Alexandria Engineering Journal (2023) 70, 61–92

H O S T E D BY
Alexandria University

Alexandria Engineering Journal

www.elsevier.com/locate/aej
www.sciencedirect.com

REVIEW

DC-based microgrid: Topologies, control schemes,

and implementations
Babangida Modu a, Md Pauzi Abdullah a,c,*, Mufutau Adewolu Sanusi a,
Mukhtar Fatihu Hamza b

a
Centre of Electrical Energy Systems, School of Electrical Engineering, Faculty of Engineering, Universiti Teknologi
Malaysia, UTM, 81310 Skudai, Johor, Malaysia
b
Department of Mechanical Engineering, College of Engineering in Alkharj, Prince Sattam Bin Abdulaziz University, Alkharj
11942, Saudi Arabia
c
Institute of Future Energy, Universiti Teknologi Malaysia, UTM, 81310 Skudai, Johor, Malaysia

Received 18 August 2022; revised 12 January 2023; accepted 13 February 2023

Available online 2 March 2023

KEYWORDS Abstract This article presents a state-of-the-art review of the status, development, and prospects of
DC microgrid; DC-based microgrids. In recent years, researchers’ focus has shifted to DC-based microgrids as a
Topology; better and more feasible solution for meeting local loads at the consumer level while complementing
Renewable energy; a given power system’s reliability, stability, and controllability. DC microgrid has an advantage in
Control structure; terms of compatibility with renewable energy systems (RESs), energy storage, modern electrical
and Energy management appliances, high efficiency, and reliability. However, the integration of different distributed gener-
system ations has complicated the control of bus voltage and current. Therefore, several efforts have been
made in the research community to further explore efficient control techniques for a reliable and
stable DC microgrid. In spite of the numerous review papers published on DC microgrid control,
so far, not any has given sufficient emphasis on the power flow analysis methods used in various DC
microgrid topologies and the hybrid control topologies discussed in this review. In general, this
paper presents a meticulous explanation of DC microgrid architecture; power flow analysis; control
strategies with comparative analysis; challenges with recommendations; as well as classical and
intelligent-based energy management strategy (EMS). Finally, suggestion for further research is pre-
sented. This review paper will go a long way in helping readers to understand the present state of
development on DC microgrid control.
Ó 2023 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Alexandria
University. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

* Corresponding author.
E-mail address: [email protected] (M.P. Abdullah).
Peer review under responsibility of Faculty of Engineering, Alexandria University.
https://doi.org/10.1016/j.aej.2023.02.021
1110-0168 Ó 2023 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Alexandria University.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
62 B. Modu et al.

Nomenclature

Gi ðsÞ Transfer function HPU Hybrid power unit

Io Output current HSR Hybrid stochastic robust
Iref Reference current HVDC High voltage direct current
rdi Droop gain IED Intelligent electronic device
Ro Output droop resistance IoT Internet of things
Ti Integral constant LCB Low communication bandwidth
Vo Output voltage LV Low voltage
Vref Reference voltage LVDC Low voltage direct current
wLpf Cutoff frequency of low pass filter MAS Multi agent system
DVo;max Maximum DC bus voltage derivative MDS Multi sample differential protection scheme
MG Microgrid
Abbreviation MILP Mix integer linear programing
AC Alternating current MIP Mix integer programing
AI Artificial intelligence MIMO Multiple-input and multiple-output
ANN Artificial neural network MPC Model predictive control
BES Battery energy storage MPPT Maximum power point tracking
CMA Covariance matrix adaptation NREL National renewable energy lab
CPLs Constant power loads PEC Power electronic converter
DC Direct current PF Power flow
DCL Digital communication line PFA Power flow analysis
DCMG DC microgrid PI Proportional integral
DER Distributed energy resources PLC Power line carrier
DESs Distributed energy sources PnP Plug and play
DG Distributed generation PSO Particle swamp optimization
DRL Deep reinforcement learning PV Photovoltaic
EMS Energy management strategy RE Renewable energy
ES Energy storage RESs Renewable energy sources
ESD Energy storage device SoC State of charge
ESS Energy storage system SS Storage system
ETAP Electrical transient analysis and program TESS Thermal energy storage
EVs Electric vehicles TT Tidal turbine
GA Genetic algorithm WBSEB West Bengal state electricity board
GHG Greenhouse gas WOS Web of science
HES Hybrid energy storage

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
1.1. Review screen method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2. DC microgrid topologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.1. Radial DC microgrid topology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
2.2. Multi bus topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
2.3. Multi terminal topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
2.4. Ring bus DC microgrid topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
2.5. Ladder topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.6. Zonal topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3. Power flow analysis in DC microgrid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.1. Review of previous research on power flow analysis in DC microgrid topologies . . . . . . . . . . . . . . . . . . . . . . . . 69
4. Control strategy in DC microgrid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.1. Centralized control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.2. Decentralized control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.3. Distributed control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4.4. Hybrid decentralized and centralized control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.5. Hybrid distributed and decentralized control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5. Hierarchical control structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.1. Primary control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.2. Secondary control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
DC-based microgrid: Topologies, control schemes, and implementations 63

5.3. Tertiary control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

5.3.1. Energy management system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
6. DC microgrid protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
7. Challenges and recommendations on DC microgrid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
8. Discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
9. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Declaration of Competing Interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

1. Introduction units, and electrical loads are all linked to the bus in DC
microgrid. Power electronic converters (PEC) connect the
The growing concern about global carbon emissions and DC microgrid to grid utility as depicted in Fig. 1. with several
energy security has necessitated the search for clean, environ- voltage levels and energy storage devices on the DC side that
mentally friendly renewable energy sources for power genera- control demand variation, a DC microgrid can deliver power
tion using the microgrid concept. A microgrid is an emerging to DC and AC loads[5].
technology that encompasses different distributed energy DC microgrid has just one voltage conversion level between
sources (DESs), storage units, power electronic converters, every dispersed sources and DC bus compared to AC micro-
and electrical load. grid, as a result, the whole system’s construction cost has been
The most recent developments in power electronics have decreased and it also simplifies the control’s implementation
enabled DC microgrids to meet the required specifications at [6,7]. Nevertheless, researchers across the world are still look-
a reasonable cost and in a smooth approach. Therefore, DC ing for a way to reduce the cost of manufacturing, improve the
loads are used extensively for industrial applications and resi- performance and maximizing the utilization of renewable
dential applications such as a Datacenter, Telecommunication, energy sources because of numerous advantages of DC micro-
Electric Vehicles (EVs), DC-powered houses, Electric ships, grid over AC microgrid. In addition to that, they are also seek-
Renewable Energy Park, rails, Hybrid Energy storage systems, ing to come up with an innovative structures, smart control
and zero-net electrical energy buildings [1,2]. Through a power techniques, stabilization methods, enhanced power flow study
electronic interface, it is also easy to effectively connect energy and standard requirement for DC microgrid [8]. A suitable
storage devices to the DC microgrid. robust control system aimed at continuous and foreseeable
The major problems of microgrids are stability, bidirec- actions is a critical condition for a microgrid utilizing any
tional power flow, modeling, less inertia, the effect of load per- bus topology. Sustaining effective and safely delivering essen-
turbation, and uncertainties [3,4]. To address all the tial power from distributed generators to the destination is
aforementioned issues, control strategies have been proposed; the primary goal of employing a robust DC microgrid con-
however, the control strategies have many limitations, includ- troller. DC microgrid controller needs to carryout numerous
ing weak dynamic response, trade-off between voltage regula- control action including voltage and current regulation as well
tion and current sharing, chattering, and inaccuracy. as energy storage synchronization [9].
Operating the DGs in accordance with the load requirement This review paper is inspired by the recent increase in the
needs suitable control techniques and power electronic con- deployment of DC microgrid systems for real-world residential
verter selection. Distributed energy sources (DESs), storage and industrial application. Consequently, the paper provides a

Fig. 1 DC microgrid topology.

64 B. Modu et al.

current review of the literature on DC microgrid topologies, scription fee are not accessible. EndNote software was used
power flow analysis, control, protection, challenges, and future to find and remove duplicates. This is followed by scanning
recommendation. Some research problems were also men- the content of the abstract and conclusion sections to evaluate
tioned and then suggestions on future research were made the relevance of the paper to the objectives of this review study.
which can be accomplished by a novel technique or an Accordingly, the publications were chosen based on the scope
improvement on the existing technologies by the research com- of this research. Also, only research studies that focused on
munity or industries. DC microgrid topologies, control strategies, power flow anal-
ysis, protection, energy management as well as the challenges
1.1. Review screen method and open research issues related to DC-based microgrid con-
trol were considered. This reduced the number of articles down
This section describes the methodology used to identify and to approximately 210.
select papers on DC microgrid control from the literature. The remainder of the sections of the paper is organized as
Two major academic databases, Scopus (https:///www.sco- follows: DC microgrid topologies as well as their applications
pus.com/) and Web of Science (WOS) (https:///www.webof- is presented in section 2. Power flow analysis is presented in
knowledge.com/), were used to search for relevant prior section 3. Overview on DC microgrid control structures
studies using a series of keywords for a detailed literature namely, centralized, decentralized, and distributed control
review. The keywords used for the search include ‘‘DC micro- each with their advantage and limitation are discussed in 4.
grid”, ‘‘control strategies”, ‘‘energy management”, ‘‘DC Hierarchical control structure, the development in primary,
microgrid protection”, ‘‘microgrid topologies”, ‘‘power flow secondary and tertiary control layer as well as energy manage-
analysis in DC microgrid”, as well as all their possible ment strategies in DC microgrid are discussed in section 5.
combinations. Protection schemes is discussed in section 6. Section 7 high-
The relevant literature search was also not restricted to a lighted the challenges of DC microgrid with suggestions on
specific timeframe. Having excluded articles relating to only the way forward. Section 8 contains a general discussion of
AC microgrid control and energy management system from the review. Finally, section 9 brings the review to a close along-
the database to provide a balanced and objective assessment side recommendation for improvement on the existing research
of only DC microgrid control and energy management strat- area. Fig. 2b shows the pictorial representation of the review
egy, the initial results yielded 256 articles in the WOS database methodology implemented in the article.
and 287 in the Scopus database. Subsequently, the data were
merged and then the articles were limited to journal articles 2. DC microgrid topologies
and conferences using certain inclusion and exclusion criteria
as shown in Fig. 2a. One filter is used to exclude article that The choice of an appropriate DC microgrid topology is critical
are not directly related to DC microgrid, articles published because it has an impact on critical aspects of a power system
in books, PhDs, Master dissertation and meta-analysis. Fur- such as flexibility, cost, reliability, controllability, robustness,
thermore, abstract, and full manuscripts that require a sub- resiliency, and scalability. The voltage level is an important

Fig. 2a The literature screen process.

DC-based microgrid: Topologies, control schemes, and implementations 65

Fig. 4 Radial topology.

the load can only receive power if all its interfaces are opera-
tional. Because the system contains other sources and energy
Fig. 2b Pictorial description of the review flow implemented. storage devices, the inability of a generation interface to func-
tion does not always result in a total power outage. Depending
on the purpose and requirements, the radial DC microgrid bus
could be unipolar or bipolar. This configuration is common in
residential buildings because majority of the appliances uses
low voltage, therefore, low DC bus voltage is preferred while
avoiding additional chopper conversion stages. In this config-
uration, loads and AC power converters can be placed nearby
each other to decrease the distribution losses. The system’s reli-
ability can be improved by employing multiple battery stacks
[9,11].

2.2. Multi bus topology

The single line diagram of a multi bus topology is depicted in

Fig. 5, it is an expanded version of radial system topology that
provides increased reliability, efficiency and energy density at
perhaps lower cost of installation and operation [12] compared
to radial system topology. It is characterized by a redundant
feature, as such, it is used in naval ship application. Power
can be exchange between two neighboring group of microgrids
known as microgrid clusters whenever there is a power deficit
Fig. 3 DC Microgrid topologies. or excess. Furthermore, depending on the topology in which
the microgrid was linked, faulty buses can be automatically
isolated from the system in the event of a failure. Power
consideration when designing the topology of a DC microgrid exchange between multi DC busses are controlled by applying
[10]. The available microgrid system topologies are classified suitable voltage deviations [13].
into six different types as shown in Fig. 3.
2.3. Multi terminal topology
2.1. Radial DC microgrid topology
Initially, multi terminal topology was proposed for use in
The concept of radial DC microgrid topology is depicted in direct current at high voltage (DC) systems as in offshore wind
Fig. 4. This type of topology is equally referred to as single power generation. Fig. 6 shows the structure of the multi-
bus structure or a feeder topology. It is characterized by a sin- terminal DC system. Several input terminals feed the DC sup-
gle DC bus and a single point of connection for generation, ply system. An AC-DC converter connects two or more AC
storage, and load in the system. In this type of configuration, grid interface to the DC grid. The flow of power in multi ter-
66 B. Modu et al.

Fig. 5 Multi bus topology.

Fig. 6 Multi terminal topology.

minal DC microgrid topology is more complicated compared 2.4. Ring bus DC microgrid topology
with the conventional radial system configuration. However,
because the system connection allows for multiple power trans- To overcome the constraints of the radial bus topology in
mission paths, it can also be flexible. The system aims to terms of flexibility and fault management, a loop or ring type
achieve balance of power sharing between system units, as well distribution system can be used. The main idea of the ring type
as lowering the scheme’s voltage and frequency deviation. topology is represented in Fig. 7. The suggested solution con-
Voltage source converters are employed to automatically bal- sists of connecting generations, storage, and load units all on a
ance power mismatch if the system available power is less. single ring. Protection switches are placed before and after
The major aspect that should be considered when designing each bus on the ring for safety reason. This implies that each
multi terminal DC system is the protection of the systems as component of the system has two possible ways of connecting
it becomes further difficult when several sources are integrated or disconnecting from the ring along the line on each bus’s left
into the system [13,14]. and right sides. Intelligent electronic device (IED) is employed
DC-based microgrid: Topologies, control schemes, and implementations 67

Fig. 7 Ring topology.

as a controller for each bus, as well as its interface with addi-

tional neighboring buses. Whenever there is a problem with
any bus, intelligent electronic device will first detects the fault
and then isolate the faulty bus and provide an alternate route
for supplying power to consumers [15]. This kind of supply
system was used in cities and industrial settings. This topology
is more dependable than the radial topology because it offers
flexibility to DC microgrid in such a way that when a fault
occurs, the corresponding switches separate it, allowing each
units excluding the faulty-one to continue to function. Because
of its flexibility, ring topology provides high resiliency, reliabil-
ity, as well as redundant operation of the system [16].

2.5. Ladder topology

A ladder topology is designed in the form of a ring, with dis-

tributed energy sources, energy storage units, and loads con-
nected in relation to ring of the laddered topology [17]. Two
buses are connected to the ring of the ladder as shown in
Fig. 8, whichever one can provide DC power to the ladder’s
other rings. In this way, network stability is significantly
improved. This type of system topology has the highest redun-
dancy compared with other type of system topology and that
has made the system capable of eliminating open circuit fault
as well as single point failure and achieve high availability
[11,18]. The topology also offers better fault management Fig. 8 Ladder topology.
because whenever a fault occur, the faulty part alone can be
isolated with the help of the switch and the remaining system age levels, the simulation results of comparison shows that the
continue to normally function [19]. [20] presents a multi-level suggested cascade topology might overcome the weaknesses of
ladder topology for producing a large amount of output volt- other topologies.
68 B. Modu et al.

2.6. Zonal topology 3. Power flow analysis in DC microgrid

In zonal topology, zones are created within the microgrid sys- Direct-current (DC) power flow analysis is a crucial technique
tem each with its protection and load center. To improve the for understanding DC microgrids. It consists of passive ele-
system’s reliability, [21] proposes a zonal electrical distribution ments, active sources, and nonlinear loads. These loads,
system for large ships. The idea behind zonal microgrid topol- which are usually constant power loads (CPLs), bring in
ogy is depicted in Fig. 9. Numerous system elements such as hyperbolic non-convexities into the power flow modelling
generations, energy storage units, power electronic converters [25]. The measurement of voltage at each network node under
and switchgears are contained in zonal DC microgrid configu- steady-state conditions is the primary goal of power flow
ration with the goal of delivering power to a collection of analysis. Notably, since the 1960 s, one of the subjects that
loads. Two redundant DC buses serve each zone, which are has received the greatest attention is the power flow analysis
powered by the grid utility as well as sources of distributed of AC grids. In contrast, power flow analysis (PFA) in DC
DC and AC. The zonal topology provides numerous options grids has only recently been investigated as a result of the res-
for delivering power to the consumer. Electrical energy can urrection of DC grids, which is being driven by advancements
be delivered simultaneously from various buses, consecutively in power electronics, renewable production, and energy stor-
or completely from a single bus. Even though, the power age systems in the active networks paradigm [26]. In the past,
drained from several buses makes distribution system design specialist software such as Digsilent, Matpower, and ETAP
and operation difficult. As a result of this, [22] proposed a was used to solve the DC PF because of its mathematical
bus selection strategy in which power is drawn from the bus and computational complication. However, high purchase
with the maximum voltage level. The load may be able to costs, complicated implementation, and processing times to
change to alternative bus, but that dependent on the required solve the DC PF are issues with the specialist software.
condition. This topology is used for ship boat energy supply Due to these constraints, traditional solution approaches that
[23,24]. don’t require specialized software were developed in the last
From the above discussion on structures of DC microgrid, decade to address the PF in DC networks [27,28]. Further-
it is obvious that each system configuration has its application. more, the non-convexities of the model render analytical solu-
The reliability of the system changes with change in system tions to power problems in DC resistive networks unfeasible.
structure. Table 1 presents the comparative analysis of differ- As a result, numerical methods are required to solve these
ent categories of microgrid structures. problems [25].

Fig. 9 Zonal topology.

DC-based microgrid: Topologies, control schemes, and implementations 69

Table 1 A comparison of different DC Microgrid topology [10,18].

DC Description Pros Cons Inherent Voltage ESS direct Possibility Reliability Application
microgrid Stability choice connection of
(DCMG) expanding
Structure to multiple
buses
Radial Enhances Increase Voltage and 12 V-  High Telecom
topology flexibility of reliability by supply from 48 V system
operation and improving ESS are
DC voltage vibrant uncontrollable.
regulation. stability and
using many
ESS.
Multi Bus Several Proper A microgrid  48 V-  Average LVDC,
topology microgrids are canalization that abruptly 380 V Interfaced
linked in of power shuts down AC Medium
sequence and among can cause a voltage
parallel to microgrid fault in the
enhance clusters. system.
efficiency and
ability for power
allocation.
Multi Many directions Flexible, Power flow is  380 V  High Radial
Terminal of energy flow reliable, and more and configuration
topology increased repetitive complicated above in DCMG
reliability. operation
Ring In fault Two-way Weakness in  24 V  High DC
topology situation, load voltage and and distribution
restructured connection power above system
arrangement approach to
gives improved the common
power flow DC bus.
Ladder Can connect to Scalability is Structure  24 V  Very high In an electric
topology any of the buses high. difficult to and vehicle, a
to deliver design and above multi-
bidirectional control. input/multi-
energy flow. output
converter is
used.
Zonal Different zones Better Appropriate  24 V  Average AC
topology improve reliability by DC voltage and distribution
reliability, and using choice is above system.
each zone is additional required.
energized by switches,
utility grid and islanding
distributed minimizes
sources. error.

3.1. Review of previous research on power flow analysis in DC nificant probabilistic power flow computation cost savings.
microgrid topologies The authors in[25] presented a novel iterative technique that
is based on the upper triangular correlation between nodal
For radial topologies, in [29], an approximated power flow and and branch currents as well as simple impedance matrix to
is suggested based on the Admittance Summation Technique. provides a quick analysis of the power flow problem for
Furthermore, for DC microgrid, an iterative algorithm previ- radial-configured DC networks with constant power loads
ously developed for AC microgrid has been adopted. The sug- (CPLs). When compared to conventional admittance-based
gested techniques were evaluated and shown effective in approaches, this method’s convergence can be enhanced in
comparison to the Newton-Raphson Technique in 33 and terms of the number of iterations and processing times needed
144 nodes in meshed as well as 33 and 906 nodes in radial by being able to avoid inverting non-diagonal matrices. In [30],
DC microgrids. The findings demonstrated that the the idea of equating the multi-bus system to a two bus radial
approaches were created with good accuracy and achieved sig- system and evaluating the required constraints of the equiva-
lent system is used to present a novel approach for assessing
70 B. Modu et al.

Table 2 Overview of power flow analysis methods.

Ref. Load flow techniques Merits Limitations
[38] Newton-Downhill & Does not depend on the initial solution. & order of convergence less than two.
& greater rate of convergence than Newton & if the Jacobian matrix is single, it fails.
Rapson (NR).
[39] Current Injection Method & Even under heavy load, there is good & If the PV DG number becomes too high, the system
convergence. fails.
& less vulnerable to the R/X ratio.
[40] Back/Forward Branch & Does not require Jacobian matrix. & Disable for networks with meshing.
Impedance Based & Linear back/forward sweep equations. & Excessive computation in large scale networks for a
high PV DG number.
[41] Backward/Forward Branch & Does not require Jacobian matrix. & Ineffective for large-scale networks with heavy
Current/Power Based & Does not depend on PV DG number. loads.
[42] Hybrid Super Position/ & Does not require Jacobian matrix. & If the PV DG number becomes too high, the system
Gauss-Seidel fails.
[43] Genetic Algorithm (GA) & easy to implement. & Large scale networks have excessive computation
based power flow & Reliable in convergence. time.
& Suitable for offline problems. & Sensitive to controller parameters of GA.
[44] Particle swamp optimization & Reliable in convergence. & Slower convergence than all the others except GA.
(PSO)-based power flow & Suitable for offline problems. & Large-scale networks have not been successful.
& Faster than GA.
[45] Artificial neural network & It has the shortest computation time. & restricted to a specified set of inputs.
(ANN) based power flow & Appropriate for online problems. & Needy to the other techniques.
[46] Probabilistic power flow & Critical uncertainties that have an impact & Computational cost is high.
analysis method on MG power flow are identified. & Need to solve a system of nonlinear equations for
different scenarios of microgrid operation.

voltage stability while taking economic criteria into account. A technique is linearly related to the number of substations. In
203 bus WBSEB and a 6-bus system as a voltage stability indi- [36], a power flow approach based on loop analysis theory
cator has been used for the implementation of the method. The and its linear formulation for DC microgrid is proposed. In
authors in [31] proposed a power flow analysis based on a this method, the branch powers, instead of the injected cur-
branch-and-bound search to discover the best switching status rents, are used to construct the suggested iterative approach
for transmission lines, while mixed-integer programming and its linear approximation. When compared to the tradi-
(MIP) relaxation is used to deal with non-convexities of power tional Newton-Raphson approach and two other current lin-
flows and converter loss equations. To get the optimal power ear methods for both radial and meshed networks,
flows inside the offshore network, an evolutionary technique simulation results showed that the proposed methods were
known as covariance matrix adaptation (CMA) is presented more accurate and faster. In [37], the power flow issue in
in [32], To evaluate the effectiveness of the control approach low-voltage direct current (LVDC) power grids is reformulated
in various case studies, the optimal power flow outcomes of for radial and mesh configuration using Taylor’s series expan-
the CMA algorithm are examined in a dynamic simulation sion. The suggested method avoids the usage of iterative
model. A complex affine arithmetic-based alternating AC/ approaches by offering an explicit solution to the system of
DC power flow algorithm for the zonal MVDC shipboard power flow equations. In Table 2, the load flow techniques
power system in the presence of power variation is proposed are highlighted with their merits and limitations.
in [33], to solve AC/DC power flows, a sequential method
based on affine analysis is used. The suggested method can 4. Control strategy in DC microgrid
be used to design, plan, and operate MVDC power systems
online when making decisions that are affected by power vari- An effective control strategy should be employed for a DC
ation. The planning of an urban distribution network using a microgrid system’s well-organized operation and stability.
2-step ladder architecture and taking joint nodes into account Converters are critical components in the operation of DG
is presented in [34], a new network configuration using a mixed microgrids as they ensure proper load sharing and harmonized
integer linear programming (MILP) method is adopted herein. interconnections between different units of DC microgrid. In
The model includes conventional electrical limitations such DC microgrid, various components are linked in parallel.
power flow balance, line thermal limit, and number of con- Hence, voltage regulation that is flexible, precise current and
sumers per feeder. Ladder circuit modelling of ungrounded power sharing between parallel-connected inverters ought to
DC rail traction system load flow numerical solution is pro- be all achieved. Control structure has become very important
posed in [35], by successively obtaining the PFA of individual because of increased generation and loads that are not linear.
substations and wayside ladder circuits, it will be possible to Therefore, the objectives of the DC microgrid control are enu-
efficiently compute the load flow and achieve numerical resili- merated as follows [18,47,48]; Effective voltage and current
ence for the entire dc traction system. In comparison to nodal control in grid linked and autonomous operation modes, com-
voltage equations, the computing overhead of the proposed parative load balancing, steady operation with continuous
DC-based microgrid: Topologies, control schemes, and implementations 71

power and loads that are not linear, synchronization amongst for DC microgrid on a small-scale application where the
different distributed energy resources and energy storage amount of data to be collected is limited [51,52]. The mas-
devices, organization of DC microgrid and utility grid, DC ter–slave relationship is a common illustration of centralized
microgrid power flow control, and network utility (if linked control in which one of the converters serves as a master–slave,
to the grid), in addition to smooth changeover from network while the others serve as slaves. The master converter is static
connected to islanded mode. and operates as voltage mode regulator in a devoted master–
DC microgrid control are characterized into two segments; slave, whereas the slave converters operate as current mode
Basic control strategy where the main element is communica- regulator [53,54]. Figs. 11 and 12 depict the architecture of cen-
tion and it is performed in the following different method tralized and master–slave control scheme respectively. A cen-
namely, centralized, decentralised or distributed control as well tralized power flow in a DC microgrid control technique
as hybrid control, and the Multilevel strategy which is carried connected with an EV was proposed by [55]. The operational
out via various control stages in a hierarchy. Fig. 10 shows the mode of every agent in the DC microgrid was defined with
diagram of microgrid control structure. respect to the information concerning the status of EV connec-
tion or disconnection, the first value of EV state of charge,
4.1. Centralized control wind power source, batteries SoC level and the availability
of grid. [56] Presented a centralized robust controller in the
In the centralized control scheme, a central regulator controls case of an autonomous DC microgrid. The controller was
the distributed sources. The central controller compares the designed to be resistant to the DG units’ parameters uncertain-
total generation and loads and then send load shedding signals ties and to reduce exogenic instabilities. Control and manage-
to the loads that are not critical if the generation is not enough ment that is effective for a grid connected and autonomous DC
to power all the loads. The centralized controller can manage microgrid was proposed by [57] to address a functional distri-
various energy sources in order to meet the demands of both bution system’s voltage regulator and energy coordination.
critical loads and loads that are not critical [49,50]. This strat- [58] suggested a robust controller for an islanded DC micro-
egy offers strong supervision in the form of simple application, grid based on communication network delay where a steady-
real-time system controllability, as well as observability. The ing, robust and interpreter-based controller to ensure the
distributed unit’s data is being processed and returned via dig- time delay DC microgrid system’s stability, it was designed
ital communication links. Communication is the prime mover using the Lyapunov-Krasovskii stability criterion. A cyber-
in centralized control strategy. However, the centralized con- attack detection and mitigation system for the control of dis-
trol strategy has limitations for example, the system’s single tributed voltage of critical loads in DC microgrids was pre-
point of failure, decreased flexibility, reliability, and scalability. sented [59], for each DER here, a distributed voltage and
Consequently, this category of control strategy is appropriate current controllers were employed to perform the voltage reg-
ulation. A communication chart is used to share each DER

Fig. 10 DC microgrid control structure.

72 B. Modu et al.

Fig. 11 Centralized control structure.

compensation, MPPT, reduction of ES degradation and con-

straint for power interchange with the utility grid.
Fig. 13 shows a block diagram of a centralised control
scheme proposed in [61]. The voltage VMG at the point of com-
mon coupling (PCC) is regulated at its rated reference voltage
V , the voltage at the PCC and the secondary level correction
term dV communicate between the central unit and the local
converters.

4.2. Decentralized control

Local controllers control the microgrid’s connected converters

in a decentralised control scheme. The input signal to the con-
trollers are locally measured signals which are managed locally
to produce gate pulses for converters. Fig. 14 depict a single
line diagram of a decentralised control structure. Droop con-
trollers are commonly used in decentralised control schemes.
The droop control has been implemented to achieve decentral-
ized control of each unit; however, it suffers from limitations
such as the current sharing, system accuracy, and system sta-
bility which highly depends on the droop parameters
[49,62,63]. For instance, high droop coefficient gives rise to a
further damped system and increased precision in current shar-
ing. Conversely, DC voltage deviation also increases with
higher droop parameters. [64] proposes a vertical battery-
centered droop control for effective power dispatch. Time-
varying parameters of energy storage system (ESS) was used
Fig. 12 Master slave control strategy. to calculate the droop controller’s virtual resistance and refer-
ence voltage. [65] Proposed a decentralized control module-
current and voltage information. [60] also proposes a power based plug and play (PnP) microgrid. In this control scheme,
flow control algorithm for centralized regulator in DC micro- only local information is required for the control of each con-
grids. The control algorithm guarantees DC bus voltage ripple verter whereas SoC balancing, sharing of power and mode
DC-based microgrid: Topologies, control schemes, and implementations 73

Fig. 13 Block diagram of a centralised control scheme.

was proposed by [68]. With uncoupled control loops, even

when the generation is dominant, the BES can still provide
dynamic compensation. [69] suggested an even-triggered
decentralized coordination control strategy in islanded
electric-hydropower based DC microgrid by developing a bet-
ter cost-effective control scheme based on the proposed struc-
ture’s cost function making the energy storage system to carry
out the matching scheme aimed at cost-effective and consistent
operation. [70] proposed a control scheme for DC-DC con-
verter parallel operation in a DC microgrid application. The
technique combines droop control and an algorithm for
instantaneous distribution of current to reduce current sharing
inaccuracy. An artificial neural network based decentralized
controller to identify and eliminate the attacks on current mea-
surement using erroneous data injection was proposed by [71].
In each unit, a neural network was implemented to evaluate
the DC current output of each converter, and a PI controller
was used to eliminate the erroneous date injected from the
attacked unit based on the assessed value. Fig. 15 shows a
block diagram of a decentralized control scheme proposed in
[72], The reference current (Is ) in the proposed control scheme
regulate the duty circle of the converter and the reference cur-
Fig. 14 Decentralized control structure. rent for the hysteresis current (Iref ) control is the output signal
of the voltage droop control. The reference current and power
of the i th power converter in the proposed scheme is expressed
acceptance can be accomplished concurrently without commu- as
nication. [66] proposed a decentralised coordinated control 
 
wLpf;i Vo;i
strategy for PV, ESS, fuel cell, and HPU in a DC microgrid Iref;i ¼ Gi ðsÞ Vo;i  Vo;i ð1Þ
system. The battery control scheme took into account the s þ wLpf;i Vs;i
SoC as well as the rapid charging and discharging power, 8  
droop control for PV MPPT mode, a PHU efficiency adaptive < Gi ðsÞ ¼ 1
1 þ 1
rd;i sTi
control scheme was developed, and the fuel cell was operated ð2Þ
: Ti ¼ wLpf;i
4
;
in constant power mode. [67] presented a decentralized droop
control with filter in other to achieve state of charge matching
in a DC microgrid by considering different SoC and capacities Pi ¼ Iref;i Vo;i ð3Þ
between battery energy storage (BES). A decentralized control Where Gi ðsÞ, Ti and wLpf;i is the transfer function, integral
strategy for a PV-BES management control to overcome volt- constant, and cutoff frequency of the low pass filter
age regulation control problems in the PV mode of dominance respectively.
74 B. Modu et al.

Fig. 15 Block diagram decentralized control structure.

4.3. Distributed control cussed the issue of network topology, robustness of network
topology in the event of link failure, communication delay sig-
This control scheme has features found in both central and nals and the role of information interchange. [74] proposed an
decentralized control topologies, including an external com- innovative method of distributed power control for converters
munications link but no central controller. The distributed interconnected in an autonomous power microgrid with objec-
controllers communicate with one another via digital commu- tive of implementing power sharing between distributed gener-
nication contacts, primarily a low bandwidth communication ators and the interconnected converters. [75] developed a
network or a power line communication network for exchang- distributed controller for islanded microgrid which provides
ing data such as DG output current, bus voltage, and so on. a data driven multi-agent framework for the regulation of volt-
Distribution control combines the advantage of both central- age and current sharing. An even-triggered distributed sec-
ized and decentralized controller, Fig. 16 depicts a single line ondary technique for radial DC microgrid is presented in
diagram of distributed control scheme, communication is lim- [76] where converter can individually choose when to transmit
ited to units that have access to a digital communication line a signal to its neighbor on a local basis. Bus voltage control
(DCL). Local controller controls each unit that is connected and equal current distribution is accomplished. A unified hier-
to the microgrid. Each unit’s controller exchanges information archical control method for power flow in autonomous DC
with its neighbors via the limited DCLs accessible. [73] has dis- microgrid collections was proposed in [77] and a distributed

Fig. 16 Distributed control structure.

DC-based microgrid: Topologies, control schemes, and implementations 75

communication based unified hierarchical is employed to real- trol methods are combined as depicted in Fig. 18. At the
ize the objective. [78] proposed a distributed-based control source level, a decentralized control approach is employed
method for voltage control and accurate sharing of current for instantaneous sharing of power making the central con-
in DC-DC converters that are parallelly connected. The pro- troller’s function to become one that is centralised coordina-
posed controller does not depend on the number of active tion rather than control of every connection in real time.
and load profile [79]. With the integration of centralized control to the decentralized
Fig. 17 shows a block diagram of a distributed control controller, it is possible to implement supply side control law.
scheme proposed in [80]. The performance of the power con- The main limitation in central control strategy is still main-
verter is determined by the input current and output voltage tained in hybrid central control because of the presence of
signal of the power converter. The reference current signal in communication line. Redundant communication lines and
the proposed control scheme is given by the following equation controllers needs to be provided in hybrid central control tech-
 Vo;i nique in other to improve the reliability of the system and
Iref;i ¼ Gi ðsÞ Vref;i  GLpf Vo;i ð4Þ doing so, will incur additional cost [99]. A hybrid central con-
Vs;i
trol strategy for charging of electric vehicle (EV) is proposed
8   by [100] in which model predictive control along with adaptive
>
> Gi ðsÞ ¼ r1d;i 1 þ sT1 i
>
< charging method was employed to define close to ideal electric
 
w
ð5Þ vehicles charging profiles instantaneously, while the decen-
Where
> GLpf ¼ sþwLpfLpf
>
> tralised controller coordinates the interaction among the elec-
:
Ti ¼ wLpf4
tric vehicles as well as charging system controller. [101]
suggested a hybrid central control method for residential
DVo;max power distribution to improve the voltage quality. Decentral-
rd;i  ð6Þ ized controllers maintain the voltage at the point of connection
irated
o;i
between the EVs and PV-wind system, then the energy storage
Where rd;i is a droop gain, wLpf is cutoff frequently of low- is remotely controlled by central controller to further improve
pass filter, and DVo;max is the maximum DC bus voltage the voltage.
deviation.
The summary of key outcomes in different control tech- 4.5. Hybrid distributed and decentralized control
niques is critically presented in Table 3. The symbols in col-
umns three and four have adequate (+), moderate (++), This is a hybrid of distributed and decentralized control tech-
and good (+++) denotation. Column five denotes complex niques, in which the control task is dispersed among the source
(+), moderate (++), and simple (+++). The design com- controllers, as in distributed control. This type of control
plexity measure is concern with the mathematical models scheme has no external communication link; instead, the sys-
required for the control algorithm’s implementation. tem power bus is used for communication among the source
controllers. Therefore, decentralized controller efficiently con-
4.4. Hybrid decentralized and centralized control trols each source by utilizing local variables. Fig. 19 depicts a
typical hybrid distributed control structure. In a hybrid dis-
Hybrid decentralized centralized control strategy is a form of tributed and decentralized control strategy, the benefits of both
multilevel method where a centralized and decentralized con- control methods are combined. When there is communication

Fig. 17 Block diagram of distributed control scheme.

76 B. Modu et al.

Table 3 Critical analysis in different control techniques.

Control Technique Current Voltage Design Characteristics Application
structure Sharing control complexity
Centralized Phase shift [81] ++ +++ ++ More adaptable, Tolerant to faults HVDC
application
PI Controller[82] ++ +++ +++ Robust, Improved efficiency but cannot handle LVDC
nonlinearity application
Decentralized Conventional PI + +++ +++ Poor disturbance rejection and cannot handle DC system
Controller[83] nonlinearity
Conventional droop +++ + +++ Deteriorate the proper power sharing LV
control[84,85] performance by owing the effects of unequal line microgrid
resistance.
Adaptive-Droop +++ ++ ++ Robust, Integrate RES such as SOFC, PVs and DC
Strategy[86] BESS microgrid
system
Model predictive based +++ + + Requires accurate dynamic models with high Microgrid
control (MPC)[87,88] computational cost and it suffers from lack of system
flexibility.
Sliding mode control +++ +++ ++ Excellent robustness but chattering is high due to DC system
[89,90] switching frequency.
Nonlinear integral ++ +++ + Reduced robustness DC
backsteppingSliding microgrid
mode control[91]
Second order sliding +++ +++ + Reduced robustness and chattering not DC
mode control[92] eliminated microgrid
system
Distributed DC bus signaling[93–95] + + ++ Deviations in voltage and current DC
microgrid
Adaptive droop control +++ + +++ The resistance between interconnected converter LVDC
[96] terminals should be known in advance, and weak microgrid
convergence grid
Discrete time sliding +++ + ++ Very weak robustness DC
mode control[97] microgrid
grid
PI Control with +++ ++ + Plug-and-Play compatibility, comparative load LVDC
Adaptive Droop Control sharing microgrid
[98] grid

line between the sources, it is possible to implement control respectively. This approach has shown to be robust in the pres-
law for the supply-side and the system being not dependent ence of both communication delay and some communication
on the central controller as well as the external communication failure scenarios. A hybrid distributed leader–follower control
link has made it to retain the reliability advantage in decentral- scheme for cascaded parallel microgrids is presented by [105],
ized control scheme. However, the limitations in distributed where one distributed generator act as a leather and the
control method are still maintained in hybrid distributed con- remaining distributed generators act as followers. Distributed
trol strategy. Hybrid distributed control employed in power control and decentralized control were employed to coordinate
line carrier (PLC) system based industrial control scheme to and control the leader and follower generators respectively.
provide communication amongst control nodes [102]. Though, Improved voltage quality and synchronization of frequency
in other for modulation and demodulation of the control sig- were achieved with low communication bandwidth at low cost.
nal to occur, extra control hardware is needed at every node [106] also recently suggested a hybrid distributed control
in the system. The use of power line as transmission intermedi- scheme where local power converters connected in series are
ate has also created signal attenuation and limited bandwidth controlled by distributed controller and the decentralized con-
issues. Nevertheless, power line carriers can still be a reliable troller handles the power sharing between parallelly arranged
control scheme for small system if perhaps it is possible to string with low communication bandwidth.
solve these problems [103]. In order to accomplish accurate In summary, there is an improvement in the two hybrid
sharing of current and improve voltage restoration, a hybrid control methods compared with the performance of individual
distributed and decentralized control strategy for a DC micro- control strategy. In hybrid centralized and decentralized con-
grid was proposed by [104]. Decentralized and distributed con- trol, for example, the advantage in centralized control is
trol strategies were implemented to accomplish enhanced retained at a lower cost because the power sharing is done
voltage restoration along with precise power distribution by the local decentralized controller. Likewise in hybrid dis-
DC-based microgrid: Topologies, control schemes, and implementations 77

Fig. 18 Hybrid decentralized centralized control.

Table 4 Summary of comparison between different control

topologies.
Topology Advantages Disadvantages
Decentralized No external Control law for supply
communication line side law cannot be
and controller, it is implemented
reliable and simple
Centralized Potential to offer best It cannot be
performance implemented practically
because real time central
control of a large multi-
input multi-output
(MIMO) system
necessitates far more
computing power than is
currently available.
Distributed Enhanced reliability External communication
compared with central line is needed
control
Fig. 19 Hybrid distributed centralized control. Hybrid Offers similar Dependency on central
decentralized performance as central coordination and
and controller at low cost communication
tributed and decentralized control strategy, it has the advan- centralized
tage of distributed controller, no central controller or commu- Hybrid Reliable and simple The size of a double
nication line. A comparative analysis of all the three basic distributed with no external direction line carrier
control and the two hybrid control techniques is presented in and controller or system is limited, and
Table 4. decentralized communication there is possibility of
requirement transmission line
5. Hierarchical control structure impedance to corrupt
the information

The modern power system which includes microgrids are very

complex, and therefore, high intelligent control strategy is
often required to accomplish the basic objectives like voltage,
current, and power control, in addition, advanced control that in energy market. Consequently, a centralized or any level
includes sharing power among distributed generations, provid- decentralized control strategy cannot control the entire system.
ing auxiliary services, reduced cost of operation, and partaking Microgrid hierarchical control consists primarily of three con-
78 B. Modu et al.

trol stages: primary, secondary, and tertiary levels according to signaling method to improve service reliability in DC micro-
[93,107–119]. Hierarchical control as depicted in Fig. 20, is grid by delivering a power supply to multiple possible loads
intended to control several energy storage devices (ESDs) not considered critical whilst ensuring uninterruptible power
and distributed generations, renewable energy resources and supply to loads that are considered critical. DC bus signaling
loads. The main local controller is the primary control. It uses is used in a power management scheme for an autonomous
local measurement to reestablish DC bus voltage and guaran- DC microgrid is reported by [95] where the DC bus voltage
tee concurrent power sharing in both steady and dynamic play a role as an information carrier between an interlink bidi-
states on the microgrid’s lowest level. Primary control is faster rectional converter and distributed generation controllers.
and the voltage variation resulting from the primary controller [123] proposed a coordinated control method for a DC micro-
is balanced by implementing secondary controller, which has a grid using DC bus signaling, an appropriate voltage require-
higher response time compared to the primary controller. Ter- ment for a microgrid system is developed and then based on
tiary control is the highest level of control, which is character- the real data acquired from meteorological agency through
ized by slow response time. The operations of different units National Renewable Energy Lab (NREL), the correlation
within the microgrid and grid utility is optimally maintained between wind and solar energy was evaluated. However, volt-
by the tertiary controller [120,121]. age variation is a major obstacle in DC bus signaling technique
[49]. A fuzzy logic controller is proposed in [124] to regulate a
5.1. Primary control bus voltage while maintaining power balance in a DC micro-
grid comprising of PV, fuel cell a battery bank. To find out
Power electronic converters are essential components in DC the efficiency of the proposed method, the system has been val-
microgrid that provides a controllable interface the sources idated under different unpredictable conditions such as 50 %
and load. In a multi-level control system, the primary stage changes in AC and DC loads, variation of irradiance and the
of control is the initial stage of control architecture and is in function of the battery bank when a change is observed in fuel
charge of voltage and current control. Inner loops for voltage cell.
and current control, as well as droop control for preliminary In the primary level of control, droop control is the most
power sharing, comprise the primary control. DC bus signal- commonly used control strategy. It is preferred over the tradi-
ing, fuzzy logic, master-level and Droop control is one of the tional techniques because it is a decentralized control and
most widely used primary level control strategies. The voltage hence, it does not require any form of communication.
operation range in DC bus signaling is divided into pre- Voltage-Current droop strategy is achieved by linearly mini-
specified sections, namely operation mode or voltage level. mizing the output voltage while output is maximized. This
DC voltage is employed as a method of transmitting data in operating characteristic has made the droop technique suitable
the DC bus signaling for power sharing between distributed for primary control to accomplish the independent operation
generations in DC power system. DC bus signaling can achieve of converters connected in parallel and PnP capability partic-
the independent control of each converter connected in parallel ularly in the absence of communication [125,126]. However,
with a particular pre-indicated control rule. [122] uses DC bus this method is characterized with shortcomings such as voltage

Fig. 20 Hierarchical control structure.

DC-based microgrid: Topologies, control schemes, and implementations 79

variation and inaccurate current sharing and as such, some

researchers have suggested a nonlinear control to overcome
these problems in [127–140]. A non-linear dynamic droop con-
trol is also suggested in [141] where power is distributed among
battery and supercapacitor by employing dynamic droop gain.
The method has outperformed the normal droop scheme.
Fig. 21 shows the schematic diagram of primary drop control
where the converters are arranged in parallel and share a com-
mon load with the aid of output virtual resistance (RO). In the
case of any voltage deviation, the output current will circulate
among the two sources. Primary control technique is used to
minimize the circulating current by employing virtual output
resistance, (RO). Reference voltage is altered by primary level
control and sent to inner control loops through virtual output
resistance.

5.2. Secondary control

This is another level of control in a multi-level control scheme

Fig. 22 Block diagram of secondary control scheme.
that is accountable for the restoration of DC bus voltage devi-
ation produced by the actions of droop control in the primary
control level to a nominal value and power quality enhance- from the secondary control. This improves the primary con-
ment. Fig. 22 shows a block diagram of secondary control troller’s output during current sharing. The secondary control
scheme presented in [9]. In the case of grid connected mode, layer can be classified into two categories according to
it is also responsible for the microgrid and grid utility synchro- [142,144,145] namely, centralised or distributed secondary con-
nization by ensuring that the distributed energy resources units troller which depends on local measurement and the central-
within the microgrid are operated in an optimal and coordi- ized controller that depends on communication. In the case
nated manner [142]. Secondary control act as a moderator of centralized method, the DC link voltage is measured and
between the tertiary and primary level which adjusts the mis- regulated by a controller [113,146], but implementation of cen-
match between the set of values measured at primary level in tral voltage regulator requires communication link between the
microgrid and the values provided in tertiary level [114]. Sec- converters and the central controller and that affects the com-
ondary level control is employed to achieve a low cost of oper- plete reliability and stability. To enhance the entire reliability
ation, reliable and a secured microgrid in both independent and stability, a decentralized controller is employed in [147–
and grid connected mode of operation [143]. Microgrid central 150] because it is relatively less complicated and with no com-
controller performs the conventional secondary stage control munication required as it operates local control task based on
based on low communication bandwidth (LCB). The local local measurement of individual converter. Distributed sec-
controller receives a reference point for voltage and current

Fig. 21 DC microgrid primary control diagram.

80 B. Modu et al.

ondary control does not require central controller and there is caused by the uncertainties in power supply from the sporadic
link of communication between neighboring converters. Dis- nature of renewable energy sources (RES) based distributed
tributed secondary approach is presented in [151–154]. It has generations (DGs) and the dynamic nature of electricity uti-
the advantage of maintaining full functionality whenever there lization. The role of energy management system (EMS) in
is failure in some communication lines, and it also overcome ensuring the stable operation of microgrid cannot be overem-
the problem of single point failure in centralized control phasized. Maintaining a smooth operation within a microgrid,
approach [113]. minimizing energy utilization, improving energy efficiency and
end users’ wellbeing are the principal objectives of EMS
5.3. Tertiary control scheme [163]. An EMS must therefore be connected to each
component of the system within a microgrid for control pur-
The tertiary control strategy is the highest and last hierarchical poses, as shown in Fig. 24. EMS can be used in a variety of
control level in microgrid that manages the power quality of control technique such as the decentralized, centralized, dis-
the system and operate multiple microgrid connected with each tributed as well as multilevel control strategies. The control
DG unit of the system [111,155–157]. Fig. 23 depict a block architecture and the adopted approach of solution of EMS
diagram of a tertiary control scheme proposed in [61]. Even determines the efficiency of EMS. Microgrid EMS performs
though the microgrid size is extremely small compared with many functions including forecasting of distributed energy
conventional grid, It is necessary to have efficient economic resources power generation, consumption of load, analyzing,
dispatch and power flow which can be accomplished through price of energy in the market, auxiliary market price and
the use of tertiary control approach [158,159]. Optimum oper- weather condition [164]. The integration of EMS in a building
ation of all microgrids is accomplished by employing tertiary is essential as the control and optimization of energy usage is
control strategy. It can be implemented in centralized or decen- needed. The primary consumption and energy bills in a build-
tralized approach too. Part of tertiary control strategy includes ing can be minimized by applying real-time energy manage-
ensuring appropriate load management between the microgrid ment. The application of intelligent energy management
in a cluster, proper transfer of power between microgrids system in microgrid (both autonomous and grid connected
which allows for maximum use of renewable energy resources mode) is also needed as it ensures optimal energy consumption
with high reliability and accessibility, communication, and and reduced cost [165]. The incorporation of renewable energy
operation [94,121]. Issues such as surplus production of power sources, as well as an energy storage system, EVs to grid with
compared with the demand of microgrid, demand of microgrid high power requirement of buildings and many more has made
with respect to voltage and frequency in the case of grid con- EMS to become more complex. However, machine learning
nected mode, the price at which the microgrid is ready to trade and internet of things (IoT) are simultaneously increasing pop-
in the energy market and the issue of reserve (spinning and ularity and they are mutually very useful in the effective imple-
non-spinning) must be taken into consideration while evaluat- mentation of the EMS in the network [166]. Inappropriate
ing microgrid economic status as a project [160]. Table 5 shows coordination of microgrid integrated with EVs, RES, and
the comparative analysis between primary, secondary, and ter- ESS might cause the system to deviate from its expected per-
tiary control strategy in hierarchical level [9,161,162]. formance. For instance, the charging and discharging of EVs
in an uncoordinated approach could lead to thermal stress
5.3.1. Energy management system on the distribution network, likewise the integration of unco-
ordinated RES might cause reactive power problem [167].
In a broad sense, an energy management system (EMS) is a
EMS objective can be single or multiple, for example, it can
system that monitors and controls various aspects of energy-
be economic, technical, technoeconomic, environmental or
related systems such as measurement, operation, scheduling,
social economic. Technical constraints are considered while
and optimization. One of the most difficult tasks in microgrid
establishing an economic objective, otherwise, the EMS opti-
application development is decreasing large energy differences
mization might deliver the best economic result, but the system

Fig. 23 Block diagram of tertiary control scheme.

DC-based microgrid: Topologies, control schemes, and implementations 81

Table 5 Microgrid control function by hierarchical layer.

Hierarchical Time scale Use of Advantage Disadvantage Application
layer communication
Primary Milliseconds- Absent Because it is decentralized, it is preferred Voltage variations and DC-DC
control seconds over other traditional methods. inaccuracies in current converters
distribution
Secondary Seconds- Present Global voltage regulation is provided, Single point of failure problem Converters of
control minutes-one and voltage deviation is eliminated. in centralised approach all kinds
hour
Tertiary Minutes- Present There are fewer power flow losses and Single point of failure problem Power and
control hours-days operating costs. in centralised approach energy
management

Fig. 24 DC-based microgrid EMS.

may be exposed to brownout, or damage the equipment oper- grid structures. As indicated in Fig. 25, there are two main
ating under the distribution network. Environmental objec- types of commonly used approaches and procedures for the
tives with a view to reduce greenhouse gas emission (GHG) energy management of microgrid systems: classical techniques
are set when several fossils fuel-based generation and RES and artificial intelligence techniques.
are programmed to accomplish low carbon footstep genera- The energy management of a DC-based microgrid has only
tion. Social economic-based events are considered while eco- been studied in a limited number of cases using classical tech-
nomic events are influenced by the social development [168]. niques. The majority of research is geared toward optimizing
In the literature, several energy management techniques the size of standalone hybrid renewable energy systems
have been put out to ensure the most optimal and effective (HES). The next subsections provide an overview of certain
running of microgrids. Table 6 summarizes several research frequently used classical approaches as outlined in Fig. 20
conducted in energy management system with different micro- respectively.
82 B. Modu et al.

Table 6 Critical analysis of EMSs based on classical and intelligent technique.

Reference Optimization Contributions Limitation Weaknesses Single/
technique Multiple
objective
[170] Iterative Novel learning-based and iterative IoT large-scale grid-connected Renewable energy Multiple
algorithm system for connected building BEM buildings and intelligent EV generation and ESS
charging in cities costs are not taken into
account
[172] Linear A heuristic algorithm based on linear Not specified High computational Single
programming programming for EVs charge and discharge difficulty
[185] Linear and A novel’ mix-mode’ energy management Battery The optimization model Multiple
mixed integer strategy and method of sizing batteries for Cost of generation did not consider battery
linear low cost microgrid operation degradation cost
programing
[174] Hybrid The improved HSR optimization method Power balance Complex mathematical Multiple
Stochastic- addresses renewable energy, load, and cost Renewable energy formulation
Robust of electricity uncertainties generation load
optimization
[186] Stochastic A simplified technique of integrating the Renewable power Modelling of battery Single
effect of scheduling in autonomous mode on generation load ageing is not considered
the grid-tied operation Balance of power The cost of DG’s
emission is not
considered
[176] Stochastic A stochastic nonlinear model for a hybrid The dynamics of FC and Batteries’ charging and Multiple
model fuel cell-energy storage-cold ironing Ship ESS operation at short time discharging rate were
predictive that takes into account the effects of intervals not taken into account
control weather and resistance of wave uncertainty Power loss not
on the ship considered
[187] Model The daily optimization that takes into Not specified Models using lithium Multiple
predictive account the batterie’s deterioration due to batteries are excluded
control corrosion and capacity losses
[188] Fuzzy logic The proposed strategy does not require Energy storage not The state of the battery Single
complex model between the source power considered and distributed sources
and converter output voltage to design the are not specified
controller.
[178] Fuzzy logic Easy implementation, enhanced the grid Battery charge and The only controllable Multiple
Quality power profile discharge component is the
battery charger and
grid-tied converter
Battery deterioration is
not taken into account
[189] Artificial Reliable sensor less control and self- Limited number of hidden Nonlinear controller
intelligent contained is implemented with less sensors layers not considered
(Neural without communication delay
Network)
[184] Multi-agents Automatically extract features, improved Modelling of the microgrid Control scheme Multiple
computing effectiveness and making it configuration in the energy complex
applicable to real-time application. management strategy not
considered
[190] Multi-agent An effective technique for managing energy Battery charge and Battery ageing Multiple
storage in real-time to best account for discharge prediction is not taken
power imbalance Power balance into account
Load scheduling
[191] Multi-agent Bayesian belief update which requires no Scalability of the algorithm Complex formulation Multiple
Bayesian direct communication and allow each agent
reinforcement to update itself on the activities of its peers
learning in accordance with the Bayesian rule

5.3.1.1. Energy management based on classical technique. Itera- energy system (HES) was optimized using an iterative
tive algorithms are used in the literature to achieve various approach to reduce the system’s overall energy cost for four
objective functions. In [169], an autonomous photovoltaic- different areas in Iran. In [170], an iterative algorithm and
wind turbine-fuel cell-battery storage-hydro turbine hybrid IoT-based application for self-scheduling energy storage sys-
DC-based microgrid: Topologies, control schemes, and implementations 83

Fig. 25 Classification for energy management techniques.

tems (ESSs), thermal energy storage systems (TESSs), and con- to the simulation results, a well-coordinated optimum power
nected building loads to maximize PV energy usage on-site, management approach for local distributed generation
flatten the load profiles of buildings, and reduce electricity (DGs), storage system (SS), and active loads in the MG
costs. Deep reinforcement learning (DRL) and a physics- improves network indices such as voltage profile, microgrid
based method were combined to first learn a good thermal line, and station overloading. The authors of this paper [176]
storage policy and then use an iterative technique to generate proposed a nonlinear model for optimum energy management
the best control commands for all connected buildings’ of emission-free ships using hybrid fuel cell-energy storage
remaining loads. A rule-based-Iterative EMS has been pro- system-cold ironing as sources while taking ageing factors of
posed in this paper [171] for an autonomous photovoltaic- the fuel cell and storage system into account. The problem
tidal turbine-diesel generator-battery storage microgrid in objectives are total operating costs, fuel cell and energy storage
order to achieve the optimal PV-TT-Battery size while mini- ageing factors. In addition, a stochastic model predictive con-
mizing energy costs, greenhouse gas emissions costs, and losses trol technique is tailored to the model in order to account for
in power due to static power converters. This paper [172] pro- uncertainties in the course of optimization horizon. The simu-
poses a linear programming-based algorithm for charging and lation results demonstrate the proposed model’s efficiency in
discharging electric vehicles. To deal with uncertain electric meeting the ageing constraints of the fuel cell and energy stor-
vehicle demand and departure times, as well as to evaluate age system while minimizing operation costs.
the effect of smart charge and discharge scheduling of electric
vehicles on peak load reduction in a building energy manage- 5.3.1.2. Energy management based on artificial intelligent
ment system, an improved two-level heuristic algorithm was technique. Intelligent technologies like IoT and AI are being
also developed. The authors in [173] proposes a mixed integer linked into renewable energy systems to develop them by mak-
linear programming-based energy management system (EMS) ing them more robust and responsive. All renewable energy
for microgrid storage systems. The charging and discharging (RE) systems are now using AI technologies. It has been
power of the battery is established in such a way that the over- demonstrated that the application of AI in RE systems has
all cost of energy consumption is reduced, taking into account made technologies such as photovoltaic panels, wind turbines,
tariff variations in the power grid, renewable energy genera- and solar PV systems more affordable and efficient, increasing
tion, as well as load requirement. The study then employs their use and perhaps displacing plants that burn fossil fuels
real-world data, and according to the simulation results, the [177]. A more thorough analysis of these research is provided
proposed real-time approach outperforms the offline optimiza- in this paragraph. A fuzzy logic-based EMS for a grid-tied res-
tion approach, lowering operating costs by 3.3 percent. To idential microgrid to obtain a smooth power profile is sug-
take care of the intermittencies from renewable energy, uncer- gested in [178]. The suggested model reduces peaks and
tainty about load and electricity prices, the authors of [174] oscillations in the energy transfer to the primary grid. Addi-
proposed a distributed stochastic-hybrid optimization method tionally, it keeps the battery’s SoC level at roughly 75 % of
for smart grid energy management problems. The suggested a its maximum capacity to lengthen its lifespan. The authors in
stochastic-robust hybrid approach combined the benefits of [179] suggested a straightforward and inexpensive control
stochastic and robust strategies, resulting in an economic approach based on fuzzy logic PI controllers which are faster
and robust solution with low computational difficulty because at recovering DC link voltage and more effective at employing
the stochastic scenarios were chosen efficiently and rationally. supercapacitor for high charge -discharge circles, and hence
The proposed method outperformed the existing distributed reducing the batterie’s stress and extending lifespan. An artifi-
algorithm developed in recent literature in terms of conver- cial intelligence (AI)-based approach was proposed by the
gence speed. This paper [175] presented a hybrid stochastic- authors of this research [180] as a way of achieving efficient
robust synchronized energy management strategy to improve microgrid EMS to reduce the whole cost of the microgrid,
MG’s flexibility, reliability, and security indices. According which includes fuel expenses, operating and maintenance costs,
84 B. Modu et al.

and generation unit emissions costs. To take DR into consid- ing. Table 6 present a critical summary the various optimiza-
eration, load is separated into four categories: critical load, tion and management techniques applied in microgrid.
controlled load, price-sensitive load, and thermal load. The
suggested method is more effective than PSO at obtaining 6. DC microgrid protection
the greatest outcome. On the basis of forecasts from Model
Predictive Control using Wavelet Neural Networks, the Despite their numerous benefits, DC microgrids have been
authors of this study [181] developed a novel energy manage- hampered by a complex protection challenge. The lack of nat-
ment technique with focus on energy losses, unfulfilled load, ural zero crossing, stochastic nature of RES, and high proba-
and component resilience. Several wavelet neural networks bility of distributed energy (DE) outages exacerbate the task of
make up the forecast models feeding the MPC, which are more fault detection and classification [192]. Furthermore, the rapid
appropriate for forecasts with limited spatial resolution. The increase in DC fault current complicates the protection strat-
simulations show that, in comparison to a rule base control, egy and increases the size and capacity of the breaker. Lack
the strategy improves all indicators that are taken into of regulations and all-inclusiveness in standards for various
account. [182] proposed a multi-agent system, incentive voltage levels on grounding is also a significant impediment
demand response, and a game theory-based energy trading to its accelerated future. As a result, a dynamic protection
model and solution algorithm for a microgrid cluster, micro- model with adequate redundancy, selectivity, and sensitivity
grid, and user. Two level are involved for the optimization pro- must be developed. However, some critical factors must be
cess, by developing an internal pricing enticement system, the considered when designing this kind of protection model,
microgrid cluster acts as a leader at the upper level and encour- namely: a) Improved converter sensitivity, particularly in
ages the microgrid to take part in intra-cluster dispatching. faulty circumstances b) Modification of fault current (short-
The microgrid, acting as the follower, decides the quantity of circuit current) under various mode of microgrid operation,
power transactions based on the realized internal price in order c) Appropriate grounding system, d) a rapid upsurge in peak
to increase its own profits. The efficiency and good conver- current thru a transient fault state, e) Arc fault analysis, and
gence of the suggested technique were confirmed by simulation f) the issue of standardization [193]. The approaches related
results, which also showed that the suggested hierarchical with Microgrid protection are summarized in brief as depicted
game strategy can increase each participant’s economic in Fig. 26. These methods were gathered from literature
rewards, which is helpful for the development of a cooperative reviews that were recently used to create microgrid protection
grid-connected microgrid cluster. Energy management system techniques.
based on a multi-agent system for a DC microgrid is proposed A novel protection scheme based on multi-sample differen-
in [183]. Here, a MAS with five major agents is employed, these tial protection scheme (MSD) to overcome the issue of insta-
agents primarily include a photovoltaic, solar, loads, and a ser- bility as a result of time synchronization error with
ver agent. The suggested system’s operation is predicated on differential relays is proposed in [194], When taking into
supplying loads as efficiently as possible using the available account latency limits from various measurement channels,
renewable sources. A multi-agent based energy management the MSD technique achieves reliable internal fault detection
architecture for microgrid is presented in [184]. The goal is while also ensuring stability for external fault conditions. A
to reduce agent’s energy losses, cost of operation. Each agent travelling wave based fault detection scheme is proposed in
uses the proposed learning methodology and its own previous [195], high frequency components of DC fault current was cal-
data to forecast the parameters and data it needs for schedul-

Fig. 26 Protection methods used for microgrid.

DC-based microgrid: Topologies, control schemes, and implementations 85

culated using the discrete wavelet form. The proposed tech- dure. The convolutional neural network classifier appropri-
nique can identify fault types and estimating fault locations. ately separates between all incidents, and its implementation
The authors in [196] proposed a protection scheme for detect- is verified by a variety of metrics.
ing fault that uses a rate of change fault current and its moving Table 7 summarizes the findings of a review of dc microgrid
average as discriminative feature. The scheme can detect fault protection schemes. The methods used to create protection
within a few microseconds of their occurrence. A fast detection schemes along with their advantages and shortcomings are
and location technique based on differential current for DC presented.
microgrid is proposed in [197], a modified cumulative sum
average method is used to classify the fault type with the help 7. Challenges and recommendations on DC microgrid
of discrete frame differential current solution. The authors in
[198] presented a novel learning based approach fault detection DC microgrid has many technical advantages over AC micro-
technique to improve detection accuracy and minimised grid, these include easy integration of renewable energy
unwanted false tripping in order to curtail the effect of DC resources, direct connection between the consumer loads and
arc faults through timely detection. This protection scheme DC bus voltage, more simply mitigated harmonics, and no fre-
enhances detection precision while decreasing the possibility quency as well as reactive power control problem. On the other
of false trips. The authors in [199] presented a novel control hand, the DC microgrid also has its downside such as complex
and protection technique based on droop and current deriva- controller, SoC and power imbalances, planning and operation
tives. The protection scheme can detect, characterize, and con- difficulty, high fault current because of short circuit, and cou-
trol fault current. Power sharing among converters is pled with uncertainties because of intermittent characteristics
accomplished using droop, and then, an adaptive droop is used of wind and solar irradiance. Therefore, in other to maintain
for fault current control. A protection scheme based on convo- stability, reliability, and flexibility of the DC microgrid, the
lution neural network is proposed by [192], The technique system planners and operators must be careful in planning
reduces the difficulty and cost of the feature extraction proce-

Table 7 Overview of DC microgrid protection schemes.

Refs. Developed techniques Strengths Weakness
[200] Inductance estimation-based & Changes in microgrid topology have no & No consideration is given to the communication
scheme effect. system.
& Used local measurement data only.
[194] Percentage differential current & Very fast response & Centralized protection approach, it is prone to
technique & When current transformer is saturated, single point failure.
the system does not malfunction. & Increase in losses due to various power
converters.
[201] Intelligent fault detection-based & Improved accuracy (95.7 %). & Considered only low impedance faults.
scheme & Reduced implementation costs. & The strategy does not address instability of
microgrid
[202] Differential current-based fault & Greater precision in calculating fault & The most terrible scenario occurs during a DC
detection approach distance. arc fault condition when proper grounding is
& Quick response not present.
& Noniterative technique is used to calcu-
late the distance of the fault
[203] Dijkstra Algorithm with Prims & determining the precise type of fault and & A large bus system is not taken into account here.
for Adaptive Protection its position.
& Microgrid is constantly monitored.
[204] Deep recurrent neural network & PV system anomaly detection, increased & An error-free environment is required for train-
technique system efficacy, and reduced maintenance ing the data on its normal behaviour.
& Faulty condition not conspired for experiment.
[205] DC microgrid protection & Microgrid is protected against communi- & The detection time for cyber-attacks is less than
schemes based on blockchain cation failure and cyber-attacks. 2.5 ms, there is need for improvement.
[206] Decentralised cuckoo artificial & The protection technique offers backup & Each classifier in the Artificial Neural Network
neural network-based protection protection. can detect faults using only local voltage and cur-
technique. & It has over 96 % precision and 99 % rent measurements.
reliability.
[207] Master slave-virtual inertia & Improve the shortcomings of converter- & Considered voltage and reactive power control
microgrid protection scheme based distributed generator systems due only.
to a lack of inertia.
& Improving controlled voltage steady-state
and transient responses in microgrids
[208] voltage-based power differential & Precise and effective detection and identi- & System voltage can only be used for
protection technique fication of faults. measurement.
& Can be implemented in both radial and & A relay is required for each node.
mesh configuration.
86 B. Modu et al.

and operating DC microgrid. Consequently, innovative plan- robustness, controllability, economic rate of the system, uti-
ning and operation strategy, effective DC bus voltage control, lization of the resources, the weather and flexibility to the
improved current and power sharing techniques have been end users. All the DC microgrid structures have their specific
developed and perhaps employed to take care of these prob- application each with advantage and disadvantage. Further
lems. However, there is still possibility of developing a better research is required to overcome the limitations and develop
planning, operation, and control techniques by conducting a standard architecture for DC microgrid.
extensive research in the following directions as summarized; Several control techniques each with their objectives have
[79,209,210]; been explained. The overall control topology is divided into
two categories: the basic control strategy and the multilevel
i. Proper modeling of a system is needed for an effective or hierarchical control strategy. The basic control strategy is
planning and operation of DC microgrid. Therefore, further classified based on communication links namely:
an improved or innovative model should be developed decentralized, centralized, distributed as well as hybrid decen-
considering the intermittent nature of DGs in microgrid tralized centralized, and distributed decentralized topologies.
in other to minimize the planning and operation cost. In decentralized control strategy, the input signal to the con-
ii. Several innovative robust, simple, cost effective and troller is locally measured signals which are locally processed
optimal control strategies can be developed or improved to generate gate pulse for the converters. The most common
with view to achieve a proper voltage regulation and method used in decentralized control strategy droop control.
accurate current sharing. However, in droop control method, there is trade-off between
iii. High level RES integration DC comes with reduced iner- current sharing accuracy and voltage deviation as high droop
tia of the overall DC microgrid which degrade the per- gain increases current sharing accuracy and voltage deviation
formance of the voltage control. Even though some increases at the same time. Communication is the heartbeat
control techniques have been developed to handle this of centralized control strategy. There is a communication link
issue, but more work is still required in this aspect since between the controllers in the neighboring converter and the
DC microgrid is becoming 100 % RES based system. central controller. The central controller compares the total
iv. Microgrid control should be able to operate the system generation and the load and then send load shading signal to
in optimum condition always. For instance, harnessing the non-critical loads in the case of insufficient generation to
of maximum renewable energy, to use energy storage supply the loads. It provides strong supervision which includes
device optimally and keeping the DC bus voltage within system real-time controllability and observability. Although,
limits. the centralized control method suffers from the limitation of
v. Cost effective, less complex, and viable communication single point failure. Distributed controllers are linked to each
techniques for the control of microgrid operation should other via a digital communication channel in distributed con-
be used in remote location. trol strategy. It combines both advantage of decentralized and
vi. Control scheme for batteries should also meet the centralized control strategy. Multilevel control with all the
extreme of over charge and undercharge when charging control levels have been explained in this paper. Primary con-
and discharging correspondingly in view of minimiza- troller which is the fastest compared to secondary and tertiary
tion of recurrent charge and discharge micro-cycles is implemented in the first control level, secondary controller
because of slight change in generation or loads. balances the voltage deviation resulting from the primary con-
vii. More work should also conduct regarding the life-span trol and the top-level control layer which is tertiary control
enhancement of storage devices whilst maintaining manage the operations of various units within the microgrid
power stability and re-establishing DC bus voltage. and the grid utility. EMS as part of tertiary control is dis-
viii. Keeping the grid-connected and islanded systems stable, cussed. Maintaining an energy balance between the intermit-
as well as ensuring seamless transfer between the two tent DGs and the stochastic nature of the electricity demand
systems through coordinated and local controls is a major issue in microgrid operation. But luckily enough,
ix. An optimal grounding that reduces the DC stray current the advancement in internet of things (IoT) or rather, informa-
and maximizes the personal security tion and communication technology has made it possible to
x. Another challenge is the cost-effective design of circuit develop various energy management approaches that ensures
breakers due to non-zero crossing of DC current. balance of power between DGs through utility or local micro-
grid and the demand side. However, there is need for the
research community to work on how to thoroughly assess
8. Discussions the potential benefits and costs of microgrid energy
management.
Several papers on DC microgrid control have been reviewed in Finally, DC microgrid has many technical advantages over
this study. The structure of DC microgrid, control methods, AC microgrid, which include easy integration of renewable
energy management as well as DC microgrid challenges are energy resources, direct connection between the consumer
discussed. loads and DC bus voltage, more simply mitigated harmonics,
DC microgrid architecture with their application, advan- and no frequency as well as reactive power control problem.
tage and disadvantage are discussed. The DC microgrid topol- However, DC microgrid also has its own shortcoming such
ogy is classified into six categories: Radial bus topology, Multi as complex controller, SoC and power imbalances, high fault
bus topology, Multi terminal bus topology, Ladder bus topol- current because of short circuit, high uncertainties because of
ogy, Ring bus topology and Zonal type bus topology. The DC intermittent characteristics of wind speed and solar irradiance
microgrid structure is a function of the following factors: are all noticed in a DC microgrid. Groundbreaking research
DC-based microgrid: Topologies, control schemes, and implementations 87

on DC microgrid control strategy can remove these [10] M. Mishra et al, A systematic review on DC-microgrid
drawbacks. protection and grounding techniques: issues, challenges and
future perspective, Appl. Energy 313 (2022) 118810.
9. Conclusions [11] V. Krishnamurthy, A. Kwasinski, Effects of power electronics,
energy storage, power distribution architecture, and lifeline
dependencies on microgrid resiliency during extreme events,
This scientific article presents a painstaking review of DC micro- IEEE J. Emerging and selected topics in Pow. Electronics 4 (4)
grid control. In terms of reliability, efficiency, power quality and (2016) 1310–1323.
perhaps cost, DC microgrid is superior and promising compared [12] M. Leng et al, Impedance-based stability evaluation for
with AC or hybrid microgrid system. However, the advance- multibus DC microgrid without constraints on subsystems,
ment in power electronic technology development and the IEEE Trans. Power Electron. 37 (1) (2022) 932–943.
rapidly growing deployment of distributed generations (DGs) [13] T. Dragičević et al, DC Microgrids—Part II: a review of power
is accompanied with technical issues such as the effective control architectures, applications, and standardization issues, IEEE
and management of the microgrid operation. Therefore, exten- Trans. Power Electron. 31 (5) (2016) 3528–3549.
[14] Y. Wang et al, A Practical DC Fault Ride-Through Method
sive research on the aspect of DC microgrid control, standards,
for MMC Based MVDC Distribution Systems, IEEE Trans.
and management system are required. The fundamental control Power Delivery 36 (4) (2021) 2510–2519.
structures namely, centralized, decentralized, distributed, and [15] G. Zou et al, A fast protection scheme for VSC based multi-
multilevel, that is, hierarchical control layers have been critically terminal DC grid, Int. J. Electr. Power Energy Syst. 98 (2018)
reviewed along with energy management system of DC micro- 307–314.
grid. Challenges of DC microgrid with feature recommenda- [16] J.-D. Park et al, DC ring-bus microgrid fault protection and
tions have been presented. This review article concluded that identification of fault location, IEEE Trans. Power Delivery 28
further research on control techniques, a standard architecture (4) (2013) 2574–2584.
for DC microgrid, and balance of power between distributed [17] P.P. Padhi, K. Deepa, DC Microgrids: architecture and
generations (DGs) and the dynamic load demand would be an Challenges, IOP Conference Series: Materials Science and
Engineering, IOP Publishing, 2021.
extraordinary contribution toward realizing a matured DC
[18] J. Kumar, A. Agarwal, V. Agarwal, A review on overall control
microgrid technology. of DC microgrids, J. Storage Mater. 21 (2019) 113–138.
[19] F. Mohan, N. Sasidharan. DC Microgrid and its Protection - A
Declaration of Competing Interest Review. in 2020 International Conference on Power,
Instrumentation, Control and Computing (PICC). 2020.
The authors declare that they have no known competing [20] R.S. Alishah et al, Optimal design of new cascaded switch-
financial interests or personal relationships that could have ladder multilevel inverter structure, IEEE Trans. Ind. Electron.
appeared to influence the work reported in this paper. 64 (3) (2017) 2072–2080.
[21] E. Tironi, M. Corti, G. Ubezio. Zonal electrical distribution
References systems in large ships: Topology and control. in 2015 AEIT
International Annual Conference (AEIT). 2015. IEEE.
[1] T.H. Pham et al, Power balancing in a DC microgrid elevator [22] R.S. Balog, P.T. Krein, Bus selection in multibus DC
system through constrained optimization, IFAC- microgrids, IEEE Trans. Power Electron. 26 (3) (2010) 860–
PapersOnLine 50 (1) (2017) 19–24. 867.
[2] D.J. Becker, B. Sonnenberg. DC microgrids in buildings and [23] P. Cairoli, R.A. Dougal, New Horizons in DC Shipboard
data centers. in 2011 IEEE 33rd International Power Systems: New fault protection strategies are essential to
Telecommunications Energy Conference (INTELEC). 2011. the adoption of dc power systems, IEEE Electrif. Mag. 1 (2)
IEEE. (2013) 38–45.
[3] G. Shahgholian, A brief review on microgrids: operation, [24] D. Kumar, F. Zare, A. Ghosh, DC microgrid technology:
applications, modeling, and control, Int. Trans. Electr. Energy system architectures, AC grid interfaces, grounding schemes,
Syst. (2021) e12885. power quality, communication networks, applications, and
[4] A. Jafari, G. Shahgholian, Analysis and simulation of a sliding standardizations aspects, IEEE Access 5 (2017) 12230–12256.
mode controller for mechanical part of a doubly-fed induction [25] O.D. Montoya, L.F. Grisales-Noreña, W. Gil-González,
generator-based wind turbine, IET Gener. Transm. Distrib. 11 Triangular matrix formulation for power flow analysis in
(10) (2017) 2677–2688. radial DC resistive grids with CPLs, IEEE Trans. Circuits Syst.
[5] M. Srinivasan, A. Kwasinski, Control analysis of parallel DC- Express Briefs 67 (6) (2019) 1094–1098.
DC converters in a DC microgrid with constant power loads, [26] O.D. Montoya, On linear analysis of the power flow equations
Int. J. Electr. Power Energy Syst. 122 (2020) 106207. for DC and AC Grids With CPLs, IEEE Trans. Circuits Syst.
[6] C. Zhang et al, Finite-time feedforward decoupling and precise Express Briefs 66 (12) (2019) 2032–2036.
decentralized control for DC microgrids towards large-signal [27] A. Garcés, On the Convergence of Newton’s Method in Power
stability, IEEE Trans. Smart Grid 11 (1) (2019) 391–402. Flow Studies for DC Microgrids, IEEE Trans. Pow. Syst. 33
[7] R. Kumar, M.K. Pathak. Control of DC Microgrid for (5) (2018) 5770–5777.
Improved Current Sharing and Voltage Regulation. in 2020 [28] A. Garces, Uniqueness of the power flow solutions in low
3rd International Conference on Energy, Power and voltage direct current grids, Electr. Pow. Syst. Res. 151 (2017)
Environment: Towards Clean Energy Technologies. 2021. 149–153.
[8] S. Jamal, N.M. Tan, J. Pasupuleti, A review of energy [29] E.N.M. Silva, A.B. Rodrigues, M. da Guia da Silva,
management and power management systems for microgrid Approximated and Iterative Power Flow Algorithms for
and nanogrid applications, Sustainability 13 (18) (2021) 10331. Islanded DC Microgrids. Electric Power Systems Research,
[9] K. Bhargavi et al, A comprehensive review on control 2023. 215: p. 108972.
techniques for power management of isolated DC microgrid [30] P. Nagendra, S.H.n. Dey, and S. Paul. OPF based voltage
system operation, IEEE Access 9 (2021) 32196–32228. stability assessment of a multi-bus power system using network
88 B. Modu et al.

equivalencing technique. in 2009 International Conference on [51] S.K. Sahoo, A.K. Sinha, N.K. Kishore, Control Techniques in
Power Systems. 2009. AC, DC, and Hybrid AC–DC Microgrid: a review, IEEE J.
[31] T. Altun, R. Madani, A. Davoudi, Topology-Cognizant Emerging and Selected Topics in Power Electronics 6 (2) (2018)
Optimal Power Flow in Multi-Terminal DC Grids, IEEE 738–759.
Trans. Power Syst. 36 (5) (2021) 4588–4598. [52] S. Sen, V. Kumar, Microgrid control: a comprehensive survey,
[32] S. Rodrigues et al, Optimal power flow control of VSC-based Annu. Rev. Control. 45 (2018) 118–151.
multiterminal DC network for offshore wind integration in the [53] P. Monica, M. Kowsalya, Control strategies of parallel
north sea, IEEE J. Emerging and Selected Topics in Power operated inverters in renewable energy application: a review,
Electronics 1 (4) (2013) 260–268. Renew. Sustain. Energy Rev. 65 (2016) 885–901.
[33] F. Lu et al, Complex affine arithmetic-based power flow [54] J. Yang et al, A novel quasi-master-slave control frame for PV-
analysis for zonal medium voltage direct current shipboard storage independent microgrid, Int. J. Electr. Power Energy
power systems in the presence of power variation, Energies 11 Syst. 97 (2018) 262–274.
(7) (2018) 1697. [55] F.A. Padhilah, K.-H. Kim, A Centralized Power Flow Control
[34] A. Bosisio, et al. Urban distribution network planning with 2-step Scheme of EV-Connected DC microgrid to satisfy multi-
ladder topology considering joint nodes. in 2017 IEEE objective problems under several constraints, Sustainability 13
Manchester PowerTech. 2017. (16) (2021) 8863.
[35] B.Y. Ku, J.S. Liu, Modelling and numerical solution of [56] M. Mehdi et al, Robust control of a DC microgrid under
ungrounded DC rail traction system load flow using ladder parametric uncertainty and disturbances, Electr. Pow. Syst.
circuit–part II: solution method, IET Gener. Transm. Distrib. Res. 179 (2020) 106074.
14 (1) (2020) 11–20. [57] S. Pannala et al, Effective control and management scheme for
[36] H. Li, L. Zhang, X. Shen, A loop-analysis theory based power isolated and grid connected DC Microgrid, IEEE Trans. Ind.
flow method and its linear formulation for low-voltage DC Appl. 56 (6) (2020) 6767–6780.
grid, Electr. Pow. Syst. Res. 187 (2020) 106473. [58] M. Mehdi, C.H. Kim, M. Saad, Robust Centralized Control
[37] O.D. Montoya et al, Linear power flow formulation for low- for DC islanded microgrid considering communication
voltage DC power grids, Electr. Pow. Syst. Res. 163 (2018) network delay, IEEE Access 8 (2020) 77765–77778.
375–381. [59] B.P. Poudel et al, Detection and mitigation of cyber-threats in
[38] Y. Hui, et al. Newton-Downhill algorithm for distribution power the DC microgrid distributed control system, Int. J. Electr.
flow analysis. in 2008 IEEE 2nd International Power and Energy Power Energy Syst. 120 (2020) 105968.
Conference. 2008. [60] K. Bzura, et al. Power flow management algorithms for
[39] B. Mohamed, P. Arboleya, C. González-Morán, Modified centralized controller in direct-current microgrid. in 2021
current injection method for power flow analysis in heavy- Progress in Applied Electrical Engineering (PAEE). 2021.
meshed DC railway networks with nonreversible substations, [61] S. Peyghami, H. Mokhtari, and F. Blaabjerg, Chapter 3 -
IEEE Trans. Veh. Technol. 66 (9) (2017) 7688–7696. Hierarchical Power Sharing Control in DC Microgrids, in
[40] H. Su, Z. Zhang, Power flow algorithm for weakly meshed Microgrid, M.S. Mahmoud, Editor. 2017, Butterworth-
distribution network with distributed generation based on Heinemann. p. 63-100.
loop-analysis in different load models, J. Electr. Eng. Technol. [62] G. Ensermu, A. Bhattacharya, N. Panigrahy, Real-time
13 (2) (2018) 608–619. simulation of smart DC microgrid with decentralized control
[41] B. Khan et al, Optimal integration of DSTATCOM using system under source disturbances, Arab. J. Sci. Eng. 44 (8)
improved bacterial search algorithm for distribution network (2019) 7173–7185.
optimization, Alex. Eng. J. 61 (7) (2022) 5539–5555. [63] H. Alrajhi Alsiraji, R. El-Shatshat, Virtual synchronous
[42] M.H. Moradi, V.B. Foroutan, M. Abedini, Power flow analysis machine/dual-droop controller for parallel interlinking
in islanded Micro-Grids via modeling different operational converters in hybrid AC–DC microgrids, Arab. J. Sci. Eng.
modes of DGs: a review and a new approach, Renew. Sustain. 46 (2) (2021) 983–1000.
Energy Rev. 69 (2017) 248–262. [64] S. Wang et al, Virtual-battery based droop control and energy
[43] T.M. Mohan, T. Nireekshana. A Genetic Algorithm for Solving storage system size optimization of a DC microgrid for electric
Optimal Power Flow Problem. in 2019 3rd International vehicle fast charging station, Appl. Energy 259 (2020) 114146.
conference on Electronics, Communication and Aerospace [65] D. Li, C.N.M. Ho, A Module-Based Plug-n-Play DC
Technology (ICECA). 2019. Microgrid With Fully Decentralized Control for IEEE
[44] U. Khaled, A.M. Eltamaly, A. Beroual, Optimal power flow Empower a Billion Lives Competition, IEEE Trans. Power
using particle swarm optimization of renewable hybrid Electron. 36 (2) (2021) 1764–1776.
distributed generation, Energies 10 (7) (2017) 1013. [66] Y. Zhang, W. Wei, Decentralized coordination control of PV
[45] S.B. Efe, M. Cebeci, Artificial neural network based power flow generators, storage battery, hydrogen production unit and fuel
analysis for micro grids, Bitlis Eren University J. Sci. Technol. cell in islanded DC microgrid, Int. J. Hydrogen Energy 45 (15)
5 (1) (2015) 42–47. (2020) 8243–8256.
[46] H. Wang et al, Probabilistic power flow analysis of microgrid [67] X. Lin, R. Zamora, C.A. Baguley, A fully filter-based
with renewable energy, Int. J. Electr. Power Energy Syst. 114 decentralized control with state of charge balancing strategy
(2020) 105393. for battery energy storage systems in autonomous DC
[47] Z. Liu et al, Stability analysis of DC microgrids with constant Microgrid Applications, IEEE Access 9 (2021) 15028–15040.
power load under distributed control methods, Automatica 90 [68] D. Li, C.N.M. Ho, Decentralized PV–BES Coordination
(2018) 62–72. Control With Improved Dynamic Performance for Islanded
[48] K. Bharath, M.M. Krishnan, P. Kanakasabapathy, A review Plug-n-Play DC Microgrid, IEEE J. Emerging and Selected
on dc microgrid control techniques applications and trends, Topics in Power Electronics 9 (4) (2021) 4992–5001.
Int. J.f Renew. Energy Res. (IJRER) 9 (3) (2019) 1328–1338. [69] L. Li et al, Event-triggered decentralized coordinated control
[49] F.S. Al-Ismail, DC Microgrid Planning, Operation, and method for economic operation of an islanded electric-
Control: a Comprehensive Review, IEEE Access 9 (2021) hydrogen hybrid DC microgrid, J. Storage Mater. 45 (2022)
36154–36172. 103704.
[50] A. Kaur, J. Kaushal, P. Basak, A review on microgrid central [70] P. Monica, M. Kowsalya, J.M. Guerrero, Logarithmic droop-
controller, Renew. Sustain. Energy Rev. 55 (2016) 338–345. based decentralized control of parallel converters for accurate
DC-based microgrid: Topologies, control schemes, and implementations 89

current sharing in islanded DC microgrid applications, IET [91] H. Armghan et al, Nonlinear integral backstepping based
Renew. Power Gener. 15 (6) (2021) 1240–1254. control of a DC microgrid with renewable generation and
[71] M.R. Habibi et al, Decentralized Coordinated Cyberattack energy storage systems, Int. J. Electr. Power Energy Syst. 117
Detection and Mitigation Strategy in DC Microgrids Based on (2020) 105613.
Artificial Neural Networks, IEEE J. Emerging and Selected [92] M. Cucuzzella et al, Sliding mode voltage control of boost
Topics in Power Electr. 9 (4) (2021) 4629–4638. converters in DC microgrids, Control Eng. Pract. 73 (2018)
[72] Shivam R. Dahiya, Robust decentralized control for effective 161–170.
load sharing and bus voltage regulation of DC microgrid based [93] A. Abhishek et al, Review of hierarchical control strategies for
on optimal droop parameters. Journal of Renewable and DC microgrid, IET Renew. Power Gener. 14 (10) (2020) 1631–
Sustainable Energy, 2017. 9(4): p. 045301. 1640.
[73] R. Olfati-Saber, J.A. Fax, R.M. Murray, Consensus and [94] A. Abedi et al, DC-bus Voltage Control based on Direct
cooperation in networked multi-agent systems. Proceedings of Lyapunov Method for a Converter-based Stand-alone DC
the IEEE, 2007. 95(1): p. 215-233. Micro-grid, Electr. Pow. Syst. Res. 187 (2020) 106451.
[74] J.W. Chang et al, A novel distributed control method for [95] M. Panda, V.B. Devara, T. Maity, A novel dc bus-signaling
interlinking converters in an islanded hybrid AC/DC based power management strategy for dc microgrid, Int. Trans.
Microgrid, IEEE Trans. Smart Grid 12 (5) (2021) 3765–3779. Electr. Energy Syst. 31 (2) (2021) e12758.
[75] Y. Xia, et al. A Distributed Control in Islanded DC Microgrid [96] J. Hu et al, Distributed adaptive droop control for optimal
based on Multi-Agent Deep Reinforcement Learning. in IECON power dispatch in DC microgrid, IEEE Trans. Ind. Electron.
2020 The 46th Annual Conference of the IEEE Industrial 65 (1) (2017) 778–789.
Electronics Society. 2020. [97] Z. Fan, B. Fan, W. Liu, Distributed Control of DC microgrids
[76] L. Xing et al, Distributed secondary control for DC microgrid for optimal coordination of conventional and renewable
with event-triggered signal transmissions, IEEE Trans. generators, IEEE Trans. Smart Grid 12 (6) (2021) 4607–4615.
Sustainable Energy 12 (3) (2021) 1801–1810. [98] Y. Jiang, et al. Line Resistance Identification-Based Adaptive
[77] S. Mudaliyar, B. Duggal, S. Mishra, Distributed tie-line power Droop Control for Distribution Power Loss Minimization of DC
flow control of autonomous DC microgrid clusters, IEEE Microgrids. in 2021 IEEE 12th Energy Conversion Congress &
Trans. Power Electron. 35 (10) (2020) 11250–11266. Exposition-Asia (ECCE-Asia). 2021. IEEE.
[78] M.S. Sadabadi, A distributed control strategy for parallel DC- [99] A.J. Wood, B.F. Wollenberg, G.B. Sheblé, Power Generation,
DC converters, IEEE Control Systems Letters 5 (4) (2021) Operation, and Control. 2013: Wiley.
1231–1236. [100] R. Wang, G. Xiao, P. Wang, Hybrid Centralized-Decentralized
[79] S. Dahale, et al. An overview of DC-DC converter topologies and (HCD) Charging Control of Electric Vehicles, IEEE Trans.
controls in DC microgrid. in 2017 7th International Conference Veh. Technol. 66 (8) (2017) 6728–6741.
on Power Systems (ICPS). 2017. [101] V. Behravesh, R. Keypour, A. Akbari Foroud, Control
[80] Shivam R. Dahiya, Stability analysis of islanded DC microgrid strategy for improving voltage quality in residential power
for the proposed distributed control strategy with constant distribution network consisting of roof-top photovoltaic-wind
power loads. Computers & Electrical Engineering, 2018. 70: p. hybrid systems, battery storage and electric vehicles, Sol.
151-162. Energy 182 (2019) 80–95.
[81] D. Chowdhury, A.K. Hasan, M.Z.R. Khan, Islanded DC [102] J.E. Newbury, K.J. Morris, Power line carrier systems for
microgrid architecture with dual active bridge converter-based industrial control applications, IEEE Trans. Power Delivery 14
power management units and time slot-based control interface, (4) (1999) 1191–1196.
IEEJ Trans. Electr. Electron. Eng. 15 (6) (2020) 863–871. [103] G. Shickhuber, O. McCarthy, Control using power lines-a
[82] H. Sarwar Khan, et al., Artificial Neural Network-Based European view, Comput. Control Eng. J. 8 (4) (1997) 180–184.
Voltage Control of DC/DC Converter for DC Microgrid [104] W.W.A.G.d. Silva, T.R. Oliveira, P.F. Donoso-Garcia, Hybrid
Applications. arXiv e-prints, 2021: p. arXiv: 2111.03207. Distributed and Decentralized Secondary Control Strategy to
[83] B. Xia et al, Decentralized control method for modular Attain Accurate Power Sharing and Improved Voltage
multilevel converters, IEEE Trans. Power Electron. 34 (6) Restoration in DC Microgrids. IEEE Transactions on Power
(2019) 5117–5130. Electronics, 2020. 35(6): p. 6458-6469.
[84] T. Dragičević et al, DC microgrids—Part I: a review of control [105] H. Han et al, Leader-distributed follower-decentralized control
strategies and stabilization techniques, IEEE Trans. Power strategy for economic dispatch in cascaded-parallel microgrids,
Electron. 31 (7) (2015) 4876–4891. Int. Trans. Electrical Energy Syst. 31 (9) (2021) e12964.
[85] T. Dragičević et al, DC microgrids—Part II: a review of power [106] Y. Sun et al, A local-distributed and global-decentralized
architectures, applications, and standardization issues, IEEE control scheme, in: Series-Parallel Converter-Based
Trans. Power Electron. 31 (5) (2015) 3528–3549. Microgrids, Springer, 2022, pp. 307–328.
[86] M.N.B. Shaheed, et al. A Novel Decentralized Adaptive Droop [107] S. Parhizi et al, State of the art in research on microgrids: a
Control Technique for DC Microgrids Based on Integrated Load review, IEEE Access 3 (2015) 890–925.
Condition Processing. in 2020 IEEE Energy Conversion [108] D.E. Olivares et al, Trends in Microgrid Control, IEEE Trans.
Congress and Exposition (ECCE). 2020. Smart Grid 5 (4) (2014) 1905–1919.
[87] Z. Karami et al, Decentralized model predictive control of DC [109] A. Kantamneni et al, Survey of multi-agent systems for
microgrids with constant power load, IEEE J. Emerging and microgrid control, Eng. Appl. Artif. Intel. 45 (2015) 192–203.
Selected Topics in Power Electr. 9 (1) (2021) 451–460. [110] L. Meng et al, Review on Control of DC Microgrids and
[88] Y. Liu et al, A novel adaptive model predictive control for Multiple Microgrid Clusters, IEEE J. Emerging and Selected
proton exchange membrane fuel cell in DC Microgrids, IEEE Topics in Power Electronics 5 (3) (2017) 928–948.
Trans. Smart Grid 13 (3) (2022) 1801–1812. [111] W. Feng et al, A review of microgrid development in the
[89] S.-C. Tan, Y.-M. Lai, K.T. Chi, General design issues of United States – a decade of progress on policies,
sliding-mode controllers in DC–DC converters, IEEE Trans. demonstrations, controls, and software tools, Appl. Energy
Ind. Electron. 55 (3) (2008) 1160–1174. 228 (2018) 1656–1668.
[90] M. Cucuzzella, et al. Decentralized Sliding Mode voltage control [112] M.H. Cintuglu, T. Youssef, O.A. Mohammed, Development
in DC microgrids. in 2017 American Control Conference and application of a real-time testbed for multiagent system
(ACC). 2017.
90 B. Modu et al.

interoperability: a case study on hierarchical microgrid control, [133] X. Li et al, A novel assorted nonlinear stabilizer for DC–DC
IEEE Trans. Smart Grid 9 (3) (2018) 1759–1768. multilevel boost converter with constant power load in DC
[113] F. Gao et al, Primary and secondary control in DC microgrids: microgrid, IEEE Trans. Power Electron. 35 (10) (2020) 11181–
a review, J. Mod Power Syst. Clean Energy 7 (2) (2019) 227– 11192.
242. [134] M.A. Hassan, Y. He, Constant Power Load Stabilization in
[114] D.Y. Yamashita, I. Vechiu, J.-P. Gaubert, A review of DC microgrid systems using passivity-based control with
hierarchical control for building microgrids, Renew. Sustain. nonlinear disturbance observer, IEEE Access 8 (2020) 92393–
Energy Rev. 118 (2020) 109523. 92406.
[115] S.H. Hanzaei, et al. Control of DC Microgrids: A Review. in [135] M. Maaruf, K.A. Khan, M. Khalid. Integrated Power
2021 31st Australasian Universities Power Engineering Management and Nonlinear-Control for Hybrid Renewable
Conference (AUPEC). 2021. IEEE. Microgrid. in 2021 IEEE Green Technologies Conference
[116] R. Sondhi, N. Bhatt, and S. Arora. Primary control in DC (GreenTech). 2021.
microgrids: a review. in 2021 International Conference on [136] M. Boukerdja et al, H1 based control of a DC/DC buck
Advance Computing and Innovative Technologies in converter feeding a constant power load in uncertain DC
Engineering (ICACITE). 2021. microgrid system, ISA Trans. 105 (2020) 278–295.
[117] H. Yu et al, An integrated and reconfigurable hybrid AC/DC [137] J. Sun et al, Voltage Regulation of DC-microgrid with PV and
microgrid architecture with autonomous power flow control Battery, IEEE Trans. Smart Grid 11 (6) (2020) 4662–4675.
for nearly/net zero energy buildings, Appl. Energy 263 (2020) [138] M. Legry et al, Non-linear primary control mapping for droop-
114610. like behavior of microgrid systems, IEEE Trans. Smart Grid 11
[118] Z.H.A. Al-Tameemi et al, Control Strategies of DC Microgrids (6) (2020) 4604–4613.
Cluster: a Comprehensive Review, Energies 14 (22) (2021) [139] A. Berzoy, et al. Non-linear Droop Control of Parallel Split-
7569. phase Inverters for Residential Nanogrids. in 2019 IEEE Applied
[119] J. Lee, H.-S. Ahn, J. Back, Robust Voltage Stabilization Power Electronics Conference and Exposition (APEC). 2019.
Controller for Uncertain DC Microgrids, IEEE Access 9 (2021) [140] U.B. Tayab et al, A review of droop control techniques for
99606–99616. microgrid, Renew. Sustain. Energy Rev. 76 (2017) 717–727.
[120] X. Hou et al, Distributed Hierarchical Control of AC [141] H. Shadabi, I. Kamwa, A decentralized non-linear dynamic
microgrid operating in grid-connected, islanded and their droop control of a hybrid energy storage system bluefor
transition modes, IEEE Access 6 (2018) 77388–77401. primary frequency control in integrated AC-MTDC systems,
[121] J.M. Guerrero et al, Hierarchical control of droop-controlled Int. J. Electr. Power Energy Syst. 136 (2022) 107630.
AC and DC microgrids—A general approach toward [142] S. Adhikari, Y. Tang, and P. Wang. Secondary control for DC
standardization, IEEE Trans. Ind. Electron. 58 (1) (2010) microgrids: A review. in 2016 Asian Conference on Energy,
158–172. Power and Transportation Electrification (ACEPT). 2016.
[122] P.-I. Hwang et al, DC microgrid operational method for [143] S. Jadidi, H. Badihi, and Y. Zhang. A Review on Operation,
enhanced service reliability using DC bus signaling, J. Electr. Control and Protection of Smart Microgrids. in 2019 IEEE 2nd
Eng. Technol. 10 (2) (2015) 452–464. International Conference on Renewable Energy and Power
[123] Y. Cao, C. Wu, X. Wang, The Coordination Control Strategy Engineering (REPE). 2019.
of a DC Microgrid based on DC Bus Signaling, in Proceedings [144] A. Alfergani, et al. Control strategies in AC microgrid: A brief
of the 3rd international conference on computer science and review. in 2018 9th International Renewable Energy Congress
application engineering, China. p, 2019. Article 67. (IREC). 2018.
[124] G. Bharathi, P. Kantharao, R. Srinivasarao, Fuzzy logic [145] S.M. Kaviri, et al. A review of AC microgrid control methods. in
control (FLC)-based coordination control of DC microgrid 2017 IEEE 8th international symposium on power electronics for
with energy storage system and hybrid distributed generation, distributed generation systems (PEDG). 2017. IEEE.
Int. J. Ambient Energy (2021) 1–17. [146] S. Peyghami, H. Mokhtari, F. Blaabjerg, Decentralized load
[125] Q. Xu et al, A decentralized dynamic power sharing strategy sharing in a low-voltage direct current microgrid with an
for hybrid energy storage system in autonomous DC adaptive droop approach based on a superimposed frequency,
microgrid, IEEE Trans. Ind. Electron. 64 (7) (2017) 5930–5941. IEEE J. Emerging and Selected Topics in Power Electronics 5
[126] Z. Shuai et al, Hierarchical structure and bus voltage control of (3) (2017) 1205–1215.
DC microgrid, Renew. Sustain. Energy Rev. 82 (2018) 3670– [147] Y. Xia et al, Decentralized coordination control for parallel
3682. bidirectional power converters in a grid-connected DC
[127] S.B. Siad et al, Nonlinear control of a DC MicroGrid for microgrid, IEEE Trans. Smart Grid 9 (6) (2018) 6850–6861.
the integration of distributed generation based on different [148] C. Wang et al, Decentralized high-performance control of DC
time scales, Int. J. Electr. Power Energy Syst. 111 (2019) microgrids, IEEE Trans. Smart Grid 10 (3) (2019) 3355–3363.
93–100. [149] M. Nabatirad, B. Bahrani, R. Razzaghi. Decentralized
[128] A. Armghan et al, Dynamical operation based robust nonlinear Secondary Controller in Islanded DC Microgrids to Enhance
control of DC microgrid considering renewable energy Voltage Regulation and Load Sharing Accuracy. in 2019 IEEE
integration, Energies 14 (13) (2021) 3988. International Conference on Industrial Technology (ICIT).
[129] S.S. Zehra et al, Artificial intelligence-based nonlinear control 2019.
of renewable energies and storage system in a DC microgrid, [150] S. Adhikari, et al. Decentralized control of DC microgrid
ISA Trans. (2021). clusters. in 2017 IEEE 3rd International Future Energy
[130] F. Perez et al, Stability analysis of a DC microgrid for a smart Electronics Conference and ECCE Asia (IFEEC 2017 - ECCE
railway station integrating renewable sources, IEEE Trans. Asia). 2017.
Control Syst. Technol. 28 (5) (2020) 1802–1816. [151] L. Xing et al, Distributed secondary control for current sharing
[131] P. Li et al, Nonlinear coordinated control of parallel and voltage restoration in DC Microgrid, IEEE Trans. Smart
bidirectional power converters in an AC/DC hybrid Grid 11 (3) (2020) 2487–2497.
microgrid, Int. J. Electr. Power Energy Syst. 122 (2020) 106208. [152] S. Thomas, et al. Distributed secondary control with reduced
[132] M. Jebali Ben Ghorbal et al, A comparison study of two DC communication in low-voltage DC microgrid. in 2016 10th
microgrid controls for a fast and stable DC bus voltage, Math. International Conference on Compatibility, Power Electronics
Comput. Simul. 184 (2021) 210–224. and Power Engineering (CPE-POWERENG). 2016.
DC-based microgrid: Topologies, control schemes, and implementations 91

[153] L. Xing et al, voltage restoration and adjustable current [173] M.B. Sigalo et al, An Energy management system for the
sharing for DC microgrid with time delay via distributed control of battery storage in a grid-connected microgrid using
secondary control, IEEE Trans. Sustainable Energy 12 (2) mixed integer linear programming, Energies 14 (19) (2021)
(2021) 1068–1077. 6212.
[154] X.K. Liu et al, Distributed supervisory secondary control for a [174] X. Chang et al, Accelerated distributed hybrid stochastic/
DC Microgrid, IEEE Trans. Energy Convers. 35 (4) (2020) robust energy management of smart grids, IEEE Trans. Ind.
1736–1746. Inf. 17 (8) (2021) 5335–5347.
[155] S. Moayedi, A. Davoudi, Distributed Tertiary Control of DC [175] A. Dini et al, Hybrid stochastic/robust scheduling of the grid-
microgrid clusters, IEEE Trans. Power Electron. 31 (2) (2016) connected microgrid based on the linear coordinated power
1717–1733. management strategy, Sustainable Energy Grids Networks 24
[156] M. Zaery, E.M. Ahmed, and M. Orabi. Consensus algorithm (2020) 100400.
based distributed control for economic operation of islanded DC [176] M. Banaei, J. Boudjadar, M.H. Khooban, Stochastic model
microgrids. in 2016 Eighteenth International Middle East Power predictive energy management in hybrid emission-free modern
Systems Conference (MEPCON). 2016. IEEE. maritime vessels, IEEE Trans. Ind. Inf. 17 (8) (2021) 5430–
[157] A. Bidram, A. Davoudi, Hierarchical structure of microgrids 5440.
control system, IEEE Trans. Smart Grid 3 (4) (2012) 1963– [177] J.T. Dellosa, E.C. Palconit. Artificial Intelligence (AI) in
1976. Renewable Energy Systems: A Condensed Review of its
[158] C. Papadimitriou, E. Zountouridou, N. Hatziargyriou, Review Applications and Techniques. in 2021 IEEE International
of hierarchical control in DC microgrids, Electr. Pow. Syst. Conference on Environment and Electrical Engineering and
Res. 122 (2015) 159–167. 2021 IEEE Industrial and Commercial Power Systems Europe
[159] A. Berrada, K. Loudiyi, R. Garde, Dynamic modeling of (EEEIC/I&CPS Europe). 2021. IEEE.
gravity energy storage coupled with a PV energy plant, Energy [178] D. Arcos-Aviles et al, Low complexity energy management
134 (2017) 323–335. strategy for grid profile smoothing of a residential grid-
[160] N.R. Merritt et al, A unified control structure for grid connected microgrid using generation and demand
connected and islanded mode of operation of voltage source forecasting, Appl. Energy 205 (2017) 69–84.
converter based distributed generation units under unbalanced [179] A.S. Nasab, et al. A new method of energy management system
and non-linear conditions, IEEE Trans. Power Delivery 35 (4) in islanded DC microgrid using fuzzy logic controller. in 2019
(2019) 1758–1768. International Power System Conference (PSC). 2019. IEEE.
[161] K. Nandini, N. Jayalakshmi, V.K. Jadoun, An overview of DC [180] T. Wang, X. He, T. Deng, Neural networks for power
Microgrid with DC distribution system for DC loads, management optimal strategy in hybrid microgrid, Neural
Proceedings, Materials Today, 2021. Comput. & Applic. 31 (7) (2019) 2635–2647.
[162] L. Ahmethodzic, M. Music, Comprehensive review of trends in [181] D. Morin, Y. Stevenin, P. Brault. Energy Management of
microgrid control, Renew. Energy Focus 38 (2021) 84–96. isolated DC microgrids with hybrid batteries-hydrogen storage
[163] M. Hamidi, O. Bouattane, and A. Raihani. Microgrid Energy system using Model Predictive Control and Wavelet Neural
Management System: Technologies and Architectures Review. in Networks based forecasts. in 2019 21st European Conference on
2020 IEEE International conference of Moroccan Geomatics Power Electronics and Applications (EPE’19 ECCE Europe).
(Morgeo). 2020. 2019. IEEE.
[164] M.F. Zia, E. Elbouchikhi, M. Benbouzid, Microgrids energy [182] X. Dong, X. Li, S. Cheng, Energy management optimization of
management systems: a critical review on methods, solutions, microgrid cluster based on multi-agent-system and hierarchical
and prospects, Appl. Energy 222 (2018) 1033–1055. Stackelberg game theory, IEEE Access 8 (2020) 206183–206197.
[165] G.S. Georgiou, P. Christodoulides, S.A. Kalogirou, Real-time [183] M. Perera, et al. Multi agent based energy management system
energy convex optimization, via electrical storage, in buildings– for microgrids. in 2020 IEEE 9th Power India International
A review, Renew. Energy 139 (2019) 1355–1365. Conference (PIICON). 2020. IEEE.
[166] S.K. Rathor, D. Saxena, Energy management system for smart [184] M. Afrasiabi et al, Multi-agent microgrid energy management
grid: an overview and key issues, Int. J. Energy Res. 44 (6) based on deep learning forecaster, Energy 186 (2019) 115873.
(2020) 4067–4109. [185] S. Sukumar et al, Mix-mode energy management strategy and
[167] R.C. Green II, L. Wang, M. Alam, The impact of plug-in battery sizing for economic operation of grid-tied microgrid,
hybrid electric vehicles on distribution networks: a review and Energy 118 (2017) 1322–1333.
outlook, Renew. Sustain. Energy Rev. 15 (1) (2011) 544–553. [186] H. Farzin, M. Fotuhi-Firuzabad, M. Moeini-Aghtaie,
[168] Z. Luo et al, Optimal coordination of plug-in electric vehicles Stochastic energy management of microgrids during
in power grids with cost-benefit analysis—Part I: Enabling unscheduled islanding period, IEEE Trans. Ind. Inf. 13 (3)
techniques, IEEE Trans. Power Syst. 28 (4) (2013) 3546–3555. (2016) 1079–1087.
[169] R. Hosseinalizadeh et al, Economic sizing of a hybrid (PV– [187] R. Dufo-López et al, Daily operation optimisation of hybrid
WT–FC) renewable energy system (HRES) for stand-alone stand-alone system by model predictive control considering
usages by an optimization-simulation model: case study of ageing model, Energ. Conver. Manage. 134 (2017) 167–177.
Iran, Renew. Sustain. Energy Rev. 54 (2016) 139–150. [188] R.R. Deshmukh, M.S. Ballal, H.M. Suryawanshi, A fuzzy
[170] Y. Gao, et al., An Iterative Optimization and Learning-based logic based supervisory control for power management in
IoT System for Energy Management of Connected Buildings. multibus DC microgrid, IEEE Trans. Ind. Appl. 56 (6) (2020)
IEEE Internet of Things Journal, 2022: p. 1-1. 6174–6185.
[171] N.R. Wee, J.J. Jamian. Rule-Based-Iterative Energy [189] A.N. Akpolat et al, Sensorless control of DC microgrid based
Management System for Islanded Hybrid Microgrid System. in on artificial intelligence, IEEE Trans. Energy Convers. 36 (3)
Proceedings of the 6th International Conference on Electrical, (2020) 2319–2329.
Control and Computer Engineering. 2022. Singapore: Springer [190] A. Anvari-Moghaddam et al, A multi-agent based energy
Singapore. management solution for integrated buildings and microgrid
[172] S. Umetani, Y. Fukushima, H. Morita, A linear programming system, Appl. Energy 203 (2017) 41–56.
based heuristic algorithm for charge and discharge scheduling [191] H. Zhou et al, Multi-agent bayesian deep reinforcement
of electric vehicles in a building energy management system, learning for microgrid energy management under
Omega 67 (2017) 115–122. communication failures, IEEE Internet Things J. (2021).
92 B. Modu et al.

[192] P. Pan, R.K. Mandal. Fault Detection and Classification in DC [202] S. Dhar, P.K. Dash, Differential current-based fault protection
Microgrid Based on Convolutional Neural Network. in 2022 2nd with adaptive threshold for multiple PV-based DC microgrid,
International Conference on Emerging Frontiers in Electrical and IET Renew. Power Gener. 11 (6) (2017) 778–790.
Electronic Technologies (ICEFEET). 2022. [203] O.V.G. Swathika, S. Hemamalini, Prims-aided dijkstra
[193] A. Chandra, G.K. Singh, V. Pant, Protection techniques for algorithm for adaptive protection in microgrids, IEEE J.
DC microgrid-A review, Electr. Pow. Syst. Res. 187 (2020) Emerging and Selected Topics in Power Electronics 4 (4)
106439. (2016) 1279–1286.
[194] C. Li et al, Multi-sample differential protection scheme in DC [204] C. Fan et al, Assessment of deep recurrent neural network-
microgrids, IEEE J. Emerging and Selected Topics in Power based strategies for short-term building energy predictions,
Electronics 9 (3) (2021) 2560–2573. Appl. Energy 236 (2019) 700–710.
[195] R. Montoya et al, DC microgrid fault detection using [205] N. Bayati, A. Hajizadeh, and M. Soltani, Chapter 11 -
multiresolution analysis of traveling waves, Int. J. Electr. Blockchain-based protection schemes of DC microgrids, in
Power Energy Syst. 135 (2022) 107590. Blockchain-based Smart Grids, M. Shafie-khah, Editor. 2020,
[196] D.D. Patil, S. Bindu. Real time protection technique for DC Academic Press. p. 195-214.
microgrid using local measurements. in 2018 Technologies for [206] J. Marı́n-Quintero et al, Micro grids decentralized hybrid data-
Smart-City Energy Security and Power (ICSESP). 2018. driven cuckoo search based adaptive protection model, Int. J.
[197] S. Dhar, R.K. Patnaik, P.K. Dash, Fault detection and Electr. Power Energy Syst. 130 (2021) 106960.
location of photovoltaic based DC microgrid using [207] K.H. Tan et al, Intelligent control of microgrid with virtual
differential protection strategy, IEEE Trans. Smart Grid 9 (5) inertia using recurrent probabilistic wavelet Fuzzy neural
(2018) 4303–4312. network, IEEE Trans. Power Electron. 35 (7) (2020) 7451–7464.
[198] P. Pan, R.K. Mandal, Learning approach based DC arc fault [208] V. Nougain, S. Mishra, A.K. Pradhan, MVDC microgrid
location classification in DC microgrids, Electr. Pow. Syst. Res. protection using a centralized communication With a localized
208 (2022) 107874. backup scheme of adaptive parameters, IEEE Trans. Power
[199] S. Augustine et al, Fault current control and protection in a Delivery 34 (3) (2019) 869–878.
standalone DC microgrid using adaptive droop and current [209] S. Sarangi, B.K. Sahu, P.K. Rout, A comprehensive review of
derivative, IEEE J. Emerging and Selected Topics in Power distribution generation integrated DC microgrid protection:
Electronics 9 (3) (2021) 2529–2539. issues, strategies, and future direction, Int. J. Energy Res. 45 (4)
[200] M. Shamsoddini et al, A novel protection scheme for low (2021) 5006–5031.
voltage DC microgrid using inductance estimation, Int. J. [210] M. Seyedmahmoudian et al, Rationale for the use of DC
Electr. Power Energy Syst. 120 (2020) 105992. microgrids: feasibility, efficiency and protection analysis,
[201] C. Cepeda et al, Intelligent fault detection system for Energy and Sustainability V: Special Contributions 206
microgrids, Energies 13 (5) (2020) 1223. (2015) 69.