Content uploaded by Xin Lin
Author content
All content in this area was uploaded by Xin Lin on Jun 19, 2022
Content may be subject to copyright.
1
Controls of hybrid energy storage systems in microgrids: critical review, case
study and future trends
Xin Lin and Ramon Zamora
Auckland University of Technology, 1142 Auckland, New Zealand
E-mail addresses: xin.lin@aut.ac.nz, ramon.zamora@aut.ac.nz.
Abstract:
In a microgrid, a hybrid energy storage system (HESS) consisting of a high energy density energy (HEDE) storage
and high power density energy (HPDE) storage is employed to suppress the power fluctuation, ensure power balance and improve
power quality. Since the HESS integrates energy storage with slow and fast dynamic characteristics, the control system design is
a challenge. Therefore, the objective of this article is to critically analyze and compare the control methods of the HESS in
microgrid as well as to identify the shortcomings of the existing co ntrol methods. The control strategies in the HESS can be divided
into three types: centralized, decentralized and distributed. In each type, a variety of the latest control systems are discussed.
Currently, the control methods of the HESS mainly focus on achieving the high and low frequency power sharing, bus voltage
recovery, state of charge (SoC) restoration for the HPDE. Therefore, more functions should be added to the controller, such as SoC
balance for the HEDE and reducing the degradation of energy storage. However, the controller with multi-functions will increase
the complexity of the system parameter design and the order of the system. In addition, this article investigates the impact of time
delay caused by communication system on centralized and distributed controllers in the HESS. Since the HESS has the energy
storage with different dynamic characteristics, the time delay deteriorates the control accuracy and system stability. This article
also presents a novel droop coordinated control method in a case study, which is used to verify the feasibility of the simplification
and multi-function of the controller. Based on the literature reviews and case study, the insights on the future development trend
of the control strategy in the HESS, including the simplification of the comprehensive multi-function controller, compensation of
the time delay, improvement of the control accuracy, automatic tracking of the dynamic performance of the load and the
combination of the multiple energy storage, are elaborated.
Keywords: Hybrid energy storage system, centralized control, distributed control, decentralized control, communication delay,
microgrid.
1. Introduction
At present, the increasing global demand for electrical energy has led to a reduction in fossil fuels and an increase in carbon
emissions [1]. In order to solve this problem, renewable energy sources (RESs), such as photovoltaic (PV) and wind, have been
installed in a large number of residential, commercial and industrial buildings[2,3]. The global generation of the RESs is shown in
Fig. 1. Hydropower generation accounts for the largest proportion of the RESs generation, followed by wind and PV power
generation [4]. However, the inherent random and intermittent characteristics of the RESs can cause power oscillations, frequency,
and voltage instability [5,6]. Therefore, an energy storage system (ESS) is an effective solution to address the issues caused by
RESs [7]. Currently, the global energy storage demand is growing rapidly. The deployment of energy storage in the grid is
summarized in Fig. 2. In 2019, the global energy storage demand is about 10 GWh. It is predicted that it will increase by 15 times
in 2030, reaching 160 GWh. During this period, China showed the largest increase in energy storage demand at 8.6 times, followed
by Europe, United States and rest of Asia (ROA) at 5-7 times [8].
Fig. 1. RESs generation from 1965 to 2019 [4]. Fig. 2. Global energy storage demand from 2018 to 2030 [8].
2
A microgrid is a small scale power system that is proposed to integrate the ESS and RESs [9,10]. The emergence of microgrid
technology is to coordinate the contradiction between the high-penetration of the distributed RESs and main grid, and to give full
play to the value and benefits of RESs [11,12]. In addition, microgrid technology can allow end users to produce, store and manage
part of the electrical energy so that consumers can participate in the operation of the grid [13]. Fig. 3 illustrates the typical structure
of the microgrid, which can be operated in islanded and grid-connected modes [14,15]. In this microgrid, each component such as
RESs and ESSs is connected to a common bus through power electronic converters such as DC/DC and DC/AC converters for
power transmission [16]. When the power demand is higher than the power generated by RESs, the ESSs operate in the discharging
mode to supply the needed power. When the power produced from the RESs is higher than the load demand, the ESSs work in the
charging mode to absorb excess power [17,18]. Therefore, the ESSs can ensure the power balance of the system, thereby improving
the system stability [19].
Fig. 3. The schematic diagram of the microgrid.
Since different categories of the ESS have different energy density and power density, they have diverse purposes [20]. High
energy density ESS (HEDE) are adopted to supply load with slow dynamic response, while high power density ESS (HPDE) are
used to compensate transient power fluctuation [21,22]. So far, no single type of ESSs satisfies all requirements. Therefore, a
hybrid energy storage system (HESS) with different characteristics of energy storage is an effective method that can meet the
requirements of various dynamic response, energy and power density [23]. Table 1 illustrates the characteristics of some ESSs
[24–26]. A supercapacitor (SC) is a HPDE, which has the characteristics of low cost and long lifespan. A lithium-ion battery can
be considered as a HED E, wh ich has higher installation c os t a nd lower lifespan. A compressed air ESS (CAESS) is a large-capacity
ESS used in power systems. The CAESS has lower installation cost and longer service life. However, its response is slower. A
superconducting Magnetic ESS (SMESS) can be considered as a HPDE, which has lower cost and higher life time. A flywheel
ESS (FESS) is electromechanical energy storage, which includes a DC-DC converter, electrical machine and massive disk.
However, the FESS has an expensive installation cost.
Table 1
The characteristics of different ESSs.
ESS type Power density (W/kg) Energy density (Wh/kg) Cost
(€/kW) Response Lifespan
(year)
SC 200-23000 0.1-5 100-300 <10 ms 10-30
Lithium-ion battery 150-300 200-350 900-1300 3-5 ms 5-15
CAESS 0.5-2 3-6 400 3-10 min 20-40
SMESS 500-2000 0.5-10 350 1-10 ms 15-20
FESS 1000 80-200 3000-10000 >10 ms 15-20
Many published articles, such as [27–33], have reviewed the research works for the HESS. In [28], the authors review the
current topology of the HESS for the electric vehicles (EVs), and suggest a new discrete topology of the HESS. In [29], the authors
illustrate the application of battery/SC system to EVs in terms of configuration, performance, energy management system, thermal
effect and driving cycle influence. In [30,31], the authors comprehensively review the advantages and disadvantages of ESS
3
technology and EV applications. The article also presents the challenges and problems faced by the ESS technology in next-
generation EVs. However, these review articles focus on the application of the HESS in electric vehicles (EVs). In [32,33], the
authors discuss the configuration, the basic energy management system and the control system for the HESS in the microgrid. In
[34], the authors review the application of HESS in standalone renewable energy power system (REPS), including topology and
control strategies. In addition, the paper uses a decision matrix to evaluate energy storage technology and economic benefits based
on the characteristics of REPS. In [35], the authors specifically investigate the ESS control system in the microgrid, and reveal
their limitations. This article also discusses a small number of HESS control strategies. Ref. [36] summarizes the application of
HESS in smart grids and EVs in recent years. This article focuses on system configuration and energy management development.
The authors in ref. [37,38] investigate the sizing methods, topology and control methods of the HESS in the microgrid. The future
development trend of HESS technology is also discussed. However, these articles do not make a reasonable classification of control
methods, and deeply analyzed their purpose and shortcomings. The impact of communication delay on the control system of the
HESS is ignored in these articles. Besides, the previous articles [34–38] do not give any cases to provide a suggestive guideline
for the control system design of the HESS in microgrid. Hence, the issues mentioned above are completely addressed in this review
article. The main contributions are summarized as follows.
a) The control strategies in the HESS can be divided into three types: centralized, decentralized and distributed. In each
method, various latest control systems are carefully discussed and studied. Then, their limitations are given.
b) The influence of the communication delay caused by the communication network on the centralized and distributed
control methods is deeply analyzed and studied. Since the decentralized method does not require any communication
system, it will not be affected by communication delay.
c) A case study of the HESS is presented, it can provide a suggestive guideline for control design. In the case study, the
novel control methods are proposed to achieve different dynamic power distribution between battery and SC, as well as
voltage recovery simultaneously without the need for an additional voltage recovery loop. The processor-in-the-loop
(PIL) simulation is used to verify the effectiveness of the proposed method.
d) Based on literature review and case analysis, the future trends of the HESS are elaborated.
The rest of this paper is organized as follows. Section 2 presents the merits and power converter structure of the HESS. Section
3 reviews and analyzes the latest HESS control strategies. Section 4 illustrates the case study. Section 5 gives the future trends.
The conclusion is described in Section 6.
2. The overview of the HESS in microgrid
2.1. The merits of the HESS
The HESS plays a vital role in improving the stability and flexibility of the power system, as well as reducing user electricity
costs. The advantages of the HESS are summarized as follows [39–42]:
Power quality improvement: The HESS can be used to solve power quality problems such as voltage instability,
frequency fluctuations and harmonics.
Power shifting: The power generated by RESs can be stored during the low demand period and transferred to the peak
period to use, thereby reducing the power supply pressure of the main grid.
Peak shaving: During the peak electricity consumption period, the HESS can supply short-term load demand to avoid
peak electricity charges.
Spinning reserve: The HESS can be used as a backup power supply in the islanded mode of the microgrid to ensure the
stability of the system.
Lifespan enhancement: In the HESS, the HPDE can provide transient power to prevent frequent cyclic discharging and
charging of the HEDE. Therefore, the lifespan of the HEDE can be extended.
2.2. The system structure of the HESS
Fig. 4. The structure of the HESS: (a) Passive structure, (b) Semi-active structure, (c) Active structure.
4
The structure for the HESS determines its control method, service life and performance. The structure of the HESS can be
divided into passive, semi-passive and active types. The choice of structure is determined by the requirements and functions of the
system [19].
The passive structure is a simple and low-cost structure, which directly connects the HESS to the DC bus without installing any
bidirectional DC/DC power converters, as shown in Fig. 4(a). Besides, the voltage of the HESS needs to be set in advance in
accordance with the bus voltage. In the passive structure, the load sharing is mainly dominated by the internal resistance and
voltage-current characteristics of the HESS [43]. In addition, under load fluctuations, the HPDE will actively absorb high-
frequency power. Therefore, the HPDE can be considered as a low-pass filter. The limitations of passive structures are summarized
as follows [44,45]:
The control flexibility is poor.
Load sharing depends on the internal resistance of energy storage.
The voltage of HESS needs to be strictly consistent with the bus voltage.
In the semi-active structure, an energy storage is connected to the DC bus through a dc/dc power converter. Then, a control
system is required to be designed to achieve power exchange and to stabilize the bus voltage. Another energy storage is directly
connected to the DC bus [46]. The principal structure is displayed in Fig. 4(b). Although the semi-active structure increases extra
cost, it provides a certain degree of control flexibility. However, the semi-active structure still has some limitations as follows [47]:
This structure can only provide partial control.
The voltage of energy storage without a power converter is still required to be consistent with the bus voltage.
In the active structure, both HPDE and HEDE are connected to the DC bus through the DC/DC bidirectional power converters,
as shown in Fig. 4(c). The control methods need to be proposed to regulate the power converters to realize the different dynamic
power distribution between HPDE and HEDE, as well as guarantee the bus voltage within a safe range [48]. Although the active
structure effectively improves the controllability of HESS, it still has some shortcomings as follows [38]:
The installation cost is higher.
The power conversion loss increases.
3. Control approaches for HESS
Fig. 5. The schematic diagram of the control structure for the HESS: (a) Centralized, (b) Distributed, (c) Decentralized.
The purpose of the HESS control is to ensure the stability of the bus voltage, limit the state of charge (SoC) of the ESSs within
a safe range, and achieve the high and low frequency power distribution between HPDE and HEDE, thereby improving the stability
5
of the system and extending the lifespan of the ESSs. The HESS control strategies can be broadly summarized into three types
[49]: (a) centralized, (b) decentralized, and (c) distributed. In the centralized control, a central controller (CC) is required to adjust
the local controllers (LC) of the HPDE and HEDE through the communication system. The CC collects system data and makes
decisions for the LCs [50,51]. In decentralized method, each converter is controlled by its LC, which only receives local data, and
is not aware of the overall status of the system as well as the working conditions of other controllers. In addition, there is no
communication network among the controllers [52,53]. In distributed multi-agent control, each controller is considered as
individual agent, which will receive local information, and will also collect data from neighbors through a sparse communication
network to achieve coordinated control [54,55]. The schematic diagram of the control types is presented in Fig. 5. It should be
noted that this article will cite a small number of articles on the applications of the HESS in electric vehicles. This is because power
distribution circuits in electric vehicles have similar topology and operation principles as in islanded microgrids, in which local
generation units have enough power to supply local loads.
3.1. Centralized control
3.1.1. Filtration based control
Fig. 6. The schematic diagram of the FBC for the HESS [56].
A filtration-based control (FBC) employs a filter to decouple the total current into high and low currents, which will serve as
the reference current for SC and battery controllers. An example of the control structure of the FBC is shown in Fig. 6. In the
controller, the bus voltage is compared with the reference voltage to generate a voltage error, which is transmitted to the
proportional-integral (PI) compensator. The compensator will calculate the total current of the HESS. The total current is divided
into high and low frequency parts through a low pass filter (LPF). The high frequency part will be used as the reference current for
the SC current control, while the low frequency part will be used as the reference current for the battery current control.
Many studies have reported the FBC for the HESS [56–61]. A novel control strategy for the HESS in [57,58] is proposed to
achieve the fast and low dynamic current sharing between the battery and SC. Besides, a current limiter and battery current
feedback loop is presented to suppress the battery peak current. In [59], a SC is used for short-term charging and discharging to
extend the battery life in small-scale wind energy systems. The supervisory control algorithm is used to transfer the high frequency
component of the system charging/discharging current to the SC by the active current filtering method. In addition, the long-term
benefits of the proposed system can be evaluated through experimental simulations. Ref. [60] proposed a unified energy
management scheme for renewable grid integrated with battery-SC system. The proposed energy management system mainly
includes power management algorithm (PMA), current generator, and switching pulse generation of each converter. The total
current is decoupled into average and transient currents through an LPF and limiter. The average current of the battery is generated
by the LPF, while the transient current of the SC is generated by subtracting the average current from the total current. The PMA
can determine the operating mode of the system based on changes in the RES power and load. In each mode, the operating target
is determined according to the current SoC state of energy storage. However, this method will cause the battery charge/discharge
rate to increase sharply during the mode switching, thereby reducing the battery lifespan.
Table 2
The critical analysis of the HESS based on the FBC.
Ref. Method ESS type Contributions Limitations
[56] FBC Battery & SC
A low-pass filter (LPF) is installed in the controller to separate the fast and low
dynamic currents, which are used as the reference current of the SC and the
b
attery.
The dynamic response of SC is slow and
cannot be used continuously. The peak
current of the battery is relatively large.
[57,58] FBC Battery & SC The proposed controller can realize power distribution and suppress the peak of
b
attery current by adding a current limiter and battery current feedback.
The SoC level of battery and SC is not
considered.
[59] FBCBattery & SC
The supervisory control algorithm is based on the active filter to realize the
power distribution of the HESS in the wind power system, and also con sider s the
long-term benefits of the system.
There is a deviation in the bus voltage. The
SC will have insufficient energy.
[60] FBC Battery & SC
The energy management system can separate the total current into high and low
frequency currents according to the LPF and current limiter. Besides, the PMA
can allow the system to operate in different modes.
During the mode switching, the peak current
of the battery will increase instantly.
[23] FBC Battery & SC The energy management system can achieve the power sharing, ensure the SoCs
of the battery and SC within a safe rage, and alleviate the bus voltage deviation.
Communication failures and delays will
cause the system to fail to operate normally.
[61] FBC Battery & SC A controller including energy management and voltage control is proposed for
the battery-SC system in photovoltaic (PV) power system. Under different
The dynamic response of SC is slow and
cannot be used continuously.
6
operating conditions, the influence of the filter constant on the current gain,
energy loss gain and load power efficiency of the battery is investigated.
[62] FBC Battery & SC
An energy management system is proposed for a remote area wind power system
including a doubly-fed induction generator (DFIG), battery-SC system, and
various loads. The power demand is divided into the low frequency and high
frequency components by adopting a high pass filter (HPF). The present method
also reduces the depth of discharge of the battery.
There is a deviation in the bus voltage. The
SC will face a lack of energy.
[63] FBC-
SMC Battery & SC
An FBC based on sliding mode control is proposed (SMC) for the battery-SC
systems. In this method, the FBC divides the total power into low and high
frequency components, which are distributed to the battery and SC. In addition,
the DC bus voltage is regulated by the SC. The traditional PI control in SC's
converter control system is replaced with SMC.
This method does not consider the SoC state
of the SC, which will cause the SC to not have
enough power for future use.
The critical analysis of the HESS based on the centralized FBC is presented in Table 2. These methods all use LPF or HPF to
decompose low and high frequency power, which are used to generate the battery and SC reference currents. However, these
methods are only suitable for a single HESS (including a battery and SC) instead of multiple HESSs. This is because the central
FBC may not be able to achieve proportional power distribution among multiple HESS systems. Besides, these methods require
the collection of the battery and SC data through a communication system. Therefore, these methods suffer from communication
delays and single points of failure. Moreover, the expansion of the HESS will require a redesign of the controller parameters,
thereby increasing the amount of calculation and complexity.
3.1.2. Fuzzy logic control
The fuzzy logic control (FLC) is usually employed in HESS to generate reference power/current for battery and SC controllers
[64,65]. An example of the control structure of the FLC is shown in Fig. 7. In the control system, the FLC can generate reference
power, which is decomposed into average and transient power by LPF. In addition, the power sharing based FLC is presented to
reasonably distribute the active power among batteries and among SCs, as well as between batteries and SCs.
Fig. 7. The schematic diagram of the FLC for the HESS [65].
Many works have studied the FLC for the HESS [65–68]. Ref. [66] proposes a real-time energy management scheme for hybrid
ac/dc microgrid including RES and HESS. The structure of the HESS is semi-active type. The battery is connected to the common
bus through a DC/DC converter, and the SC is directly connected to the common bus. The proposed method can deal with the
uncertainty of RES, effectively manage the balance of energy supply and demand, and achieve optimal economic operation. The
FLC only intervenes in the system when the power demand is greater than the power supply. In this case the battery is in discharge
mode. The FLC manages the output power of the battery, and considers the ratio of the currently available power of the battery to
the peak power consumption. Ref. [67] proposes a multi-mode fuzzy logic power allocator for HESS with battery and SC in the
PV system to compensate for the imbalance between supply and demand. The fuzzy logic power allocator consists of three power
sharing modes. The FLC adaptively allocates power to the battery and SC under different conditions according to their SoC levels,
and makes full use of the characteristics of the battery and SC to improve power compensation efficiency. Another FLC is used to
perform the energy exchange between the battery and the SC, thereby preventing excessive use of energy storage. Therefore, the
SoC of the battery and SC can be maintained within a safe range. In [68], an optimal control method for the PV power system
consisting of the battery-SC system is proposed to reduce the high transient current demand and dynamic stress of the battery. The
proposed method includes both FBC and FLC. The FBC divides the total current into low and high frequency parts. The FLC
reduces the peak current of the battery while constantly considering the SoC state of the SC. The particle swarm optimization (PSO)
algorithm is used to optimize the membership function of the FLC to maximize the reduction of battery peak current.
Table 3
The critical analysis of the HESS based on the FLC.
Ref. Method ESS type Contributions Limitations
[65]FLCBattery & SC
An energy management system based on the FLC is proposed for the medium
voltage dc shipboard power system. In addition, an FLC-based power
allocation strategy is proposed to realize the expansion of the HESS.
The battery voltage is strictly required to be
consistent with the bus voltage. The
p
erformance of the battery is low.
[66] FLCBattery & SC A real-time energy management scheme for hybrid ac/dc microgrid including
RES and HESS. The proposed method can deal with the uncertainty of RES,
The SC will not have insufficient energy for
future use. The bus voltage will deviate.
7
effectively manage the balance of energy supply and demand, and achieve
optimal economic operation. The FLC only intervenes in the system when the
p
ower demand is greater than the power supply.
[67] FLCBattery & SC
A multi-mode fuzzy logic power allocator for HESS with battery and SC in the
PV system to compensate for the imbalance between supply and demand. The
FLC adaptively allocates power to the battery and SC under different conditions
according to their SoC levels. Another FLC is used to perform the energy
exchange between the battery and the SC, thereby preventing excessive use of
energy storage.
The bus voltage restoration system is ignored.
[68] FLC-
PSO Battery & SC
An optimal control for the PV power system consisting of the battery-SC
system is proposed to reduce the high transient current demand and dynamic
stress of the battery. The FBC divides the total current into low and high
frequency parts. The FLC reduces the peak current of the battery while
constantly considering the SoC state of the SC. The PSO algorithm is used to
optimize the membership function of the FLC to maximize the reduction of
b
attery peak current.
The SoC state of the battery and the bus voltage
restoration system are ignored.
[69] FLC Battery & FC
A power management system based on the FLC with the water cycle algorithm
is proposed for green power generation. The proposed method determines the
power generated by the FC. The water cycle algorithm is used to optimize the
membership function of the FLC to minimize power loss and operating cost.
The battery voltage is strictly required to be
consistent with the bus voltage. The battery will
supply the transient power. The controller
calculation time is longer.
[70] FLC Battery & SC
A power management system based on multiple optimally-designed FLC is
proposed to reduce the HESS current fluctuations, system operating costs and
power supply probability losses in the autonomous hybrid green power
systems.
The controller is more complicated. The SoC
level of battery and the SC is not considered.
[71] FLC Battery & SC
An energy management scheme based on FLC is proposed for a semi-active
HESS including a battery and SC. The SC is connected to the DC bus through
a DC/DC converter, and the battery is directly connected to the DC bus. The
proposed control method can effectively suppress the peak current of the
b
attery and ensure that the voltage of the SC is within the set range.
The SC will have insufficient energy and cannot
be used.
[72] FLC Battery & SC
The online FLC based on PSO is presented to regulate the battery-SC system
in the wind-diesel power system. Under wind speed and load changes, the
proposed method can effectively prevent the instantaneous high current of the
battery, and ensure that the system voltage and frequency are within a safe
range.
The recovery system of the bus voltage is
ignored.
[73] FLCBattery & SC
The FLC is presented to suppress the instantaneous current of the batter y un der
pulse load conditions in the naval microgrid. The experimental setup is
established at the University of Texas to verify the feasibility of the proposed
method.
The bus voltage restoration system is ignored.
The SoC level of battery and the SC is not
considered.
The critical analysis of the HESS based on the FLC is presented in Table 3. In these methods, the FLC is mainly used to generate
reference power for different energy storages, and reduce the peak current of the battery. Therefore, the life of the battery can be
extended. However, this method is more suitable for a single HESS including a battery and a SC. For multiple HESSs, additional
power distribution controllers need to be developed. Therefore, the entire control design becomes more complicated. In addition,
this method suffers from communication delays and single points of failure.
3.1.3. Rule based control
Rule-based control (RBC) is achieved by formulating a series of predefined and empirical control rules [74]. An example of the
control structure of the RBC can be presented in Fig. 8. In the controller, the RBC contains a series of predetermined control rules,
which generate corresponding control signals according to the system state. Then, the control signal will be sent to the battery and
UC controllers through the communication network for mode switching.
Many works have studied the RBC for the HESS [75–78]. Ref. [76] proposes a novel hierarchical optimized energy management
system for the battery-SC system. This method is based on the vehicle-to-cloud connectivity and is developed using dynamic
programming to achieve optimal power distribution and reduce battery degradation. At the local level, model predictive control
(MPC) is developed to deal with system constraints and uncertainties, as well as reduce energy loss. At the system level, a rule-
based energy management strategy is proposed to optimize battery power supply, thereby maximizing battery life. The SC is
considered as a buffer to reduce the power peak of the battery. Ref. [77] proposes a novel hybrid energy management strategy
integrated with the PV, FC, electrolyzer, battery and SC for a remote house. The proposed energy management system can
effectively control the power balance in the system and determine the power supply of each power source. In [78], an advanced
two-layer control strategy for a renewable hybrid power system in islanded mode is proposed. This hybrid system includes a wind
turbine (WT), battery, FC and electrolyzer. The top layer is the power management and power regulation system, which generate
reference dynamic operating points to the low-level control system based on wind and load conditions. In addition, the top-level
system can also perform load dispatch when wind energy and energy storage are insufficient, so as to avoid system power outages.
The low-level control adjusts the output power of the WT, battery and FC based on the reference dynamic operating point.
The critical analysis of the HESS based on the RBC is given in Table 4. In these methods, the RBC is to control the HESS by
adopting a predetermined rule that is usually generated based on system operation and mathematical models. However, this method
will lead to inaccurate power output of the HPDE and HEDE in the HESS. In addition, the RBC needs to send instructions to the
8
HESS through the communication network. This means that communication delays and single points of failure will cause the
performance of the control system to decrease. In addition, the scalability of the system is poor.
Fig. 8. The schematic diagram of the RBC for the HESS [75].
Table 4
The critical analysis of the HESS based on the RBC.
Ref. Method ESS type Contributions Limitations
[75]RBCBattery & SC
A novel optimal energy management strategy (NOEMS) is proposed for a hybrid
po wer g ene rat ion syst em t hat comb ine s a H ESS, off sho re w ind e nergy and ocean
current energy. The NOEMS can ensure power balance, and regulate the power
flow between the battery and the SC by minimizing the power fluctuation of the
system.
The proposed energy management system
ignores the SoC level of the SC, which may
result in that the SC does not have enough
power for the next use.
[76] RBC Battery & SC
A novel hierarchical optimized energy management system is proposed to
achieve optimal power distribution and reduce battery degradation fo r the HESS .
At the local level, a MPC is developed to deal with system constraints and
uncertainties, as well as reduce energy loss. At the system level, a rule-based
energy management strategy is proposed to optimize battery power supply,
thereby maximizing battery life.
The control system is more complicated and
does not consider the SoC state of the SC.
[77] RBC Battery & SC
The proposed energy management system can effectively control the power
balance in the system and determine the power supply of each power source for
a remote house.
The SoC status of the battery and SC are
ignored. The bus voltage will deviate.
[78] RBC Battery & FC
An advanced two-layer control strategy is proposed for a renewable hybrid
power system in islanded mode. The top layer is the power management and
power regulation system, which generate reference dynamic operating points to
the low-level control system based on wind and load conditions. The low-level
control adjusts the output power of the WT, battery and FC based on the
reference dynamic operating point.
The battery will be exposed to instantaneous
release and absorption of excessive current,
thereby reducing its life.
9
[79] RBC Battery & SC
A load prediction energy management system based on a support vector machine
(SVM) is proposed to control the power interaction among the PV, battery-SC
ESS and load. The proposed energy management system based on the SVM can
predict the PV power generation and load demand in advance, thereby
determining the discharge capacity of energy storage equipment.
The controller has a large amount of calculation
and a slow response speed.
[80] RBC Battery & SC
An effective energy management system is proposed for DC microgrid that
consists of the RES, variable load, HESS and standby diesel generators. The
proposed energy management system determines the charge and discharge of the
battery based on the power generation of the RES and the SoC level of the
battery. In addition, the current controller acts as the main controller of the DC
b
us.
The DC bus voltage cannot be restored to the
initial value. The power distribution between
the battery and the SC is not precise.
[81] RBC Battery & SC
An energy management scheme is proposed for the HESS including a battery
and SC in microgrid. The proposed method can adjust the operation of the HESS
according to the real-time status of the system. Therefore, the balance between
power generation and load demand can be ensured not to be disrupted. In
addition, the HESS has a longer service life, thereby reducing the operating cost
of the microgrid.
The SC is always in the state of charge and
discharge, thereby reducing the cycle of use.
There is an error in the bus voltage.
3.1.4. Model predictive control
MPC is a control technology based on online optimization, which can be used for power allocation and reference current/power
determination in the HESS according to different constraint conditions [82,83]. An example of the control structure of the MPC is
shown in Fig. 9. In the control system, the measured parameters are sent to the MPC as input. Then, the MPC will calculate the
power allocation in the HESS according to the preset constraint conditions and generate a modulated signal. The advantage of this
method is to effectively limit the current and SoC of the battery and UC within a safe range.
Fig. 9. The schematic diagram of the MPC for the HESS [83].
Table 5
The critical analysis of the HESS based on the MPC.
Ref. Method ESS type Contributions Limitations
[83]MPCBattery & SC
A novel MPC is applied to the battery-UC HESS and carries out
experimental verification. The objective function of the MPC refers
to minimizing the voltage change of the UC and the peak value of the
battery current. Therefore, the SoC of the UC and the output current
of battery can be limited within the preset range.
When the system load changes, the bus voltage
will deviate and will not automatically return to
the initial value.
[84]
Nonlinear
receding &
quadratic MPCs
Battery & SC
A two-layer energy management system is proposed for microgrids
with the HESS to achieve high system robustness with minimal
operating costs. The high-level system is a nonlinear receding MPC
used to minimize the total operating cost of the microgrid. The low-
level system is a quadratic MPC used to adjust the power output of
the HESS and eliminate system fluctuations caused by prediction
errors.
The bus voltage recovery system and the SoC
compensation control of the SC are not
considered.
[85] Real-time MPC Battery & SC
A real-time MPC is proposed for a shipboard microgrid under the
conditions of propulsion load fluctuation and pulse power load. The
proposed control can optimize the operating efficiency of the system,
extend the cycle time of the battery, and ensure the power quality and
system reliability.
The uncertainty of the parameters of the hybrid
system is not considered.
[86] Adaptive MPC Battery & FC
An online parameter recognition system is developed to reduce
parameter uncertainty. The MPC is used to optimize the power
distribution between the battery and SC and to handle different
dynamic responses.
The controller is more complicated and
computationally expensive.
10
[87] MPC Battery &
SMESS
A fast MPC is proposed to achieve voltage control and optimal power
allocation strategy in the HESS. The proposed method can adjust the
DC bus voltage by simplifying the switching state and one-step
prediction, thereby enhancing the anti-interference ability of the
system.
The proposed method will cause the power
supply of SMESS to be longer, thereby reducing
its continuous use.
[88] MPC Battery & SC
& FC
An MPC based on mixed integer quadratic programming is proposed
for the microgrid including a WT, PV power generation, electrolyzer,
FC, batteries and SC. The MPC can maximize the economic benefits
of the microgrid under different constraints, while minimizing the
degradation of each energy storage system.
The proposed method does not consider the
power distribution between different energy
storages. In addition, the MPC system is subject
to a number of constraints.
[89] MPC Battery & SC
& FC
A novel power management system based on the MPC is proposed
for the FC-battery-SC system. The purpose of this control is to
optimize the load sharing among the battery, FC, and SC while
ensuring the maximum economic benefits of the microgrid.
The controller has many constraints. In
addition, the controller is more complicated.
[90] MPC Battery & SC
& FC
A new MPC-based energy management system is presented for
islanded microgrid integrating the PV, load, battery, SC and FC. In
the proposed method, the maximization of operating efficiency and
the minimization of the degradation of HESS are formulated as an
optimization problem in MPC. In addition, the proposed energy
management system can adjust the power supply of the HESS,
thereby increasing the utilization rate of the PV system.
The proposed method ignores the system
voltage recovery system and the SC's SoC
restoration control.
Several works have studied the MPC for the HESS [84–86]. Ref. [84] proposes a two-layer predictive energy management
system for microgrids with the HESS to achieve high system robustness with minimal operating costs. The HESS consists of a
battery and SC. Considering the degradation cost of the HESS, the long-term cost model of the battery and SC is developed and
transformed into a short-term cost model related to real-time operation. In the proposed energy management system, the high-level
system is a nonlinear receding MPC used to minimize the total operating cost of the microgrid. The low-level system is a quadratic
MPC used to adjust the power output of the HESS and eliminate system fluctuations caused by prediction errors. Ref. [85] proposes
a real-time MPC applied to a shipboard microgrid under the conditions of propulsion load fluctuation and pulse power load. The
HESS combining a battery and SC acts as a buffer to suppress the load fluctuation of the system. Three control objectives are
considered in the proposed control strategy. The first is to ensure the power quality and system reliability. The second is to optimize
the operating efficiency of the system. This means that the power loss of the HESS must be minimized. The third is to extend the
cycle time of the battery. Experimental equipment is built to verify the feasibility of the proposed method. However, this method
does not consider the uncertainty of the parameters of the HESS. To solve this problem, an adaptive MPC is proposed for the
shipboard microgrid in [86]. An online parameter recognition system is developed to reduce parameter uncertainty. The MPC is
used to optimize the power distribution between the battery and SC and to handle different dynamic responses.
The critical analysis of the HESS based on the MPC is given in Table 5. These methods are mainly to optimize the power
distribution between different energy storages, the operation of the microgrid and the economic benefits. However, these methods
usually require the design of an accurate mathematical model for the MPC including many constraints that increase the
computational burden of the controller. In addition, these methods are not suitable for multiple HESSs. This is because the control
system has to consider more constraints, and the system design is more complicated.
3.1.5. Optimization based control
Optimization based control (OBC) is used to optimize power distribution, increase energy storage life and reduce investment
costs in the HESS [91,92]. An example of the control structure of the OBC is given in Fig. 10. In the controller, an LPF is used to
divide the total power into high and low frequency power, and they will generate the optimal reference power for the battery and
SC through the multi-objective optimization problem (MOP) based power sharing algorithm. Several works have studied the OBC
for the HESS [92–95]. In [92], an intelligent energy management system based on particle swarm optimization combined with
Nelder-Mead (PSO-NM) for the HESS is proposed to minimize the power stress of the battery and to extend its life. The key idea
of this method is the active change of the power sharing limits based on the SoC of the SC. In addition, the aging prediction of the
battery is estimated by building a mathematical model, and its parameters are completed through accelerated aging tests in the
laboratory. In [93], a two-stage optimal power control framework is proposed for the battery-SC HESS. The two stages in this
method are used to calculate the reference voltage of the SC and to optimize the output power of the HESS. In the proposed
framework, a method based on real-time load dynamics is used to calculate the SC reference voltage. In addition, the HESS power
control is represented as a convex optimization problem while minimizing battery power and power loss. Ref. [94] proposes a
simulated annealing particle swarm optimization (SAPSO) algorithm to optimize the capacity of a HESS in a wind/solar power
generation system. This method searches for the optimal HESS capacity allocation according to the load demand change, thereby
effectively reducing the operating and investment costs of the system. The advantage of the proposed SAPSO is that it improves
the global search ability of the PSO algorithm, prevents the algorithm from falling into local extremes, and increasing the diversity
of particle swarms. Ref. [95] presents an optimized configuration based on an improved genetic algorithm (GA) and an expert
system for a HESS in wind power generation. The HESS consists of a battery and SC. This method is based on the grid energy-
11
saving efficiency and static voltage stability index to obtain the wind power generation curve. Subsequently, the expert system was
added to the improved GA to obtain optimized configuration parameters.
The critical analysis of the HESS based on the OBC is given in Table 6. These methods can usually achieve multi-objective
optimization, such as HESS's energy consumption, operating and investment costs. However, with the increase of the optimization
target, the constraint conditions increase accordingly, so that the computational burden of the controller becomes heavier. In
addition, these methods require the establishment of an accurate mathematical model of the system. Otherwise, the accuracy of the
optimization results will be reduced.
Fig. 10. The schematic diagram of the OBC for the HESS [91].
Table 6
The critical analysis of the HESS based on the OBC.
Ref. Method ESS type Contributions Limitations
[91] OBC Battery & SC
An energy management system based on MOP is proposed for the
battery-SC system. The power loss, the SoC level and power limitation
of the SC are considered as the optimization objectives. Besides, a linear
weighted summation algorithm based on variable weights is developed
to solve the MOP to obtain the reference current of the battery and SC.
This method does not optimize the power distribution
between the battery and the SC. In addition, the system
voltage recovery control is ignored.
[92] OBC-
PSO-NN
Battery & SC
& FC
An intelligent energy management system based on PSO-NM for the
HESS is proposed to minimize the power stress of the battery and to
extend its life.
The proposed method does not achieve the optimal
power distribution between the battery and SC.
[93] OBC Battery & SC
A two-stage optimal power control framework is proposed to calculate
the reference voltage of the SC, minimize battery power loss, as well as
optimize the output power of the HESS.
The proposed method ignores the optimal power
distribution between battery and SC.
[94] OBC-
SAPSO Battery & FC
A SAPSO algorithm is proposed to optimize the capacity of a HESS in
a wind/solar power generation system based on the load demand change,
thereby effectively reducing the operating and investment costs of the
system.
This method does not consider the bus voltage
recovery system, and the voltage of the SC.
[95] OBC-
GA Battery & SC
An optimized configuration based on an improved GA and an expert
system is proposed to optimize the system parameters for a HESS in
wind power generation.
This method does not consider the degradation
problem of hybrid ESS. In addition, the optimal power
b
etween the battery and the SC is not considered.
[96] OBC-RP Battery & SC
A multi-objective optimization energy management system based on
Radau pseudospectrum (RP) is proposed for the battery-SC system. The
proposed scheme transforms the optimal control problem into a
nonlinear programming problem by approximating the global
interpolation polynomial to the state and control variables in the system.
Then, the coefficient nonlinear optimizer is used to solve this problem.
The energy consumption of the system and battery lifespan are
expressed as the objective function to obtain the Pareto solution set.
The battery voltage requirement is consistent with the
bus voltage. In addition, the constraint condition does
not consider the SoC level of the SC.
[97] OBC Battery & SC
A capacity optimization algorithm is proposed for the microgrid that
integrates the battery-SC and PV-WT. The goal of this method is to
improve the reliability of the system, minimize the operating cost and
the emission of greenhouse gases.
This method does not consider the degradation cost of
the battery-SC system.
3.1.6. Artificial neural network
Artificial neural network (ANN) is an intelligent algorithm that uses collected data for training and verification to optimize the
system parameters of the HESS [98,99]. An example of the control structure of the ANN for the HESS is displayed in Fig. 11. In
the controller, a supervisory control includes a back propagation (BP) network to train measured data such as current to generate
a reference current for the current controller. Many research works, such as [100–103], have studied the ANN for the HESS. Ref.
[100] proposes a neural network control system based on online training for a hybrid microgrid. The microgrid is composed of a
PV-WT, FC and battery. In the proposed scheme, the feedforward neural network (FNN) is used to adjust the DC bus voltage to
ensure that it is within the preset range. Two controllers based on the Elman neural network (ENN) are used to realize the two-
way flow of active power and the compensation of reactive power of AC load. In addition, an energy management system based
on the FLC is developed to minimize the purchase of electricity from the main grid, thereby reducing the operating cost of the
microgrid. Ref. [101] proposes an intelligent control system for grid-connected microgrid composed of a PV, FC and battery. In
the proposed control system, an ENN-based controller based on online training was developed to track the optimal operating point
12
of the PV power supply. In addition, a fuzzy gain coordinator based on Takagi-Sugeno-Kang is proposed to adjust the PI controller
parameters of the FC and battery. Finally, a neuro-fuzzy-based gain coordinator is used to manage the power flow among the
microgrid, local load and main grid. Ref. [102] proposes an advanced method to determine the capacity of the battery-SC system
in a microgrid. In the proposed method, an empirical mode decomposition technology is used to analyze historical wind power
data to establish the intrinsic mode function of wind power. According to the instantaneous frequency and time distribution of the
intrinsic mode function, the wind energy is decoupled into the high and low frequency power part. The high frequency power is
allocated to the SC, and the low frequency part is taken care of by the battery. Therefore, smooth power transmission of the system
can be achieved. Then, the latest smoothness standard is used to evaluate the smoothness of the power transmitted to the load.
Then, the back propagation neural network (BPNN) is used to determine the minimum capacity of the battery-SC, thereby reducing
the cost of the system.
The critical analy sis of the HESS based on the ANN is given in Table 7. These methods are focused on adjusting the bus voltage,
determining the size of the HESS, extending battery life and reducing system operating costs. However, these methods ignore the
current leakage and degradation problems of the SC. In addition, the ANN requires a large amount of historical data for training
and learning, which leads to an increase in the computational burden of the controller.
Fig. 11. The schematic diagram of the ANN for the HESS [99].
Table 7
The critical analysis of the HESS based on the ANN.
Ref. Method ESS type Contributions Limitations
[99] BPNN Battery & SC
A real-time supervisory control combining a BPNN and dynamic
programming is proposed for the battery-SC HESS. The optimization
goal of the controller is to simultaneously minimize the power loss of
the HESS and the peak value of the battery current, thereby prolonging
the battery life and reducing costs.
The proposed method ignores the service life of SC.
[100] FNN &
ENN Battery & FC
A neural network control system including a FNN and ENN is
proposed to achieve the adjustment of the DC bus voltage, the two-
way flow of active power and the compensation of reactive power of
AC load for a hybrid microgrid consisting of a PV-WT, FC and battery.
In addition, a FLC is developed to minimize the purchase of electricity
from the main grid, thereby reducing the operating cost of the
microgrid.
Since the system does not have HPDE such as SC, the
transient power of the battery is larger, and its service
life will be shortened.
[101] ENN Battery & FC
An intelligent control system integrating a ENN and fuzzy gain
coordinator is proposed to achieve the tracking of the optimal
operating point of the PV power supply, adjustment of the PI controller
parameters and power balance for grid-connected microgrid composed
of a PV, FC and battery.
Since the system does not have HPDE such as SC, the
transient power of the battery is larger, and its service
life will be shortened.
[102] BPNN Battery & SC
An advanced method including an empirical mode decomposition and
BPNN is proposed to achieve the power sharing and determine the
capacity of the battery-SC system in a microgrid.
This method does not take into account the increase
in cost caused by battery degradation.
[103] HBSAN
N Battery & SC
An energy management system based on hybrid bat search and ANN
(HBSANN) is proposed to maintain the b us vo ltag e wit hin a safe range
and achieve power sharing for the battery-SC HESS in a DC microgrid.
Therefore, the power supply pressure of the battery is reduced and life
is extended.
When the SC provides transient power, the SoC of the
SC cannot automatically return to the set value.
[104] ANFIS Battery & FC
An energy management system based on Adaptive Neuro-Fuzzy
Inference System (ANFIS) is proposed for microgrid consisting of a
WT, PV, FC, electrolyzer and battery. The proposed ANFIS is used to
determine the power produced and stored by energy storage, while also
considering the power demand and the SoC level of the battery.
The proposed method does not consider the peak
current of the battery.
[105] WNN Battery & SC
A rolling optimization strategy based on the wavelet neural network
(WNN) and dynamic programming is proposed to optimize the state of
the HESS to determine its power output, thereby improving system
efficiency and extending battery life.
This method only focuses on extending the battery
life cycle, while ignoring the life of the SC.
13
3.1.7. The communication delay on centralized method
In the centralized method, the control command is sent to the LCs through the communication network. The communication
time delay of the signal transmission between the CC and the LCs is usually ignored in research studies. The time delay can cause
voltage and current fluctuations in distributed energy systems [106,107]. In addition, the time delay reduces the convergence speed
of the system and may deteriorate the dynamic performance of the system in practice [108,109]. Considering that the HESS has
different dynamic performance, time delay may weaken the performance and stability of centralized control. Therefore, the time
delay in the centralized control needs to be studied and discussed in the HESS. An example of the control structure by considering
the time delay is given in Fig. 12. In the control system, a frequency sharing control algorithm by considering the time delay is
proposed to quickly respond to load fluctuations in the HESS [110]. The time delay decreases the bandwidth of the current control,
thereby resulting in an error in the output current of the battery during transient period. In order to solve this problem, the proposed
algorithm uses the SC current loop to compensate the error current of the battery. Besides, energy management algorithm (EMA)
is used to limit the SC voltage, battery SoC level, battery and SC peak currents.
In [111], a time-delay small-signal stability model is proposed for the time-delay stability analysis of a hierarchical control in
the HESS. Then, an advanced method is used to determine the index of maximum delayed time (MDT), which is used to evaluate
the time delay stability margin in the HESS. In the simulation results, the proposed method can determine the maximum time delay
that the HESS can withstand without losing stability. In addition, the results also show that the HPDE is more sensitive to time
delay stability than the HEDE. In [112], a battery-SC HESS based on semi-active topology is presented. In this system, a small
signal model based on time delay is proposed to determine the MDT of hierarchical control in the HESS, so as to ensure the
stability of the system. In [113], a new type of frame with double-layer structure for the HESS is proposed to solve the optimization
problem of aggregate slope capacity, that is, energy storage ramping optimization (ESRO). In this framework, the upper system is
used to solve the optimization problem of aggregation slopes of multiple HESSs. The lower system is used to optimize the LC of
a single HESS according to the energy storage characteristics. Then, the result of the lower-level optimization will be fed back to
the upper-level system, thereby updating the optimization algorithm parameters in the upper-level system. In addition, a
communication delay is considered in the proposed system.
The critical analysis of the communication time delay for the centralized control in the HESS is given in Table 8. These methods
are mainly to determine the maximum delay time and to analyze the impact of time delay on system stability. However, time delay
compensation has not been discussed for maintaining stability an d powe r loss, espec ially for the HPD E that is h igh ly tim e-s ensitive.
Fig. 12. The schematic diagram of the control structure with considering the time delay for the HESS [110].
Table 8
The critical analysis of the communication time delay in the centralized method for the HESS.
Ref. Method ESS type Contributions Limitations
[110]EMABattery &
UC
A frequency sharing control algorithm is proposed for the HESS to
compensate the error current caused by the time delay. The EMA is used
to limit the SC voltage, battery SoC level, battery and SC peak currents.
This method will increase the SC's power supply
burden and neglect the maximum delay time and
the SC's delay compensation. In addition, the
communication delay between the EMA and the
LC is ignored .
[111] MTD Battery & SC
A time-delay small-signal stability model is proposed for the time-delay
stability analysis of a hierarchical control in the HESS. Then, an advanced
method is used to determine the index of MDT, which is used to evaluate
the time delay stability margin in the HESS.
This method does not consider time delay
compensation for the controller.
[112] MTD Battery & SC
A small signal model based on time delay is proposed to determine the
MDT of hierarchical control in the HESS, so as to ensure the stability of
the system.
This method does not consider time delay
compensation for the controller.
[113] ESRO Battery & SC
A new type of frame with double-layer structure for the HESS is proposed
to solve the optimization problem of aggregate slope capacity, that is,
ESRO. The communication delay is considered in the proposed system.
This method just adds the time delay parameter to
the controller and does not consider the influence
of the time delay on the stable line and delay
compensation.
[114] LKF Battery & SC
An optimized control algorithm with input saturation and sampling time
delay is proposed for the battery-SC HESS. In this method, a state feedback
system based on the state space average model is developed to track the
DC bus voltage and battery current. A Lyapunov-Krasovskii functional
This method does not consider the maximum
delay time and delay compensation. In addition,
this method focuses on the sampling delay
b
etween the controller and the actuator, rather
14
(LKF) is built to evaluate the delay information. Then, a Wirtinger-based
integral inequality is used to establish a time delay stability stan dard f or the
battery-SC system. In addition, an optimal control algorithm based on
linear matrix inequality for actuators is presented to reduce the transition
time of load switching.
than the communication delay between the CC
and the LC.
3.2. Decentralized control
3.2.1. Droop based control
In the droop-controlled HESS, power distribution method is achieved by adjusting the reference voltage through the change of
output power [115]. Since the traditional droop based control (DBC) cannot meet the different dynamic characteristics of the HESS,
the DBC needs to be further studied and developed [37,116]. An example of the control structure of the DBC for the HESS is
displayed in Fig. 13. In this control scheme, an integral droop (ID) control derived from the characteristics of the capacitor is used
to regulate the SC converter to compensate for the transient power, while the traditional V-P control is used to adjust the battery
converter to deliver the average power. The droop coefficient in the integral droop control is designed according to the nominal
slope of the battery converter control, which can extend the cycle period of the HESS.
Fig. 13. The schematic diagram of the control structure of the DBC for the HESS [116].
Several researches have studied the DBC for the HESS [117–121]. In [117], an extended droop control (EDC) method is
proposed to achieve the different dynamic current distribution autonomously for the HESS during load sudden changes and
resource changes. This method consists of virtual capacitance droop control (VCDC) and virtual resistance droop control (VRDC).
The VCDC is derived from the voltage-current characteristics of the capacitor. From the perspective of the frequency domain, a
capacitor is considered as a short circuit at high frequency and as an open circuit at low frequency. Therefore, the VCDC is used
to adjust the SC converter to provide the transient current, and the VRDC is used to regulate the battery converter to deliver the
average power. In addition, this article also gives the detailed design process of the controller's parameters. However, the proposed
method ignores the SoC level of the SC, resulting in insufficient p ow er for the SC for future us e. In addition, w he n the load changes
suddenly, the DC bus voltage cannot return to the preset point. By extending the research study in [117], a decentralized power
management system for the HESS in [118] is proposed to realize different dynamic power sharing and SoC recovery at the same
time. In this method, a VCDC with an SoC recovery control loop is proposed to regulate the SC converter, while the traditional
VRDC is employed to adjust the battery converter. The high-frequency power and low-frequency power in the system are allocated
to the SC and the battery, respectively. In addition, the SoC of the SC can automatically restore to the pre-set value without any
control mode switching, so that it can have enough power to work continuously. However, the proposed method does not consider
the recovery of the DC bus voltage. In order to address this issue, ref. [119] proposes a decentralized power management method
for the HESS to achieve different dynamic power sharing, automatic bus voltage restoration and SoC recovery. In this solution, a
HPF based VRDC is used to control the battery converter to compensate for low frequency power, while VCDC is used to control
the SC converter to suppress high frequency power fluctuations. During operation, the DC bus voltage can automatically restore
to the initial value. In addition, the recovery of the SC’s SoC is automatically realized when the bus voltage is restored, without
the need for an additional controller. This means that the controller is simplified, and the order of the system is reduced.
Table 9
The critical analysis of the HESS based on DBC.
Ref. Method ESS type Contributions Limitations
[116] DBC Battery & SC
An integral droop (ID) control derived from the characteristics of the capacitor
is used to regulate the SC converter to compensate for the transient power, while
the traditional V-P control is used to adjust the battery converter to deliver the
average power.
The proposed method causes the DC bus
voltage deviation. The SoC of the SC does
not consider.
[117] EDC Battery & SC
The VCDC is proposed to regulate the SC converter to supply the transient
power, and the VRDC is used to control the battery converter to supply the
average power.
The proposed EDC will cause deviations in
the bus voltage. In addition, the SoC of SC is
not considered.
15
[118] EDC Battery & SC
The VCDC with an SoC recovery control loop is proposed to regulate the SC
converter to supply the transient power and restore the SoC of the SC. The
VRDC is used to control the battery converter to supply the average power.
The proposed EDC will cause deviations in
the bus voltage.
[119] EDC-
HPF Battery & SC
A HPF based VRDC is used to control the battery converter to provide the
average power and maintain the bus voltage. The VCDC is used to control the
SC converter to suppress high frequency power fluctuations. The SoC of the SC
is automatically recovery when the bus voltage is maintained within a safe range.
This method does not take into account the
operating limitations of the SoC of the SC.
[120] EMDC FC & SC
An enhanced mixed droop control (EMDC) is proposed for the FC-SC auxiliary
power unit of large electric vehicles to realize dynamic power sharing. The SC
converter is controlled by the adjustable virtual impedance droop method, and
the VRDC is used to control the FC converter. The proposed method can
appropriately adjust the output impedance of the SC converter according to the
SoC level of the SC to realize the adjustment of dynamic power distribution, and
effectively extend the service life of the SC.
The proposed method does not consider the
influence of line impedance on power
distribution and the lossless of regenerative
energy.
[121] DBC FC & SC
A decentralized energy management system based on improved droop control is
proposed to realize the lossless regulation of regenerative energy, and SC’s SoC
compensation and protection for FC and SC auxiliary power units. In addition,
the proposed method can minimize the influence of line impedance on dynamic
p
ower distribution.
The control system is more complicated.
[122] VRDC Battery & SC
& FESS
A novel virtual resistor and capacitor droop (VRCD) control is proposed for the
HESS in shipboard medium voltage DC (MVDC) system. The proposed control
method can realize the low frequency, intermediate frequency and high
frequency dynamic power distribution of the HESS. In addition, the secondary
SoC recovery control of the SC can effectively avoid the energy shortage caused
b
y the current leakage of the SC.
The proposed method will lead to bus
voltage deviation. In addition, the control is
more complicated.
[123] DBC Battery & SC
A decentralized control system is proposed to ensure power balance, maintain
the bus voltage and achieve dynamic power sharing between the battery and the
SC for the HESS with battery and SC in MVDC shipboard. In addition, an
adaptive droop control is used to determine the reference current for different
energy storage. The field programmable gate array is used to verify the proposed
controller.
The proposed method does consider the SC
SoC recovery.
[124] DBC Battery & SC
An adaptive droop control based on a fuzzy logic algorithm is used to solve the
influence of line impedance in DC microgrid. The proposed method
automatically adjusts the droop coefficient by detecting voltage deviation and
unbalanced power to reduce the influence of line impedance, thereby improving
the accuracy of power distribution.
The proposed method will cause bus voltage
deviation. In addition, the SoC of SC is not
considered.
[125] DBC Battery & SC
A composite finite time controller is proposed for the FC and SC hybrid systems
with constant power loads to achieve power sharing and ensure system stability.
The ID is used to control the SC converter to suppress the transient power, and
the proportional droop (PD) regulates the FC converter to provide the average
power of the load. A finite time observer is proposed to provide feedforward
compensation to mitigate the effects of disturbances.
When multiple energy storages are
integrated, the bus voltage will deviate.
[126] DBC Battery &
SMESS
A new primary frequency control strategy is proposed for use in a renewable
energy microgrid with SMESS and battery. The proposed method can realize fast
frequency adjustment and prolong the service life of the battery. Then, a dynamic
droop factor control was developed to realize the charge and discharge priority
and power sharing between the SMESS and the battery.
The proposed method does not consider the
bus voltage deviation.
[127] DBC Battery & SC
A virtual droop control based on secondary voltage control (SVC) is proposed
for the HESS composed of the battery and SC. In this method, the SC converter
is adjusted by VCDC to deliver high-frequency power fluctuations, while the
VRDC is used to provide low-frequency power. The SVC is used to compensate
the bus voltage deviation caused by virtual droop control.
When the system integrates multiple
ba tter ies and SC, the S oC o f th e ba tter ies wil l
be unbalanced.
[128] DBC Battery & SC
An energy management system based on adaptive droop control is proposed for
the battery and SC HESS in microgrid. This method can ensure that the SC acts
as a buffer for high-frequency power. In addition, the SoC compensation control
with mode switching has been developed to be applied to SC converters to ensure
that the SC can have sufficient energy.
When the SC is operating in SoC
compensation mode, it cannot provide
transient power.
[129] DBC Battery & SC
A decentralized control algorithm based on DBC is proposed for HESS in
MVDC shipboard microgrid. The proposed algorithm can adaptively adjust the
droop coefficient according to the change rate of the DC bus voltage. Besides,
this method can achieve the SoC balance between batteries and the SoC r ecov ery
of the SC.
The controller is more complicated.
[130]DBCBattery & SC
A completely decentralized control strategy without communication is proposed
for the HESS in islanded microgrid. In addition, the proposed control system can
achieve the SoC balance between the batteries, thereby avoiding excessive
charging and discharging of a single battery.
The DC bus voltage cannot be adjusted, and
SC will have current leakage.
The critical analysis of the HESS based on the DBC is given in Table 9. The DBC is mainly used for energy storage with
different characteristics to realize dynamic power distribution without any communication system. This means that the controller
needs to add additional control loops to implement multiple functions, which leads to the complexity and higher order of the control
system, as the method mentioned involves the bus voltage regulation, SoC balance between batteries and SC’s SoC compensation.
16
In addition, the frequency of different control loops should be distinguished in the design of the controller. The accuracy of power
distribution between energy storage is not high.
3.2.2. Other decentralized control methods (DCM)
Ref. [131] proposes a frequency-coordinated virtual impedance control system for DC microgrid. The control structure is given
in Fig. 14. In this system, the virtual impedance is equivalent to a HPF and LPF loop, respectively, which are installed in the SC
and the battery converter controllers. The proposed method can effectively make the battery and SC respond to low and high
frequency power fluctuations. In addition, the SC’s SoC restoration system is used to ensure that its SoC is within the set range.
However, the proposed method will produce steady-state errors, resulting in the bus voltage cannot be adjusted. In addition, SC
will have current leakage in steady state. Ref. [132] presents an online optimization control method based on reinforcement learning
(RL), which is used in the battery-UC system in the AC/DC microgrid of the PV system and diesel generator. This method can
minimize the disturbance caused by the charging and discharging of the battery-UC system. In addition, considering that the
internal impedance of the energy storage is unknown, an ANN is designed to online estimate the system dynamics. Then, based
on the estimated system dynamics, another ANN is hired to calculate the optimal control input for the battery-UC system. The
proposed method is a fully decentralized control method and only needs to measure local data. The control system is given in Fig.
15. Ref. [133] presents a hierarchical control system for a shipboard microgrid that includes diesel generators, HESS and bow
thrusters. The HESS is composed of the battery and SC. In the primary control, a HPF and LPF are installed in the SC and battery
converter control system respectively, so as to realize the power distribution of different frequencies. The secondary control is to
ensure that the output power from the diesel generator to the bow thruster is equal to the power demand.
Fig. 14. The control structure of the HESS in ref. [131]. Fig. 15. The control structure of the HESS in ref [132].
Table 10
The critical analysis of the HESS based on DCM.
Ref. Method ESS type Contributions Limitations
[131] HPF/LPF Battery & SC
The HPF and LPF are used to make the battery and SC respond to
low and high frequency power fluctuations. Besides, the SC's SoC
recovery system is used to restore the SoC to the pre-set value.
The DC bus voltage cannot be adjusted, and SC will
have current leakage.
[132] RL-ANN Battery &
UC
An online optimization control method based on reinforcement
learning (RL) is proposed for the AC/DC microgrid. This method can
minimize the disturbance caused by the charging and discharging of
the battery-UC system. An ANN is designed to online estimate the
system dynamics and calculate the optimal control input for the
b
attery-UC system.
The SoC restoration system of the SC is ignored, and
the controller has a large computational burden.
[133] HPF/LPF Battery & SC
A hierarchical control system is proposed for a shipboard microgrid
that includes diesel generators, HESS and bow thrusters. The primary
control uses the HPF and LPF to achieve power sharing in the SC and
battery converter control system. The secondary control is to ensure
the power demand.
The proposed method does not take into account the
SoC recovery system of the SC.
[134]DCMFC & SC
A decentralized control system is proposed to achieve the power
sharing in the HESS. The SC can effectively reduce the transient
p
ower fluctuation of the FC.
The proposed control still has to measure the total
power, so it is not a completely decentralized control,
and the SoC restoration of the SC is not considered.
The critical analysis of the HESS based on the DCM is given in Table 10. These methods are mainly used to complete dynamic
power sharing, bus voltage regulation, the SoC restoration of the SC. However, the above methods have disadvantages. For
example, HPF/LPF requires accurate calculation of the filtering frequency to distinguish high and low frequency components; the
controller based on RL-ANN requires a large amount of data, and the calculation burden is relatively large. Besides, these methods
do not consider the SoC state of the batteries, which may lead to overuse of the battery.
17
3.3. Distributed multi-agent control
3.3.1. Consensus algorithm
The idea of the consensus algorithm (CA) is to utilize the information exchanged between neighboring agents to achieve the
global control goal through a sparse communication network [135,136]. The CA has been widely applied in the distributed HESS
[137–139]. An example of the control structure of the CA for the HESS is drawn in Fig. 16. In the control system, the VCDC and
VRDC are used to control the SC and battery converters to achieve high and low frequency power distribution. Then, a voltage
restoration system based on a CA is developed for batteries to average the voltages of multiple buses regardless of the influence
of the line impedance between multiple buses. In addition, the current compensation loop is designed to restore the SoC of the SC
to its initial value, thereby suppressing the self-discharge of the SC.
Fig. 16. The schematic diagram of the control structure of the CA for the HESS [138].
Ref. [139] proposes a distributed coordinated control strategy based on a sparse communication networks for the HESSs to
achieve power distribution and energy balance. In the primary control, a virtual impedance droop control is used to complete
dynamic power sharing. In the secondary control, the voltage reduction system based on the average consensus protocol is used to
suppress the bus voltage deviation. In addition, a distributed SoC balancing strategy is developed to eliminate the SoC imbalance
between the batteries, thereby preventing the overuse of a single battery. The SoC of the SC can be restored to the pre-set value.
Ref. [140] proposes a distributed control method based on the leaderless consensus protocol for the multiple HESSs to achieve
dynamic power distribution and bus voltage regulation. This method is to complete global control through data exchange between
controllers. The internal power sharing of a single HESS with the battery and SC is realized by setting the outer loop voltage
controller. When the system has power oscillations, the SC can immediately suppress the transient power fluctuations, and the
battery compensates for the average power fluctuations. In addition, the external coordination of multiple HESSs is achieved
through battery state consistency and SC terminal voltage consistency. Ref. [141] proposes a novel consensus-based control system
for the HESS with cascaded multi-port converters in DC microgrids. The cascaded multi-port converter is composed of two stages:
the first stage is the SC converter, and the second stage is the battery converter. It is worth noting that the second stage can contain
multiple battery converters. The proposed method can solve the SoC imbalance problem of the batterie and realize plug-and-play
operation. In addition, the bus voltage can also be stabilized at the initial value.
Table 11
The critical analysis of the HESS based on distributed control.
Ref. Method ESS type Contributions Limitations
[138] CA Battery & SC
An advanced secondary voltage recovery control based on virtual
impedance droop is proposed to achieve the power sharing, bus voltage
maintenance and the SoC recovery of the SC.
The proposed method does not take into account
the SoC balance between batteries.
[139] CA Battery & SC
A distributed coordinated control strategy based on a CA is proposed for
the HESSs to achieve power distribution, the SoC balance of the batteries
and the SoC restoration of the SC.
The SC still faces the problem of current leakage.
The control system is complex.
[140] CA Battery & SC
A distributed control method based on the leaderless consensus protocol
is proposed to achieve dynamic power sharing and bus voltage regulation
in the HESSs.
A communication failure between the controllers
will reduce the performance of the proposed
method.
18
[141] CA Battery & SC
A novel consensus-based control system is proposed to solve the SoC
imbalance of the batteries and bus voltage deviation as well as realize
plug-and-play operation for the HESS with cascaded multi-port
converters in DC microgrid.
The SoC of the SC is ignored.
[142] CA Battery & SC
A semi-consensus-based multi-functional control strategy is proposed to
achieve bus voltage regulation as well as the SoC balance and recovery
for the batteries and SCs for the HESS in DC microgrids. The ID is hired
to adjust the SC to provide high-frequency power, while the battery uses
the traditional V-P droop control to transmit low-frequency power. In the
semi-consensus protocol, only the battery converter exchanges data
through a sparse communication network, while the SC does not need to
exchange data, thus saving operating costs.
A communication failure between the battery
controllers will reduce the performance of the
proposed method.
[143] CA Battery & SC
& FC
A multi-source coordinated energy management strategy based on the
SoC consensus is proposed for large-scale hybrid trams. In the proposed
method, the self-convergency droop control is used to adjust the power
requirements of multiple power supplies, so as to ensure the SoC
consistency of the battery and the SC. Real-time simulation is used to
verify the feasibility and effectiveness of the proposed method.
The proposed method ignores the self-discharge of
SC.
[144] CA Battery & SC
A multi-agent system based on consensus control is proposed to realize
power distribution among multiple microgrids. The FBC is used to
regulate the battery and SC converters to provide low and high frequency
power. The proposed method can achieve DC voltage stability between
all microgrids.
The proposed method does not consider the SoC
restoration of the SC in a single microgrid.
[145] Multi-
agent Battery & SC
A multi-agent coordinated control system is proposed to achieve the
dynamic power allocation, bus voltage regulation, SoC balance and SoC
recovery in the HESS. This method introduces the concept of leader and
follower. The SC leader is used to adjust the bus voltage, while the battery
leader is used to control the SoC level of the SC leader. The SC followers
collect data from their neighbors to track the SoC level of the SC leader.
The battery followers try to track the SoC level of the battery leader based
on neighbors’ data.
The adjacent energy storage has a communication
failure, the performance of the proposed method
will be reduced. The SCs have the problem of self-
discharge.
[146] Multi-
agent Battery & SC
A hierarchical control strategy is proposed for HESS in a low-voltage
microgrid. In this control strategy, primary control is used to achieve
dynamic active power sharing. The secondary control is a multi-agent
system, which can achieve the SoC balance between batteries, reactive
power distribution, frequency and voltage recovery. The tertiary control
is an energy management coordination system, which takes into account
the characteristics of various types of the ESSs to make the energy
management system expandable and also optimizes the self-discharge
and degradation of the ESS.
The proposed method is not completely distributed
control, and it does not consider the SoC recovery
of SC.
[147] CA-MPC Battery & SC
A hierarchical distributed coordinated control is proposed for the
optimized operation of the battery-SC system in the microgrid, and
prolongs the service life of the battery. In the lower-level distributed
system, a weighted discrete consensus algorithm based on the MPC is
proposed to realize adaptive power sharing between battery and SC. In
the upper system, two time-scale models based on the MPC are proposed
to achieve the optimal power allocation between the battery and the SC,
and to keep their SoC levels within a safe range.
The proposed method is not all distributed control,
and the computational burden of the controller is
relatively large. The SC has current leakage.
The critical analysis of the HESS based on the distributed control is given in Table 11. Most of the works in Table 11 focus on
the bus voltage adjustment and power sharing between different characteristics of energy storage. Very few works consider the
SoC balance of the battery and the SoC recovery of the SC. However, the methods from [139], [141] and [147] integrate multiple
control loops, which lead to the complexity of the design of system control parameters. Besides, the mentioned methods in Table
11 all require data exchange between controllers. If there is a communication failure between the controllers, the overall
performance of the controllers will be reduced. Besides, since the CA requires higher calculations, it complicates the controller.
3.3.2. The communication delay on the distributed method
The distributed control method is realized by data exchange based on peer-to-peer communication network. Therefore, it suffers
from the time delay issue. However, the aforementioned methods ignore the time delay, which may make the system lose stability
and deteriorate the convergence and dynamic performance of the system. Therefore, it needs further investigation and research.
An example of the control structure with the time delay is presented in Fig. 17. In the control system, a delay compensation method
is proposed for the microgrid to suppress the influence of the communication delay on the small-signal stability of the distributed
secondary frequency and voltage control system. This method first establishes a time-delayed small-signal dynamic model for the
secondary frequency and voltage control. The small signal model is used to study the influence of the communication delay,
communication network topology and control parameters on system stability. Finally, a delay compensation loop and gain model
with lead-lag compensation are developed to enhance the stability of the system to communication delay.
Ref. [148] presents an advanced sliding mode control system for a DC microgrid to achieve the SoC balance between battery
ESSs. The proposed method can eliminate the circulating current, thereby increasing the service life of the battery. The battery
19
ESSs can prevent overuse during periods of high load. In addition, the simulation results consider the communication delay factor,
and plug and play. In [149], the voltage control is expressed as an optimization problem, and it is integrated into the power
distribution control system based on a PI consensus algorithm to solve the error caused by the voltage observer under the time
delay. This method can realize bus voltage regulation and power sharing in a controller at the same time, thereby simplifying the
complexity of the controller, and the control accuracy is not affected by the time delay. In addition, the Lyapunov function is used
to analyze and ensure the stability of the control algorithm under time delay. Scattering transformation is proposed to solve the
negative effects of the time delay.
Fig. 17. The schematic diagram of the control structure with time delay compensator for the microgrid [150].
Table 12
The critical analysis of the communication time delay in the centralized method for the HESS.
Ref. Method ESS type Contributions Limitations
[150] Multi-
agent None
A delay compensation method is proposed to suppress the influence of
the communication delay on the small-signal stability of the distributed
secondary frequency and voltage control system in the microgrid.
The presented method does not consider the HESS.
[148] Multi-
agent Battery
An advanced sliding mode control system with considering
communication delay is proposed to eliminate the circulating current and
achieve the SoC balance between battery ESSs in DC microgrid.
The proposed method only considers the time delay in the
simulation but does not analyze the influence of the time
delay on the stability of the control system.
[149] CA Battery
The power distribution control system integrating a voltage control and
PI consensus algorithm to solve the error caused by the voltage observer
under the time delay. Besides, the Lyapunov function is used to analyze
and ensure the stability of the control algorithm under time delay.
Scattering transformation is proposed to solve the negative effects of the
time delay.
The proposed method is only for a single ESS, not a HESS.
[151] Multi-
agent Battery
A new finite time control strategy with time delay is proposed to solve
the voltage deviation and SoC imbalance between batteries in the DC
microgrid. The Artstein transformation can convert a time-delay system
into a delay-free system. The finite-time Lyapunov method is used to
analyze the stability and accuracy of the controller under time delay.
The proposed method is only for a single ESS, not a HESS.
[152] Multi-
agent Battery
A distributed control method considering communication delay non-
periodic sampling data is proposed to realize frequency restoration,
voltage restoration and SoC balance between batteries in the microgrid.
The Lyapunov–Krasovskii functional and linear matrix inequality are
used to calculate the control system parameters and to ensure its stability,
thereby reducing the adverse effects of time delay on the control system.
In addition, since the stability of the system is determined by the
maximum and minimum eigenvalues of the Laplacian matrix of the
communication graph, the control system has better scalability.
The proposed method is only for a single ESS, not a HESS.
The critical analysis of the communication delay for the distributed control in the HESS is given in Table 12. The mentioned
methods mainly focus on learning the influence of communication delay on the stability and convergence of the distributed control
system and give the corresponding delay compensation scheme. However, the proposed methods are only applied to the microgrid
with a single energy storage, not a HESS. Unfortunately, there is no evidence that any research has studied the impact of
communication delays on the distributed control of the HESS consisting of the HEDE and HPDE. Therefore, the communication
delay in distributed control for the HESS can be further studied.
4. Case study: a coordinated droop control for the HESS
As discussed in section 3.2, the decentralized method for the HESS is mainly implemented based on the droop control. However,
this control method can only achieve high and low frequency power sharing [116,117]. To realize more functions, more control
20
loops need to be added, so that the entire control system is more complicated and the order becomes higher [122]. To solve these
drawbacks, an advanced droop control for the HESS with the battery and SC in dc microgrid is proposed to simplify the control
system.
The advantage of the proposed control method is that it can achieve dynamic power sharing and DC bus voltage regulation
simultaneously, without the need for an extra voltage restoration system. A V-dP/dt droop control is proposed to adjust the battery
converter, while the conventional power droop method is employed to adjust the SC converter. In the presented method, the total
power is decomposed into low and high frequency components, and allocated to the battery and SC converters, respectively. The
DC bus voltage is regulated by the SC converter. Therefore, the control system order and parameter design complexity are reduced.
The proposed control method will be verified in the PIL simulation and compared with the popular method from [116].
4.1. Proposed a fully decentralized control
4.1.1. Droop based control for the battery and SC converters
The V-dP/dt droop control is derived from the inductor characteristic. From the frequency domain, an inductor operates as a
short circuit at low-frequency and as an open-circuit at high frequency. Therefore, the battery converter is regulated to deliver the
average power instead of transient power at steady state. According to inductor characteristics, the equation about the inductor
current and voltage is presented as (1).
iio1
LvLdt (1)
where L, io, i, and vL are inductance, inductor initial current, the output current and voltage of the inductor, respectively. Then, (1)
is further written as (2).
vLvoLdi
dt (2)
where voLdio/dt. Similar to (2), the V-dP/dt control in (3) can be developed by replacing current with power and preserving the
equivalent format as (2).
vbvrmdPb
dt (3)
where v
r
, m, vb, and Pb are the reference voltage for the dc bus voltage, droop coefficient, output voltage and power of the battery
converter terminal, respectively.
The conventional power droop control of the SC converter can be written by (4) based on [17].
vsc vrnPsc (4)
where vsc and Psc are the output voltage and power of the SC converter terminal, respectively. n is the droop coefficient that can
be calculated by (5).
nvmax vmin
Psc
max (5)
where vmax and vmin are the maximum and minimum dc bus voltage; Psc
max is the maximum high frequency power of the SC
converter terminal. Since the battery and SC simultaneously provide the different dynamic power for the system load, the total
output power Pt of the HESS is the sum of the Pb and Psc, as presented in (6).
PtPbPsc (6)
By considering (3), (4), and (6), the power decomposition in the HESS is shown as (7).
PbGLPF(s)∙Ptn
ms nPt
Psc GHPF(s)∙Ptms
ms nPt
(7)
where s is the Laplace operator; GLPF(s) and GHPF(s) are low-pass filter (LPF) and high-pass filter (HPF). From (7), the system
power can be automatically separated into low frequency and high frequency components, and then allocated each of them to the
battery converter and SC converter, respectively. The desired dynamic power-sharing is realized by reasonably adjusting the corner
frequency ωcn/m of both LPF and HPF.
4.1.2. Voltage recovery method
The conventional power droop control can cause the bus voltage deviation during power fluctuations. If the power remains
constant, there will be no deviation in the bus voltage. The SC converter only provides transient power and does not supply any
21
Fig. 18. The proposed droop coordinated controls for battery and SC converters.
power in the steady state. Therefore, the bus voltage can be restored to the initial value under the control of the SC converter in the
steady state, without the need to design an additional voltage recovery loop. In the case of a sudden power change, the SC converter
immediately responds to suppress the power oscillation. Based on (5), the maximum voltage deviation ∆v is expressed as (8).
∆vvmax vmin nPsc
max 0(8)
It can be seen from (8) that the voltage drop caused by the SC converter releasing/absorbing transient power is less than the
maximum voltage deviation.
After eliminating the power fluctuation of the system, the SC converter slowly reduces the output power until it reaches zero.
This means that the SC converter is not providing any power. Based on (4), the output voltage of the SC converter is equal to the
reference bus voltage, as shown in (9).
vsc vr (9)
Therefore, the bus voltage deviation is equal to zero, as given in (10).
∆vvmax vmin nPsc
max 0(10)
Finally, the bus voltage can be kept within a safe range, and no additional voltage compensation loop is required. The entire control
system is simplified.
The existence of the line resistance can reduce the performance of the control system. To solve this problem, an extra control
loop in [153] is adopted to eliminate the influence of the line resistance, as given in (11).
vx
LRx
LPx
vxvxvbus (11)
where Rx
L is the line resistance of the battery or SC; vx
L is the voltage of the line resistance of the battery or SC converter; vbus is the
dc bus voltage. Specifically, xb denotes the battery, while xsc denotes the SC component. The total control system is shown
in Fig. 18.
4.2. PIL simulation verification
To verify the effectiveness of the proposed method in real control system, the PIL platform including a microcontroller (MC)
and PC, is established [154]. The MC is implemented by using STM32F429ZIT6, connected in a closed loop with the controlled
object, as shown in Fig. 19.
Fig. 19. PIL simulation.
In this case, the proposed method is compared with the popular method from [116]. In [116], the traditional power droop control
and ID control are presented to regulate the battery and SC converters, respectively. The relevant equations can be written as (12)
and (13). The relevant control structure is given in Fig. 13.
vbvrmPb (12)
vsc vrnPscdt (13)
In Fig. 20(a), the HESS initially provides 1 kW of power. At 5 s, the load suddenly increases from 1 kW to 2 kW. The SC
responds immediately to compensate for the transient power, and then gradually reduces the output power until it reaches zero.
The battery has a slower response and provides constant power in a steady state. In Fig. 20(b), the DC bus voltage is maintained
22
within the desired range through the SC converter control. According to the traditional power droop control characteristics, when
the SC provides transient power, the DC bus voltage has a deviation. Once the SC does not provide any power, the DC bus voltage
automatically restores to the initial value.
Fig. 20. Simulation results: (a) Dynamic power-sharing, (b) DC bus voltage, (c) Dynamic power-sharing from [116], and (d) DC bus voltage from [116].
It can be seen from Fig. 20(c) that the power sharing simulation results generated by the method from [116] are similar to the
proposed method. The SC is responsible for providing high-frequency power, while the battery supplies low-frequency power.
However, the DC bus voltage has a steady-state error in 0-5 s and 10-15 s, as shown in Fig. 20(d). In addition, when the system
power changes suddenly at 5 s, the bus voltage has voltage drop and does not return to the set value. This is because the DC bus
voltage in [116] is regulated by the battery converter control, which employs the traditional power droop control, as given in (12).
This means that if the battery continues to supply power, there will always be a voltage deviation in the DC bus voltage. Therefore,
ref. [116] requires a voltage restoration control loop to eliminate steady-state errors. Hence, the controller becomes complicated.
Based on the comparison of the two methods, the proposed method can simultaneously realize dynamic power distribution and bus
voltage regulation without any voltage control system. Therefore, the proposed method is significantly more advanced.
5. The future development trends
In the decentralized and distributed methods, the designed control systems are mainly to realize the high and low frequency
power sharing, bus voltage regulation, SoC balance and recovery for the battery and SC. Therefore, a comprehensive controller
containing multiple functions is required, such as reducing the degradation of energy storage, and suppressing the peak current of
the battery and SC. More functions means that the control system becomes more complex, and the order of the system becomes
higher. Then, the control system parameters need to be carefully designed to avoid the potential influence between control loops
with different functions, thereby avoiding the instability of the system. Therefore, simplifying the control system while ensuring
multi-functionality can be a research and development trend, and it is also a challenge. Section 4 provides a case study for the
design of the control system. In the case, a new coordinated droop control for the HESS is proposed to achieve dynamic power
distribution and bus voltage regulation at the same time, without the need for additional voltage recovery control. Therefore, the
control system is simplified. The proposed method also compares the popular method from [116] to show its advancement.
The centralized and distributed method achieve the exchange of the system information through the communication network,
so as to ensure that the controllers can operate stably. However, the time delay generated by the communication network will
reduce the dynamic performance of the control system, control accuracy and system stability. Unfortunately, most centralized and
distributed methods ignore the time delay. So far, only very few works have investigated the impact of time delay on the centralized
control system of the HESS. The methods in Table 8 focus on calculating the maximum delayed time, the effect of time delay on
the stability and convergence of the system. However, no works propose a method to eliminate the time delay and give a
corresponding time delay compensation scheme. In addition, the current works in the distributed approach focus on solving the
effect of time delay on the control system of a single energy storage, rather than the HESS. Since the HPDE in the HESS is sensitive
to time, it is more susceptible to delay. Therefore, the time delay for distributed control in the HESS can be further studied.
The current control strategy is mainly for two different types of energy storage, such as battery-SC, FC-SC and battery-FC. The
control method proposed in a very small number of articles can be used for three types of energy storage such as battery-SC-FESS
and battery-SC-FC. However, these methods do not take into account the cost and economic benefits of energy storage in the
microgrid. In addition, most of the energy storage considered in these methods are batteries, SC and FC, and more energy storage
combinations should be developed. Therefore, the low-cost combination and high economic benefits of multiple types of energy
storage can be considered as the future development trend.
Under different load conditions, the traditional control methods used for the HESS, such as FBC and DBC, may need to redesign
the system parameters to obtain effective power distribution. Therefore, the intelligent control methods such as deep reinforcement
learning can be used to automatically track the dynamic performance of different loads to complete the optimal power allocation
in the HESS, while achieving more functions. In addition, the system model establishing and stability analysis based on deep
reinforcement learning are still a challenge. Hence, it can be considered as a future research trend.
23
6. Conclusion
A microgrid is an emerging small-scale power system that includes RESs, ESS and loads. In the microgrid, the ESS is usually
a HEDE, which can be used to ensure power balance and to improve power quality. Since the HEDE has low power density
characteristics, it cannot effectively suppress the transient power fluctuations of the system, thereby reducing the stability of the
system. To solve this issue, the HESS combining the HEDE and HPDE is proposed. Since HESS contains energy storage with
different dynamic characteristics, the control system design faces some challenges. The objective of this article is to critically
review and study the control approaches of the HESS in the microgrid to discover the limitations of the existing control methods.
This article divides the control methods of the HESS into three types: centralized, decentralized and distributed. In each type,
various latest control approaches have been carefully analyzed and discussed. Existing works are mainly focused on achieving the
dynamic power sharing for energy storages with different capacities, bus voltage recovery, and SoC recovery of the HPDE.
Therefore, the HESS control system should consider more functions. However, multi-function control will complicate the
parameter design of the system and increase the order of the system. Besides, this article points out that the time delay will reduce
the control accuracy and stability of the centralized and distributed control system in the HESS. This article also presents a novel
coordinated droop control method in the case study to validate the feasibility of the simplification and multi-function of the
controller. The simulation results show that the proposed method outperforms the currently popular method. Finally, this article
discusses and suggests the future development trend of the control technology of the HESS in the microgrid that include
simplification of the multi-functional controller, compensation of the time delay, improvement of the control accuracy, automatic
tracking of the dynamic performance of the load and the combination of the multiple energy storage.
References
[1] Sebestyén V. Renewable and Sustainable Energy Reviews : Environmental impact networks of renewable energy power plants. Renew Sustain Energy
Rev 2021;151:111626. https://doi.org/10.1016/j.rser.2021.111626.
[2] Pina EA, Lozano MA, Serra LM. Assessing the influence of legal constraints on the integration of renewable energy technologies in polygeneration
systems for buildings. Renew Sustain Energy Rev 2021;149:111382. https://doi.org/10.1016/j.rser.2021.111382.
[3] Ellabban O, Abu-Rub H, Blaabjerg F. Renewable energy resources: Current status, future prospects and their enabling technology. Renew Sustain Energy
Rev 2014;39:748–64. https://doi.org/10.1016/j.rser.2014.07.113.
[4] BP. Statistical Review of World Energy globally consistent data on world energy markets . 2020:66.
[5] Kyritsis A, Voglitsis D, Papanikolaou N, Tselepis S, Christodoulou C, Gonos I, et al. Evolution of PV systems in Greece and review of applicable
solutions for higher penetration levels. Renew Energy 2017;109:487–99. https://doi.org/10.1016/j.renene.2017.03.066.
[6] Bullich-Massagué E, Díaz-González F, Aragüés-Peñalba M, Girbau-Llistuella F, Olivella-Rosell P, Sumper A. Microgrid clustering architectures. Appl
Energy 2018;212:340–61. https://doi.org/10.1016/j.apenergy.2017.12.048.
[7] Hannan MA, Faisal M, Jern P, Begum RA, Dong ZY, Zhang C. Review of optimal methods and algorithms for sizing energy storage systems to achieve
decarbonization in microgrid applications. Renew Sustain Energy Rev 2020;131:110022. https://doi.org/10.1016/j.rser.2020.110022.
[8] U.S. Department of Energy DOE. Energy Storage Grand Challenge : Energy Storage Market Report 2020.
[9] Zamora R, Srivastava AK. Controls for microgrids with storage: Review, challenges, and research needs. Renew Sustain Energy Rev 2010;14:2009–18.
https://doi.org/10.1016/j.rser.2010.03.019.
[10] Jin M, Feng W, Marnay C, Spanos C. Microgrid to enable optimal distributed energy retail and end-user demand response. Appl Energy 2018;210:1321–
35. https://doi.org/10.1016/j.apenergy.2017.05.103.
[11] Lasseter RH. MicroGrids 2002:305–8.
[12] Nichols DK, Stevens J, Lasseter RH, Eto JH, Vollkommen HT. Validation of the CERTS microgrid concept the CEC/CERTS microgrid testbed. 2006
IEEE Power Eng Soc Gen Meet PES 2006:5–7. https://doi.org/10.1109/pes.2006.1709248.
[13] Guerrero JM, Vasquez JC, Matas J, De Vicuña LG, Castilla M. Hierarchical control of droop-controlled AC and DC microgrids - A general approach
toward standardization. IEEE Trans Ind Electron 2011;58:158–72. https://doi.org/10.1109/TIE.2010.2066534.
[14] Kaur A, Kaushal J, Basak P. Areview on microgrid central controller. Renew Sustain Energy Rev 2016;55:338–45.
https://doi.org/10.1016/j.rser.2015.10.141.
[15] Martin-Martínez F, Sánchez-Mirall es A, Rivier M. A literature review of Microgrids: A functional layer based classification. Renew Sustain Energy Rev
2016;62:1133–53. https://doi.org/10.1016/j.rser.2016.05.025.
[16] Farrokhabadi M, Lagos D, Wies RW, Paolone M, Liserre M, Meegahapola L, et al. Microgrid Stability Definitions, Analysis, and Examples. IEEE Trans
Power Syst 2020;35:13–29. https://doi.org/10.1109/TPWRS.2019.2925703.
[17] Lu X, Sun K, Guerrero JM, Vasquez JC, Huang L. State-of-charge balance using adaptive droop control for distributed energy storage systems in DC
microgrid applications. IEEE Trans Ind Electron 2014;61:2804–15. https://doi.org/10.1109/TIE.2013.2279374.
[18] Hoang KD, Lee HH. Accurate power sharing with balanced battery state of charge in distributed DC microgrid. IEEE Trans Ind Electron 2019;66:1883–
93. https://doi.org/10.1109/TIE.2018.2838107.
[19] Zhang Z, Ding T, Zhou Q, Sun Y, Qu M, Zeng Z, et al. A review of technologies and applications on versatile energy storage systems 2021;148.
https://doi.org/10.1016/j.rser.2021.111263.
[20] Farhadi M, Mohammed O. Energy Storage Technologies for High-Power Applications. IEEE Trans Ind Appl 2016;52:1953–62.
24
https://doi.org/10.1109/TIA.2015.2511096.
[21] Qiu X, Nguyen TA, Crow ML. Heterogeneous Energy Storage Optimization for Microgrids. IEEE Trans Smart Grid 2016;7:1453–61.
https://doi.org/10.1109/TSG.2015.2461134.
[22] Abeywardana DBW, Hredzak B, Agelidis VG, Demetriades GD. Supercapacitor sizing method for energy-controlled filter-based hybrid energy storage
systems. IEEE Trans Power Electron 2017;32:1626–37. https://doi.org/10.1109/TPEL.2016.2552198.
[23] Manandhar U, Ukil A, Gooi HB, Tummuru NR, Kollimalla SK, Wang B, et al. Energy management and control for grid connected hybrid energy storage
system under different operating modes. IEEE Trans Smart Grid 2019;10:1626–36. https://doi.org/10.1109/TSG.2017.2773643.
[24] Aneke M, Wang M. Energy storage technologies and real life applications – A state of the art review. Appl Energy 2016;179:350–77.
https://doi.org/10.1016/j.apenergy.2016.06.097.
[25] Luo X, Wang J, Dooner M, Clarke J. Overview of current development in electrical energy storage technologies and the application potential in power
system operation. Appl Energy 2015;137:511–36. https://doi.org/10.1016/j.apenergy.2014.09.081.
[26] Hannan MA, Wali SB, Ker PJ, Rahman MSA, Mansor M, Ramachandaramurthy K, et al. Battery energy-storage system : A review of technologies ,
optimization objectives , constraints , approaches , and outstanding issues. J Energy Storage 2021;42:103023. https://doi.org/10.1016/j.est.2021.103023.
[27] Chemali E, Preindl M, Malysz P, Emadi A. Electrochemical and Electrostatic Energy Storage and Management Systems for Electric Drive Vehicles:
State-of-the-Art Review and Future Trends. IEEE J Emerg Sel Top Power Electron 2016;4:1117–34. https://doi.org/10.1109/JESTPE.2016.2566583.
[28] Zimmermann T, Keil P, Hofmann M, Horsche MF, Pichlmaier S, Jossen A. Review of system topologies for hybrid electrical energy storage systems. J
Energy Storage 2016;8:78–90. https://doi.org/10.1016/j.est.2016.09.006.
[29] Kouchachvili L, Yaïci W, Entchev E. Hybrid battery/supercapacitor energy storage system for the electric vehicles. J Power Sources 2018;374:237–48.
https://doi.org/10.1016/j.jpowsour.2017.11.040.
[30] Hannan MA, Hoque MM, Mohamed A, Ayob A. Review of energy storage systems for electric vehicle applications: Issues and challenges. Renew
Sustain Energy Rev 2017;69:771–89. https://doi.org/10.1016/j.rser.2016.11.171.
[31] Hemmati R, Saboori H. Emergence of hybrid energy storage systems in renewable energy and transport applications – A review. Renew Sustain Energy
Rev 2016;65:11–23. https://doi.org/10.1016/j.rser.2016.06.029.
[32] Bocklisch T. Hybrid energy storage approach for renewable energy applications. J Energy Storage 2016;8:311–9.
https://doi.org/10.1016/j.est.2016.01.004.
[33] Jing W, Lai CH, Wong SHW, Wong MLD. Battery-supercapacitor hybrid energy storage system in standalone DC microgrids: Areview. IET Renew
Power Gener 2017;11:461–9. https://doi.org/10.1049/iet-rpg.2016.0500.
[34] Chong LW, Wong YW, Rajkumar RK, Rajkumar RK, Isa D. Hybrid energy storage systems and control strategies for stand-alone renewable energy
power systems. Renew Sustain Energy Rev 2016;66:174–89. https://doi.org/10.1016/j.rser.2016.07.059.
[35] Khodadoost Arani AA, B. Gharehpetian G, Abedi M. Review on Energy Storage Systems Control Methods in Microgrids. Int J Electr Power Energy
Syst 2019;107:745–57. https://doi.org/10.1016/j.ijepes.2018.12.040.
[36] Zhang L, Hu X, Wang Z, Ruan J, Ma C, Song Z, et al. Hybrid electrochemical energy storage systems: An overview for smart grid and electrified vehicle
applications. Renew Sustain Energy Rev 2021;139:110581. https://doi.org/10.1016/j.rser.2020.110581.
[37] Hajiaghasi S, Salemnia A, Hamzeh M. Hybrid energy storage system for microgrids applications: A review. J Energy Storage 2019;21:543–70.
https://doi.org/10.1016/j.est.2018.12.017.
[38] Babu TS, Vasudevan KR, Ramachandaramurthy VK, Sani SB, Chemud S, Lajim RM. A Comprehensive Review of Hybrid Energy Storage Systems:
Converter Topologies, Control Strategies and Future Prospects. IEEE Access 2020;8:148702–21. https://doi.org/10.1109/ACCESS.2020.3015919.
[39] Bullich-Massagué E, Cifuentes-García FJ, Glenny-Crende I, Cheah-Mañé M, Aragüés-Peñalba M, Díaz-González F, et al. A review of energy storage
technologies for large scale photovoltaic power plants. Appl Energy 2020;274:115213. https://doi.org/10.1016/j.apenergy.2020.115213.
[40] Morstyn T, Hredzak B, Agelidis VG. Control Strategies for Microgrids with Distributed Energy Storage Systems: An Overview. IEEE Trans Smart Grid
2018;9:3652–66. https://doi.org/10.1109/TSG.2016.2637958.
[41] Mbungu NT, Naidoo RM, Bansal RC, Siti MW, Tungadio DH. An overview of renewable energy resources and grid integration for commercial building
applications. J Energy Storage 2020;29:101385. https://doi.org/10.1016/j.est.2020.101385.
[42] Wang Y, Wang L, Li M, Chen Z. A review of key issues for control and management in battery and ultra-capacitor hybrid energy storage systems.
ETransportation 2020;4:100064. https://doi.org/10.1016/j.etran.2020.100064.
[43] Dougal RA, Liu S, White RE. Power and life extension of battery-ultracapacitor hybrids. IEEE Trans Components Packag Technol 2002;25:120–31.
https://doi.org/10.1109/6144.991184.
[44] Ma T, Yang H, Lu L. Development of hybrid battery-supercapacitor energy storage for remote area renewable energy systems. Appl Energy 2015;153:56–
62. https://doi.org/10.1016/j.apenergy.2014.12.008.
[45] Vazquez S, Lukic SM, Galvan E, Franquelo LG, Carrasco JM. Energy storage systems for transport and grid applications. IEEE Trans Ind Electron
2010;57:3881–95. https://doi.org/10.1109/TIE.2010.2076414.
[46] Song Z, Hofmann H, Li J, Han X, Zhang X, Ouyang M. A comparison study of different semi-active hybrid energy storage system topologies for electric
vehicles. J Power Sources 2015;274:400–11. https://doi.org/10.1016/j.jpowsour.2014.10.061.
[47] Zhuang W, Ye J, Song Z, Yin G, Li G. Comparison of semi-active hybrid battery system configurations for electric taxis application. Appl Energy
2020;259:114171. https://doi.org/10.1016/j.apenergy.2019.114171.
[48] Momayyezan M, Abeywardana DBW, Hredzak B, Agelidis VG. Integrated reconfigurable configuration for battery/ultracapacitor hybrid energy storage
systems. IEEE Trans Energy Convers 2016;31:1583–90. https://doi.org/10.1109/TEC.2016.2589933.
25
[49] Olivares DE, Mehrizi-Sani A, Etemadi AH, Cañizares CA, Iravani R, Kazerani M, et al. Trends in microgrid control. IEEE Trans Smart Grid
2014;5:1905–19. https://doi.org/10.1109/TSG.2013.2295514.
[50] Yamashita DY, Vechiu I, Gaubert JP. A review of hierarchical control for building microgrids. Renew Sustain Energy Rev 2020;118:109523.
https://doi.org/10.1016/j.rser.2019.109523.
[51] Levron Y, Guerrero JM, Beck Y. Optimal power flow in microgrids with energy storage. IEEE Trans Power Syst 2013;28:3226–34.
https://doi.org/10.1109/TPWRS.2013.2245925.
[52] Abhishek A, Ranjan A, Devassy S, Kumar Verma B, Ram SK, Dhakar AK. Review of hierarchical control strategies for DC microgrid. IET Renew
Power Gener 2020;14:1631–40. https://doi.org/10.1049/iet-rpg.2019.1136.
[53] Meng L, Shafiee Q, Trecate GF, Karimi H, Fulwani D, Lu X, et al. Review on Control of DC Microgrids and Multiple Microgrid Clusters. IEEE J Emerg
Sel Top Power Electron 2017;5:928–48. https://doi.org/10.1109/JESTPE.2017.2690219.
[54] Han Y, Zhang K, Li H, Guerrero JM. MAS-Based Distributed Coordinated Control and Optimization in Microgrid and Microgrid Clusters :
2018;33:6488–508.
[55] Rajesh KS, Dash SS, Rajagopal R, Sridhar R. A review on control of ac microgrid. Renew Sustain Energy Rev 2017;71:814–9.
https://doi.org/10.1016/j.rser.2016.12.106.
[56] Sathishkumar R, Kollimalla SK, Mishra MK. Dynamic energy management of micro grids using battery super capacitor combined storage. 2012 Annu
IEEE India Conf INDICON 2012 2012:1078–83. https://doi.org/10.1109/INDCON.2012.6420777.
[57] Kollimalla SK, Mishra MK, Narasamma NL. Design and analysis of novel control strategy for battery and supercapacitor storage system. IEEE Trans
Sustain Energy 2014;5:1137–44. https://doi.org/10.1109/TSTE.2014.2336896.
[58] Kollimalla SK, Mishra MK, Ukil A, Gooi HB. DC Grid Voltage Regulation Using New HESS Control Strategy. IEEE Trans Sustain Energy 2017;8:772–
81. https://doi.org/10.1109/TSTE.2016.2619759.
[59] Gee AM, Robinson FVP, Dunn RW. Analysis of battery lifetime extension in a small-scale wind-energy system using supercapacitors. IEEE Trans
Energy Convers 2013;28:24–33. https://doi.org/10.1109/TEC.2012.2228195.
[60] Tummuru NR, Mishra MK, Srinivas S. Dynamic Energy Management of Renewable Grid Integrated Hybrid Energy Storage System. IEEE Trans Ind
Electron 2015;62:7728–37. https://doi.org/10.1109/TIE.2015.2455063.
[61] Cabrane Z, Ouassaid M, Maaroufi M. Analysis and evaluation of battery-supercapacitor hybrid energy storage system for photovoltaic installation. Int J
Hydrogen Energy 2016;41:20897–907. https://doi.org/10.1016/j.ijhydene.2016.06.141.
[62] Mendis N, Muttaqi KM, Perera S. Active power management of a super capacitor-battery hybrid energy storage system for standalone operation of DFIG
based wind turbines. Conf Rec - IAS Annu Meet (IEEE Ind Appl Soc 2012:1–8. https://doi.org/10.1109/IAS.2012.6374045.
[63] Etxeberria A, Vechiu I, Camblong H, Vinassa JM. Comparison of sliding mode and PI control of a hybrid energy storage system in a microgrid application.
Energy Procedia 2011;12:966–74. https://doi.org/10.1016/j.egypro.2011.10.127.
[64] Cabrane Z, Ouassaid M, Maaroufi M. Battery and supercapacitor for photovoltaic energy storage : a fuzzy logic management 2017;11:1157–65.
https://doi.org/10.1049/iet-rpg.2016.0455.
[65] Khan MMS, Faruque O, Newaz A. Fuzzy Logic Based Energy Storage Management System for MVDC Power System of All Electric Ship. IEEE Trans
Energy Convers 2017;32:798–809.
[66] Mohamed A, Salehi V, Mohammed O. Real-time energy management algorithm for mitigation of pulse loads in hybrid microgrids. IEEE Trans Smart
Grid 2012;3:1911–22. https://doi.org/10.1109/TSG.2012.2200702.
[67] Feng X, Gooi HB, Chen SX. Hybrid energy storage with multimode fuzzy power Allocator for PV systems. IEEE Trans Sustain Energy 2014;5:389–97.
https://doi.org/10.1109/TSTE.2013.2290543.
[68] Chong LW, Wong YW, Rajkumar RK, Isa D. An optimal control strategy for standalone PV system with Battery-Supercapacitor Hybrid Energy Storage
System. J Power Sources 2016;331:553–65. https://doi.org/10.1016/j.jpowsour.2016.09.061.
[69] Sarvi M, Avanaki IN. An optimized Fuzzy Logic Controller by Water Cycle Algorithm for power management of Stand-alone Hybrid Green Power
generation. Energy Convers Manag 2015;106:118–26. https://doi.org/10.1016/j.enconman.2015.09.021.
[70] Zahedi R, Ardehali MM. Power management for storage mechanisms including battery, supercapacitor, and hydrogen of autonomous hybrid green power
system utilizing multiple optimally-designed fuzzy logic controllers. Energy 2020;204:117935. https://doi.org/10.1016/j.energy.2020.117935.
[71] Zhang Q, Li G. Experimental study on a semi-active battery-supercapacitor hybrid energy storage system for electric vehicle application. IEEE Trans
Power Electron 2020;35:1014–21. https://doi.org/10.1109/TPEL.2019.2912425.
[72] Shayeghi H, Monfaredi F, Dejamkhooy A, Shafie-khah M, Catalão JPS. Assessing hybrid supercapacitor-battery energy storage for active power
management in a wind-diesel system. Int J Electr Power Energy Syst 2021;125. https://doi.org/10.1016/j.ijepes.2020.106391.
[73] Cohen IJ, Wetz DA, McRee BJ, Dong Q, Heinzel JM. Fuzzy logic control of a hybrid energy storage module for use as a high rate prime power supply.
IEEE Trans Dielectr Electr Insul 2017;24:3887–93. https://doi.org/10.1109/TDEI.2017.006647.
[74] Teleke S, Member S, Baran ME, Member S, Bhattacharya S, Huang AQ. Rule-Based Control of Battery Energy Storage for Dispatching Intermittent
Renewable Sources 2010;1:117–24.
[75] Aktaş A, Kırçiçek Y. A novel optimal energy management strategy for offshore wind/marine current/battery/ultracapacitor hybrid renewable energy
system. Energy 2020;199. https://doi.org/10.1016/j.energy.2020.117425.
[76] Hou J, Song Z. A hierarchical energy management strategy for hybrid energy storage via vehicle-to-cloud connectivity. Appl Energy 2020;257:113900.
https://doi.org/10.1016/j.apenergy.2019.113900.
[77] Alnejaili T, Drid S, Mehdi D, Chrifi-Alaoui L, Belarbi R, Hamdouni A. Dynamic control and advanced load management of a stand-alone hybrid
26
renewable power system for remote housing. Energy Convers Manag 2015;105:377–92. https://doi.org/10.1016/j.enconman.2015.07.080.
[78] Osman Haruni A, Negnevitsky M, Haque ME, Gargoom A. A novel operation and control strategy for a standalone hybrid renewable power system.
IEEE Trans Sustain Energy 2013;4:402–13. https://doi.org/10.1109/TSTE.2012.2225455.
[79] Chia YY, Lee LH, Shafiabady N, Isa D. A load predictive energy management system for supercapacitor-battery hybrid energy storage system in solar
application using the Support Vector Machine. Appl Energy 2015;137:588–602. https://doi.org/10.1016/j.apenergy.2014.09.026.
[80] Ahmed M, Kuriry S, Shafiullah MD, Abido MA. DC Microgrid Energy Management with Hybrid Energy Storage Systems. 2019 23rd Int Conf
Mechatronics Technol ICMT 2019 2019. https://doi.org/10.1109/ICMECT.2019.8932147.
[81] Wang J, Wang M, Li H, Qin W, Wang L. Energy Management Strategy for Microgrid Including Hybrid Energy Storage. 2018 Asian Conf En ergy, Power
Transp Electrif ACEPT 2018 2019:1–6. https://doi.org/10.1109/ACEPT.2018.8610715.
[82] Amin, Bambang RT, Rohman AS, Dronkers CJ, Ortega R, Sasongko A. Energy management of fuel cell/battery/supercapacitor hybrid power sources
using model predictive control. IEEE Trans Ind Informatics 2014;10:1992–2002. https://doi.org/10.1109/TII.2014.2333873.
[83] Hredzak B, Agelidis VG, Jang M. A model predictive control system for a hybrid battery-ultracapacitor power source. IEEE Trans Power Electron
2014;29:1469–79. https://doi.org/10.1109/TPEL.2013.2262003.
[84] Ju C, Wang P, Goel L, Xu Y. A two-layer energy management system for microgrids with hybrid energy storage considering degradation costs. IEEE
Trans Smart Grid 2018;9:6047–57. https://doi.org/10.1109/TSG.2017.2703126.
[85] Hou J, Song Z, Park H, Hofmann H, Sun J. Implementation and evaluation of real-time model predictive control for load fluctuations mitigation in all-
electric ship propulsion systems. Appl Energy 2018;230:62–77. https://doi.org/10.1016/j.apenergy.2018.08.079.
[86] Hou J, Song Z, Hofmann H, Sun J. Adaptive model predictive control for hybrid energy storage energy management in all-electric ship microgrids.
Energy Convers Manag 2019;198:111929. https://doi.org/10.1016/j.enconman.2019.111929.
[87] Ni F, Zheng Z, Xie Q, Xiao X, Zong Y, Huang C. Enhancing resilience of DC microgrids with model predictive control based hybrid energy storage
system. Int J Electr Power Energy Syst 2021;128:106738. https://doi.org/10.1016/j.ijepes.2020.106738.
[88] Garcia-Torres F, Bordons C. Optimal Economical Schedule of Hydrogen-Based Microgrids With Hybrid Storage Using Model Predictive Control. IEEE
Trans Ind Electron 2015;62:5195–207. https://doi.org/10.1109/TIE.2015.2412524.
[89] Garcia-Torres F, Valverde L, Bordons C. Optimal Load Sharing of Hydrogen-Based Microgrids with Hybrid Storage Using Model-Predictive Control.
IEEE Trans Ind Electron 2016;63:4919–28. https://doi.org/10.1109/TIE.2016.2547870.
[90] Nair UR, Costa-Castello R. A Model Predictive Control-Based Energy Management Scheme for Hybrid Storage System in Islanded Microgrids. IEEE
Access 2020;8:97809–22. https://doi.org/10.1109/ACCESS.2020.2996434.
[91] Jiang W, Zhang L, Zhao H, Huang H, Hu R. Research on power sharing strategy of hybrid energy storage system in photovoltaic power station based on
multi-objective optimisation. IET Renew Power Gener 2016;10:575–83. https://doi.org/10.1049/iet-rpg.2015.0199.
[92] Mesbahi T, Rizoug N, Bartholomeüs P, Sadoun R, Khenfri F, Le Moigne P. Optimal energy management for a Li-ion battery/supercapacitor hybrid
energy storage system based on a particle swarm optimization incorporating nelder-mead simplex approach. IEEE Trans Intell Veh 2017;2:99–110.
https://doi.org/10.1109/TIV.2017.2720464.
[93] Choi ME, Lee JS, Seo SW. Real-time optimization for power management systems of a battery/supercapacitor hybrid energy storage system in electric
vehicles. IEEE Trans Veh Technol 2014;63:3600–11. https://doi.org/10.1109/TVT.2014.2305593.
[94] Zhou T, Sun W. Optimization of battery-supercapacitor hybrid energy storage station in Wind/solar generation system. IEEE Trans Sustain Energy
2014;5:408–15. https://doi.org/10.1109/TSTE.2013.2288804.
[95] Jiang RZ, Lin ZY, Teng H. Optimal configuration method of hybrid energy storage system for wind farm. Adv Mater Res 2013;724–725:576–81.
https://doi.org/10.4028/www.scientific.net/AMR.724-725.576.
[96] Liu Y, Li Z, Lin Z, Zhao K, Zhu Y. Multi-objective optimization of energy management strategy on hybrid energy storage system based on radau
pseudospectral method. IEEE Access 2019;7:112483–93. https://doi.org/10.1109/ACCESS.2019.2935188.
[97] Akram U, Khalid M, Shafiq S. An innovative hybrid wind-solar and battery-supercapacitor microgrid system—development and optimization. IEEE
Access 2017;5:25897–912. https://doi.org/10.1109/ACCESS.2017.2767618.
[98] Raza MQ, Mithulananthan N, Li J, Lee KY. Multivariate Ensemble Forecast Framework for Demand Prediction of Anomalous Days. IEEE Trans Sustain
Energy 2020;11:27–36. https://doi.org/10.1109/TSTE.2018.2883393.
[99] Shen J, Khaligh A. A Supervisory Energy Management Control Strategy in a Battery/Ultracapacitor Hybrid Energy Storage System. Ieee Trans Transp
Electrif 2015;1:223–31. https://doi.org/10.1109/TTE.2015.2464690.
[100] Chettibi N, Mellit A, Sulligoi G, Pavan AM. Adaptive Neural Network-Based Control of a Hybrid AC/DC Microgrid. IEEE Trans Smart Grid
2018;9:1667–79. https://doi.org/10.1109/TSG.2016.2597006.
[101] Chettibi N, Mellit A. Intelligent control strategy for a grid connected PV/SOFC/BESS energy generation system. Energy 2018;147:239–62.
https://doi.org/10.1016/j.energy.2018.01.030.
[102] Yuan Y, Sun C, Li M, Choi SS, Li Q. Determination of optimal supercapacitor-lead-acid battery energy storage capacity for smoothing wind power using
empirical mode decomposition and neural network. Electr Power Syst Res 2015;127:323–31. https://doi.org/10.1016/j.epsr.2015.06.015.
[103] Singh P, Lather JS. Dynamic power management and control for low voltage DC microgrid with hybrid energy storage system using hybrid bat search
algorithm and artificial neural network. J Energy Storage 2020;32:101974. https://doi.org/10.1016/j.est.2020.101974.
[104] García P, García CA, Fernández LM, Llorens F, Jurado F. ANFIS-Based control of a grid-connected hybrid system integrating renewable energies,
hydrogen and batteries. IEEE Trans Ind Informatics 2014;10:1107–17. https://doi.org/10.1109/TII.2013.2290069.
[105] Wang Y, Yang Z, Lin F, An X, Zhou H, Fang X. A Hybrid Energy Management Strategy based on Line Prediction and Condition Analysis for the Hybrid
27
Energy Storage System of Tram. IEEE Trans Ind Appl 2020;56:1793–803. https://doi.org/10.1109/TIA.2020.2967312.
[106] Xiao J, Wang P, Setyawan L. Hierarchical Control of Hybrid Energy Storage System in DC Microgrids. IEEE Trans Ind Electron 2015;62:4915–24.
https://doi.org/10.1109/TIE.2015.2400419.
[107] Morstyn T, Hredzak B, Demetriades GD, Agelidis VG. Unified distributed control for DC microgrid operating modes. IEEE Trans Power Syst
2016;31:802–12. https://doi.org/10.1109/TPWRS.2015.2406871.
[108] Lai J, Zhou H, Lu X, Yu X, Hu W. Droop-Based Distributed Cooperative Control for Microgrids With Time-Varying Delays 2016;7:1775–89.
[109] Dou C, Yue D, Guerrero JM, Xie X, Hu S. Multiagent System-Based Distributed Coordinated Control for Radial DC Microgrid Considering Transmission
Time Delays 2017;8:2370–81.
[110] Joshi MC, Samanta S, Srungavarapu G. Energy management with improved frequency sharing based control for battery/ ultracapacitor hybrid energy
system in the presence of delay. IEEE Trans Veh Technol 2019;13:10571–84. https://doi.org/10.1049/iet-pel.2018.5118.
[11 1] Don g C, Jia H, Xu Q, X iao J, Xu Y, T u P, et al. Ti me-de lay stability analysis for hybrid energy storage system with hierarchical control in DC Microgrids.
IEEE Trans Smart Grid 2018;9:6633–45. https://doi.org/10.1109/TSG.2017.2717504.
[112] Kozhushko Y, Pavković D, Karbivska T, Bondarenko O. Stability analysis of battery-supercapacitor energy storage system for resistance welding. 2019
IEEE 2nd Ukr Conf Electr Comput Eng UKRCON 2019 - Proc 2019:349–54. https://doi.org/10.1109/UKRCON.2019.8879850.
[113] Luo Y, Chen C, Member S, Kadavil R, Liaw B, Muljadi E, et al. A Novel Framework for Optimizing Ramping Capability of Hybrid Energy Storage
Systems 2021;12:1651–62.
[114] Zhong Q, Xie C, Jin S, Li B, Shi K. New Optimal Control Algorithms for Battery-Supercapacitor HESS Based on Wirtinger-Based Integral Inequality
2021:17707–16. https://doi.org/10.1109/ACCESS.2021.3053065.
[115] Lu X, Guerrero JM, Sun K, Vasquez JC. An improved droop control method for dc microgrids based on low bandwidth communication with dc bus
voltage restoration and enhanced current sharing accuracy. IEEE Trans Power Electron 2014;29:1800–12. https://doi.org/10.1109/TPEL.2013.2266419.
[116] Lin P, Wang P, Xiao J, Wang J, Jin C, Tang Y. An Integral Droop for Transient Power Allocation and Output Impedance Shaping of Hybrid Energy
Storage System in DC Microgrid. IEEE Trans Power Electron 2018;33:6262–77. https://doi.org/10.1109/TPEL.2017.2741262.
[117] Xu Q, Hu X, Wang P, Xiao J, Tu P, Wen C, et al. A Decentralized Dynamic Power Sharing Strategy for Hybrid Energy Storage System in Autonomous
DC Microgrid. IEEE Trans Ind Electron 2017;64:5930–41. https://doi.org/10.1109/TIE.2016.2608880.
[118] Xu Q, Xiao J, Wang P, Pan X, Wen C. A Decentralized Control Strategy for Autonomous Transient Power Sharing and State-of-Charge Recovery in
Hybrid Energy Storage Systems. IEEE Trans Sustain Energy 2017;8:1443–52. https://doi.org/10.1109/TSTE.2017.2688391.
[119] Xu Q, Xiao J, Hu X, Wang P, Lee M Y. A Dec entralized Power Management Strategy for Hybrid Energy Storage System with Autonomous Bus Voltage
Restoration and State-of-Charge Recovery. IEEE Trans Ind Electron 2017;64:7098–108. https://doi.org/10.1109/TIE.2017.2686303.
[120] Chen J, Song Q. A Decentralized Dynamic Load Power Allocation Strategy for Fuel Cell/Supercapacitor-Based APU of Large More Electric Vehicles.
IEEE Trans Ind Electron 2019;66:865–75. https://doi.org/10.1109/TIE.2018.2833031.
[121] Chen J, Song Q. A decentralized energy management strategy for a fuel cell/supercapacitor-based auxiliary power unit of a more electric aircraft. IEEE
Trans Ind Electron 2019;66:5736–47. https://doi.org/10.1109/TIE.2018.2866042.
[122] Chen X, Zhou J, Shi M, Yan L, Zuo W, Wen J. A Novel Virtual Resistor and Capacitor Droop Control for HESS in Medium-Voltage DC System. IEEE
Trans Power Syst 2019;34:2518–27. https://doi.org/10.1109/TPWRS.2019.2894754.
[123] Faddel S, Saad AA, Youssef T, Mohammed O. Decentralized Control Algorithm for the Hybrid Energy Storage of Shipboard Power System. IEEE J
Emerg Sel Top Power Electron 2020;8:720–31. https://doi.org/10.1109/JESTPE.2019.2899287.
[124] Ni F, Yan L, Liu J, Shi M, Zhou J, Chen X. Fuzzy logic-based virtual capacitor adaptive control for multiple HESSs in a DC microgrid system. Int J
Electr Power Energy Syst 2019;107:78–88. https://doi.org/10.1016/j.ijepes.2018.11.023.
[125] Xu Q, Zhang C, Xu Z, Lin P, Wang P. A Composite Finite-Time Controller for Decentralized Power Sharing and Stabilization of Hybrid Fuel
Cell/Supercapacitor System with Constant Power Load. IEEE Trans Ind Electron 2021;68:1388–400. https://doi.org/10.1109/TIE.2020.2967660.
[126] Li J, Xiong R, Yang Q, Liang F, Zhang M, Yuan W. Design/test of a hybrid energy storage system for primary frequency control using a dynamic droop
method in an isolated microgrid power system. Appl Energy 2017;201:257–69. https://doi.org/10.1016/j.apenergy.2016.10.066.
[127] Cheng L, Zhang Z, Zhang F. Secondary voltage controls of virtual-droop-controlled bidirectional DC/DC converters in hybrid energy storage system.
IET Power Electron 2020;13:3001–8. https://doi.org/10.1049/iet-pel.2019.1566.
[128] Zhao X, Li YW, Tian H, Wu X. Energy Management Strategy of Multiple Supercapacitors in a DC Microgrid Using Adaptive Virtual Impedance. IEEE
J Emerg Sel Top Power Electron 2016;4:1174–85. https://doi.org/10.1109/JESTPE.2016.2601097.
[129] Lin Y, Fu L, Xiao X. Decentralised power distribution and SOC management algorithm for the hybrid energy storage of shipboard integrated power
system. IET Gener Transm Distrib 2020;14:6493–503. https://doi.org/10.1049/iet-gtd.2020.1345.
[130] Pinthurat W, Hredzak B. Fully decentralized control strategy for heterogeneous energy storage systems distributed in islanded DC datacentre microgrid.
Energy 2021;231:120914. https://doi.org/10.1016/j.energy.2021.120914.
[131] Gu Y, Li W, He X. Frequency-coordinating virtual impedance for autonomous power management of DC microgrid. IEEE Trans Power Electron
2015;30:2328–37. https://doi.org/10.1109/TPEL.2014.2325856.
[132] Duan J, Yi Z, Shi D, Lin C, Lu X, Wang Z. Reinforcement-Learning-Based Optimal Control of Hybrid Energy Storage Systems in Hybrid AC–DC
Microgrids. IEEE Trans Ind Informatics 2019;15:5355–64. https://doi.org/10.1109/tii.2019.2896618.
[133] Xiao ZX, Li HM, Fang HW, Guan YZ, Liu T, Hou L, et al. Operation Control for Improving Energy Efficiency of Shipboard Microgrid including Bow
Thrusters and Hybrid Energy Storages. IEEE Trans Transp Electrif 2020;6:856–68. https://doi.org/10.1109/TTE.2020.2992735.
[134] Madani O, Bhattacharjee A, Das T. Decentralized Power Management in a Hybrid Fuel Cell Ultracapacitor System. IEEE Trans Control Syst Technol
28
2016;24:765–78. https://doi.org/10.1109/TCST.2015.2464295.
[135] Olfati-saber BR, Fax JA, Murray RM. Consensus and Cooperation in Networked Multi-Agent Systems 2007;95.
[136] Olfati-saber R, Murray RM. Consensus Problems in Networks of Agents With Switching Topology and Time-Delays 2004;49:1520–33.
[137] Wang Y, Xu Y, Tang Y, Member S, Liao K, Syed MH, et al. Aggregated Energy Storage for Power System Frequency Control : A Finite-Time Consensus
Approach 2019;10:3675–86.
[138] S hi M, Chen X, Zh ou J, Chen Y, Wen J , He H . Adv anced S econdary Voltage Recovery Control for Multiple HESSs in a Droop-Controlled DC Microgrid.
IEEE Trans Smart Grid 2019;10:3828–39. https://doi.org/10.1109/TSG.2018.2838108.
[139] Zhang R, Hredzak B, Morstyn T. Distributed Control with Virtual Capacitance for the Voltage Restorations, State of Charge Balancing and Load
Allocations of Heterogeneous Energy Storages in a DC Datacenter Microgrid. IEEE Trans Energy Convers 2018;34:1296–308.
https://doi.org/10.1109/TEC.2018.2889065.
[140] Chen X, Shi M, Zhou J, Chen Y, Zuo W, Wen J, et al. Distributed Cooperative Control of Multiple Hybrid Energy Storage Systems in a DC Microgrid
Using Consensus Protocol. IEEE Trans Ind Electron 2020;67:1968–79. https://doi.org/10.1109/TIE.2019.2898606.
[141] Wang B, Wang Y, Xu Y, Zhang X, Gooi HB, Ukil A, et al. Consensus-Based Control of Hybrid Energy Storage System with a Cascaded Multiport
Converter in DC Microgrids. IEEE Trans Sustain Energy 2020;11:2356–66. https://doi.org/10.1109/TSTE.2019.2956054.
[142] Lin P, Zhao T, Wang B, Wang Y, Wang P. A Semi-Consensus Strategy Toward Multi-Functional Hybrid Energy Storage System in DC Microgrids.
IEEE Trans Energy Convers 2020;35:336–46. https://doi.org/10.1109/TEC.2019.2936120.
[143] Han Y, Li Q, Wang T, Chen W, Ma L. Multisource coordination energy management strategy based on SOC consensus for a PEMFC-battery-
supercapacitor hybrid tramway. IEEE Trans Veh Technol 2018;67:296–305. https://doi.org/10.1109/TVT.2017.2747135.
[144] Gamage D, Zhang X, Ukil A, Swain A. Multi-Agent System Based Coordinated Consensus Control for Distributed Multi-Micro-grids. IECON Proc
(Industrial Electron Conf 2020;2020-Octob:1639–44. https://doi.org/10.1109/IECON43393.2020.9255328.
[145] Morstyn T, Hredzak B, Agelidis VG. Cooperative Multi-Agent Control of Heterogeneous Storage Devices Distributed in a DC Microgrid. IEEE Trans
Power Syst 2016;31:2974–86. https://doi.org/10.1109/TPWRS.2015.2469725.
[146] Morstyn T, Savkin A V., Hredzak B, Tuan HD. Scalable Energy Management for Low Voltage Microgrids Using Multi-Agent Storage System
Aggregation. IEEE Trans Power Syst 2018;33:1614–23. https://doi.org/10.1109/TPWRS.2017.2734850.
[147] Zhao H, Guo W. Hierarchical Distributed Coordinated Control Strategy for Hybrid Energy Storage Array System. IEEE Access 2019;7:2364–75.
https://doi.org/10.1109/ACCESS.2018.2886595.
[148] Morstyn T, Savkin A V., Hredzak B, Agelidis VG. Multi-agent slidin g mode control for state of char ge balancin g betwe en battery energy storage systems
distributed in a DC Microgrid. IEEE Trans Smart Grid 2018;9:4735–43. https://doi.org/10.1109/TSG.2017.2668767.
[149] Shi M, Chen X, Zhou J, Chen Y, Wen J, He H. Distributed Optimal Control of Energy Storages in a DC Microgrid with Communication Delay. IEEE
Trans Smart Grid 2020;11:2033–42. https://doi.org/10.1109/TSG.2019.2946173.
[150] Wu X, Xu Y, He J, Shen C, Chen G, Vasquez JC, et al. Delay-Dependent Small-Signal Stability Analysis and Compensation Method for Distributed
Secondary Control of Microgrids. IEEE Access 2019;7:170919–35. https://doi.org/10.1109/ACCESS.2019.2955090.
[151] Zhang R, Hredzak B. Distributed finite-time multiagent control for DC microgrids with time delays. IEEE Trans Smart Grid 2019;10:2692–701.
https://doi.org/10.1109/TSG.2018.2808467.
[152] Zhang R, Hredzak B. Distributed control system with aperiodic sampled time-delayed data for batteries and renewable energy sources in microgrid. IEEE
Trans Sustain Energy 2020;11:1013–22. https://doi.org/10.1109/TSTE.2019.2916206.
[153] Xin L, Zamora R, Baguley C. A Fully Filter-Based Decentralized Control With State of Charge Balancing Strategy for Battery Energy Storage Systems
in Autonomous DC Microgrid Applications 2021;9. https://doi.org/10.1109/ACCESS.2021.3052924.
[154] Lopez-flores DR, Duran-gomez JL, Chacon-murguia MI. A Mechanical Sensorless MPPT Algorithm for a Wind Energy Conversion System based on a
Modular Multilayer Perceptron and a Processor-in-the- Loop Approach. Electr Power Syst Res 2020;186:106409.
https://doi.org/10.1016/j.epsr.2020.106409.
