, the electricity
tributed control algorithms. The distributed nature of the smart consumption (i.e., load profile) can be managed, too.
grid approach makes IEC 61499 particularly also useful for this The smart grid ideas have appeared to enhance the capa-
domain. This was demonstrated by Hegny et al. [129], where bilities of the power distribution systems as reaction to grow-
an IEC 61499 application was developed for energy systems ing requirements on better exploration of (distributed) energy
focusing on the reduction in the downtime of reconfiguration resources. The smart grid visions are supported by the latest
tasks in energy management systems. achievements in the field of AI and software engineering,
The integration of the IEC 61850 and IEC 61499 stan- namely, on holonic, MAS, and SOA principles. The trend is to
dards is explored in the literature as well. As an example, use holonic control principles on the lowest, near to the physi-
Higgins et al. [130] introduced an approach to power sys- cal equipment automation, level to ensure very fast reactions,
tems automation based on distributed intelligence combining particularly reconfigurations of the network. Both the MAS
the use of IEC-61850-compliant devices with the IEC-61499- and SOA approaches are used to support services and decision
compliant “glue logic.” According to [131], the IEC 61499 ref- making on a higher automation level and requires an additional
erence model providing a distributed function-block-oriented communication infrastructure. From the philosophical point of
architecture is a proper technology to implement the advanced view, the latter two approaches converge and are in some sense
functions of holons, particularly the bioinspired negotiation complementary [96]. As a rule, the MAS or the SOA architec-
behaviors based on the stigmergic approach. ture is being considered to ensure higher level optimization
Strasser et al. [132] discussed the usage of the IEC 61850 based on negotiation and cooperation principles in a highly dis-
standard for power utility automation together with IEC 61499 tributed communication environment. The trials to explore the
for distributed automation and the development of a related semantic web or big data technologies to support the distributed
2434 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 62, NO. 4, APRIL 2015

decision making in the MAS or the SOA architecture represent [5] L. Nian, C. Jinshan, Z. Lin, Z. Jianhua, and H. Yanling, “A key
an up-to-date trend [96]. management scheme for secure communications of advanced metering
infrastructure in smart grid,” IEEE Trans. Ind. Electron., vol. 60, no. 10,
In contrast to the “classical,” hierarchically, and ‘rigidly’ pp. 4746–4756, Oct. 2013.
organized quite expensive power distribution infrastructure, the [6] D. De Silva, X. Yu, D. Alahakoon, and G. Holmes, “A data mining
ideas of smart grids complement the power distribution ele- framework for electricity consumption analysis from meter data,” IEEE
ments by highly distributed autonomous decision making units Trans. Ind. Informat., vol. 7, no. 3, pp. 399–407, Aug. 2011.
[7] P. Palensky and D. Dietrich, “Demand side management: Demand re-
with highly decentralized way of communication, negotiation, sponse, intelligent energy systems, and smart loads,” IEEE Trans. Ind.
and cooperation on different levels in the system as pointed Informat., vol. 7, no. 3, pp. 381–388, Aug. 2011.
out in this paper. This strong contrast between hierarchically [8] X. Yu, C. Cecati, T. Dillon, and M. Simoes, “The new frontier of smart
grids,” IEEE Ind. Electron. Mag., vol. 5, no. 3, pp. 49–63, Sep. 2011.
organized physical infrastructure of power distribution systems [9] C. Cecati, G. Hancke, P. Palensky, P. Siano, and X. Yu, “Guest Editorial
and decentralized communication and decision making control Special Section on Information Technologies in Smart Grids,” IEEE
infrastructure does represent the key challenge for bringing Trans. Ind. Informat., vol. 9, no. 3, pp. 1380–1383, Aug. 2013.
smart grids into reality. [10] Technology Roadmap Smart Grids, IEA, Paris, France, Tech. Rep., 2011.
[Online]. Available: http://www.iea.org
The first step is to achieve standardization everywhere to [11] H. Farhangi, “The path of the smart grid,” IEEE Power Energy Mag.,
guarantee interoperability and scalability. The standards in the vol. 8, no. 1, pp. 18–28, Jan./Feb. 2010.
area of automation of power distribution systems are well [12] V. C. Gungor et al., “A survey on smart grid potential applications and
communication requirements,” IEEE Trans. Ind. Informat., vol. 9, no. 1,
developed and available (i.e., IEC 61850). The same is true pp. 28–42, Feb. 2013.
for holonic systems (i.e., IEC 61499). On the other hand, the [13] M. Sechilariu, W. Baochao, and F. Locment, “Building integrated
standardization in the field of MAS and SOA is in the early photovoltaic system with energy storage and smart grid communi-
stage and needs substantial enhancements. cation,” IEEE Trans. Ind. Electron., vol. 60, no. 4, pp. 1607–1618,
Apr. 2013.
In summary, intelligent operation and optimization are re- [14] V. Gungor et al., “Smart grid technologies: Communication technologies
quired on different levels in the whole smart grid system. On and standards,” IEEE Trans. Ind. Informat., vol. 7, no. 4, pp. 529–539,
the subcomponent and component levels, mainly research and Nov. 2011.
development related to new PC topologies, advanced control [15] M. Specht, S. Rohjans, J. Trefke, M. Uslar, and J. M. Gonzalez, “Inter-
national smart grid roadmaps and their assessment,” ICST EAI Endorsed
algorithms, and the definition of ancillary services are the Trans. Energy Web, vol. 13, no. 1, pp. 1–18, 2013.
main drivers. The subsystem level has to integrate the provided [16] NIST Framework and Roadmap for Smart Grid Interoperability Stan-
component functions in a smart way. Central and distributed dards Release 2.0, NIST, US Dept. Commerce, Gaithersburg, MD, USA,
Tech. Rep. NIST Publication 1108R2, 2012. [Online]. Available: http://
control concepts enable intelligent behavior on this level. On www.nist.gov
the system level, the optimization and control of the whole [17] IEC Smart Grid Standardization Roadmap, IEC, Geneva, Switzerland,
smart grid with all its components is in the focus of advanced Tech. Rep., Ed. 1.0, 2010. [Online]. Available: http://www.iec.ch
research supporting the necessary functions and services, as [18] IEEE Guide for Smart Grid Interoperability of Energy Technology
and Information Technology Operation With the Electric Power System
stated in this paper. (EPS), End-Use Applications, and Loads, USA, IEEE Std. 2030-2011,
Despite the growing awareness and the ongoing research and Sep. 10, 2011.
technology developments in smart grids, there are still many [19] The German Roadmap e-Energy/Smart Grids 2.0—Smart Grids Stan-
dardization Status, Trends and Prospects, German Commission for
open issues, which have to be covered in future research work. Electrical, Electronic and Information Technologies of DIN and VDE,
The most important points in the context of this review paper Frankfurt, Germany, Tech. Rep., 2013. [Online]. Available: http://www.
are as follows: 1) advanced control functions provided by DER, vde.com
ESS, and flexible loads; 2) integration of all ICT/automation [20] A. Chaouachi, R. M. Kamel, R. Andoulsi, and K. Nagasaka, “Multi-
objective intelligent energy management for a microgrid,” IEEE Trans.
concepts into one smart grid control solution taking all as- Ind. Electron., vol. 60, no. 4, pp. 1688–1699, Apr. 2013.
pects into account (i.e., generation, distribution, storage, and [21] European Smartgrids Technology Platform: Vision and Strategy for
consumption) and the technology developed in the AI and Europe’s Electricity Networks of the Future, Directorate-General for
software engineering areas; and 3) the provision of proper Research-Sustainable Energy Systems, Eur. Comm., Brussel, Belgium,
Tech. Rep., 2006. [Online]. Available: http://cordis.europa.eu
design, development, and validation methods for component [22] M/490 Standardization Mandate to European Standardization Organiza-
design and integrated ICT/automation solutions. There is still tions (ESOs) to Support European Smart Grid Deployment, Eur. Comm.,
a lot of work necessary in order to implement a real smart grid Brussel, Belgium, 2012. [Online]. Available: http://ec.europa.eu
[23] S. McArthur et al., “Multi-agent systems for power engineering
system. Industrial electronics and informatics together with the applications—Part I: Concepts, approaches, and technical challenges,”
AI techniques will play an important role of such an integrated IEEE Trans. Power Syst., vol. 22, no. 4, pp. 1743–1752, Nov. 2007.
approach in the near future. [24] C. Younghoon and L. Jih-Sheng, “Digital plug-in repetitive controller for
single-phase bridgeless PFC converters,” IEEE Trans. Power Electron.,
vol. 28, no. 1, pp. 165–175, Jan. 2013.
R EFERENCES [25] A. Prasai and D. M. Divan, “Control of dynamic capacitor,” IEEE Trans.
[1] World Energy Outlook 2013, IEA, Paris, France, Tech. Rep., 2013. Ind. Appl., vol. 47, no. 1, pp. 161–168, Jan./Feb. 2011.
[Online]. Available: http://www.iea.org [26] M. J. Hossain, H. R. Pota, and R. A. Ramos, “Improved low-
[2] C. Cecati, C. Citro, A. Piccolo, and P. Siano, “Smart operation of wind voltage-ride-through capability of fixedspeed wind turbines using decen-
turbines and diesel generators according to economic criteria,” IEEE tralised control of STATCOM with energy storage system,” IET Gener.
Trans. Ind. Electron., vol. 58, no. 10, pp. 4514–4525, Oct. 2011. Transmiss. Distrib., vol. 6, no. 8, pp. 719–730, Aug. 2012.
[3] M. Liserre, T. Sauter, and J. Hung, “Future energy systems: Integrating [27] EN 50160 Voltage Characteristics of Electricity Supplied by Public
renewable energy sources into the smart power grid through indus- Electricity Networks, CENELEC, Brussels, Belgium, 2012. [Online].
trial electronics,” IEEE Ind. Electron. Mag., vol. 4, no. 1, pp. 18–37, Available: http://www.cenelec.eu
May 2010. [28] “Harmonics, characteristic parameters, methods of study, estimates of
[4] G. Spagnuolo et al., “Renewable energy operation and conversion existing values in the network,” Electra, no. 77, pp. 35–54, Jul. 1981.
schemes: A summary of discussions during the seminar on renewable [29] J. Rodriguez, J.-S. Lai, and F. Z. Peng, “Multilevel inverters: A survey
energy systems,” IEEE Ind. Electron. Mag., vol. 4, no. 1, pp. 38–51, of topologies, controls, and applications,” IEEE Trans. Ind. Electron.,
Mar. 2010. vol. 49, no. 4, pp. 724–738, Aug. 2002.
STRASSER et al.: REVIEW OF ARCHITECTURES AND CONCEPTS FOR INTELLIGENCE IN SYSTEMS 2435

[30] M. Malinowski, K. Gopakumar, J. Rodriguez, and M. A. Perez, “A [53] J. Napoles et al., “Selective harmonic mitigation technique for cascaded
survey on cascaded multilevel inverters,” IEEE Trans. Ind. Electron., H-Bridge converters with non-equal dc link voltages,” IEEE Trans. Ind.
vol. 57, no. 7, pp. 2197–2206, Jul. 2010. Electron., vol. 60, no. 5, pp. 1963–1971, May 2013.
[31] C. Buccella, C. Cecati, and H. Latafat, “Digital control of power [54] J. Napoles, J. I. Leon, R. Portillo, L. G. Franquelo, and M. A. Aguirre,
converters—A survey,” IEEE Trans. Ind. Informat., vol. 8, no. 3, “Selective harmonic mitigation technique for high-power converters,”
pp. 437–447, Aug. 2012. IEEE Trans. Ind. Electron., vol. 57, no. 7, pp. 2197–2206, Jul. 2010.
[32] S. R. Pulikanti, G. Konstantinou, and V. G. Agelidis, “Hybrid seven-level [55] L. G. Franquelo, J. Napoles, R. Portillo, J. I. Leon, and M. A. Aguirre,
cascaded active neutral-point-clamped-based multilevel converter under “A flexible selective harmonic mitigation technique to meet grid codes in
SHE-PWM,” IEEE Trans. Ind. Electron., vol. 60, no. 11, pp. 4794–4804, three-level PWM converters,” IEEE Trans. Ind. Electron., vol. 54, no. 6,
Nov. 2013. pp. 3022–3029, Dec. 2007.
[33] M. Narimani and G. Moschopoulos, “Three-phase multimodule VSIs [56] Electrical energy storage, IEC, Geneva, Switzerland, Tech. Rep., 2011.
using SHE-PWM to reduce zero-sequence circulating current,” IEEE [Online]. Available: http://www.iec.ch
Trans. Ind. Electron., vol. 61, no. 4, pp. 1659–1668, Apr. 2014. [57] A European Strategic Energy Technology Plan (SET-Plan), Eur. Comm.,
[34] Z. Ming, H. Long, Y. Wenxi, and L. Zhengyu, “Circulating harmonic Brussel, Belgium, No. 723, 2007. [Online]. Available: http://ec.europa.eu
current elimination of a CPS-PWM-based modular multilevel converter [58] European white book on grid-connected storage, DERlab e.V., Kassel,
with a plug-in repetitive controller,” IEEE Trans. Power Electron., Germany, Version 1.11.2010, 2010, Tech. Rep., 2010. [Online]. Avail-
vol. 29, no. 4, pp. 2083–2097, Apr. 2014. able: http://www.der-lab.net
[35] K. Gopakumar et al., “A medium voltage inverter fed IM drive using [59] J. V. Appen, M. Braun, and R. Estrella, “A framework for different
multilevel 12-sided polygonal vectors, with nearly constant switching storage use cases in distribution systems,” in Proc. CIRED, Lisbon,
frequency current hysteresis controller,” IEEE Trans. Ind. Electron., Portugal, May 29–30, 2012, pp. 1–4.
vol. 61, no. 4, pp. 1700–1709, Apr. 2014. [60] EPRI-DOE Handbook of Energy Storage for Transmission & Distribution
[36] M. Sztykiel et al., “Modular multilevel converter modeling, control and Applications, U.S. DOE, Washington, DC, USA, Tech. Rep., No. 1001834,
analysis under grid frequency deviations,” in Proc. 15th EPE, Lille, EPRI, Palo Alto, CA, 2003. [Online]. Available: http://www.epri.com
France, Sep. 3–5, 2013, pp. 1–11. [61] J. Kathan, “Increasing the hosting capacity of photovoltaics with elec-
[37] G. Konstantinou, M. Ciobotaru, and V. Agelidis, “Selective harmonic tric storage—Simulation and hardware-in-the-loop concept,” M.S thesis,
elimination pulse-width modulation of modular multilevel converters,” Univ. Appl. Sci. Technikum Vienna, Vienna, Austria, 2011.
IET Power Electron., vol. 6, no. 1, pp. 96–107, Jan. 2013. [62] P. Wang et al., “Integrating electrical energy storage into coordinated
[38] C. Buccella, C. Cecati, M. G. Cimoroni, and K. Razi, “Analytical method voltage control schemes for distribution networks,” IEEE Trans. Smart
for pattern generation in five-level cascaded H-bridge inverter using Grid, vol. 5, no. 2, pp. 1018–1032, Mar. 2014.
selective harmonic elimination,” IEEE Trans. Ind. Electron., vol. 61, [63] M. Kabir, Y. Mishra, G. Ledwich, Z. Dong, and K. Wong, “Coordinated
no. 11, pp. 5811–5819, Nov. 2014. control of grid connected photo voltaic reactive power and battery energy
[39] S. Qiang et al., “A steady-state analysis method for a modular multilevel storage systems to improve the voltage profile of a residential distribu-
converter,” IEEE Trans. Power Electron., vol. 28, no. 8, pp. 3702–3713, tion feeder,” IEEE Trans. Ind. Informat., vol. 10, no. 2, pp. 967–977,
Aug. 2013. May 2014.
[40] J. Kaniewski, Z. Fedyczak, and G. Benysek, “AC voltage sag/swell [64] D. Velasco de la Fuente, C. L. T. Rodriguez, G. Garcera, E. Figueres,
compensator based on three-phase hybrid transformer with buck–boost and R. O. Gonzalez, “Photovoltaic power system with battery backup
matrix-reactance chopper,” IEEE Trans. Ind. Electron., vol. 61, no. 8, with grid-connection and islanded operation capabilities,” IEEE Trans.
pp. 3835–3846, Aug. 2014. Ind. Electron., vol. 60, no. 4, pp. 1571–1581, Apr. 2013.
[41] S. Xu, A. Q. Huang, and R. Burgos, “Review of solid-state transformer [65] P. Eder-Neuhauser, “Technical comparison of provision for increasing
technologies and their application in power distribution systems,” IEEE the hosting capacity in rural power distribution grids,” M.S thesis, Univ.
J. Emerging Sel. Topics Power Electron., vol. 1, no. 3, pp. 186–198, Appl. Sci. Technikum Vienna, Vienna, Austria, 2013.
Sep. 2013. [66] J. von Appen, T. Stetz, M. Braun, and A. Schmiegel, “Local voltage
[42] Q. Hengsi and J. W. Kimball, “Solid-state transformer architecture using control strategies for PV storage systems in distribution grids,” IEEE
ac–ac dual-active-bridge converter,” IEEE Trans. Ind. Electron., vol. 60, Trans. Smart Grid, vol. 5, no. 2, pp. 1002–1009, Mar. 2014.
no. 9, pp. 3720–3730, 2013. [67] Communicating the Customer Benefits of Information Standards—A
[43] S.-H. Hwang, X. Liu, J.-M. Kim, and H. Li, “Distributed digital con- Guide to Defining How Standards Reduce Cost and Eliminate Risk
trol of modular-based solid-state transformer using DSP+FPGA,” IEEE in Solar Installations, SunSpec Alliance, White Paper, 2012. [Online].
Trans. Ind. Electron., vol. 60, no. 2, pp. 670–680, Feb. 2013. Available: http://www.sunspec.org
[44] R. Salehi, N. Farokhnia, M. Abedi, and S. H. Fathi, “Elimination of low [68] F. Marra, G. Yang, C. Traeholt, J. Ostergaard, and E. Larsen, “A decen-
order harmonics in multilevel inverter using genetic algorithm,” J. Power tralized storage strategy for residential feeders with photovoltaics,” IEEE
Electron., vol. 11, no. 2, pp. 132–139, Mar. 2011. Trans. Smart Grid, vol. 5, no. 2, pp. 974–981, Mar. 2014.
[45] M. T. Hagh, H. Taghizadeh, and K. Razi, “Harmonic minimization in [69] S. T. Cha, H. Zhao, Q. Wu, A. Saleem, and J. Ostergaard, “Coordinated
multilevel inverters using modified species-based particle swarm opti- control scheme of battery energy storage system (BESS) and distributed
mization,” IEEE Trans. Power Electron., vol. 24, no. 10, pp. 2259–2267, generations (DGs) for electric distribution grid operation,” in Proc. IEEE
Oct. 2009. IECON, Montreal, QC, Canada, Oct. 25–28, 2012, pp. 4758–4764.
[46] A. Kaviani, S. H. Fathi, N. Farokhnia, and A. Ardakani, “PSO, an effective [70] H. Sugihara, K. Yokoyama, O. Saeki, K. Tsuji, and T. Funaki, “Eco-
tool for harmonics elimination and optimization in multilevel inverters,” nomic and efficient voltage management using customer-owned energy
in Proc. 4th IEEE ICIEA, Xi’an, China, May 25–27, 2009, pp. 2902–2907. storage systems in a distribution network with high penetration of pho-
[47] R. Salehi, B. Vahidi, N. Farokhnia, and M. Abedi, “Harmonic elim- tovoltaic systems,” IEEE Trans. Power Syst., vol. 28, no. 1, pp. 102–111,
ination and optimization of stepped voltage of multilevel inverter by Feb. 2013.
bacterial foraging algorithm,” J. Electron. Eng. Technol., vol. 5, no. 4, [71] P. Siano, “Demand response and smart grids—A survey,” Renew. Sus-
pp. 545–551, 2010. tain. Energy Rev., vol. 30, pp. 461–478, Feb. 2014.
[48] Y. X. Xie, L. Zhou, and H. Peng, “Homotopy algorithm research of the [72] C. Cecati, C. Citro, and P. Siano, “Combined operations of renewable
inverter harmonic elimination PWM model,” in Proc. Chin. Soc. Elect. energy systems and responsive demand in a smart grid,” IEEE Trans.
Eng., 2000, vol. 20, pp. 23–26. Sustain. Energy, vol. 2, no. 4, pp. 468–476, Oct. 2011.
[49] K. Sundareswaran, K. Jayant, and T. N. Shanavas, “Inverter harmonic [73] National assessment of demand response, FERC, Washington, DC, USA,
elimination through a colony of continuously exploring ants,” IEEE Tech. Rep., 2009. [Online]. Available: http://www.ferc.gov
Trans. Ind. Electron., vol. 54, no. 5, pp. 2558–2565, Oct. 2007. [74] Assessment of Demand Response and Advanced Metering: Staff Report,
[50] J. N. Chiasson, L. M. Tolbert, K. J. McKenzie, and Z. Du, “Control of FERC, Washington, DC, USA, Tech. Rep., 2012. [Online]. Available:
a multilevel converter using resultant theory,” IEEE Trans. Control Syst. http://www.ferc.gov
Technol., vol. 11, no. 3, pp. 345–353, May 2003. [75] P. Cappers, C. Goldman, and D. Kathan, “Demand response in U.S.
[51] J. N. Chiasson, L. M. Tolbert, K. J. McKenzie, and Z. Du, “A com- electricity markets: Empirical evidence,” Energy, vol. 35, no. 4,
plete solution to the harmonic elimination problem,” IEEE Trans. Power pp. 1526–1535, 2010.
Electron., vol. 19, no. 2, pp. 491–499, Mar. 2004. [76] B. Shen et al., “The role of regulatory reforms, market changes, and
[52] C. Buccella, C. Cecati, M. G. Cimoroni, and K. Razi, “Harmonic miti- technology development to make demand response a viable resource
gation technique for multilevel inverters in power systems,” in Proc. Int. in meeting energy challenges,” Appl. Energy, vol. 130, pp. 814–823,
SPEEDAM, Ischia, Italy, Jun. 18–20, 2014, pp. 73–77. Oct. 2014.
2436 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 62, NO. 4, APRIL 2015

[77] P. Siano and G. Mokryani, “Assessing wind turbines placement in a [101] G. Peppink et al., “INTEGRAL: ICT-platform based distributed control
distribution market environment by using particle swarm optimization,” in electricity grids with a large share of distributed energy resources and
IEEE Trans. Power Syst., vol. 28, no. 4, pp. 3852–3864, Nov. 2013. renewable energy sources,” in Energy-Efficient Computing and Network-
[78] J. Torriti, M. G. Hassan, and M. Leach, “Demand response experience ing, N. Hatziargyriou et al., Ed. Berlin, Germany: Springer-Verlag,
in Europe: Policies, programmes and implementation,” Energy, vol. 35, 2011, pp. 215–224.
no. 4, pp. 1575–1583, Apr. 2010. [102] S. Karnouskos et al., “Monitoring and control for energy efficiency
[79] A. Ipakchi and F. Albuyeh, “Grid of the future,” IEEE Power Energy in the smart house,” in Energy-Efficient Computing and Networking,
Mag., vol. 7, no. 2, pp. 52–62, Mar./Apr. 2009. N. Hatziargyriou, A. Dimeas, T. Tomtsi, and A. Weidlich, Eds. Berlin,
[80] J. C. Fuller, K. Schneider, and D. Chassin, “Analysis of residential de- Germany: Springer-Verlag, 2011, pp. 197–207.
mand response and double-action markets,” in Proc. IEEE Power Energy [103] J. M. Jorgensen, S. H. Sorensen, K. Behnke, and P. B. Eriksen, “EcoGrid
Soc. Gen. Meet., Detroit, MI, USA, Jul. 24–28, 2011, pp. 1–7. EU—A prototype for European smart grids,” in Proc. IEEE Power En-
[81] P. Siano, G. Graditi, M. Atrigna, and A. Piccolo, “Designing and ergy Soc. Gen. Meet., Detroit, USA, Jul. 24–29, 2011, pp. 1–7.
testing decision support and energy management systems for smart [104] G. Zhabelova and V. Vyatkin, “Multiagent smart grid automation ar-
homes,” J. Ambient Intell. Human. Comput., vol. 4, no. 6, pp. 651–661, chitecture based on IEC 61850/61499 intelligent logical nodes,” IEEE
Dec. 2013. Trans. Ind. Electron., vol. 59, no. 5, pp. 2351–2362, May 2012.
[82] M. G. Ippolito, G. Zizzo, A. Piccolo, and P. Siano, “Definition and appli- [105] IEC 61850 Communication Networks and Systems for Power Utility
cation of innovative control logics for residential energy optimization,” Automation, Std., Ed. 2.0, Mar. 2013. [Online]. Available: http://www.
in Proc. Int. SPEEDAM, Ischia, Italy, Jun. 18–20, 2014, pp. 1–6. iec.ch
[83] M. G. Ippolito, E. R. Sanseverino, and G. Zizzo, “Impact of building [106] IEC 61499-1 Function Blocks—Architecture, Std., Ed. 2.0, Nov. 7, 2012.
automation control systems and technical building management systems [Online]. Available: http://www.iec.ch
on the energy performance class of residential buildings: An Italian case [107] G. Zhabelova, Y. Chen-Wei, and V. Vyatkin, “SysGrid: IEC 61850/IEC
study,” Energy Buildings, vol. 69, pp. 33–40, Feb. 2014. 61499 based engineering process for smart grid automation design,” in
[84] S. Yin, S. X. Ding, X. Xie, and H. Luo, “A review on basic data- Proc. 11th IEEE Int. Conf. INDIN, Bochum, Germany, Jul. 29–31, 2013,
driven approaches for industrial process monitoring,” IEEE Trans. Ind. pp. 364–369.
Electron., vol. 61, no. 11, pp. 6418–6428, Nov. 2014. [108] C.-W. Yang, G. Zhabelova, V. Vyatkin, A. Apostolov, and
[85] S. Yin, X. Li, H. Gao, and O. Kaynak, “Data-based techniques focused N.-K. C Nair, “Smart grid automation distributed protection application
on modern industry: An overview,” IEEE Trans. Ind. Electron., vol. 62, with IEC 61850/IEC 61499,” in Proc. 10th IEEE Int. Conf. INDIN,
no. 1, pp. 657–667, Jan. 2015. Beijing, China, Jul. 25–27, 2012, pp. 364–369.
[86] A. Q. Huang, M. L. Crow, G. T. Heydt, J. P. Zheng, and S. J. Dale, “The [109] D. Issicaba, N. J. Gil, and J. A. Peas Lopes, “Islanding operation of active
future renewable electric energy delivery and management (FREEDM) distribution grids using an agent-based architecture,” in IEEE PES ISGT
system: The energy internet,” Proc. IEEE, vol. 99, no. 1, pp. 133–148, Europe, Chalmers, Sweden, Oct. 10–13, 2010, pp. 1–8.
Jan. 2011. [110] M. Postina, S. Rohjans, U. Steffens, and M. Uslar, “Views on service
[87] M. Xiaolin, S. Falcones, and R. Ayyanar, “Energy-based control design oriented architectures in the context of smart grids,” in Proc. 1st IEEE
for a solid state transformer,” in Proc. IEEE Power Energy Soc. Gen. Int. Conf. SmartGridComm, Gaithersburg, MD, USA, Oct. 4–6, 2010,
Meet., Minneapolis, MN, USA, Jul. 25–29, 2010, pp. 1–7. pp. 25–30.
[88] G. Zhabelova, V. Vyatkin, and N. Nair, “Standard-based engineering [111] P. Brédillet, E. Lambert, and E. Schultz, “CIM, 61850, COSEM stan-
and distributed execution framework for intelligent fault management dards used in a model driven integration approach to build the smart grid
for FREEDM system,” in Proc. IEEE IECON, Melbourne, Australia, service oriented architecture,” in Proc. 1st IEEE Int. Conf. SmartGrid-
Nov. 7–10, 2011, pp. 2724–2729. Comm, Gaithersburg, MD, USA, Oct. 4–6, 2010, pp. 467–471.
[89] G. Zhabelova, Z. Zhang, V. Vyatkin, and M.-Y. Chow, “Agent-based [112] S. Lehnhoff, S. Rohjans, M. Uslar, and W. Mahnke, “OPC Unified
distributed consensus algorithm for decentralized economic dispatch in Architecture: A service-oriented architecture for smart grids,” in Proc.
smart grid,” in Proc. IEEE IECON, Vienna, Austria, Nov. 10–13, 2013, Int. Workshop SE4SG, Zurich, Switzerland, Jun. 3, 2012, pp. 1–7.
pp. 1968–1973. [113] G. A. Pagani and M. Aiello, “Towards a service-oriented energy market:
[90] B. McMillin et al., “Architecture of a smart microgrid distributed operat- Current state and trend,” in Proc. 1st Int. Workshop SEE, vol. 6568,
ing system,” in Proc. IEEE/PES PSCE, Phoenix, AZ, USA, Mar. 20–23, Lecture Notes in Computer Science, San Francisco, CA, USA, Dec. 7,
2011, pp. 1–5. 2010, pp. 203–209.
[91] Z. Zhang and M.-Y. Chow, “Incremental cost consensus algorithm in a [114] K. Takagiwa, R. Kubo, S. Ishida, K. Inoue, and H. Nishi, “Feasibility
smart grid environment,” in Proc. IEEE Power Energy Soc. Gen. Meet., study of service-oriented architecture for smart grid communications,”
San Diego, CA, USA, Jul. 24–28, 2011, pp. 1–6. in Proc. IEEE ISIE, Taipei, Taiwan, May 28–31, 2013, pp. 1–7.
[92] R. Akella, F. Meng, D. Ditch, B. McMillin, and M. Crow, “Distributed [115] C. Warmer, K. Kok, S. Karnouskos, A. Weidlich, and D. Nestle, “Web
power balancing for the FREEDM system,” in Proc. 1st IEEE Int. Conf. services for integration of smart houses in the smart grid,” in Proc. Grid-
SmartGridComm, Gaithersburg, MD, USA, Oct. 4–6, 2010, pp. 7–12. Interop Forum, Denver, CO, USA, Nov. 17–19, 2009, pp. 1–5.
[93] F. Meng, R. Akella, M. L. Crow, and B. McMillin, “Distributed grid [116] S. Karnouskos, “The cooperative Internet of things enabled smart grid,”
intelligence for future microgrid with renewable sources and storage,” in in Proc. IEEE ISCE, Braunschweig, Germany, Jun. 7–10, 2010, pp. 1–6.
Proc. NAPS, Arlington, TX, USA, Sep. 26–28, 2010, pp. 1–6. [117] S. Karnouskos, O. Terzidis, and P. Karnouskos, “An advanced metering
[94] S. Patil, V. Vyatkin, and B. McMillin, “Implementation of FREEDM infrastructure for future energy networks,” in Proc. IFIP/IEEE 1st Int.
smart grid distributed load balancing using IEC 61499 function Conf. NTMS, Paris, France, May 2–4, 2007, pp. 597–606.
blocks,” in Proc. IEEE IECON, Vienna, Austria, Nov. 10–13, 2013, [118] A. Koestler, The Ghost in the Machine. London, U.K.: Arkana Books,
pp. 8154–8159. 1969.
[95] P. Leitao, V. Mařík, and P. Vrba, “Past, present, and future of industrial [119] M. H. Moghadam and N. Mozayani, “A novel information exchange
agent applications,” IEEE Trans. Ind. Informat., vol. 9, no. 4, pp. 2360– model in IT infrastructure of smart grid,” Res. J. Appl. Sci., Eng.
2372, Nov. 2013. Technol., vol. 6, no. 23, pp. 4399–4404, 2013.
[96] P. Leitao, P. Vrba, and T. Strasser, “Multi-agent systems as automation [120] E. Negeri, N. Baken, and M. Popov, “Holonic architecture of the
platform for intelligent energy systems,” in Proc. IEEE IECON, Vienna, smart grid,” Smart Grid Renew. Energy, vol. 4, no. 2, pp. 202–212,
Austria, Nov. 10–13, 2013, pp. 66–71. May 2013.
[97] G. J. Schaeffer and H. Akkermand, CRISP Final Summary Report, [121] E. Negeri and N. Baken, “Architecting the smart grid as holarchy,” in
CRISP, Petten, The Netherlands, 2006. [Online]. Available: http://www. Proc. 1st Int. Conf. SMARTGREENS, Porto, Portugal, Apr. 19–20, 2012,
crisp.ecn.nl/documents.html pp. 73–78.
[98] N. Hatziargyriou, “Smart agent technology to help DER integrate [122] S. Frey, A. Diaconescu, M. David, and I. Demeure, “A holonic control
markets: Experiments in Greece,” in Proc. 3rd IRED, Nice, France, architecture for a heterogeneous multi-objective smart micro-grid,” in
Dec. 10–12, 2010, pp. 1–21. Proc. 7th IEEE Int. Conf. SASO, Philadelphia, USA, Sep. 9–13, 2013,
[99] A. L. Dimeas and N. D. Hatziargyriou, “Agent based control of virtual pp. 21–30.
power plants,” in Proc. Int. Conf. ISAP, Niigata, Japan, Nov. 5–8, 2007, [123] A. Pahwa et al., “Holonic multi-agent control of power distribution
pp. 1–6. systems of the future,” in Proc. Grid Future Symp., CIGRE US Nat.
[100] J. K. Kok, C. J. Warmer, and I. G. Kamphuis, “PowerMatcher: Multia- Com., Kansas, MO, USA, Oct. 28–30, 2012, pp. 1–7.
gent control in the electricity infrastructure,” in Proc. 4th Int. Joint Conf. [124] IEC 61131-3 Programmable Logic Controllers—Programming Lan-
AAMAS, Utrecht, The Netherlands, Jul. 25–29, 2005, pp. 75–82. guages, Std., Ed. 2.0, 2012. [Online]. Available: http://www.iec.ch
STRASSER et al.: REVIEW OF ARCHITECTURES AND CONCEPTS FOR INTELLIGENCE IN SYSTEMS 2437

[125] W. Mahnke, S. H. Leitner, and M. Damm, OPC Unified Architecture. Johannes Kathan received the B.Sc. degree
Berlin, Germany: Springer-Verlag, 2009. in energy and environmental management,
[126] S. Rohjans, M. Uslar, and H. Appelrath, “OPC UA and CIM: Semantics specializing in electric energy systems, from
for the smart grid,” in Proc. IEEE PES Transmiss. Distrib. Conf. Expo., the University of Applied Sciences, Pinkafeld,
New Orleans, LA, USA, Apr. 19–22, 2010, pp. 1–8. Austria, in 2009 and the M.Sc. degree in urban
[127] S. Sucic, B. Bony, L. Guise, F. Jammes, and A. Marusic, “Integrating renewable energy systems from the University
DPWS and OPC UA device-level SOA features into IEC 61850 applica- of Applied Sciences Technikum Vienna, Vienna,
tions,” in Proc. IEEE IECON, Montreal, QC, Canada, Oct. 25–28, 2012, Austria.
pp. 5773–5778. Since 2009, he has been a Researcher with
[128] V. Vyatkin, IEC 61499 Function Blocks for Embedded and Distributed the Business Unit for Electric Energy Systems,
Control Systems Design., 2nd ed. Champaign, IL, USA: ISA, 2011. AIT Austrian Institute of Technology, Vienna. He
[129] I. Hegny et al., “A distributed energy management approach for au- is active in the networking activities of the European Electricity Grid
tonomous power supply systems,” in Proc. IEEE Int. Conf. INDIN, Initiative and the European Energy Research Alliance. His research
Vienna, Austria, Jul. 23–27, 2007, pp. 1065–1070. interests include the integration electrical storage systems into the
[130] N. Higgins, V. Vyatkin, N.-K. Nair, and K. Schwarz, “Distributed power distribution system and their applications.
system automation with IEC 61850, IEC 61499, and intelligent control,”
IEEE Trans. Syst., Man, Cybern. C, Appl. Rev., vol. 41, no. 1, pp. 81–92,
Jan. 2011.
[131] V. Vyatkin et al., “Standards-enabled smart grid for the future energy
web,” in Proc. 1st IEEE Int. Conf. SmartGridComm, Gaithersburg, MD, Carlo Cecati (M’90–SM’03–F’06) received the
USA, Oct. 4–6, 2010, pp. 7–12. Dr.Ing. degree in electrotechnical engineering
[132] T. Strasser, F. Andren, V. Vyatkin, G. Zhabelova, and C. W. Yang, from the University of L’Aquila, L’Aquila, Italy,
“Towards an IEC 61499 compliance profile for smart grids review and in 1983.
analysis of possibilities,” in Proc. IEEE IECON, Montreal, QC, Canada, Since then, he has been with the Univer-
Oct. 25–28, 2012, pp. 3750–3757. sity of L’Aquila, where he is a Professor of
[133] A. Zoitl, T. Strasser, and A. Valentini, “Open source initiatives as ba- industrial electronics and drives. Since 2014, he
sis for the establishment of new technologies in industrial automation: has also been a Chief International Academic
4DIAC a case study,” in Proc. IEEE ISIE, Bari, Italy, Jul. 4–7, 2010, Adviser with Harbin Institute of Technology,
pp. 3817–3819. Harbin, China. In 2007, he was a Cofounder
[134] T. Strasser, M. Stifter, F. Andren, D. Burnier de Castro, and W. Hribernik, of DigiPower Ltd., L’Aquila. His research and
“Applying open standards and open source software for smart grid appli- technical interests cover several aspects of power electronics, electrical
cations: Simulation of distributed intelligent control of power systems,” drives, digital control, distributed generation, and smart grids.
in Proc. IEEE Power Energy Soc. Gen. Meet., San Diego, CA, USA, Prof. Cecati currently serves as the Editor-in-Chief of the IEEE
Jul. 24–29, 2011, pp. 1–8. T RANSACTIONS ON I NDUSTRIAL E LECTRONICS.
[135] T. Strasser, F. Andren, F. Lehfuss, M. Stifter, and P. Palensky, “Online
reconfigurable control software for IEDs,” IEEE Trans. Ind. Informat.,
vol. 9, no. 3, pp. 1455–1464, Aug. 2013.
[136] L. Zhu, D. Shi, and X. Duan, “Standard function blocks for flexible IED
in IEC 61850-based substation automation,” IEEE Trans. Power Del., Concettina Buccella (M’92–SM’03) received
vol. 26, no. 2, pp. 1101–1111, Apr. 2011. the Dr.Eng. degree in electrical engineering
from the University of L’Aquila, L’Aquila, Italy,
and the Ph.D. degree in electrical engineering
from the University of Rome “La Sapienza,”
Thomas Strasser (M’09–SM’13) received the Rome, Italy.
Ph.D. degree in mechanical engineering, with From 1988 to 1989, she was with Italtel
a focus on automation and control theory, S.p.A. She then joined the University of L’Aquila,
from Vienna University of Technology, Vienna, where she has been an Associate Professor
Austria, in 2003. since 2001. She is currently also the C.E.O.
He was a Senior Researcher with PROFAC- of DigiPower Ltd., L’Aquila, a university spin-off
TOR Research, Steyr, Austria, working in the dealing with industrial electronics and renewable energies. Her research
field of reconfigurable automation for six years. interests include power converters, power systems, smart grids, elec-
He is currently a Senior Scientist with the AIT tromagnetic compatibility, electrostatic processes, and ultrawideband
Austrian Institute of Technology, Vienna, work- signal interference.
ing in the domain of smart grids, with a focus on Prof. Buccella was a corecipient of the 2012 and 2013 IEEE T RANS -
power utility automation. ACTIONS ON I NDUSTRIAL I NFORMATICS Best Paper Awards.
Dr. Strasser is a member of the International Electrotechnical
Commission—Subcommittee 65B/Working Group 15—Function Blocks
(IEC C65B/WG15) maintaining the IEC 61499 standard. In addition, he
is a Senior Member of the IEEE Industrial Electronics Society, IEEE
Systems, Man, and Cybernetics Society, and IEEE Power and Energy Pierluigi Siano (M’09–SM’14) received the
Society. M.Sc. degree in electronic engineering and the
Ph.D. degree in information and electrical engi-
neering from the University of Salerno, Salerno,
Italy, in 2001 and 2006, respectively.
Filip Andrén (M’12) received the Master’s de- He is currently an Aggregate Professor of
gree in applied physics and electrical engi- electrical energy engineering with the Depart-
neering, with a thematic focus on control and ment of Industrial Engineering, University of
information systems, from Linköping University, Salerno. In 2013, he received the Italian Na-
Linköping, Sweden, in 2009. tional Scientific Qualification as a Full Professor
Since 2009, he has been a Scientist with in the competition sector of electrical energy
the Energy Department, AIT Austrian Institute engineering. He has coauthored over 160 papers, including over 70
of Technology, Vienna, Austria. He specializes published in international journals. His research activities are centered
in smart grids and power utility automation. on the integration of distributed energy resources in smart distribution
His main research interests are automation and systems and on planning and management of power systems.
control systems, communication and automa- Dr. Siano is an Associate Editor of the IEEE T RANSACTIONS ON
tion standards, and modeling, simulation, and development of intelligent I NDUSTRIAL I NFORMATICS and a member of the Editorial Boards of over
grid components. 30 international journals.
2438 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 62, NO. 4, APRIL 2015

Paulo Leitão (M’98–SM’08) received the Ph.D. Pavel Vrba (M’05–SM’14) received the Ph.D.
degree in electrical and computer engineering degree in applied sciences and informatics, cy-
from the University of Porto, Porto, Portugal, bernetics, from the University of West Bohemia,
in 2004. Pilsen, Czech Republic, in 2001.
He is currently a Professor and the Head of In 2001–2012, he was with the Rockwell Au-
the Department of Electrical Engineering with tomation Research Center, Prague, where he
the Polytechnic Institute of Bragança, Bragança, led the Distributed Intelligent Control Laboratory
Portugal, and also a member of the Artificial in 2005–2012. He is currently a Researcher
Intelligence and Computer Science Laboratory with the Czech Institute of Informatics, Robotics,
(LIACC), Porto. He has authored or coauthored and Cybernetics, Czech Technical University
over 130 papers in high-ranked international sci- in Prague, Prague, Czech Republic, where he
entific journals and conference proceedings. He has coauthored three leads the Intelligent Systems for Industry Group in research and de-
patents. His research interests include industrial informatics, collabora- velopment of intelligent distributed solutions in various industrial fields.
tive factory automation, reconfigurable production systems, intelligent His main research interests are intelligent industrial control systems;
supervisory control, agent-based and holonic control, and bioinspiration holonic and multiagent systems; cyberphysical systems; systems of
engineering. systems; Internet of things; semantic technologies and ontologies;
Dr. Leitão is currently the Chair of the IEEE Industrial Electronics service-oriented architectures; and agent-based modeling, simulation,
Society Technical Committee on Industrial Agents. He served as a and visualization. He has authored or coauthored over 80 conference
General Cochair of several international conferences, including the papers, journal articles, and book chapters related to his research areas.
International Federation of Automatic Control Intelligent Manufacturing He has five filed U.S. patents, with one pending.
Systems (IFAC IMS’10) and the International Conference on Industrial
Applications of Holonic and Multi-Agent Systems (HoloMAS 2011).
Vladimír Mařík (M’95–SM’13) received the
Ph.D. degree in cybernetics from Czech Techni-
Gulnara Zhabelova received the Diploma cal University in Prague (CTU), Prague, Czech
degree in robotics and mechatronics and Republic, in 1979.
the M.E. degree in automation and control In 1990, he was appointed as a Full Professor
from Karaganda State Technical University, at CTU, where he is currently the Director of the
Karaganda, Kazakhstan, in 2006 and 2008, re- newly established Czech Institute of Informatics,
spectively, and the M.E. degree in computer Robotics, and Cybernetics. From 1999 to 2013,
systems and the Ph.D. degree in electrical and he was the Head of the Department of Cyber-
electronics engineering from The University of netics, CTU. From 1992 to 2009, he served as
Auckland, Auckland, New Zealand, in 2009 and the Founder and the Director of the Rockwell
2013, respectively. Automation Research Center, Prague. His research interests include
From 2009 to 2013, she was with the In- artificial intelligence, multiagent systems, knowledge-based systems,
dustrial Informatics Laboratory, The University of Auckland. She is soft computing, production planning, and scheduling applications.
currently a Research Engineer with Luleâ University of Technology, Mr. Mařík is currently serving as the Vice-President of the IEEE Sys-
Luleâ, Sweden. Her research interests include agent technology and tems, Man, and Cybernetics Society. He served as the Editor-in-Chief
its formal definition, theory, and application in wide practical domains, of the IEEE T RANSACTIONS ON S YSTEMS , M AN , AND C YBERNETICS —
including automation and control; protection in energy generation, trans- PART C: A PPLICATIONS AND R EVIEW from 2005 to 2013.
mission, distribution, and consumption; building automation; demand-
side management; advanced metering infrastructure; energy markets;
and policies.

Valeriy Vyatkin (M’03–SM’04) received the

Ph.D. degree from the Taganrog State Univer-
sity of Radio Engineering, Taganrog, Russia,
in 1992.
He is on joint appointment as a Chaired
Professor of Dependable Computation and
Communication Systems, Luleå University of
Technology, Luleå, Sweden, and a Professor
of information and computer engineering in au-
tomation with Aalto University, Helsinki, Finland.
Previously, he was a Visiting Scholar with the
University of Cambridge, Cambridge, U.K. He has permanent academic
appointments with The University of Auckland, Auckland, New Zealand;
Martin Luther University of Halle-Wittenberg, Halle, Germany; and
Taganrog State University of Radio Engineering. He was a Postdoc-
toral Fellow with the Nagoya Institute of Technology, Nagoya, Japan.
His research interests include dependable distributed automation and
industrial informatics, software engineering for industrial automation,
and distributed architectures and multiagent systems applied in various
industry sectors.
Dr. Vyatkin was a recipient of the Andrew P. Sage Award for the Best
IEEE Transactions Paper in 2012.