CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 8, NO.

2, MARCH 2022 329

Review of Operational Control Strategy for DC

Microgrids with Electric-hydrogen Hybrid
Storage Systems
Wei Pei, Member, IEEE, Xue Zhang, Wei Deng, Member, IEEE, Chenghong Tang, Senior Member, IEEE,
and Liangzhong Yao, Fellow, IEEE

Abstract—Hydrogen production from renewable energy development of renewable energy, many countries have pro-
sources (RESs) is one of the effective ways to achieve carbon mulgated various renewable energy policies and energy devel-
peak and carbon neutralization. In order to ensure the efficient, opment plans [3]–[6]. In 2016, the National Development and
reliable and stable operation of the DC microgrid (MG) with
an electric-hydrogen hybrid energy storage system (ESS), the Reform Commission (NDRC) issued the “Energy Production
operational constraints and static dynamic characteristics of and Consumption Revolution Strategy (2016–2030)” and the
a hydrogen energy storage system (HESS) needs to be fully “13th Five-Year Plan for National Science and Technology
considered. First, different hydrogen production systems, using Innovation Plan,” which proposes that by 2030, non-fossil
water electrolysis are compared, and the modeling method of the energy power generation will account for 50% of all power
electrolyzer is summarized. The operational control architecture
of the DC MG with electric-hydrogen is analyzed. Combined generation [7]–[9]. This policy can ensure the long-term and
with the working characteristics of the alkaline electrolyzer, the stable development of the renewable energy industry.
influence of hydrogen energy storage access on the operational In September 2020, China has put forward the development
mode of the DC MG is analyzed. The operational control goals of “carbon peak” and “carbon neutralization,” striving
strategies of the DC MG with electric-hydrogen hybrid ESS are to achieve the peak of carbon dioxide by 2030 and carbon
classified and analyzed from four different aspects: static and
dynamic characteristics of the hydrogen energy storage system, neutralization by 2060. Vigorously developing and utilizing
power distribution of the electric-hydrogen hybrid ESS and the the renewable energy grid connected power generation is
efficiency optimization of hydrogen energy storage. Finally, the an important measure to achieve carbon peak and carbon
advantages of a modular hydrogen production system (HPS) are neutralization. According to relevant research statistics, the
described, and the technical problems and research directions in global cumulative installed wind power and solar photovoltaic
the future are discussed.
(PV) power capacity has reached 733.276 GW and 707.495
Index Terms—Hydrogen energy, DC microgrid, modeling GW by the end of 2020 [10]. With the increasing of the
method, operation control, renewable energy sources. installed capacity of renewable energy, it cannot be used
effectively, which reduces the utilization rate of renewable
energy sources (RESs). According to statistics, approximately
17.2% of wind power and 10.3% of solar power was curtailed
I. I NTRODUCTION in 2016. Although the utilization rate of RESs is improved by
early warning and a guarantee mechanism, approximately 4%
E NERGY crisis and environmental pollution are important
factors restricting the rapid economic development of
all countries throughout the world. Optimizing energy con-
of wind power and 2% of solar power was still curtailed in
2019 [11].
figurations, promoting energy transformation and realizing Because RESs have the characteristics of multiple time
clean, low-carbon and sustainable development have become scale, wide power range fluctuation, intermittence and uncer-
important development goals in future energy fields [1], [2]. tainty, it is difficult to dynamically match the output charac-
In order to reduce carbon emissions and promote industrial teristics of RESs and load characteristics. The energy storage
system (ESS) must be configured to compensate unbalanced
Manuscript received September 16, 2021; revised November 21, 2021; power between RESs and load, and reduce the adverse impact
accepted January 20, 2022. date of online publication, Date of current version of RESs on power system stability and power quality [12]–
February 25, 2022. This work was supported by the Major Science and
Technology Project in Inner Mongolia Autonomous Region (2021ZD0040). [14]. Currently, supercapacitors (SC) and batteries are the most
W. Pei is with the Institute of Electrical Engineering, Chinese Academy of widely used energy storage units. The SC and batteries com-
Sciences, Beijing 100190, and University of Chinese Academy of Sciences, plement each other and can obtain better performance [15]–
Beijing 100049, China.
X. Zhang (corresponding author, e-mail: zhangxue@mail.iee.ac.cn) and W. [17]. However, from the perspective of storage capacity, bat-
Deng are with the Institute of Electrical Engineering, Chinese Academy of teries and SC belong to short-term energy storage systems
Sciences, Beijing 100190, China. (ESSs), so they cannot continuously absorb and compensate
C. H. Tang is with the NARI Group Corporation, Nanjing 211106, China.
L. Z. Yao is with the School of Electrical Engineering and Automation, the unbalanced power in the microgrid (MG) [18], [19]. As
Wuhan University, Wuhan 430072, China. an energy carrier, hydrogen energy has the characteristics of
DOI: 10.17775/CSEEJPES.2021.06960 high energy density, large capacity, long service life and easy
2096-0042 © 2021 CSEE
330 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 8, NO. 2, MARCH 2022

storage and transmission. The application and development of includes AC and DC type structures. Fig. 1 shows the typical
hydrogen energy has attracted extensive attention [20], [21]. structure of the AC MG.
Hydrogen energy is usually used as a long-term ESS, but AC bus
it has a slower time response, which is difficult to apply
to microgrids requiring high dynamic regulation speed and AC DC Grid
frequent start-stop in a short time period. Therefore, the hydro- DC AC AC DC + -
WT DC DC
gen energy storage system (HESS) should be integrated with DC DC Electrolyzer HST
batteries or SC to form an electric-hydrogen hybrid ESS [22], DC AC
PV AC DC HRSs
[23]. The batteries or SC are used to compensate transient DC DC
unbalanced energy, while the HESS is used to compensate DC DC FC
DC AC AC DC
long-term and steady-state unbalanced power. Battery DC DC
At present, the initial investment of hydrogen production EV
equipment is high, and producing hydrogen by water elec-
Fig. 1. Typical structure of the electric-hydrogen AC MG.
trolysis requires a large amount of electricity, therefore the
economy of this hydrogen production method is very poor. The
From Fig. 1, various units are connected to a common
combination of the hydrogen production unit (HPU) and RESs
AC bus by power electronic devices. When the static switch
cannot only increase the utilization of RESs, but also promote
is closed, the grid connected hydrogen production mode is
the economy of the hydrogen production system (HPS) [24]–
selected. When the static switch is off, the MG works in an off
[26]. The generated hydrogen cannot only be used directly and
grid hydrogen production mode. Since the PV cell, battery, FC
efficiently, but also provide hydrogen for fuel cells. With the
and electrolyzer are DC units, more power electronic equip-
rapid development of hydrogen fuel cell vehicles and the rapid
ment is required to achieve energy and voltage conversion for
layout of hydrogen refueling stations (HRSs), the strategic
AC MG. Therefore, the overall investment cost is higher and
position of hydrogen energy development is further promoted.
the efficiency is lower. In addition, the problems of frequency
In order to reduce the production cost, renewable energy
regulation and reactive power compensation also need to be
power generation and HRSs are combined to form a hydrogen
solved for AC MG.
production hydrogenation integrated business model, reducing
The typical structure of the electric-hydrogen DC MG is
the loss caused by power transmission and distribution, and
shown in Fig. 2.
improving the system efficiency [27]–[31]. In addition, the
HPU, hydrogen storage tank (HST) and fuel cell (FC) can DC bus
form a HESS, which cannot only absorb and compensate
DC
the unbalanced power, but also provide power for the MG, AC
AC
realizing the complementary conversion of electricity and gas. Grid
DC
Although the HPU improves the flexible regulation ability of WT + -
DC
MG, it also puts forward new technical requirements for the DC
DC
operational control strategy of MG. The operational control DC Electrolyzer HST
strategy of DC MG with electric-hydrogen hybrid ESS is one PV
of the key technical challenges to ensure the reliable operation DC HRSs
DC
of the system. The scope of this paper is to provide a status DC FC
overview and discuss operational control strategy for electric- DC
Battery DC
hydrogen DC MG. This paper summarizes the research on
DC
operational control strategy of electric-hydrogen DC MG. EV
Initially, the characteristics of different hydrogen production
systems (HPSs) by water electrolysis are introduced, and the Fig. 2. Typical structure of the electric-hydrogen DC MG.
modeling method of electrolyzer is summarized. Secondly,
the operational control architecture and operational mode of From Fig. 2, various units are connected to common DC bus
electric-hydrogen DC MG are analyzed. Thirdly, the oper- by power electronic devices. The RES units, battery energy
ational control strategies of electric-hydrogen DC MG are storage system (BESS) and HESS only need one energy
classified and analyzed from four different aspects: static conversion, which reduces the investment costs and improves
and dynamic characteristics of HESS, power distribution of the overall efficiency. In addition, DC MG only needs to
electric-hydrogen hybrid ESS and efficiency optimization of control the DC bus voltage without considering frequency
hydrogen energy storage. Finally, the technical problems and and reactive power compensation, which simplifies the control
research directions for the future are discussed. structure. Therefore, DC hydrogen production from RESs is a
promising proposal for the future.

II. S YSTEM S TRUCTURE OF T HE E LECTRIC III. C LASSIFICATION AND C HARACTERISTICS OF

H YDROGEN M ICROGRID H YDROGEN P RODUCTION FROM WATER E LECTROLYSIS
An electrolyzer is the core component of HPU. Different
The electric-hydrogen MG system architecture primarily electrolyzers have larger differences in static and dynamic
PEI et al.: REVIEW OF OPERATIONAL CONTROL STRATEGY FOR DC MICROGRIDS WITH ELECTRIC-HYDROGEN HYBRID STORAGE SYSTEMS 331

characteristics, energy consumption, cost, lifespan and site of electrolysis temperature is high, and an external heat source
demand. Therefore, mastering the characteristics of different with a large power source is required to keep the temperature
electrolyzers is beneficial to the design of renewable energy stable.
hydrogen production control proposal.
The HPSs by water electrolysis can be divided into three IV. M ODELING M ETHOD OF AN E LECTROLYZER
forms according to the type of electrolyzer: alkaline elec-
trolyzer (AE), proton exchange membrane (PEM), solid oxide In order to provide full play to the flexible regulation ability
electrolyzer (SOE) [32]–[34]. Comparison of different HPSs of HPU, you need to deeply tap into the complementary
by water electrolysis is shown in Table I. According to characteristics of electric hydrogen hybrid ESS and ensure
Table I, the characteristics of different types of HPSs by water the efficient, reliable and stable operation of electric hydrogen
electrolysis are analyzed as follows: DC MG, static and dynamic characteristics of the electrolyzer
should be deeply understood and mastered. Mathematical
1) The main advantages of alkaline water electrolysis are
modeling is an important means to obtain the working char-
mature technology, large hydrogen production, and relatively
acteristics of an electrolyzer. Therefore, the research progress
low investment cost. At present, the alkaline water electrolysis
of modeling methods of different types of electrolyzers is
system has had a large scale promotion and application. The
summarized.
main defects of an alkaline water electrolysis system are a
large occupation area, corrosive electrolyte, high maintenance A. Modeling Method of Alkaline Electrolyzer
costs in the later stages and large demand for electric energy.
In addition, AE has slow dynamic response and needs to At present, AE modeling methods can be divided into
cooperate with other ESSs for hydrogen production from a linear model, empirical and semi-empirical model, and
RESs. physical model. Different modeling methods are summarized
as follows.
2) The main advantages of a PEM water electrolysis system
are high current density, small occupied area, good dynamic 1) Linear Model
response characteristics and start-stop performance, and a wide Linear model is a simplified equivalent modeling method.
operational range. A PEM electrolyzer has good matching with The voltage of the electrolyzer is simulated by constant volt-
RESs. At present, it has been commercialized and applied in age source and resistance in series, which cannot accurately
a small scale. The anode and cathode catalysts of a PEM elec- describe the characteristics of the electrolyzer [35]. Therefore,
trolyzer uses precious metals, so the cost of a PEM electrolyzer the linear model is usually not used.
is high. In addition, the technology of domestic manufacturers 2) Empirical and Semi-empirical Model
is immature, there is a large gap with international advanced The empirical model does not need to establish a the-
technology, and the key equipment depends on importing. oretical model and is only fitted according to the experi-
3) The main advantages of a SOE water electrolysis system mental measured data. Therefore, the model cannot reveal
are high efficiency and high current density. At present, a SOE the electrochemical reaction mechanism and has no practi-
electrolyzer is primarily in the laboratory stage, so it has not cal physical meaning [36]. For solving these problems, the
been commercialized. The main defect is that the requirement method of combining theoretical modeling with experimental

TABLE I
C OMPARISON OF D IFFERENT H YDROGEN P RODUCTION S YSTEMS BY WATER E LECTROLYSIS
Specification Alkaline PEM SOE
Charge carrier OH− H+ O2−
Cell temperature 60–90◦ C 50–80◦ C 700–1000 ◦ C
Electrolyte 25∼30% KOH Pure water Y2 O3 -Zr O2 , Sc2 O3 -ZrO2
Overall reaction H2 O=H2 +1/2O2 H2 O=H2 +1/2O2 H2 O=H2 +1/2O2
Anode reaction 2OH− =1/2O2 +H2 O+2e− H2 O=1/2O2 +2e− +2H+ O2− −2e− =1/2O2
Anode catalyst Ni2 CoO4 , La-Sr-CoO3 Ir/Ru oxide (La, Sr) MnO3 , Ni-YSZ
Cathode reaction 2H2 O+2e− =H2 +2OH− 2H+ +2e− =H2 2H2 O+2e− =H2 +O2−
Cathode catalyst Ni-Mo/ZrO2 -TiO2 Platinum Ni-YSZ/Ni-GDC
Dynamic response capability Comparatively good Good Weak
Electrolyzer efficiency 63–71% 60–68% 100%
System efficiency 51–60% 46–60% 76–81%
Electrolyzer energy consumption 4.2–4.8 kWh/Nm3 4.4–5.0 kWh/Nm3 3 kWh/Nm3
System energy consumption 5.0–5.9 kWh/Nm3 5.0–6.5 kWh/Nm3 3.7–3.9 kWh/Nm3
Cell pressure 0–30 bar 0–30 bar 0–30 bar
Operating range 20%–110% 0%–160% 20%–100%
Hydrogen purity >99.8% 99.999 99.999
Current density 0.2–0.5 A/cm2 1.0–2.0 A/cm2 0.3–2.0 A/cm2
Hydrogen production per stack <1400 Nm3 /h <400 Nm3 /h <10 Nm3 /h
System lifetime 55–120 kh 60–100 kh 20–80 kh
Stop/go cycling Comparatively good Good Weak
Cold start-up time >60 min >5 min >60 min
Warm start-up time 1–5 min <5 s >15 min
Technology maturity Widespread commercialization Commercialization Research &Development
Investment costs 3000–12000 ¥/kW 10000–16000 ¥/kW >16000 ¥/kW
332 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 8, NO. 2, MARCH 2022

data is adopted. First, the theoretical mathematical model is electrolyzer can be divided into analytical, semi-empirical
established according to fundamental thermodynamics, heat model and mechanistic model [48]. In the research of electric
transfer theory and empirical electrochemical relationships. hydrogen DC MG operational control, more attention is paid to
The main parameters in the theoretical model are determined the characteristics of voltage, current and efficiency. Therefore,
by using the experimental data through the fitting algorithm. it is an effective way to obtain the static and dynamic charac-
The parameters are modified by comparing the model with the teristics of a PEM electrolyzer through the empirical and semi-
experimental results, and then a semi-empirical mathematical empirical models. The classification of a PEM electrolyzer
model is obtained. At present, the semi-empirical model is the modeling method is shown in Fig. 3.
most commonly used modeling method, which can be divided
into a static model and dynamic model. The static model can Modeling
better simulate the electrolyzer voltage, hydrogen production,
efficiency and hydrogen purity under specific temperature
and pressure conditions [37]–[39]. However, the dynamic Empirical Semi-empirical
model pays more attention to the dynamic behavior of the
electrolyzer, including the dynamic response characteristics of
electrolyzer voltage, current, temperature, purity, liquid level, Dynamic Static
pressure and other parameters during the start-up process and
the change of current or power reference value [40], [41].
3) Physical Model System Cell/Stack

To ensure the high accuracy of the semi-empirical model,

Fig. 3. Modeling methods of the PEM electrolyzer.
the key parameters need to be determined by long-term
experimental data. At the same time, the semi-empirical
model is only applicable to one specific electrolyzer and not Similar to the AE modeling method, the empirical model
to other electrolyzers, so the universality is poor. To deal of PEM electrolyzer is also directly fitted through the exper-
with these problems, a multi-physical model of an alkaline imental data to obtain the static characteristics of the PEM
electrolyzer is proposed. This model considers the changes electrolyzer [49], [50]. The semi-empirical model can also be
of all structural parameters and operational parameters of divided into a static model and dynamic model. The difference
the electrolyzer, which can enhance the accurateness of the from the AE model is primarily reflected in the physical
model [42]. Furthermore, the reduced order model based on structure of the electrolysis cell and electrochemical reaction
equivalent circuit is proposed, which reduces the number of mechanism. In the static modeling, the theoretical model is still
parameter requirements and the complexity of modeling [43]. established by thermodynamic theory, heat transfer theory and
A semi-physical model based on phenomenological theory is electrochemistry. However, there are main differences in the
proposed to describe the dynamic characteristics of HPS [44]. impact analysis of operating parameters and structural param-
The advantages and disadvantages of different modeling eters, focusing more on the impact of different temperatures
methods are shown in Table II. on the exchange current and charge transfer coefficient [51],
[52]. In the voltage modeling of the electrolyzer, not only
TABLE II
C OMPARISON OF D IFFERENT AE M ODELING M ETHODS
the open circuit voltage, activation overvoltage and ohmic
overvoltage are considered, but also the diffusion overvoltage
Modeling methods Description of advantages and disadvantages
Linear model Simple and easy to implement, but it cannot
and mass transfer effects are also taken into account [53].
simulate the working and operating characteristics A new ohmic loss model is proposed, and the resistance of
of electrolyzer, so it is usually not used. different components are fully considered, such as the bipolar
Empirical and The static and dynamic characteristics of the
semi-empirical electrolyzer can be better simulated, but the
plate, electrode and membrane thicknesses [54].
model model has poor universality and cannot clearly Because the PEM electrolyzer has the advantage of fast
describe the internal physical and chemical dynamic response and good matching with RESs, the PEM
reaction mechanism.
Physical model Good universality and clear mechanism, but the electrolyzer is more suitable for hydrogen production from
modeling method is complex. RESs. Relevant scholars have carried out in-depth research
on the accurate dynamic modeling of the PEM electrolyzer.
From the model scale, it can be divided into PEM cell/stack
B. Modeling Method of PEM Electrolyzer and PEM HPS. To describe the dynamic behavior of the
For the mathematical modeling of a PEM electrolyzer, dif- PEM electrolyzer, the dynamic model is established by using
ferent classification methods are provided in relevant literature. Simulink simulation software. The simulation model is applied
Literature [45] only divides the voltage and efficiency models, to analyze the dynamic performance of electrolyzer voltage,
which is not comprehensive. Literature [46] summarizes the electrolyzer current, energy efficiency, energy consumption
research on thermal effects, semi-empirical model and two- and temperature, etc., [55]–[57]. However, the previously
phase flow effects of the PEM electrolyzer. Literature [47] dynamic model is developed under the condition of a fixed
classifies the modeling methods of electrochemical model, parameter, so the accuracy of the model will be reduced for
thermal model, mass transfer model and fluidic model of a other input currents. To deal with this problem, an adaptive
low-temperature PEM electrolyzer. The models of the PEM static-dynamic model is proposed, the parameters of the model
PEI et al.: REVIEW OF OPERATIONAL CONTROL STRATEGY FOR DC MICROGRIDS WITH ELECTRIC-HYDROGEN HYBRID STORAGE SYSTEMS 333

can be adjusted independently under the input current change conditions. Therefore, empirical models are often used in
conditions [58]. Similarly, an adaptive cell voltage static- situations where the accuracy of the model is not high.
dynamic model is developed to investigate the degradation 2) Semi-empirical Model
and wear effects caused by dynamic operations and current The semi-empirical model is established based on the mech-
ripple [59]. The existing modeling research primarily focuses anism model and the parameters that are difficult to obtain
on the behavior of cells or stacks. Literature [60] developed in the model are determined through experimental data or
the dynamic model of PEM HPS by using Simulink software, parameter identification methods. The semi-empirical model is
which can better evaluate the efficiency and loss proportion of a simplification of the mechanism model. A simple dynamic
PEM HPS. model is proposed to describe the transient output character-
istics of PEMFC [71]. A semi-empirical model-based prog-
C. Modeling Method of SOE
nostics method is developed to achieve degradation prediction
At present, SOE is primarily in the laboratory research and and evaluate the remaining service life. The electrochemical
development stage. Although it has high energy efficiency, surface area, equivalent resistance and recovery factor are
the start-stop of SOE is inconvenient and the response is introduced to predict the degradation trend and performance
slow. The modeling methods of SOE can be divided into a recovery of PEMFC [72]. A semi-empirical model is proposed
macro-scale model and micro-scale model [61]. In this macro- for PEMFC, and the whale optimization algorithm is used
scale model, the electrochemical characteristics of SOE and for obtain unidentified parameters [73]. A one-dimensional,
SOE stack were researched, and the effects of fabrication semi-empirical, and steady-state model of PEMFC is devel-
parameters and operating conditions on SOE performance oped, and unknown parameters are obtained by using the
were analyzed through parametric simulations [62]–[64]. In experimental results [74]. A semi-empirical model based on
this micro-scale model, the research primarily focused on the online identification is proposed to improve the performances
transport and electrochemical reactions. The purpose of the of PEMFC [75].
micro-scale model is to optimize the internal structure and Although the semi empirical model is built on the basis of
process parameter design of SOE [65], [66]. the mechanism model, it is still unable to describe the internal
mechanism characteristics of PEMFC.
V. M ODELING M ETHOD OF F UEL C ELL 3) Mechanism Model
FC is the core component of HESS. Its static and dynamic The mechanism model is primarily used to describe the in-
characteristics will also put forward new requirements and ternal electrochemical and physical properties of the PEMFC,
challenges for the operational control of the electric-hydrogen such as the capacitance of double-layer charge effect, mass dif-
DC microgrid. According to the operating temperature, FC can fusion, material conservation, thermodynamic characteristics,
be divided into low-temperature and high-temperature types. and voltage drops inside the FC [76]–[78]. For the calculation
Among all fuel cells, PEMFC are most likely to be applied of the mechanism model, numerical solution is generally
for distributed generation and microgrid applications. PEMFC used [79].
can provide reliable power in steady state. However, when the The mechanism model is primarily used for the optimization
load rapidly changes, they cannot respond quickly due to their design of the internal structure and material selection of
slow internal electrochemical and thermodynamic responses. PEMFC. Therefore, the mechanism model is not suitable for
Accurate modeling methods are needed to predict and evaluate the design and optimization of the PEMFC control system.
steady-state and dynamic responses of PEMFC. Therefore, the 4) Data-driven Model
modeling methods of PEMFC are summarized. Although the mechanism model can obtain higher accuracy,
At present, PEMFC modeling methods can be divided into the modeling is very complex and the obtained model is
an empirical model, semi-empirical model, mechanism model difficult to solve. In order to avoid complex modeling and
and data-driven model [67]. Different modeling methods are ensure high accuracy of the model, the data-driven modeling
summarized as follows. method represents an effective solution.
1) Empirical Model The data-driven modeling method is primarily based on a
The empirical model of PEMFC does not need to establish large number of experimental data to obtain a high-precision
a theoretical model. The unknown coefficients of the em- model through training and learning. The data-driven mod-
pirical model are only fitted based on experimental data. A els primarily include non-physics-based and physics-informed
following correlation model is proposed to predict the cell data-driven results. The data-driven (non-physics-based) mod-
voltage. However, this model is inaccurate for high current els do not need to know the internal physical parameters,
density regions because the voltage is overestimated [68]. A but rather construct the model through training, learning
mathematical correlation is established by FC testing data, and and statistics according to a large number of experimental
the voltage losses are evaluated [69]. An empirical equation data [80]–[83]. The physics informed data-driven model com-
is developed to fit the experimental cell voltage, and the bines physical knowledge with the data-driven model, which
exponential term is introduced to compensate for the mass can increase constraints through physical knowledge and avoid
transport over-potential [70]. generating incorrect models [84].
Due to the limited experimental data, the fitting models Although the data-driven modeling method does not need
obtained from these data are difficult to apply to all operating internal parameter information, in order to obtain an accurate
334 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 8, NO. 2, MARCH 2022

model, a large amount of available data is required, so the cost so its dynamic response is slow. Due to the mismatch of
of the data-driven modeling method is high. dynamic characteristics, it often cooperates with other ESS
The comparison of different PEMFC modeling methods is to form a hybrid ESS. The dynamic power distribution of
shown in Table III. hybrid ESS and coordinated control of multi-units will have a
TABLE III
significant impact on system reliability, mode switching times,
C OMPARISON OF D IFFERENT PEMFC M ODELING M ETHODS energy storage life, system efficiency and start-stop times.
Modeling methods Description of advantages and disadvantages Therefore, the power allocation strategy of hybrid ESS and
Empirical model Simple and easy to implement, but it cannot coordinated control of multi-units are the main difficulty of
accurately describe internal mechanism. the short-time control scale. This paper pays more attention
Semi-empirical It can describe part of the internal mechanism,
model but the model still cannot describe the complete to the coordination control of multiple units in a short time
internal reaction mechanism. scale.
Mechanism model It can accurately describe internal mechanism,
but the modeling method is complex. A. Operational Control Structure of the Electric-hydrogen
Data-driven model It does not require knowledge about internal
system parameter information, but it needs a DC MG
large amount of experimental data. The electric-hydrogen DC MG can be divided into two
working modes: grid connected and off grid. During grid con-
nected operations, its main task is to give priority to the output
VI. O PERATION C ONTROL S TRATEGY OF power of RESs to meet the power demand of the power grid.
E LECTRIC - HYDROGEN DC M ICROGRID The HESS is similar to the control objectives of other ESSs.
According to the current research progress and engineering It is usually regarded as an auxiliary equipment to absorb or
demonstration application of renewable energy coupled HPS, compensate the unbalanced power between RESs and power
the most widely used approach is still alkaline water electroly- demand of the power grid, and enhance the utilization of RESs.
sis hydrogen production, such as the Guyuan dongxinying hy- In the off-grid operational mode, because the system loses
drogen production station (10 MW alkaline water electrolysis the external power grid, it is necessary to coordinate multiple
HPS, and the annual hydrogen production capacity can reach types of sources, loads and ESSs in the DC MG to ensure the
17.52 × 105 Nm3 ) and Chongli Wind-PV-storage-hydrogen voltage stability of the DC bus. At this time, more emphasis is
demonstration project (3 MW alkaline water electrolysis HPS, placed on the coordinated control of each unit interface device
and the hydrogen production capacity can reach 400 Nm3 /h) in multiple operational scenarios, which puts forward higher
invested and built by Hebei Construction & Investment Group control requirements for multi-type equipment in the local
New Energy Co., Ltd.. Therefore, the following primarily layer. Its main task is to absorb or compensate the unbalanced
takes the integration of alkaline water electrolysis HPU as an power between the RESs and local load demand through HESS
example to study and analyze an operational control strategy and other ESSs, fully tap the complementary characteristics of
for the electric-hydrogen DC MG. HESS and other ESSs, ensure the system power balance and
Currently, the research on operational control strategy pri- DC voltage stability under multiple operational modes, and
marily covers two time scales. In the research on the long- enhance the utilization of RESs.
term operational scale, it does not pay attention to the control The electric-hydrogen DC MG primarily includes two
and mode switching of interface devices at the local layer, control architectures: the two-layer control architecture and
but focuses on the optimal dispatching strategy of the electric decentralized control architecture.
hydrogen DC MG. The main goal is to improve system 1) Two-layer Control Architecture
economy and reduce cost and energy loss [85]–[90]. In the At present, the developed two-layer control architecture is
existing proposals, most studies primarily focus on improving shown in Fig. 4, including upper layer energy management
the economy of HPS and reducing investment cost. However, strategy and local layer controller. The energy management
the actual HPS and FC systems usually include multiple parts. system (EMS) interacts with the local equipment through com-
The HESS is a power-hydrogen-heat multi-energy coupling munication technology. The upper layer EMS generates the
system. In the traditional HPS, the hydrogen production
power is relatively stable without significant change, and
the performance evaluation of HPS is relatively easy. In the Energy flow
EMS Information flow
renewable energy hydrogen production system, the power of Hydrogen flow
the electrolyzer and FC changes in real time, and it is diffi-
cult to evaluate the impact of wide-range power fluctuations
on the performance of the electrolyzer, such as Life cycle,
performance degradation and remaining life. Therefore, the DC DC DC DC DC DC
AC AC DC DC DC DC
economic optimization and scheduling control, considering the
performance of HESS, represent the major difficulties of long- + 
time scale control. The control goal of short-time operational
scale is to balance the system power and operate stably. In the Grid WT PV Battery AE HST FC
internal working process of HESS, there is mutual conversion
among thermal energy, electrical energy and chemical energy, Fig. 4. Typical structure of two layer control.
PEI et al.: REVIEW OF OPERATIONAL CONTROL STRATEGY FOR DC MICROGRIDS WITH ELECTRIC-HYDROGEN HYBRID STORAGE SYSTEMS 335

control mode and power command of each unit according to control and artificial neural network are combined to form
the received local information, and then sends the instructions artificial neural networks fuzzy inference system (ANNFIS).
to the local device through communications. The local layer The data is trained by the artificial neural network, and the
device executes the upper layer commands. membership function can be tuned by the training data [108].
In the grid connected operational control, the current re- The control objective of ECMS is to ensure the minimum
search primarily adopts the two-layer control architecture. The instantaneous hydrogen consumption. A two-level EMS is
grid-connected converter is used to keep the DC bus voltage developed, and the ECMS is applied to distribute power
stability. The EMS distributes the power of electric-hydrogen between the battery pack and FC system [130]. Three energy
hybrid ESS according to the unbalanced power between the management strategies (basic rule-base strategy (RB-EMS),
RESs and the power grid and operational state of each unit, so frequency separation rule-based strategy (FSRB-EMS), and
as to ensure the power balance. In grid connected operations, ECMS) are proposed and compared. Through contrastive
more emphasis is placed on the research of the upper layer analysis, the ECMS achieves a 3% improvement in hydrogen
power management strategy (PMS), while the control method use [131]. A hierarchical EMS is developed to enhance the
of the local layer device is relatively simple. The power only economy of this DC MG, and the fuel consumption is reduced
needs to be controlled according to the power command issued by the ECMS [146].
by the EMS, so the control mode of each interface device does The comparison of different energy management strategies
not need to be switched [91]–[112]. During off grid operational is shown in Table IV.
control, due to the limited capacity of the ESS, once the TABLE IV
constraints are met, it is necessary to switch the control C OMPARISON OF D IFFERENT E NERGY M ANAGEMENT S TRATEGIES
modes of different units to maintain the system power balance. Existing
Description of advantages and disadvantages
Therefore, the control strategies of the local layer equipment methods
are more complex in island operational mode [113]–[151]. SMC Simple, reliable and easy to operate. But SMC will become
complex with the increase of the types of units and the
In the two-layer control architecture, more research focuses number of constraints.
on upper layer energy/power management control. The energy MPC It can effectively deal with uncertainties and constraints.
management strategy primarily includes: state machine control However, it is difficult to determine the multi-objective
weight coefficient.
(SMC), model predictive control (MPC), fuzzy control, equiv- Fuzzy It does not need accurate model of the system, and the
alent consumption, minimization strategy (ECMS). The EMS design of energy management is simple. However, the
based on SMC is one of the most commonly used and mature control effect depends on engineering experience.
ECMS ECMS can minimize hydrogen consumption, but this method
strategies. In SMC strategy, the definition of various modes is limited to the optimization between fuel cell and battery.
is based on the status of state of charge (SOC) and hydrogen
storage capacity. The power reference value and control mode
of each unit are generated by SMC strategy [91]–[101]. The 2) Decentralized Control Architecture
complexity of SMC is closely related to the type of units and The reliability of the two layer control architecture depends
the number of constraints in the system. The SMC will become heavily on communication. If the communication fails, the
complex with the increase of the types of units and the number bottom devices cannot receive the power command and control
of constraints. mode, resulting in the collapse of the whole system. To
MPC methodology is a good control choice for multi- deal with these problems, the decentralized control without
parameter plant systems. The uncertainties and constraints communication is a common solution, and this method has
can be handled effectively. In the MPC optimization, power better reliability. Fig. 5 provides the typical structure of the
reference values of the HESS and BESS can be flexibly con- decentralized control. All units in the MG adopt droop control,
trolled by adjusting the proper weight factors of the objective
function [151]. A PMS based on distributed explicit model
predictive control (DeMPC) is developed, the constraints on Battery WT
the current ramp rates are considered to better match the
dynamic characteristics of the FC and the electrolyzer [120]. Droop DC AC MPPT
A decentralized MPC is proposed to effectively avoid frequent CP DC DC Droop
start and stop [136].
The fuzzy control method is very suitable for nonlinear
systems. Fuzzy control does not need an accurate mathematical Droop Droop
DC DC
model, and is insensitive to the change of system parameters, DC DC
CP CP
and has strong robustness. In addition, energy management
strategies can be constructed only according to simple lan- + −
guage rules. Compared with the SMC, it can simplify the
design process of energy management [137]–[141]. The tra- AE HST FC
ditional fuzzy energy management strategy uses the empirical
method to design the membership function and fuzzy rules, Modes Selection Scheme Based on DC Bus Signal

so this strategy may not be optimal. To further improve the

system performance and realize the optimal control, the fuzzy Fig. 5. Typical structure of Decentralized control.
336 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 8, NO. 2, MARCH 2022

and use the DC bus voltage signal for hierarchical division to cell current is large. The voltage of the electrolyzer is greatly
determine the control mode of each unit. The output power affected by temperature and less by pressure.
of all the different devices will be adjusted independently 2) The purity of hydrogen is closely related to the current.
according to the droop characteristics. To ensure the safety of HPS, the operational range of the
In the decentralized control architecture, more research electrolyzer needs to be limited. Hydrogen purity is affected, at
focuses on coordinated control, smooth switching and stability the same time, by temperature and pressure changes. Hydrogen
control. How to combine hydrogen energy characteristics with purity can be adjusted by properly adjusting temperature and
droop control is the focus of decentralized control research. pressure.
To consider the characteristics of HESS, some improved 3) The hydrogen production is primarily related to current
droop controls are developed [152]–[157]. A decentralized and is less affected by temperature.
energy management strategy based on a mode-triggered droop 4) The efficiency of AE is sensitive to temperature and
proposal is considered, the droop characteristic curve of each current changes. Under specific temperature conditions, there
unit is designed according to the three states of SOC high, is a unique current corresponding to the optimal efficiency,
medium and low [152]. The droop control based on the so the hydrogen production efficiency can be optimized by
efficiency characteristic curve is proposed, and the efficiency adjusting the current.
of the HPU and FC units can be dynamically adjusted [153]– Based on the above analysis, it is necessary to restrict the
[155]. In addition, an active coordinated control strategy is working range of HPU. In addition, since the HESS also
proposed, and the power voltage reverse droop control is used includes a hydrogen storage tank and FC, it is necessary to
to highlight the controllability of the electrolyzer [157]. consider the pressure constraints of the hydrogen storage tank
and the working range constraints of FC. The integration of
B. Influence of Hydrogen Energy Storage Integration on the HESS introduces three additional constraints. Compared with
System Operational Mode the traditional DC MG, its operational mode is significantly
To analyze the influence of HPU on the DC MG operational increased and more complex.
mode, first, the static characteristics of the AE electrolyzer
are analyzed, and the constraints of HPU participating in the C. Operational Control Strategy Considering Static Charac-
operational control are extracted. The working characteristics teristics of Hydrogen Energy Storage
of AE can be obtained according to the parameters in refer- From Fig. 6, the static characteristics of AE and the con-
ence [32], as shown in Fig. 6. straints of HESS should be fully considered in the operational
Figure 6, can be summarized as follows: strategy of electric hydrogen DC MG to meet the actual
1) The change of cell voltage is small, while the change of requirements. The operational control method of the grid-

2.2 2.2 99.5

2 2
Ucell (V)

Ucell (V)

H2p (%)

99
1.8 1.8
98.5
1.6 1.6
1.4 1.4 98
100 10 90
500 8 500 80 500
80 400 6 400 400
300 4 300 70 300
200 pae (bar) 2 200 Tae (ć) 200
Tae (ć) 60 100 I (A) 100 Iae (A) 60 100 I (A)
ae ae

(a) (b) (c)

3
99.5 70
nH2 (Nm3/h)

2
H2p (%)

99 60
ηae (%)

98.5 1
50

98 0 40
9 600 100
8 500 80 500
7 400 400 400
300 200 80 300
pae (bar) 6 5 200
100 I (A) Tae (ć) 60 0 Tae (ć) 100
200
ae Iae (A) 60 Iae (A)

(d) (e) (f)

Fig. 6. Working characteristics of AE. (a) U-I characteristics of AE under temperature change conditions. (b) U-I characteristics of AE under pressures change
conditions. (c) Hydrogen purity of AE under temperature change conditions. (d) Hydrogen purity of AE under pressures change conditions. (e) Hydrogen
production. (f) Efficiency of AE.
PEI et al.: REVIEW OF OPERATIONAL CONTROL STRATEGY FOR DC MICROGRIDS WITH ELECTRIC-HYDROGEN HYBRID STORAGE SYSTEMS 337

connected system of wind/PV/electrolyzer/SC is proposed, and regulation ability needs to be further explored. The coordinated
four typical operating conditions are analyzed. However, only control proposal of a wind-to-hydrogen is considered, the DC
HPU is considered in the system, and the above constraints bus voltage of the WT generator is regulated by a machine-
are not considered [91]–[94]. A PMS for PV/FC/hybrid energy side converter and HESS. However, the FC and HPU needs
storage DC MG is developed to maintain the DC voltage sta- frequent start-up and shut-down, which eventually degrades
bility and power balance [95]. However, the electrolyzer is not their performance and lifespan [110]. A fuzzy PMS based
introduced into the system. The coordinated control proposal on hydrogen-priority is proposed, a large amount of power
is developed for the grid-connected hybrid renewable power in the system is used to produce hydrogen, which reduces
system based on hydrogen energy storage [96]. The control the service life of the battery. But the round trip efficiency
method of HESS is proposed, and the main grid is used to is very low [119]. A PMS is designed for island DC MG to
compensate unbalanced power [97]. The operational control of maintain DC voltage stability and power balance [127]. The
the grid-connected RES-hydrogen DC microgrid is proposed, control method is proposed for the microhydro power system
and five operational modes are considered [98]. However, the with FC/electrolyzer/ultracapacitor. The FC and electrolyzer
power grid is applied to keep DC voltage stability, and the are used to maintain long-term energy balance, whereas the SC
control strategy has been greatly simplified [96]–[98]. The acts as an energy buffer for the transient compensation [133].
control strategy of the hybrid system of wind/hydrogen/FC/SC But the pressure constraint of the HST and SOC constraints
is proposed, ten operational modes are divided under the of SC are not considered. A control strategy is proposed for a
conditions of low wind speed and high wind speed, the wind/FC/hydrogen/battery system for a standalone operation,
control strategies of different units are given, but the pressure and the constraints of BESS and HESS are fully considered. In
constraints of HST were not considered in the energy manage- the energy management and power regulation systems, eight
ment strategy [99]. A coordinated control strategy is proposed operational modes are divided by the status of wind speed and
for the hybrid system of wind/FC/hydrogen/battery, and the load [134].
pressure constraints of HST are considered, so as to increase
the operational modes of the system [100]. The test platform D. Operational Control Strategy Considering Dynamic Char-
of the wind/PV/hydrogen hybrid system is established and acteristics of Hydrogen Energy Storage
an online energy control strategy is developed, and the con- Due to the activation polarization overvoltage effect in the
straints of the electrolyzer, FC and hydrogen storage tank are voltage model of AE, the dynamic response characteristics
considered [101]. The interior point algorithm is proposed to of AE voltage are slow, resulting in slow power dynamic
find the optimal references for the voltage source converter response of HPU. In the process of load step increase, there is a
(VSC) and Energy Hub. However, the energy management fuel starvation phenomenon, resulting in instantaneous voltage
strategy only considers the current state and battery voltage reduction, even negative, which will seriously damage the
constraints, and other constraints are not considered [102]. A performance and service life of the FC. Therefore, the current
semi-decentralized control strategy is proposed for electric ve- or power slope must be limited by the FC controller [158],
hicle fast charging and hydrogen production. The decentralized [159]. Due to the slow dynamic response speed of HPU
control strategy based on the virtual battery model and DC and FC power generation systems, the actual power cannot
bus-signaling is used for photovoltaic/BESS/EV units, while quickly track the power command, resulting in a short-term
decentralized control and power-based control are used for power imbalance. To solve the adverse impact of frequent
the electrolyzer to reach its hydrogen production target [109]. changes in hydrogen production power on system life and
A HESS-priority PMS of DC MG with PV/hydrogen/FC/SC efficiency, and considering the slow response characteristics of
is proposed, the six operational states are considered [113]. HPU, literature [103] shows how the HESS can operate stably
The EMS based on a multi-agent technology is developed with constant power. However, this method lacks flexibility.
and the hydrogen-priority control proposal is used for au- The coordinated control strategy of the hybrid system of
tonomous hybrid system [114]. However, the supercapacitor PV/FC/hydrogen/SC is proposed, seven operational modes are
is assumed to compensate any unbalanced power and meet divided, and the SC is used to quickly compensate the transient
the system stability, which simplifies the energy manage- unbalanced power to enhance the dynamic response perfor-
ment [113], [114]. A battery-priority PMS of DC MG with mance, so as to realize fast dynamic compensation and steady-
Wind/PV/hydrogen/FC/battery is proposed, but the pressure state power stabilization [104]. Similarly, literature [105] and
constraints of HST were not considered [115]. A PMS for literature [106] still use SC to quickly compensate the unbal-
household solar-hydrogen power plants is designed to reduce anced power caused by the response delay of HPU and FC
costs [116]. A control method is proposed for an autonomous for the wind-hydrogen system. An adaptive dynamic power
electric-hydrogen hybrid system [117]. The fuzzy controller is management and control strategy is proposed for HESS and
used to regulate the DC bus voltage, and the low pass filter is SC ESS, the power rate limit of HPU and FC are considered,
adopted to realize the power frequency division of HESS and the eight operational modes are divided by the status of
SC. However, the inherent constraints of the electric-hydrogen HST and SC, and the set-reset flip-flop based fixed frequency
hybrid ESS are not taken into account. The energy man- current controller is used to accurately track the reference
agement algorithm of PV/hydrogen/FC/SC is proposed, six currents [135].
operational modes are divided in the system [118]. However, The comparison of different operational control strategies
there is only one control mode in each unit, and the flexible considering dynamic characteristics are shown in Table V.
338 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 8, NO. 2, MARCH 2022

TABLE V and the power of BESS and HESS are shared by a preset
C OMPARISON OF O PERATIONAL C ONTROL S TRATEGIES C ONSIDERING
DYNAMIC C HARACTERISTICS OF HESS fitting curve [132]. But this power allocation method lacks
Existing methods Description of advantages and disadvantages
a theoretical basis and completely depends on the defined
[103] Simple and easy to implement, but lack of curve. An EMS based on an adaptive neuro-fuzzy controller
flexibility, which is not conducive to the is proposed for the grid-connected electric-hydrogen hybrid
dynamic consumption of renewable energy.
[104]–[106], [135] The power of HPU and FC can be adjusted
system, the membership functions are obtained by training and
dynamically. The proposal has high testing data, and the fuzzy EMS is used to optimize the power
flexibility, but other ESSs must be configured, distribution of the electric-hydrogen hybrid ESS, which can
with high cost and control complexity.
continuously ensure that the SOC of the battery and hydrogen
storage levels are maintained within a reasonable range [108].
E. Operational Control Strategies Considering Start-stop Based on both real-time and long-term predicted data of the
Characteristics of Hydrogen Energy Storage energy generation and consumption, a fuzzy EMS is proposed
to select the control mode [137]. The PMS based on a fuzzy
Frequent start-up and shut-down actions for the electrolyzer logic controller (FLC) is developed to intelligently manage
and the FC will eventually degrade their performance and the output power of the FC system [138]. A PMS based on
reduce their lifespan. In order to solve this problem, a hys- FLC is proposed to enhance the hydrogen production, and
teresis band control (HBC) method is developed for BESS reduce the usage of the battery and promote the lifespan of
and HESS, the on/off switching of FC and the electrolyzer the battery [139]. But the hydrogen storage level has not
can be controlled according to the status of SOC [121]. Two been effectively controlled in [138], [139]. A PMS based on
PMSs based on HBC are developed to prevent excessive use intelligent FLC is developed to ensure a continuous power
of the battery, and the influence of the hysteresis band size supply and maximize hydrogen production [140]. A power
on system performance is also discussed [123]. Similarly, a allocation strategy based on FLC is proposed for islanding
PMS with HBC is also applied for Hydrogen-Based MG [124]. DC MG, and this method can cause the SOC of the bat-
A decentralized MPC is proposed to effectively avoid the tery and hydrogen storage level to approach a reasonable
frequent turning on and off of the electrolyzer [136]. The range [141]. A hierarchical control strategy is proposed, and
EMS with HBC is proposed to reduce ON/OFF events of the the fuzzy EMS is adopted for the master level to achieve power
FC system and enhance the economy of the system [125]. allocation of HESS and superconducting magnetic energy
An EMS based on HBC is proposed, the control methods of storage, while the nonlinear sliding mode control is used
HESS, BESS, and RES unit can be designed by the status for the slave level to resist external disturbance [142]. A
of SOC for the battery and hydrogen storage level [126]. decentralized EMS is proposed, the power distribution mode
The start-up control strategy of modular HPU is proposed in of FC and BESS is determined based on three states of SOC.
the energy management strategy design of the wind-hydrogen However, this method lacks a power allocation proposal for
system. This method can not only increase the hydrogen the charging process [152]. The previous research primarily
production, but also reduce the start-stop times of HPU [107]. focuses on the system power balance without considering
the system economy. A hierarchical self-regulation control
F. Power Allocation Strategy of Electric-hydrogen Hybrid ESS method is proposed to achieve economic operations based on
In the electric-hydrogen DC MG, the battery or SC need the charge and discharge costs of HESS and BESS [143].
to be configured to compensate for the short-term unbalanced An EMS based on minimum utilization costs and an energy
power caused by the slow response characteristics of HESS. In storage state balance is proposed to reduce running cost and
the steady-state power regulation stage, the power allocation of optimize the energy storage state and promote the efficiency
electric-hydrogen hybrid ESS should be optimized depending of the hydrogen energy system [144]. A hierarchical EMS
on the SOC of the battery and hydrogen storage level, and with ECMS is developed to enhance the economy of DC
then reduce the duration of deep charge and discharge of the MG [146]. A hierarchical SMC based on the minimum cost
lithium battery, the number of operational mode switching and algorithm is developed to achieve economic operations [145].
the number of start-stop of HESS. To deal with this problem, An EMS based on multi-FLC is proposed, the FLC1 makes
three PMSs are developed to make sure the load requirements, decisions about the exchange with the main power grid, while
and the control performance of PMSs are evaluated [122]. Four the FLC2 is used for distributing the power of BESS and
energy management strategies are evaluated, and 10% hystere- HESS. The technical and economic targets are considered
sis bandwidth is selected to avert frequent switching [128]. simultaneously [111]. A fuzzy PMS with power prediction and
Similarly, the battery-priority and hydrogen-priority energy uncertainty is proposed to improve the HESS lifetime [147].
management strategies are compared. Hydrogen-priority EMS A methodological foundation is applied to develop a general
achieves a smaller loss of load probability, while battery- control-oriented model. The technical and economic param-
priority EMS obtains a better global storage efficiency [129]. eters are determined both in the short-term and long-term.
However, the above proposals adopt BESS and HESS to Apart from optimizing electrical performance, economical
share the unbalanced power separately, while ignoring the parameters are also optimized [112].
coordination of the two ESSs. To solve the problem of optimal The previous research primarily focuses on the power
power allocation of electric-hydrogen hybrid ESS, a two-level allocation of HESS and other ESSs in a microgrid, but did
EMS is developed, the nine operational states are considered, not consider the power distribution among multiple microgrids
PEI et al.: REVIEW OF OPERATIONAL CONTROL STRATEGY FOR DC MICROGRIDS WITH ELECTRIC-HYDROGEN HYBRID STORAGE SYSTEMS 339

with HESS. To solve this problem, a hierarchical method is is proposed for both HPU and FC [155]. An efficiency
proposed, which gives priority to the power distribution of the coordination and optimization control strategy of multi-stack
hybrid energy storage system, and then distributes the obtained FC systems was proposed to ensure the whole multi-stack
power between the BESS and HESS [148]. An inverse droop system was in the optimal operational state [160]. A restriction
control strategy based on the super-twisting algorithm (STA) is based on a safe operating area is developed to obtain the opti-
developed to realize the power distribution of inverters [149]. mal efficiency operational trajectory of the PEMFC system.
An adaptive droop control strategy based on the hydrogen And the disturbance estimation sliding mode control (DE-
storage level is developed to achieve equilibrium control of SMC) is adopted, which realizes optimal efficiency control
HST [150]. of the PEMFC system [161]. An online extremum seeking-
based optimized EMS is proposed to seek the maximum
G. Operational Control Strategy Considering Efficiency Opti- power and maximum efficiency points. This strategy can
mization of Hydrogen Energy Storage save hydrogen consumption and improve the stack efficiency
Efficiency optimization should be considered to improve the respectively [162]. A hierarchical performance improvement
system efficiency of electric-hydrogen DC MG. As shown in control method is developed, the upper level is to seek optimal
Fig. 6, when the temperature is constant, there is a unique trajectory for efficient and stable operations, while the lower
current corresponding to the optimal efficiency of the HPU. control level is to track optimal trajectory [163].
Therefore, the hydrogen production efficiency can be opti-
H. Performance Comparison and Prospects
mized by dynamically adjusting the current. To solve the
above problems, the efficiency of HPU is evaluated, and The current operational control proposal of the electric-
an efficiency adaptive control is proposed to obtain optimal hydrogen DC MG has been analyzed in detail. The comparison
efficiency control [153]. Similarly, the coordinated control results of different control proposals are shown in Table VI.
strategy is extended to AC MG, the operational mode is In the existing proposals, from the perspective of control
divided by frequency information, and an efficiency-frequency architecture, most control proposals adopt the two-layer con-
adaptive droop control method is proposed for HPU [154]. On trol architecture with communication technology, while there
the basis of reference [154], the power generation efficiency is less research on the operational control proposal without
of FC is evaluated, and a power-frequency adaptive droop communication. The energy management strategy and stability
control strategy with efficiency dynamic adjustment capability performance of the two-layer control architecture are greatly

TABLE VI
C OMPARISON WITH E XISTING M ETHODS
Existing methods Communication SOC SOCH Efficiency Power Dynamic Start-stop Operation
management management optimization characteristics strategy range
√ √ allocation
[91]–[95], [115] √ ×
√ × × × × ×
√
[96], [98] √ × × × × ×
[97] √ ×
√ × × × × × ×
√
[99] √ √ ×
√ × × × ×
[100], [103], [109], × × × × ×
[114] [116], [118], √ √ √ √
[101], [134] √ √ × ×
√ ×
√ ×
[102] √ √ ×
√ × √ × ×
[104]–[106], [120] √ √ × × ×
√ ×
√
[107] √ √ ×
√ × ×
√ × √
[108], [111], × × ×
[130]–[132] √ √ √
[110], [113] √ ×
√ √ ×
√ ×
√ × ×
√
[112] √ √ × ×
[117] √ ×
√ ×
√ × √ × × ×
[119], [128], [129] √ √ × √ × ×
√ ×
√
[121]–[124] √ √ ×
√ × √ × √ √
[125]–[127] √ × ×
√
[133] √ ×
√ ×
√ × ×
√ √ × ×
[135] √ √ × ×
√ ×
[136] √ ×
√ √ × ×
√ ×
√ ×
[137] √ √ × √ × ×
[138], [139] √ √ ×
√ × √ × × ×
√
[140], [141], [151] √ √ √ × √ × ×
[142], [143], [145], × × × ×
[146] [147], [150], √ √ √ √ √
[144], [148] √ √ √ × √ √ × ×
√
[149] √ √ × √ √ ×
[152] × √ ×
√ √ × ×
[153]–[155] × √ ×
√ × × ×
[156] × √ √ × × × × ×
√
[157] ×
√ √ √ ×
√ ×
√ ×
√ ×
[162] × ×
340 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 8, NO. 2, MARCH 2022

affected by communication, and the system reliability is poor. the current research primarily assumes that the efficiency curve
The decentralized control proposal has higher reliability, but its of the electrolyzer is fixed. However, the temperature of the
energy management and optimization ability are weak. From electrolyzer will change in real time under different operating
the perspective of control structure, the integration of hydrogen conditions, and the efficiency of the electrolyzer is very sensi-
energy does not change the control structure of the DC MG. tive to the temperature change, so the efficiency curve of the
Therefore, the control architecture of the electric-hydrogen DC electrolyzer is time-varying. To ensure the effectiveness of the
MG can still use the existing control architecture of DC MG. efficiency optimization control of HPU, the efficiency curve
In the future, it is a better choice to combine the existing two should be adjusted in real time according to the operating
control structures to form a multi-layer control architecture. conditions. In addition, the current research is limited to the
In terms of static response characteristics of hydrogen efficiency optimization of the electrolyzer. However, the actual
energy storage, the existing proposals consider more aspects, hydrogen production system consists of many parts, and the
such as operational range of hydrogen energy storage and efficiency estimation and efficiency optimization control of the
SOCH management of hydrogen storage tanks, but there are whole system still need to be further explored.
still some literature that have not fully considered the above In addition to the above discussion, in the current research,
operational constraints. At present, the selection principle the HPU or FC system is usually equivalent to a power
of static constraints is still based on experience, such as electronic device and a controlled source. However, in practice,
thresholds of SOC and hydrogen storage capacity. However, the hydrogen production unit and fuel cell also include many
the selection of upper and lower limits of SOC and hydrogen other levels of control, such as gas flow rate control, liquid
storage levels has a greater impact on system operational level control, temperature control, pressure control and so
reliability, cost, lifespan and benefits. Therefore, how to better on. Therefore, these control objects and control systems also
formulate constraints in the future needs further evaluation and need to be taken into account to simulate the real operational
analysis. scenario.
In terms of dynamic response characteristics of hydrogen
energy storage, the existing literature proposes to use lithium VII. F UTURE R ESEARCH T RENDS AND E XPLORATION
batteries or SC to solve the problem of the slow dynamic
response of HESS. However, the real-time change of the power A. Advantages of Modular Parallel HPS
of the HESS has an adverse impact on the performance and According to the current research updates, the previous
service life of the electrolyzer. Therefore, in order to avoid research was primarily limited to a single HPU. In the future,
the degradation of HESS performance, the correlation be- with the continuous increase of hydrogen production capacity,
tween power change and hydrogen energy storage performance the single module hydrogen production power supply is diffi-
should be further explored. cult to be applied in low-voltage and high current situations,
In the existing operational control proposals, the startup and so the modular parallel HPS is usually used. The advantages
shutdown characteristics of HPU have been widely addressed. of modular parallel HPS are not only reflected in the hydrogen
The startup and shutdown strategy based on HBC is usually production power supply, but also in the hydrogen production
used. In this proposal, the capacity of the BESS is assumed electrolyzer, as follows:
to be large enough. However, the BESS capacity is limited 1) Increase the operational range of HPU. According to
in view of economy and other factors. In addition, PMSs the purity characteristics of the alkaline electrolyzer, the HPU
based on battery-priority and hydrogen-priority are used, and cannot operate at low power for ensuring hydrogen energy
the power allocation proposal may be poor for reducing the safety, and its lower limit is about 40% of the rated power
number of frequent starts and stops. Therefore, the start-stop of the system. Therefore, the operational range of a single
strategy and advanced power allocation proposal should be HPU is narrow and cannot match the wide power fluctuation
further integrated. characteristics of RESs. If the modular HPS is adopted, the
In terms of power allocation of electric-hydrogen hybrid lower limit of operations can be reduced to 1/N (N is the
ESS, the most existing proposals adopt a serial sequential number of modules), thus increasing the overall operational
allocation mode, such as hybrid battery-priority and hybrid range of the HPS.
hydrogen-priority. Although there have been relevant studies 2) Increase the startup speed of HPU. After adopting the
on the parallel power allocation mode, these power distribution modular hydrogen production structure, due to the increase of
proposals are primarily based on SOC of battery, hydrogen the operating range, the lower limit of the starting power of
storage level and economical parameters. In the future, more the whole HPS is reduced. Therefore, the HPS can be started
factors should be considered, such as bulk properties of fuel with less power, thus shortening the starting time.
cell and battery. 3) Improve the operational efficiency of HPS. If a cen-
In terms of efficiency optimization of HPU and FC, for tralized HPS is adopted, according to the energy efficiency
the efficiency optimization control of the fuel cell system, characteristic curve of the electrolyzer, when the input power is
there have been some valuable research results. The real- certain, the operational efficiency of a single HPS is fixed, and
time tracking control of optimal efficiency and the power the efficiency cannot be optimized and adjusted. If the modular
allocation strategy of the modular FC system based on optimal parallel HPS is adopted, multiple HPUs can optimize and
efficiency have been studied and addressed. Although the adjust the operational efficiency point of the system through
efficiency optimization control of HPU have also been studied, the power allocation method.
PEI et al.: REVIEW OF OPERATIONAL CONTROL STRATEGY FOR DC MICROGRIDS WITH ELECTRIC-HYDROGEN HYBRID STORAGE SYSTEMS 341

B. Power Allocation Strategy of Modular Parallel HPS and dynamic characteristics of HESS, power allocation of the
For the modular parallel HPS, how to optimize the power electric-hydrogen hybrid ESS and efficiency optimization of
allocation of multiple HPUs is the key technical problem. the hydrogen energy storage. From the summary and analysis
In the process of power allocation, on the one hand, it is of the current research, compared with the traditional DC MG,
necessary to consider the capacity ratio of multiple HPUs. the integration of HESS will not affect the system control ar-
On the other hand, it is necessary to consider the energy chitecture, but the static and dynamic characteristics of HESS
consumption and energy efficiency curves of different HPUs. will have a great impact on the operational mode, energy
The power of modular HPSs is allocated depending on the management strategy and the coordinated control method of
overall efficiency index of the hydrogen production system. In DC MG. The static dynamic characteristics of HESS and the
addition, SOCH management of HST shall also be considered complementary characteristics of the electric-hydrogen hybrid
in power allocation. Once the pressure of HST reaches the ESS should be fully considered in the operational control and
upper limit, the HPU needs to be stopped to ensure hydrogen energy management proposal, which is conducive to improv-
safety. Shutdown and restart will not only bring economic ing the adaptability of HPU to the wide power fluctuation of
losses, but also bring complexity to the operational control RESs, and the service life and operational efficiency of the
of DC MG. Therefore, the pressure of multiple hydrogen system. In addition, the control strategy of electric hydrogen
storage tanks needs to be intelligently managed to avoid DC MG is analyzed and discussed. Finally, the development
frequent start-up and shut-down actions. Based on the above and advantages of the modular hydrogen production system
analysis, in the future, the power allocation strategy of modular are analyzed, and the technical problems and future research
parallel HPUs can be studied from the aspects of system directions of DC MG with modular parallel HPS are explored.
efficiency optimization, SOCH intelligent management and
R EFERENCES
balance control.
[1] X. X. Zhou, S. Y. Chen, Z. X. Lu, Y. H. Huang, S. C. Ma, and Q. Zhao,
C. Operational Control Strategy Considering the Start-stop “Technology features of the new generation power system in China,”
Proceedings of the CSEE, vol. 38, no. 7, pp. 1893–1904, Apr. 2018.
Characteristics of Modular Parallel HPS [2] W. X. Sheng, M. Wu, Y. Ji, L. F. Kou, J. Pan, H. F. Shi, G. Niu, and
Z. G. Wang, “Key techniques and engineering practice of distributed
According to the research progress at home and abroad, renewable generation clusters integration,” Proceedings of the CSEE,
the current research is limited to the start-stop strategy of vol. 39, no. 8, pp. 2175–2186, Apr. 2019.
single hydrogen production equipment, while the start-stop [3] Z. X. Lu, H. Huang, B. G. Shan, Y. H. Wang, S. H. Du, and J. H.
Li, “Morphological evolution model and power forecasting prospect
strategy of modular parallel HPS is rarely involved. The of future electric power systems with high proportion of renewable
response time of the cold start of HPS is usually minute, energy,” Automation of Electric Power Systems, vol. 41, no. 9, pp. 12–
which is slow. Meanwhile, the shutdown of HPUs will also 18, May 2017.
[4] S. S. Akadiri, A. A. Alola, A. C. Akadiri, and U. V. Alola, “Renew-
affect the hydrogen production and cause hydrogen losses. able energy consumption in EU-28 countries: Policy toward pollution
For the modular parallel HPS, because the system contains mitigation and economic sustainability,” Energy Policy, vol. 132, pp.
multiple HPUs, when the power command is small, only one 803–810, Sep. 2019.
[5] B. Ameyaw, Y. Li, Y. K. Ma, J. K. Agyeman, J. Appiah-Kubi, and A.
HPU is started in the system, while other HPUs may be in Annan, “Renewable electricity generation proposed pathways for the
a shutdown or hot standby state. During the real-time change US and China,” Renewable Energy, vol. 170, pp. 212–223, Jun. 2021.
of the power command, multiple HPUs will switch frequently [6] X. X Zhou, Q. Zhao, Y. Q. Zhang, and L. Sun, “Integrated Energy
Production Unit: An Innovative Concept and Design for Energy Tran-
in hot standby, shutdown and operational states. In order to sition Toward Low-carbon Development,” CSEE Journal of Power and
improve the response characteristics of HPUs, the cold and hot Energy Systems, vol. 7, no. 6, pp. 1133–1139, Nov. 2021.
state transition time of different HPUs should be considered. [7] Z. Wang, X. G. Zhao, and Y. Zhou, “Biased technological progress and
total factor productivity growth: From the perspective of China’s re-
In addition, the power loss caused by the hot standby state of newable energy industry,” Renewable and Sustainable Energy Reviews,
the HPU needs to be considered to evaluate the operational vol. 146, pp. 111136, Aug. 2021.
efficiency of the whole system. Based on the above analysis, [8] X. H. Liu, T. Zhao, C. T. Chang, and C. J. Fu, “China’s renewable
energy strategy and industrial adjustment policy,” Renewable Energy,
in the future, it is necessary to comprehensively consider the vol. 170, pp. 1382–1395, Jun. 2021.
control objectives, such as hydrogen production, startup and [9] H. Y. Zheng, M. L. Song, and Z. Y. Shen, “The evolution of renewable
shutdown times, energy consumption and dynamic response energy and its impact on carbon reduction in China,” Energy, vol. 237,
pp. 121639, Dec. 2021.
performance, to study the start-stop strategy of modular par- [10] International Renewable Energy Agency (IRENA). (2021, Mar.). Re-
allel HPS. newable capacity statistics 2021. [Online]. Available: https://www.
irena.org/-/media/Files/IRENA/Agency/Publication/2021/Apr/IRENA
RE Capacity Statistics 2021.pdf
VIII. C ONCLUSION [11] G. P. Chen, Z. F. Liang, and Y. Dong, “Analysis and reflection
on the marketization construction of electric power With Chinese
This paper introduces the typical structure of an electric- characteristics based on energy transformation,” Proceedings of the
hydrogen MG, and analyzes the modeling methods of different CSEE, vol. 40, no. 2, pp. 369–378, Jan. 2020.
types of electrolyzers. The operational control architecture of [12] C. Q. Kang and L. Z. Yao, “Key scientific issues and theoretical re-
search framework for power systems with high proportion of renewable
the electric-hydrogen DC MG is analyzed. Combined with energy,” Automation of Electric Power Systems, vol. 41, no. 9, pp. 1–11,
the working characteristics of an alkaline electrolyzer, the May 2017.
influence of hydrogen energy storage access on the operational [13] H. Z. Cheng, J. Li, Y. W. Wu, H. Y. Chen, N. Zhang, and L.
Liu, “Challenges and prospects for AC/DC transmission expansion
mode of DC MG is also analyzed. The operational control planning considering high proportion of renewable energy,” Automation
strategies are compared and analyzed from four aspects: static of Electric Power Systems, vol. 41, no. 9, pp. 19–27, May 2017.
342 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 8, NO. 2, MARCH 2022

[14] Z. X. Lu, H. B. Li, and Y. Qiao, “Flexibility evaluation and sup- [33] S. A. Grigoriev, V. N. Fateev, D. G. Bessarabov, and P. Millet, “Current
ply/demand balance principle of power system with high-penetration status, research trends, and challenges in water electrolysis science and
renewable electricity,” Proceedings of the CSEE, vol. 37, no. 1, pp. technology,” International Journal of Hydrogen Energy, vol. 45, no. 49,
9–19, Jan. 2017. pp. 26036–26058, Oct. 2020.
[15] S. Kotra and M. K. Mishra, “A supervisory power management system [34] A. Buttler and H. Spliethoff, “Current status of water electrolysis for
for a hybrid microgrid with HESS,” IEEE Transactions on Industrial energy storage, grid balancing and sector coupling via power-to-gas
Electronics, vol. 64, no. 5, pp. 3640–3649, May 2017. and power-to-liquids: A review,” Renewable and Sustainable Energy
[16] U. Manandhar, A. Ukil, H. B. Gooi, N. R. Tummuru, S. K. Kollimalla, Reviews, vol. 82, pp. 2440–2454, Feb. 2018.
B. F. Wang, and K. Chaudhari, “Energy management and control for [35] R. Takahashi, H. Kinoshita, T. Murata, J. Tamura, M. Sugimasa, A.
grid connected hybrid energy storage system under different operating Komura, M. Futami, M. Ichinose, and K. Ide, “Output power smoothing
modes,” IEEE Transactions on Smart Grid, vol. 10, no. 2, pp. 1626– and hydrogen production by using variable speed wind generators,”
1636, Mar. 2019. IEEE Transactions on Industrial Electronics, vol. 57, no. 2, pp. 485–
[17] T. S. Babu, K. R. Vasudevan, V. K. Ramachandaramurthy, S. B. Sani, 493, Feb. 2010.
S. Chemud, and R. M. Lajim, “A comprehensive review of hybrid [36] W. Hug, H. Bussmann, and A. Brinner, “Intermittent operation and
energy storage systems: Converter topologies, control strategies and operation modeling of an alkaline electrolyzer,” International Journal
future prospects,” IEEE Access, vol. 8, pp. 148702–148721, Aug. 2020. of Hydrogen Energy, vol. 18, no. 12, pp. 973–977, Dec. 1993.
[18] G. S. Pan, W. Gu, H. Y. Zhang, and Y. Qiu, “Electricity and hydrogen [37] Ø. Ulleberg, “Modeling of advanced alkaline electrolyzers: a system
energy system towards accomodation of high proportion of renewable simulation approach,” International Journal of Hydrogen Energy, vol.
energy,” Automation of Electric Power Systems, vol. 44, no. 23, pp. 28, no. 1, pp. 21–33, Jan. 2003.
1–10, Dec. 2020. [38] D. Jang, H. S. Cho, and S. Kang, “Numerical modeling and analysis
[19] F. Cao, T. T. Guo, K. Y. Chen, X. L. Jin, L. Zhang, J. H. Yang, and of the effect of pressure on the performance of an alkaline water
A. M. Yin, “Progress and development prospect of coupled wind and electrolysis system,” Applied Energy, vol. 287, pp. 116554, Apr. 2021.
hydrogen systems,” Proceedings of the CSEE, vol. 41, no. 6, pp. 2187– [39] D. Jang, W. Choi, H. S. Cho, W. C. Cho, C. H. Kim, and S. Kang,
2200, Mar. 2021. “Numerical modeling and analysis of the temperature effect on the
[20] G. W. Cai, L. G. Kong, Y. Xue, and B. Z. Sun, “Overview of performance of an alkaline water electrolysis system,” Journal of Power
research on wind power coupled with hydrogen production technology,” Sources, vol. 506, pp. 230106, Sep. 2021.
Automation of Electric Power Systems, vol. 38, no. 21, pp. 127–135, [40] M. Maruf-ul-Karim and M. T. Iqbal, “Dynamic modeling and simula-
Nov. 2014. tion of alkaline type electrolyzers,” in 2009 Canadian Conference on
[21] G. W. Cai and L. G. Kong, “Techno-economic analysis of wind Electrical and Computer Engineering, St. John’s, NL, Canada, 2009,
curtailment/hydrogen production/fuel cell vehicle system with high pp. 711–715.
wind penetration in China,” CSEE Journal of Power and Energy [41] A. Ursúa and P. Sanchis, “Static-dynamic modelling of the electrical
Systems, vol. 3, no. 1, pp. 44–52, Mar. 2017. behaviour of a commercial advanced alkaline water electrolyser,”
[22] A. Fathy, M. Al-Dhaifallah, and H. Rezk, “Recent coyote algorithm- International Journal of Hydrogen Energy, vol. 37, no. 24, pp. 18598–
based energy management strategy for enhancing fuel economy of 18614, Dec. 2012.
hybrid FC/Battery/SC system,” IEEE Access, vol. 7, pp. 179409– [42] M. Hammoudi, C. Henao, K. Agbossou, Y. Dubé, and M. L. Doumbia,
179419, Dec. 2019. “New multi-physics approach for modelling and design of alkaline
[23] X. Zhang, W. Pei, C. X. Mei, W. Deng, J. X. Tan and Q. Q. Zhang, electrolyzers,” International Journal of Hydrogen Energy, vol. 37, no.
“Transform from gasoline stations to electric-hydrogen hybrid refueling 19, pp. 13895–13913, Oct. 2012.
stations: An islanding DC microgrid with electric-hydrogen hybrid [43] J. Milewski, G. Guandalini, and S. Campanari, “Modeling an alkaline
energy storage system and its control strategy ,” International Journal electrolysis cell through reduced-order and loss estimate approaches,”
of Electrical Power and Energy Systems , vol. 136, pp. 107684, Mar. Journal of Power Sources, vol. 269, pp. 203–211, Dec. 2014.
2022. [44] M. David, H. Alvarez, C. Ocampo-Martinez, and R. Sánchez-Peña,
[24] H. X. Sun, Z. Li, A. B. Chen, Y. Zhang, and C. X. Mei, “Current “Dynamic modelling of alkaline self-pressurized electrolyzers: a
status and development trend of hydrogen production technology by phenomenological-based semiphysical approach,” International Jour-
wind power,” Transactions of China Electrotechnical Society, vol. 34, nal of Hydrogen Energy, vol. 45, no. 43, pp. 22394–22407, Sep. 2020.
no. 19, pp. 4071–4083, Oct. 2019. [45] Á. Hernández-Gómez, V. Ramirez, and D. Guilbert, “Investigation of
[25] Z. Li, R. Zhang, H. X. Sun, W. D. Zhang, and C. X. Mei, “Review PEM electrolyzer modeling: Electrical domain, efficiency, and specific
on key technologies of hydrogen generation, storage and transportation energy consumption,” International Journal of Hydrogen Energy, vol.
based on multi-energy complementary renewable energy,” Transactions 45, no. 29, pp. 14625–14639, May 2020.
of China Electrotechnical Society, vol. 36, no. 3, pp. 445–462, Feb. [46] D. S. Falcão and A. M. F. R. Pinto, “A review on PEM electrolyzer
2021. modelling: Guidelines for beginners,” Journal of Cleaner Production,
[26] A. Mohammadi and M. Mehrpooya, “A comprehensive review on vol. 261, pp. 121184, Jul. 2020.
coupling different types of electrolyzer to renewable energy sources,” [47] P. Olivier, C. Bourasseau, and P. B. Bouamama, “Low-temperature
Energy, vol. 158, pp. 632–655, Sep. 2018. electrolysis system modelling: A review,” Renewable and Sustainable
[27] N. Chrysochoidis-Antsos, M. R. Escudé, and A. J. M. van Wijk, “Tech- Energy Reviews, vol. 78, pp. 280–300, Oct. 2017.
nical potential of on-site wind powered hydrogen producing refuelling [48] A. H. A. Rahim, A. S. Tijani, S. K. Kamarudin, and S. Hanapi, “An
stations in the Netherlands,” International Journal of Hydrogen Energy, overview of polymer electrolyte membrane electrolyzer for hydrogen
vol. 45, no. 46, pp. 25096–25108, Sep. 2020. production: Modeling and mass transport,” Journal of Power Sources,
[28] X. Xu, W. H. Hu, D. Cao, Q. Huang, W. Liu, M. Z. Jacobson, vol. 309, pp. 56–65, Mar. 2016.
and Z. Chen, “Optimal operational strategy for an offgrid hybrid [49] M. Santarelli, P. Medina, and M. Calı̀, “Fitting regression model
hydrogen/electricity refueling station powered by solar photovoltaics,” and experimental validation for a high-pressure PEM electrolyzer,”
Journal of Power Sources, vol. 451, pp. 227810, Mar. 2020. International Journal of Hydrogen Energy, vol. 34, no. 6, pp. 2519–
[29] X. Wu, H. Y. Li, X. L. Wang, and W. C. Zhao, “Cooperative operation 2530, Mar. 2009.
for wind turbines and hydrogen fueling stations with on-site hydrogen [50] O. Atlam and M. Kolhe, “Equivalent electrical model for a proton
production,” IEEE Transactions on Sustainable Energy, vol. 11, no. 4, exchange membrane (PEM) electrolyser,” Energy Conversion and
pp. 2775–2789, Oct. 2020. Management, vol. 52, no. 8–9, pp. 2952–2957, Aug. 2011.
[30] Y. Gu, Q. Q. Chen, J. L. Xue, Z. Y. Tang, Y. H. Sun, and Q. [51] C. Y. Biaku, N. V. Dale, M. D. Mann, H. Salehfar, A. J. Peters, and
Wu, “Comparative techno-economic study of solar energy integrated T. Han, “A semiempirical study of the temperature dependence of
hydrogen supply pathways for hydrogen refueling stations in China,” the anode charge transfer coefficient of a 6 kW PEM electrolyzer,”
Energy Conversion and Management, vol. 223, pp. 113240, Nov. 2020. International Journal of Hydrogen Energy, vol. 33, no. 16, pp. 4247–
[31] D. Apostolou and G. Xydis, “A literature review on hydrogen refu- 4254, Aug. 2008.
elling stations and infrastructure. Current status and future prospects,” [52] N. V. Dale, M. D. Mann, and H. Salehfar, “Semiempirical model
Renewable and Sustainable Energy Reviews, vol. 113, pp. 109292, Oct. based on thermodynamic principles for determining 6 kW proton
2019. exchange membrane electrolyzer stack characteristics,” Journal of
[32] M. David, C. Ocampo-Martı́nez, and R. Sánchez-Peña, “Advances in Power Sources, vol. 185, no. 2, pp. 1348–1353, Dec. 2008.
alkaline water electrolyzers: A review,” Journal of Energy Storage, vol. [53] F. Marangio, M. Santarelli, and M. Calı̀, “Theoretical model and
23, pp. 392–403, Jun. 2019. experimental analysis of a high pressure PEM water electrolyser for
PEI et al.: REVIEW OF OPERATIONAL CONTROL STRATEGY FOR DC MICROGRIDS WITH ELECTRIC-HYDROGEN HYBRID STORAGE SYSTEMS 343

hydrogen production,” International Journal of Hydrogen Energy, vol. degradation prediction of fuel cells,” Journal of Power Sources, vol.
34, no. 3, pp. 1143–1158, Feb. 2009. 488, pp. 229435, Mar. 2021.
[54] B. Han, S. M. Steen III, J. Mo, and F. Y. Zhang, “Electrochemical [73] A. A. El-Fergany, H. M. Hasanien, and A. M. Agwa, “Semi-empirical
performance modeling of a proton exchange membrane electrolyzer PEM fuel cells model using whale optimization algorithm,” Energy
cell for hydrogen energy,” International Journal of Hydrogen Energy, Conversion and Management, vol. 201, pp. 112197, Dec. 2019.
vol. 40, pp. 7006–7016, Jun. 2015. [74] Y. Nalbant, C. O. Colpan, and Y. Devrim, “Development of a one-
[55] V. Ruuskanen, J. Koponen, K. Huoman, A. Kosonen, M. Niemelä, and dimensional and semi-empirical model for a high temperature proton
J. Ahola, “PEM water electrolyzer model for a power-hardware-in-loop exchange membrane fuel cell,” International Journal of Hydrogen
simulator,” International Journal of Hydrogen Energy, vol. 42, no. 16, Energy, vol. 43, no. 11, pp. 5939–5950, Mar. 2018.
pp. 10775–10784, Apr. 2017. [75] K. Ettihir, L. Boulon, M. Becherif, K. Agbossou, and H. S. Ramadan,
[56] M. Espinosa-López, C. Darras, P. Poggi, R. Glises, P. Baucour, A. “Online identification of semi-empirical model parameters for PEM-
Rakotondrainibe, S. Besse, and P. Serre-Combe, “Modelling and exper- FCs,” International Journal of Hydrogen Energy, vol. 39, no. 36, pp.
imental validation of a 46 kW PEM high pressure water electrolyzer,” 21165–21176, Dec. 2014.
Renewable Energy, vol. 119, pp. 160–173, Apr. 2018. [76] H. W. Wu, “A review of recent development: Transport and perfor-
[57] D. Guilbert and G. Vitale, “Dynamic emulation of a PEM electrolyzer mance modeling of PEM fuel cells,” Applied Energy, vol. 165, pp.
by time constant based exponential model,” Energies, vol. 12, no. 4, 81–106, Mar. 2016.
pp. 750, Feb. 2019. [77] A. Kulikovsky, “Analytical model for PEM fuel cell concentration
[58] Á. Hernández-Gómez, V. Ramirez, D. Guilbert, and B. Saldivar, impedance,” Journal of Electroanalytical Chemistry, vol. 899, pp.
“Development of an adaptive static-dynamic electrical model based 115672, Oct. 2021.
on input electrical energy for PEM water electrolysis,” International [78] D. M. Fadzillah, M. I. Rosli, M. Z. M. Talib, S. K. Kamarudin,
Journal of Hydrogen Energy, vol. 45, no. 38, pp. 18817–18830, Jul. and W. R. W. Daud, “Review on microstructure modelling of a gas
2020. diffusion layer for proton exchange membrane fuel cells,” Renewable
[59] Á. Hernández-Gómez, V. Ramirez, D. Guilbert, and B. Saldivar, and Sustainable Energy Reviews, vol. 77, pp. 1001–1009, Sep. 2017.
“Cell voltage static-dynamic modeling of a PEM electrolyzer based [79] F. J. Asensio, J. I. San Martı́n, I. Zamora, G. Saldaña, and O. Oñederra,
on adaptive parameters: Development and experimental validation,” “Analysis of electrochemical and thermal models and modeling tech-
Renewable Energy, vol. 163, pp. 1508–1522, Jan. 2021. niques for polymer electrolyte membrane fuel cells,” Renewable and
[60] T. Yigit and O. F. Selamet, “Mathematical modeling and dynamic Sustainable Energy Reviews, vol. 113, pp. 109283, Oct. 2019.
Simulink simulation of high-pressure PEM electrolyzer system,” In- [80] J. Zhao, X. G. Li, C. Shum, and J. McPhee, “A review of physics-based
ternational Journal of Hydrogen Energy, vol. 41, no. 32, pp. 13901– and data-driven models for real-time control of polymer electrolyte
13914, Aug. 2016. membrane fuel cells,” Energy and AI, vol. 6, pp. 100114, Dec. 2021.
[61] M. Ni, M. K. H. Leung, and D. Y. C. Leung, “Technological develop- [81] A. Guarino, R. Trinchero, F. Canavero, and G. Spagnuolo, “A fast
ment of hydrogen production by solid oxide electrolyzer cell (SOEC),” fuel cell parametric identification approach based on machine learning
International Journal of Hydrogen Energy, vol. 33, no. 9, pp. 2337– inverse models,” Energy, vol. 239, pp. 122140, Jan. 2022.
2354, May 2008. [82] K. Sun, I. Esnaola, O. Okorie, F. Charnley, M. Moreno, and A. Tiwari,
[62] M. Ni, M. K. H. Leung, and D. Y. C. Leung, “Electrochemical “Data-driven modeling and monitoring of fuel cell performance,”
modeling of hydrogen production by proton-conducting solid oxide International Journal of Hydrogen Energy, vol. 46, no. 66, pp. 33206–
steam electrolyzer,” International Journal of Hydrogen Energy, vol. 33217, Sep. 2021.
33, no. 15, pp. 4040–4047, Aug. 2008. [83] W. Zou, D. Froning, X. J. Lu, and W. Lehnert, “An online spatiotem-
[63] A. A. AlZahrani and I. Dincer, “Thermodynamic and electrochemi- poral temperature model for high temperature polymer electrolyte fuel
cal analyses of a solid oxide electrolyzer for hydrogen production,” cells,” Energy Conversion and Management, vol. 199, pp. 111974, Nov.
International Journal of Hydrogen Energy, vol. 42, no. 33, pp. 21404– 2019.
21413, Aug. 2017. [84] L. Vichard, N. Y. Steiner, N. Zerhouni, and D. Hissel, “Hybrid fuel
[64] F. R. Bianchi, A. Baldinelli, L. Barelli, G. Cinti, E. Audasso, and B. cell system degradation modeling methods: A comprehensive review,”
Bosio, “Multiscale modeling for reversible solid oxide cell operation,” Journal of Power Sources, vol. 506, pp. 230071, Sep. 2021.
Energies, vol. 13, no.19, pp. 5058, Sep. 2020. [85] F. R. Wei, Q. Sui, X. N. Lin, Z. T. Li, B. Zhao, and C. Xu, “A new
[65] A. Demin, E. Gorbova, and P. Tsiakaras, “High temperature elec- equity mode and scheduling strategy of hydrogen production equipment
trolyzer based on solid oxide co-ionic electrolyte: A theoretical model,” in the multi-subject scene of the grid,” Proceedings of the CSEE, vol.
International Journal of Hydrogen Energy, vol. 171, no. 1, pp. 205– 38, no. 11, pp. 3214–3225, Jun. 2018.
211, Sep. 2007. [86] T. F. Ma, W. Pei, H. Xiao, D. X. Li, X. Y. Lv, and K. Hou, “Cooperative
[66] M. Ni, “Computational fluid dynamics modeling of a solid oxide operation method for wind-solar-hydrogen multi-agent energy system
electrolyzer cell for hydrogen production,” International Journal of based on Nash bargaining theory,” Proceedings of the CSEE, vol. 41,
Hydrogen Energy, vol. 34, no. 18, pp. 7795–7806, Sep. 2009. no. 1, pp. 25–39, Jan. 2021.
[67] R. Ma, Z. J. Ren, R. Y. Xie, D. D. Zhao, and Y. G. Huangfu, [87] F. Garcia-Torres and C. Bordons, “Optimal economical schedule of
“A comprehensive review for proton exchange membrane fuel cell hydrogen-based microgrids with hybrid storage using model predictive
modeling based on model feature analysis,” Proceedings of the CSEE, control,” IEEE Transactions on Industrial Electronics, vol. 62, no. 8,
vol. 41, no. 22, pp. 7712–7729, Nov. 2021. pp. 5195–5207, Aug. 2015.
[68] S. Srinivasan, E. A. Ticianelli, C. R. Derouin, and A. Redondo, [88] N. A. El-Taweel, H. Khani, and H. E. Z. Farag, “Hydrogen storage
“Advances in solid polymer electrolyte fuel cell technology with low optimal scheduling for fuel supply and capacity-based demand response
platinum loading electrodes,” Journal of Power Sources, vol. 22, no. program under dynamic hydrogen pricing,” IEEE Transactions on
3–4, pp. 359–375, Mar./Apr. 1988. Smart Grid, vol. 10, no. 4, pp. 4531–4542, Jul. 2019.
[69] J. C. Amphlett, R. M. Baumert, R. F. Mann, B. A. Peppley, P. R. [89] G. W. Cai, L. G. Kong, A. X. Xu, and Z. X. Li, “Economical evaluation
Roberge, and T. J. Harris, “Performance modeling of the Ballard of wind/hydrogen/fuel cell grid-connected system power smoothing
Mark IV solid polymer electrolyte fuel cell II. Empirical model based on improved chemical reaction optimization algorithm,” Trans-
development,” Journal of The Electrochemical Society, vol. 142, no. actions of China Electrotechnical Society, vol. 32, no. 20, pp. 251–260,
1, pp. 9–15, Jan. 1995. Oct. 2017.
[70] J. Kim, S. M. Lee, S. Srinivasan, and C. E. Chamberlin, “Modeling [90] Y. C. Pu, Q. Li, X. L. Zou, R. R. Li, L. Y. Li, W. R. Chen, and
of proton exchange membrane fuel cell performance with an empirical H. Liu, “Optimal sizing for an integrated energy system considering
equation,” Journal of The Electrochemical Society, vol. 142, no. 8, pp. degradation and seasonal hydrogen storage,” Applied Energy, vol. 302,
2670–2674, Jan. 1995. pp. 117542, Nov. 2021.
[71] H. I. Kim, C. Y. Cho, J. H. Nam, D. Shin, and T. Y. Chung, “A [91] G. W. Cai, C. Chen, L. G. Kong, L. Peng, and H. Zhang, “Modeling
simple dynamic model for polymer electrolyte membrane fuel cell and control of grid-connected system of wind/PV/electrolyzer and SC,”
(PEMFC) power modules: Parameter estimation and model prediction,” Power System Technology, vol. 40, no. 10, pp. 2982–2990, Oct. 2016.
International Journal of Hydrogen Energy, vol. 35, no. 8, pp. 3656– [92] S. Q. Han, W. X. Li, C. Chen, and L. Z. Liang, “Modeling and control
3663, Apr. 2010. of controllable D-PMSG based on hybrid energy storage of wind power
[72] M. Y. Ou, R. F. Zhang, Z. F. Shao, B. Li, D. J. Yang, P. W. Ming, and hydrogen production and supercapacitor,” Guangdong Electric Power,
C. M. Zhang, “A novel approach based on semi-empirical model for vol. 32, no. 5, pp. 1–12, May 2019.
344 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 8, NO. 2, MARCH 2022

[93] S. S. Yang and P. P. Lou, “Research on power coordination control tional Journal of Hydrogen Energy, vol. 41, no. 2, pp. 857–865, Jan.
strategy of photovoltaic grid-connected system with hybrid energy 2016.
storage,” Modern Electric Power, vol. 36, no. 1, pp. 37–44, Feb. 2019. [114] S. Nasri, S. B. Slama, I. Yahyaoui, B. Zafar, and A. Cherif, “Au-
[94] L. G. Kong, J. M. Yu, and G. W. Cai, “Modeling, control and simulation tonomous hybrid system and coordinated intelligent management ap-
of a photovoltaic /hydrogen/ supercapacitor hybrid power generation proach in power system operation and control using hydrogen storage,”
system for grid-connected applications,” International Journal of Hy- International Journal of Hydrogen Energy, vol. 42, no. 15, pp. 9511–
drogen Energy, vol. 44, no. 46, pp. 25129–25144, Sep. 2019. 9523, Apr. 2017.
[95] J. S. Kang, H. Fang, and L. Y. Yun, “A control and power management [115] Y. Sahri, Y. Belkhier, S. Tamalouzt, N. Ullah, R. N. Shaw, M. S.
scheme for photovoltaic/fuel cell/hybrid energy storage DC microgrid,” Chowdhury, and K. Techato, “Energy management system for hybrid
in 2019 14th Conference on Industrial Electronics and Applications PV/wind/battery/fuel cell in microgrid-based hydrogen and economical
(ICIEA), Xi’an, China, 2019, pp. 1937–1941. hybrid battery/super capacitor energy storage,” Energies, vol. 14, no.
[96] Z. Li, H. Dong, S. D. Hou, L. Y. Cheng, and H. X. Sun, “Coordinated 18, pp. 5722, Sep. 2021.
control scheme of a hybrid renewable power system based on hydrogen [116] Z. U. Bayrak, G. Bayrak, M. T. Ozdemir, M. T. Gencoglu, and M.
energy storage,” Energy Reports, vol. 7, pp. 5597–5611, Nov. 2021. Cebeci, “A low-cost power management system design for residential
[97] A. Ganeshan, D. G. Holmes, L. Meegahapola, and B. P. McGrath, hydrogen & solar energy based power plants,” International Journal
“Enhanced control of a hydrogen energy storage system in a micro- of Hydrogen Energy, vol. 41, no. 29, pp. 12569–12581, Aug. 2016.
grid,” in 2017 Australasian Universities Power Engineering Conference [117] R. Bhosale and V. Agarwal, “Control of fuel cell and electrolyzer based
(AUPEC), Melbourne, VIC, Australia, 2017, pp. 1–6. hydrogen storage system with ultra-capacitor for voltage stability and
[98] J. E. V. Londono, A. Mazza, E. Pons, H. Lok, and E. Bompard, enhanced transient stability of a DC micro grid,” in 2018 International
“Modelling and control of a grid-connected RES-hydrogen hybrid Conference on Power, Instrumentation, Control and Computing (PICC),
microgrid,” Energies, vol. 14, no. 6, pp. 1540, Mar. 2021. Thrissur, India, 2018, pp. 1–6.
[99] G. W. Cai, C. Chen, L. G. Kong, and L. Peng, “Control of hybrid [118] D. N. Luta and A. K. Raji, “Energy management system for a hybrid
system of wind/hydrogen/fuel cell/supercapacitor,” Transactions of hydrogen fuel cell-supercapacitor in an islanded microgrid,” in 2019
China Electrotechnical Society, vol. 32, no. 17, pp. 84–94, Sep. 2017. Southern African Universities Power Engineering Conference/Robotics
[100] L. G. Kong, G. W. Cai, C. Chen, and L. Xing, “Modeling and grid- and Mechatronics/Pattern Recognition Association of South Africa
connected control of proactive permanent magnet direct-driven wind (SAUPEC/RobMech/PRASA), Bloemfontein, South Africa, 2019, pp.
turbine based on energy storage of hydrogen,” Transactions of China 611–615.
Electrotechnical Society, vol. 32, no. 18, pp. 276–285, Sep. 2017. [119] F. Zhang, K. Thanapalan, A. Procter, S. Carr, J. Maddy, and G. Premier,
[101] L. G. Kong, G. W. Cai, L. F. Li, and R. F. Ji, “Online energy control “Power management control for off-grid solar hydrogen production and
strategy and experimental platform of integrated energy system of wind, utilisation system,” International Journal of Hydrogen Energy, vol. 38,
photovoltaic and hydrogen,” Transactions of China Electrotechnical no. 11, pp. 4334–4341, Apr. 2013.
Society, vol. 33, no. 14, pp. 3371–3384, Jul. 2018. [120] F. K/bidi, C. Damour, D. Grondin, M. Hilairet, and M. Benne, “Power
[102] S. Mudaliyar and S. Mishra, “Coordinated voltage control of a grid Management of a Hybrid Micro-Grid with Photovoltaic Production and
connected ring DC microgrid with energy hub,” IEEE Transactions on Hydrogen Storage,” Energies, vol. 14, no. 6, pp. 1628, Mar. 2021.
Smart Grid, vol. 10, no. 2, pp. 1939–1948, Mar. 2019.
[121] Ø. Ulleberg, “The importance of control strategies in PV-hydrogen
[103] D. F. R. Melo and L. R. Chang-Chien, “Synergistic control between
systems,” Solar Energy, vol. 76, no. 1–3, pp. 323–329, Jan./Mar. 2004.
hydrogen storage system and offshore wind farm for grid operation,”
[122] D. Ipsakis, S. Voutetakis, P. Seferlis, F. Stergiopoulos, and C. Elma-
IEEE Transactions on Sustainable Energy, vol. 5, no. 1, pp. 18–27,
sides, “Power management strategies for a stand-alone power system
Jan. 2014.
using renewable energy sources and hydrogen storage,” International
[104] G. W. Cai, L. Peng, L. G. Kong, C. Chen, and L. Xing, “Power coor-
Journal of Hydrogen Energy, vol. 34, no. 16, pp. 7081–7095, Aug.
dinated control of photovoltaic and hydrogen hybrid power generation
2009.
system,” Automation of Electric Power Systems, vol. 41, no. 1, pp.
109–116, Jan. 2017. [123] D. Ipsakis, S. Voutetakis, P. Seferlis, F. Stergiopoulos, S. Papadopoulou,
[105] G. W. Cai, C. Chen, L. G. Kong, L. Peng, and Z. X. Li, “Control and C. Elmasides, “The effect of the hysteresis band on power
management strategies in a stand-alone power system,” Energy, vol.
strategy of hybrid grid-connected system of wind and hydrogen,” Acta
Energiae Solaris Sinica, vol. 39, no. 10, pp. 2970–2980, Oct. 2018. 33, no. 10, pp. 1537–1550, Oct. 2008.
[106] H. Deng, J. Chen, D. D. Jiao, and Y. Q. Li, “Control strategy for energy [124] L. Valverde, F. Rosa, and C. Bordons, “Design, planning and manage-
management of hybrid grid-connected system of wind and hydrogen,” ment of a hydrogen-based microgrid,” IEEE Transactions on Industrial
High Voltage Engineering, vol. 46, no. 1, pp. 99–106, Jan. 2020. Informatics, vol. 9, no. 3, pp. 1398–1404, Aug. 2013.
[107] R. M. Fang and Y. Liang, “Control strategy of electrolyzer in a [125] K. Kumar, M. Alam, S. Verma S, and V. Dutta, “Effect of hysteresis
wind-hydrogen system considering the constraints of switching times,” band control strategy on energy efficiency and durability of solar-
International Journal of Hydrogen Energy, vol. 44, no. 46, pp. 25104– hydrogen storage based microgrid in partial cloudy condition,” Journal
25111, Sep. 2019. of Energy Storage, vol. 32, pp. 101926, Dec. 2020.
[108] P. Garcı́a, C. A. Garcı́a, L. M. Fernández, F. Llorens, and F. Jurado, [126] A. Kafetzis, C. Ziogou, K. D. Panopoulos, S. Papadopoulou, P. Seferlis,
“ANFIS-based control of a grid-connected hybrid system integrating and S. Voutetakis, “Energy management strategies based on hybrid
renewable energies, hydrogen and batteries,” IEEE Transactions on automata for islanded microgrids with renewable sources, batteries and
Industrial Informatics, vol. 10, no. 2, pp. 1107–1117, May 2014. hydrogen,” Renewable and Sustainable Energy Reviews, vol. 134, pp.
[109] S. Q. Wang, J. Dang, X. B. Han, L. G. Lu, H. W. Wang, M. G. 110118, Dec. 2020.
Ouyang, and K. Sun, “A semi-decentralized control strategy of a PV- [127] M. Alam, K. Kumar, S. Verma, and V. Dutta, “Renewable sources based
based microgrid with battery energy storage systems for electric vehicle DC microgrid using hydrogen energy storage: Modelling and experi-
charging and hydrogen production,” in 2021 IEEE 4th International mental analysis,” Sustainable Energy Technologies and Assessments,
Electrical and Energy Conference (CIEEC), Wuhan, China, 2021, pp. vol. 42, pp. 100840, Dec. 2020.
1–6. [128] A. M. Ferrario, A. Bartolini, F. S. Manzano, F. J. Vivas, G. Comodi,
[110] J. Kim, E. Muljadi, and R. M. Nelms, “Modelling and control coor- S. J. McPhail, and J. M. Andujar, “A model-based parametric and
dination scheme of a wind-to-hydrogen set for future renewable-based optimal sizing of a battery/hydrogen storage of a real hybrid microgrid
power systems,” IET Renewable Power Generation, vol. 14, no. 17, supplying a residential load: Towards island operation,” Advances in
pp. 3317–3326, Dec. 2020. Applied Energy, vol. 3, pp. 100048, Aug. 2021.
[111] F. J. Vivas, F. Segura, J. M. Andújar, A. Palacio, J. L. Saenz, [129] A. M. Ferrario, F. J. Vivas, F. S. Manzano, J. M. Andújar, E. Bocci,
F. Isorna, and E. López, “Multi-objective fuzzy logic-based energy and L. Martirano, “Hydrogen vs. battery in the long-term operation.
management system for microgrids with battery and hydrogen energy A comparative between energy management strategies for hybrid
storage system,” Electronics, vol. 9, no. 7, pp. 1074, Jun. 2020. renewable microgrids,” Electronics, vol. 9, no. 4, pp. 698, Apr. 2020.
[112] F. J. Vivas, F. Segura, J. M. Andújar, and J. J. Caparrós, “A suitable [130] Y. Han, W. R. Chen, Q. Li, H. Q. Yang, F. Zare, and Y. K. Zheng,
state-space model for renewable source-based microgrids with hydro- “Two-level energy management strategy for PV-Fuel cell-battery-based
gen as backup for the design of energy management systems,” Energy DC microgrid,” International Journal of Hydrogen Energy, vol. 44, no.
Conversion and Management, vol. 219, pp. 113053, Sep. 2020. 35, pp. 19395–19404, Jul. 2019.
[113] S. Nasri, B. S. Sami, and A. Cherif, “Power management strategy for [131] M. N. Boukoberine, M. F. Zia, M. Benbouzid, Z. B. Zhou, and
hybrid autonomous power system using hydrogen storage,” Interna- T. Donateo, “Hybrid fuel cell powered drones energy management
PEI et al.: REVIEW OF OPERATIONAL CONTROL STRATEGY FOR DC MICROGRIDS WITH ELECTRIC-HYDROGEN HYBRID STORAGE SYSTEMS 345

strategy improvement and hydrogen saving using real flight test data,” [150] M. Alam, K. Kumar, and V. Dutta, “Droop based control strategy for
Energy Conversion and Management, vol. 236, pp. 113987, May 2021. balancing the level of hydrogen storage in direct current microgrid
[132] Q. Li, Y. C. Pu, Y. Han, and W. R. Chen, “Hierarchical Energy application,” Journal of Energy Storage, vol. 33, pp. 102106, Jan. 2021.
management for electric-hydrogen island direct current micro-grid,” [151] K. Kumar, M. Alam, and V. Dutta, “Energy management strategy
Journal of Southwest Jiaotong University, vol. 55, no. 5, pp. 912–919, for integration of fuel cell-electrolyzer technologies in microgrid,”
Oct. 2020. International Journal of Hydrogen Energy, vol. 46, no. 68, pp. 33738–
[133] N. Gyawali and Y. Ohsawa, “Integrating fuel 33755, Oct. 2021.
cell/electrolyzer/ultracapacitor system into a stand-alone microhydro [152] Y. Han, H. Q. Yang, Q. Li, W. R. Chen, F. Zare, and J. M. Guerrero,
plant,” IEEE Transactions on Energy Conversion, vol. 25, no. 4, pp. “Mode-triggered droop method for the decentralized energy man-
1092–1101, Dec. 2010. agement of an islanded hybrid PV/hydrogen/battery DC Microgrid,”
[134] A. M. O. Haruni, M. Negnevitsky, M. E. Haque, and A. Gargoom, “A Energy, vol. 199, pp. 117441, May 2020.
novel operation and control strategy for a standalone hybrid renewable [153] Y. Zhang and W. Wei, “Decentralized coordination control of PV
power system,” IEEE Transactions on Sustainable Energy, vol. 4, no. generators, storage battery, hydrogen production unit and fuel cell in
2, pp. 402–413, Apr. 2013. islanded DC Microgrid,” International Journal of Hydrogen Energy,
[135] B. R. Naidu, G. Panda, and P. Siano, “A self-reliant DC microgrid: vol. 45, no. 15, pp. 8243–8256, Mar. 2020.
sizing, control, adaptive dynamic power management, and experimental [154] Y. Zhang and W. Wei, “Decentralised coordination control strategy
analysis,” IEEE Transactions on Industrial Informatics, vol. 14, no. 8, of the PV generator, storage battery and hydrogen production unit in
pp. 3300–3313, Aug. 2018. islanded AC microgrid,” IET Renewable Power Generation, vol. 14,
[136] M. Trifkovic, M. Sheikhzadeh, K. Nigim, and P. Daoutidis, “Modeling no. 6, pp. 1053–1062, Apr. 2020.
and control of a renewable hybrid energy system with hydrogen [155] Y. Zhang and W. Wei, “Model construction and energy management
storage,” IEEE Transactions on Control Systems Technology, vol. 22, system of lithium battery, PV generator, hydrogen production unit and
no. 1, pp. 169–179, Jan. 2014. fuel cell in islanded AC microgrid,” International Journal of Hydrogen
[137] M. Jafari, Z. Malekjamshidi, D. D. C. Lu, and J. G. Zhu, “Development Energy, vol. 45, no. 33, pp. 16381–16397, Jun. 2020.
of a fuzzy-logic-based energy management system for a multiport [156] H. Matayoshi, M. Kinjo, S. S. Rangarajan, G. G. Ramanathan, A. M.
multioperation mode residential smart microgrid,” IEEE Transactions Hemeida, and T. Senjyu, “Islanding operation scheme for DC microgrid
on Power Electronics, vol. 34, no. 4, pp. 3283–3301, Apr. 2019. utilizing pseudo Droop control of photovoltaic system,” Energy for
[138] D. Abd-El Baset, H. Rezk, and M. Hamada, “Fuzzy logic control based Sustainable Development, vol. 55, pp. 95–104, Apr. 2020.
energy management strategy for renewable energy system,” in 2020 [157] W. Q. Yang, X. W. Xing, S. Y. Qin, and Y. J. Fan, “Study of
International Youth Conference on Radio Electronics, Electrical and coordination control strategy of large-scale wind/PV hybrid hydrogen
Power Engineering (REEPE), Moscow, Russia, 2020, pp. 1–5. energy system,” Power Electronics, vol. 54, no. 12, pp. 24–27, 55, Dec.
2020.
[139] F. Zhang, K. Thanapalan, A. Procter, J. Maddy, and A. Guwy, “Fuzzy
logic control for solar powered hydrogen production, storage and [158] P. Thounthong, S. Raël, and B. Davat, “Analysis of supercapacitor as
utilisation system,” in Proceedings of 2012 UKACC International second source based on fuel cell power generation,” IEEE Transactions
on Energy Conversion, vol. 24, no. 1, pp. 247–255, Mar. 2009.
Conference on Control, Cardiff, UK, 2012, pp. 912–917.
[159] H. C. Chen, X. Zhao, T. Zhang, and P. C. Pei, “The reactant starvation
[140] I. Abadlia, T. Bahi, and H. Bouzeria, “Energy management strategy
of the proton exchange membrane fuel cells for vehicular applications:
based on fuzzy logic for compound RES/ESS used in stand-alone
A review,” Energy Conversion and Management, vol. 182, pp. 282–
application,” International Journal of Hydrogen Energy, vol. 41, no.
298, Feb. 2019.
38, pp. 16705–16717, Oct. 2016.
[160] Y. N. Zhu, Q. Li, W. Q. Huang, W. L. Shang, W. R. Chen, and
[141] X. Zhang, W. Pei, C. X. Mei, J. X. Tan, Q. Q. Zhang, and W. Deng.
Y. Ding, “Efficiency coordination and optimization control method
(2021, Jun.). Fuzzy power allocation strategy and coordinated control
of multi-stack fuel cell systems based on power adaptive allocation,”
method of islanding DC microgrid with electricity/hydrogen hybrid Proceedings of the CSEE, vol. 39, no. 6, pp. 1714–1722, Mar. 2019.
energy storage systems. High Voltage Engineering. [Online]. Available:
[161] W. Yang, Q. Li, Q. Liu, S. Li, L. Z. Yin, and W. R. Chen. (2021, Oct.).
https://kns.cnki.net/kcms/detail/42.1239.TM.20210603.1133.007.html
Efficiency optimization control method of PEMFC power generation
[142] S. S. Zehra, A. Ur Rahman, and I. Ahmad, “Fuzzy-barrier sliding system based on safe operating area constraints. Proceedings of the
mode control of electric-hydrogen hybrid energy storage system in DC CSEE. [Online]. https://kns.cnki.net/kcms/detail/11.2107.TM.202110
microgrid: Modelling, management and experimental investigation,” 28.1557.022.html
Energy, vol. 239, pp. 122260, Jan. 2022. [162] Q. Li, T. H. Wang, S. H. Li, W. R. Chen, H. Liu, E. Breaz, and F.
[143] H. Q. Yang, Q. Li, S. D. Zhao, W. R. Chen, and H. Liu, “A Gao, “Online extremum seeking-based optimized energy management
hierarchical self-regulation control for economic operation of AC/DC strategy for hybrid electric tram considering fuel cell degradation,”
hybrid microgrid with hydrogen energy storage system,” IEEE Access, Applied Energy, vol. 285, pp. 116505, Mar. 2021.
vol. 7, pp. 89330–89341, Jun. 2019. [163] Q. Li, L. Z. Yin, H. Q. Yang, T. H. Wang, Y. B. Qiu, and W. R.
[144] Y. C. Pu, Q. Li, W. R. Chen, W. Q. Huang, B. B. Hu, Y. Han, and X. Chen, “Multiobjective optimization and data-driven constraint adaptive
Wang, “Energy management for islanded DC microgrid with hybrid predictive control for efficient and stable operation of PEMFC system,”
electric-hydrogen energy storage system based on minimum utilization IEEE Transactions on Industrial Electronics, vol. 68, no. 12, pp.
cost and energy storage state balance,” Power System Technology, vol. 12418–12429, Dec. 2021.
43, no. 3, pp. 918–927, Mar. 2019.
[145] Y. C. Pu, Q. Li, W. R. Chen, and H. Liu, “Hierarchical energy
management control for islanding DC microgrid with electric-hydrogen
hybrid storage system,” International Journal of Hydrogen Energy, vol.
44, no. 11, pp. 5153–5161, Feb. 2019.
[146] Y. Han, G. R. Zhang, Q. Li, Z. Y. You, W. R. Chen, and H. Liu,
“Hierarchical energy management for PV/hydrogen/battery island DC Wei Pei (M’14) received the B.S. and M.S. degrees
microgrid,” International Journal of Hydrogen Energy, vol. 44, no. 11, in Electrical Engineering from Tianjin University,
pp. 5507–5516, Feb. 2019. Tianjin, China, in 2002 and 2005, respectively, and
[147] M. H. Cano, S. Kelouwani, K. Agbossou, and Y. Dubé, “Power man- Ph.D. degree from the Institute of Electrical En-
agement system for off-grid hydrogen production based on uncertainty,” gineering, Chinese Academy of Sciences, Beijing,
International Journal of Hydrogen Energy, vol. 40, no. 23, pp. 7260– China, in 2008, where he is currently working as
7272, Jun. 2015. a Professor with the Institute of Electrical Engi-
[148] B. Li, H. Z. Miao, and J. C. Li, “Multiple hydrogen-based hybrid neering, Chinese Academy of Sciences. And he is
storage systems operation for microgrids: A combined TOPSIS and the Director of the Power System Laboratory. His
model predictive control methodology,” Applied Energy, vol. 283, pp. research interests include integrated energy system,
116303, Feb. 2021. and AC/DC micro-grid. He is also an Editor of IET
[149] Q. Li, R. R. Li, Y. C. Pu, S. Li, C. Sun, and W. R. Chen, “Coordinated Smart Grid, IET Energy Systems Integration, and the CSEE Journal of Power
control of electric-hydrogen hybrid energy storage for multi-microgrid and Energy Systems.
with fuel cell/ electrolyzer/ PV/ battery,” Journal of Energy Storage,
vol. 42, pp. 103110, Oct. 2021.
346 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 8, NO. 2, MARCH 2022

Xue Zhang received the B.S. and M.S. degrees in Chenghong Tang is currently the Deputy Chief
Electrical Engineering from Yanshan University, Qin Engineer of the Technology Research Center, NARI
huangdao, China, in 2010 and 2013, respectively, Research Institute. She has been primarily involved
and a Ph.D. degree from the Institute of Electrical in Technical Research and Product Research and
Engineering, Chinese Academy of Sciences, Beijing, Development in the field of distribution networks,
China, in 2021. He is an assistant researcher with micro-grids, and integrated energy systems. She has
the Power System Technology Laboratory, Insti- received 11 provincial, ministerial and state grid
tute of Electrical Engineering, Chinese Academy of company-level awards, and six prefecture-level sci-
Sciences. His research interests include hydrogen ence and technology progress awards; which have
energy technology, DC microgrid, and energy man- been applied in many areas of the country, and
agement and operation control. achieved good economic and social benefits.

Wei Deng (M’18) received the B.S. degree in Com- Liangzhong Yao received the M.S and Ph.D degrees
puter Science and Technology from North China in Electrical Engineering from Tsinghua University,
Electric Power University, Beijing, China, in 2004, Beijing, China, in 1989 and 1993 respectively. He
and a Ph.D. degree in Electrical Engineering from worked as a Senior Power System Analyst at ABB
the Chinese Academy of Sciences, Beijing, in 2010. in the UK from 1999 to 2004, and the Depart-
He is currently an Associate Professor with the In- ment Manager and Senior Expert at Alstom Grid
stitute of Electrical Engineering, Chinese Academy Research and Technology Centre in the UK from
of Sciences, and with the University of Chinese 2004 until 2011. From 2011 to 2018, he served as
Academy of Sciences. His research interests include Vice President, respectively, at State Grid Electric
distributed generation, microgrids, active distribution Power Research and China Electric Power Research
networks, AC/DC hybrid grids, and smart grids. Institute. In 2019, he joined Wuhan University and
is now a Professor at the School of Electrical Engineering and Automation,
and also the director of Smart Grid Institute at Wuhan University. His
current research interests include the integration technologies of large-scale
renewable energy, control and operation of DC grids, and DC transmission
key technologies. He has published over 250 journal papers, has developed
over 40 patents, and authored 5 books and book chapters.