IEEJ TRANSACTIONS ON ELECTRICAL AND ELECTRONIC ENGINEERING

IEEJ Trans 2022; 17: 1330–1338

Published online in Wiley Online Library (wileyonlinelibrary.com). DOI:10.1002/tee.23623

Paper

Three-Layer Management Strategy of V2G, Renewable Energy and Energy

Storage Based on Fuzzy Control
Qihui Liu* a , Non-member
Ruohan Tian* , Non-member
Chenwei Hong** , Non-member
Simin Pang* , Non-member
This paper proposes a three-layer coordinated optimization strategy based on fuzzy control of renewable energy sources (RESs),
vehicle-to-grid (V2G) and energy storage. Firstly, the overall structure of this strategy is introduced, which is composed of V2G
management system, renewable energy dispatching system and energy storage (ES) real-time balancing system. Secondly, the
control strategy is proposed for different systems, among them, the V2G management system considers shortage of power grid,
forecasted maximum output of RESs, and state of charge (SOC) of electric vehicles (EVs) to calculate the charging plan for
each EV in next time interval by fuzzy control algorithm; The renewable energy management system gets the expected shortage
state of the power grid in the next period according to the charging plan and the current state of the power grid, then combined
with the battery state of the energy storage system, the renewable energy guiding power generation in the next period can be
calculated by the fuzzy control algorithm; The energy storage real-time balancing system can charge or discharge according
to the real-time frequency and shortage of power grid to ensure balance between supply and demand of power grid. Finally,
the simulation proves that the three-layer management strategy proposed in this paper can help grid to cut peak and fill valley
effectively, improve the load characteristics of the regional power grid and reduce the amounts of the rejected renewable energy.
© 2022 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.

Keywords: electric vehicle; renewable energy sources; hierarchical architecture; fuzzy control; coordinated optimization

Received 27 December 2021; Revised 17 March 2022

1. Introduction In the past research on V2G management system, a dispatching

method that electric vehicles be used as energy storage components
With the increasingly serious problems of energy and environ- to participate in V2G dispatching operation of power grid is
ment, the economic and environmental costs of traditional fossil proposed, but the method is not suitable for the situation when
energy are increasing, and information technology is evolving. In a large number of electric vehicles participating in V2G service
this background, renewable energy sources (RESs) such as EVs, together [4]. In Ref. [5], a charging and discharging control
wind power and photovoltaics have developed rapidly. But the method is proposed for EVs based on synchronous inverters,
large random load generated by large-scale electric vehicle charg- which can adjust grid frequency by EVs, but the scheduling
ing will have negative impacts on the power grid [1, 2]. And strategy lacked constraints, and ignored the interaction of newly
by the influence of natural conditions, wind power and photo- connected electric vehicles, so it lead to charging peak when the
voltaics’ output has great uncertainty, which has directly influence power grid frequency is high. Chen [6] proposed a multi-objective
on the ability of renewable energy to access the power grid [3]. collaborative optimization scheduling model, which aiming at
However, due to the flexible and adjustable charging requirements reducing the peak-valley difference rate of power grid load and the
of EVs, their batteries can be used as potential energy storage power consumption cost of electric vehicle users, at the same time,
devices by using V2G technology. In addition, coordinated opti- the particle swarm optimization algorithm is used for optimization,
mization of EVs under reasonable charge and discharge schedul- when the number of electric vehicles is large, the model runs
ing can effectively improve the consumption rate of renewable slowly and inefficiently. In Refs. [7], an optimal and efficient
energy and increase the overall economic benefits of system hybrid prediction model is proposed, which predicts the usage
operation. demand of pedelecs in every 48 h on a scale of millions of pedelecs
and captures both the short-term and long-term factors affecting
a Correspondence to: Qihui Liu. E-mail: [email protected] the pedelec usage. In Refs. [8–10], a control strategy of electric
vehicle charging and discharging station based on fuzzy controller
* Key Lab of Power Systems with Renewable Energy Sources, North China
is proposed, by the scheduling method based on fuzzy control,
Electric Power University, Beijing 102206, China
** Suzhou Power Supply Company, State Grid Jiangsu Electric Power Co.,
EVs are controlled based on the goal of reducing peak-to-valley
Ltd., Suzhou, 215004 Jiangsu, Province, China difference and improving voltage stability of power grid [8–10].
[Correction added on 20 May 2022, after first online publication: At present, there are few literatures on coordinated optimization
Authors’ affiliations have been amended.] of EVs and RESs. Yu [11] presented a collaborative dispatching

© 2022 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.

 19314981, 2022, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tee.23623 by Cochrane Thailand, Wiley Online Library on [27/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
THREE-LAYER MANAGEMENT STRATEGY OF V2G, RENEWABLE ENERGY AND ENERGY STORAGE BASED ON FUZZY CONTROL

model for EVs and wind turbines, but does not consider V2G Information of EVs
Real-time vacancy
technology. Aiming at minimum load variance, Jiang [12] pro- V2G Management
System of EV EV Charging pile
posed a wind power and photovoltaic power generation model, charging power of
EVs next interval
and established the model of EV charging and discharging and Prediction result of
maximal output power EV charging plan
the coordinated scheduling model of RESs, which improved the
next interval Prediction result of
utilization of RESs. Cheng [13] proposed a distribution elec- maximal output power
tric vehicles V2G multi-objective optimization method considered next interval Wind farm and
Renewable Energy
Power Grid photovoltaic power
renewable energy access, network loss, voltage deviation, volt- Dispatching System
generation system
Guiding generat ed
age stability and other indicators, which was calculated by genetic Real-time vacancy output power
State of energy
algorithm. An EV double-layer intelligent charging and discharg- storage system
ing strategy is established in Ref. [14] that considered the satis- Real-time vacancy Real-time Balance
faction of wind power and electric vehicle users. System of Energy
Storage
At present, most researches on the coordination and optimization
of EV and RESs choose multi-objective optimization model of Fig. 1. Three-tier management framework for electric vehicles,
EV and RESs combined system, and the planning or heuristic RESs and energy storage system
algorithm is used to optimize the charge and discharge power of
EVs to obtain the overall optimal solution; The more complex the are: control the charging and discharging power of each energy
optimization objectives of multi-objective optimization problems storage battery in the area reasonably to guarantee the frequency
are, the higher the requirements for processing and solving are, stability of the power grid. The distribution network structure of
and the planning and various heuristic algorithms may take much three-layer management architecture is shown in Fig. 2.
time to solve, which is not suitable for real time control when a The renewable energy forecasting system forecasts the short-
large number of electric vehicles are connected. term maximum output power of wind power and photovoltaic
Therefore, this paper proposed a management framework with systems, respectively, and obtains the prediction result of maxi-
V2G, RESs and ES based on fuzzy control, which including mum output of RESs in the next time interval, then transmits it
V2G management system, renewable energy dispatching system to V2G management system. According to the EV charging plan
and energy storage real-time balancing system. Among them, calculated by the EV management system, power vacancy of the
the V2G management system calculates EVs’ charging plan in grid, and state of energy storage, the renewable energy guidance
next time interval considering RESs’ forecasting maximum output output power is obtained.
results, the grid shortage and the EVs’ state of charge (SOC); The V2G management system collects the output power and
The renewable energy management system calculates the guide load data from power distribution network, the prediction result of
generation power of RESs in next time interval referring to the maximum output of RESs in next time interval, and information of
grid power shortage, the EV charging schedule and the state of accessed EVs reported by charging piles, to calculate the charging
energy storage system batteries; And both of them are solved by and discharging power of each charging pile connected with EV,
fuzzy control algorithm. Referring to the grid frequency, the energy then delivered the power instructions to charging piles. The energy
storage management system adjusts the charging and discharging storage management system adjusts the total charging/discharging
power in real time to achieve the supply and demand balance power of energy storage system in real time according to the
of power grid. On the premise of meeting the needs of electric actual frequency deviation of the power grid, to realize the balance
vehicle users, the three-layer framework proposed in this paper between supply and demand of the regional power grid.
can realize the auxiliary function of peak-cutting and valley-filling The overall control process of V2G management system is as
of the power grid, avoid new peak load of power grid caused by too follows:
many electric vehicles connected in the low load, and effectively
improve the consumption rate of RESs. (1) Renewable energy forecasting system collects historical
data of wind power maximum output, numerical weather
2. Three-Layer Management Architecture forecast and photovoltaic output to predict the output of
wind power and photovoltaic system respectively, and
of Electric Vehicle, RESs and Energy Storage
adds them together to get the prediction results of total
The three-layer management architecture of EVs, RESs and ES maximum output power of RESs in next time interval.
proposed in this paper is shown in Fig. 1. It consists of V2G Then the results will be transmitted to V2G management
management system, renewable energy dispatching system and system.
energy storage real-time balancing system. The V2G management (2) V2G management system collects information of EVs
system is responsible for managing the power distribution of each connected to charging pile in the current time interval, and
charging pile in a certain area, its main tasks are: arrange the obtains the power generation and load from the regional
charging and discharging power of each charging pile rationally, power grid dispatching center.
improve the load characteristics of the regional power grid, (3) V2G management system calculates the EVs’ electricity
and realize the auxiliary function of peak shaving and valley demand, the allowed discharging power and the latest
filling. The renewable energy dispatching system is responsible charging time based on EVs’ information, which considers
for controlling the output of RESs, its main tasks are: arrange the the situation of current grid’s supply and demand, to
active output power of RESs rationally and guarantee the load calculate the reference charging and discharging power of
demand of the regional power grid to be satisfied. The energy EVs, and send instructions to each charging pile.
storage real-time balancing system is responsible for the charging (4) The energy storage management system adjusts the charg-
and discharging power of the energy storage system, its main tasks ing/discharging power of the energy storage battery in

1331 IEEJ Trans 17: 1330–1338 (2022)

 19314981, 2022, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tee.23623 by Cochrane Thailand, Wiley Online Library on [27/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Q. LIU ET AL.

State of energy flow

EV
energy charging information flow
storage plan
system
Energy storage DC/AC
T 10kV

Power Prediction result

Load
Supply of maximal output
demand
Charging pile 1 output power next Charging pile n
power interval
power
instructions EV instructions
EV
information V2G information
EV Management EV
system
EV
Power
charging
instructions
plan

Fig. 2. Structure of power grid with three-layer management system

real time according to the real-time frequency deviation residence time

of EV 2
of power grid. Computing Point
residence time
of EV 3
At present, there are many research results on short-term forecast
for maximum output of RESs [15–17], and the participation 5 6 7 8 9 ĂĂ288
of energy storage systems in power grid frequency modulation
[18–20], due to the limitation of space, these parts are not covered
in this paper. Instead, this paper focuses on renewable energy Fig. 3. The schematic diagram of control time interval
dispatching systems and V2G management.
From the above control process, it can be seen that the charging in current time interval and the prediction results of maximum
and discharging power of EVs is related to EVs’ SOC, the grid load output of RESs, it figures out the charging and discharging power
condition, and the prediction result of maximum output of RESs; command for each EV connected to charging pile, to improve
The renewable energy dispatching system needs to consider the regional load characteristics, achieve peak shaving and valley
condition of grid load, the charging/discharging plan of EVs, and filling, and improve consumption rate of RESs.
the state of energy storage system, so both systems are difficult
to describe with an analytical model. Fuzzy control is based on
operational experience of human of controlled object, which is not 3.1. EV status measurement When EV is connected
necessary to know the internal structure and analytical model of to charging station, the charging pile type (DC charging pile is
the controlled object, so the robustness and adaptability can meet taken as an example in this paper), time of accessing to charging
the requirements for control. Therefore, V2G management system stationT 1 , the initial state of charge SOC 0 , and the departure time
and renewable energy dispatching system adopted the fuzzy control set by EV user T 2 can be collected. The metering unit can figure
method in this paper. out the electricity that EVs need-Q c , and the allowed discharging
At the same time, considering the randomness of EVs connected, capacity of EV-Q d .
it is impossible to predict the exact access time and charging 
Qc = Q(1 − SOC) SOC < 1
demand of EVs in the future, then the day is divided into 288 (1)
0 SOC = 1
time intervals, each time interval is 5 min, and the grid operation
status is updated at the beginning of each time interval, which is 
also the operation calculation point of each management system, Qd = Q (SOC − SOCmin ) SOC > SOCmin
(2)
the schematic diagram of control time interval is shown in Fig. 3 0 SOC = SOCmin
[10]. If the arrival time of EVs is in the middle part of the time
where Q is the battery capacity of EV, SOC is the current state
period, the EVs needs to wait for the next time period to charging
of charge of EV, and SOC min is the minimum limit of state of
and discharging.
battery during EV charging and discharging; In this paper, the
value SOC min is 0.2.
3. Electric Vehicle V2G Management System At the same time, in order to meet the charging demand of EVs
and make them full charge before leaving, the latest charging time
The V2G management system is responsible for managing the of EV is defined. The calculation formula is as follows:
charging and discharging power of each EV connected to charging
pile in regional distribution network. Considering power grid load T3 = T2 − (1 − SOCmin ) × Q/Pn (3)

1332 IEEJ Trans 17: 1330–1338 (2022)

 19314981, 2022, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tee.23623 by Cochrane Thailand, Wiley Online Library on [27/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
THREE-LAYER MANAGEMENT STRATEGY OF V2G, RENEWABLE ENERGY AND ENERGY STORAGE BASED ON FUZZY CONTROL

where P n is the rated charging power of EV; Taking into NB NM NS ZE PS PM PB

Grade of membership
1.0
account the uncertainty of EV charging or discharging, the battery
minimum limit is used in (3).
0.6
3.2. EV ordered scheduling system based on fuzzy
control The core idea of EV orderly scheduling is that: if grid
is still at the peak of load though the output power of RESs is the 0
highest, the charging of a large number of EVs would inevitably -600 -400 -200 0 200 400 600
aggravate the peak-to-valley difference of power grid, which is not ΔP1/kW
conducive to grid’s stable operation. At this situation, EVs can be (a)
controlled to reduce the consumption of electric energy from the
Z VS S M B VB

Grade of membership
power grid or even to supply electric energy to the power grid; 1.0
If the power grid is at the valley of load, and the consumption of
RESs is little, EVs can be controlled to consumed electricity to 0.6
improve the consumption of RESs. At the same time, the charging
power of EVs is related to their electricity demand, when demand
is high, EVs can be charged at higher power; When demand is
low, the charging power should be set to a lower level in order to 0
0 0.2 0.4 0.6 0.8 1
prevent overcharging.
Qc
Pe is the power shortage between current power generation
and load of the grid, PRESs is the forecasted maximum output of (b)
RESs in next time interval predicted by RESs forecasting system, ZE L M L RP

Grade of membership
and the estimated power shortage of grid in the next time interval 1.0
P1 is figured as follows:
0.6
P1 = Pe + PRESs (4)

The EV discharge fuzzy controller has two inputs and single

output, the two inputs are the estimated grid shortage P1 and 0
the EV demand power in the next time interval, respectively, the 0 3 6 9 12
output is the EV charging power. The fuzzy subset of estimated Pc /kW
power grid shortage P1 is set to [NB , NM , NS , ZE , PS , (c)
PM , PM ]. Among them, when P1 = ZE, the power shortage
is 0, which represents that considering RESs, the grid supply and Fig. 4. Fuzzy membership function curves of V2G management
demand basically reach equilibrium; P1 = NB indicates that even system. (a) Fuzzy membership function curves of P 1 . (b) Fuzzy
if RESs is on the maximize output power, the power shortage of membership function curves of Qc. (c) Fuzzy membership function
grid is still very high, which represents that it is at the peak of curves of P c
power consumption and all the RESs are consumed; P1 = PM
indicates that there is a lot of residual power generated by the Table I. Fuzzy control rules of EV charging
current power grid, which represents that it is at the valley of power P 1
consumption and RESs consumption; And other states indicate
different states of the grid estimated shortage. Qc NB g NM g NS g ZE g PS g PM g PB g
The fuzzy subset of EV charging demand Qc is set to [Z , Zc ZE ZE ZE ZE ZE ZE ZE
VS, S , M , B , VB ]. Among them, Qc = Z indicates that the VS c ZE ZE ZE ZE L L H
charging demand of EV is close to zero, and the battery is almost Sc ZE ZE ZE ZE M M H
full; Qc = VB represents that the charging demand of the EV is Mc ZE ZE ZE ZE M H H
extremely high, and its SOC is very low. Other states indicate Bc ZE ZE ZE ZE RP RP RP
different states of EV demand, and fuzzy subset of Qd is similar VB c ZE ZE ZE ZE RP RP RP
with Qc .
Considering that the charging and discharging power of the DC
Table II. Fuzzy control rules of EV discharging
charging pile can be continuously adjusted, the fuzzy subset of
the EV charging power PcDC is set to [ZE, L, M, H, RP ], which P 1
indicates that the conditions of allowed charging power of EVs
is zero, lower, medium, higher and rated respectively. The fuzzy Qd NB NM NS ZE PS PM PB
subset of EV discharging power PdDC is similar with PcDC . Part Z ZE ZE ZE ZE ZE ZE ZE
of the membership function curves in this paper are shown in VS M L L ZE ZE ZE ZE
Fig. 4, the membership function of PdDC is similar with PcDC and the S M H H ZE ZE ZE ZE
membership function of Qc is similar with Qd . The fuzzy control M RP RP H ZE ZE ZE ZE
rules of charging and discharging are shown in Tables I and II. B RP RP RP ZE ZE ZE ZE
Set pc and pd as the charging power and discharging power of VB RP RP RP ZE ZE ZE ZE
electric vehicle, respectively, and the controller can be described

1333 IEEJ Trans 17: 1330–1338 (2022)

 19314981, 2022, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tee.23623 by Cochrane Thailand, Wiley Online Library on [27/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Q. LIU ET AL.

by the above N = 42 fuzzy rules. If the output of the i-th fuzzy Start
rule is u i , the output expression is shown in (5).
⎧ N
⎪
⎪ = i =1 vic uic P DC /5
⎪
⎨ pc  N nc t=1
i =1 vic
N
(5)
⎪
⎪
⎪
⎩pd = i =1

vid uid DC
Pnd /5
N t<=288?
i =1 vid N
where, uic and uid are the output results of the i-th rule in the Y
charge and discharge controller; vic and vid are the weight of the Renewable Energy Dispatching
System collect the data needed
i -th rule in the charge and discharge controller of the total output, for forecasting, and forecast the
the calculation formula is shown in (6). maximum output of RESs in the
next interval
⎧
⎪ 
2
⎪
⎨ vic = Aic (xc )
⎪ i
Update the current grid information, read
i =1 the current charging information of N
(6) t=t+1
⎪
⎪ 
2 EVs, and calculate the latest charging
⎪
⎩vid = Aiid (xd ) time of each EV.
i =1

where, xc and xd are the input of charge and discharge controller; Sort electric vehicles
according to arrival
Aiic and Aiid are the membership of the fuzzy subset corresponding time,and,make id=1
to rule i -th of charge and discharge controller. When l = 1
Transfer the
representing the fuzzy set (Qc /Qd ), and l = 2 representing the charge plan Pref
fuzzy set (pc /pd ). N of all EV in next
id≤ N
The ambiguity resolution of fuzzy controller adopts the center of interval to Energy
Y Dispatching
gravity method to obtain the charging power pc and the discharging System
power pd of EVs, then the power command Pref can be obtained by
Y
coupling output pc and pd of the charging and discharging fuzzy The EV is marked full
controller, the formula is shown as Equ.7.
N
Pref = (pc − pd ) (7)
N
Not reaching the EV
where, P ref is the power consumption reference of electric vehicle the latest charging time?
in the next period which calculated by fuzzy control method.
Charging with
Y rated charging
3.3. Flow chart of orderly charging scheduling sys- power
tem The operation flow of the orderly charging scheduling Calculate the charging and
discharging power of the EV
system is shown in Fig. 5. In this figure, the number of N electric
vehicles are sorted according to the arrival time, and the identifier Update the EV's SOC
is recorded as i d . and grid load

SOC< 1
4. Renewable Energy Dispatching System Based N
on Fuzzy Control Y

The renewable energy dispatching system is responsible for Mark that the EV Mark that the EV
is not full is full
managing the guiding output power of RESs in the region, by
integrating the current grid load, the charging schedule of EVs
next time interval, and the state of energy storage battery, which
can calculate and issue the guiding output power for regional wind id=id+1

power and photovoltaic to ensure the basic balance between supply

and demand of the regional power grid.
End
The core idea of the RESs dispatching system is that when
the EV is charged/discharged according to the plan, if power
grid would still be at the peak of load, the power generation Fig. 5. Flow chart of EV charging and discharging control
of RESs should be increased as much as possible to meet the
load demand; If the power grid would still be at the valley
of load, which represents the consumptive of RESs is low, the The fuzzy controller of RESs dispatching system has double
output of RESs needs to be reduced. The output of RESs is also inputs and single output, and the inputs are predicted grid
related to the state of energy storage system. When the state of power shortage P2 and energy storage system state SOC b , the
energy storage batteries is high, the RESs can reduce the output. calculation formula of P2 is shown in (8).
Instead, when the state is low, the RESs can increase the output
appropriately. P2 = Pe + Pref (8)

1334 IEEJ Trans 17: 1330–1338 (2022)

 19314981, 2022, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tee.23623 by Cochrane Thailand, Wiley Online Library on [27/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
THREE-LAYER MANAGEMENT STRATEGY OF V2G, RENEWABLE ENERGY AND ENERGY STORAGE BASED ON FUZZY CONTROL

where Pe is the power shortage between current power gener- NB NM M NS ZE PS S PM PB

Grade of membership
1.0
ation and the load, Pref is the charging power of all EVs in next
time interval, and the calculation formula is shown in (7).
The fuzzy subset of predicted power shortage P2 is set to [NB , 0.6
NM , M , NS , ZE , PS , S , PM , PM ]. Among them, when P2 =
ZE, the power shortage is 0, and the supply and demand of power
grid are basically balanced after the EV is charged/discharged as
planed; P2 = NB represents the power shortage is still high after 0
-300 -200 -100 0 100 200 300
the EVs are charged/discharged as planned, and it is necessary ΔP2
to increase the power generation of RESs as much as possible;
P1 = PM means that there is a lot of electric energy left by the
(a)

Grade of membership
current grid, which represents that the demand for RESs is very 1.0 Z VS S M B F
little; Other states indicate different states of the grid’s predicted
shortage. 0.6
The fuzzy subset of the battery state of energy storage system
SOC b is set to [ZE , VS, S , M , B , F ]. Among them, SOCb = ZE
represents that state of charge of energy storage system battery is
extremely low, and it needs to absorb energy from power grid; 0
0 0.2 0.4 0.6 0.8 1
SOCb = F represents that the battery of energy storage system SOCb
is almost full, and it’s ready to discharging to power grid. Other
(b)
states indicate different states of the energy storage system battery.
The fuzzy subset of RESs guidance output power PRESs is set 1.0
ZE SL L M SM H RP

Grade of membership
to [ZE , SL, L, M , SM , H , RP ], where PRESs = ZE represents that
RESs need to stop outputting; PRESs = RP represents that RESs
should output in rated power, and the maximum possible power 0.6
should be achieved when the generating conditions do not meet the
conditions of rated power; And the other states indicate different
states of RESs guidance output power. The membership function
curve and fuzzy rules of the fuzzy controller of RESs dispatching 0
system are shown in Fig. 6 and Table III. 0 30 60 100 150 200 300
PRESs /kW

5. Simulation and Verification (c)

5.1. Basic data of the example Figure 7 shows the total Fig. 6. Fuzzy membership function curves of renewable energy
base load curve of distribution transformer in a certain area, the dispatching system. (a) Fuzzy membership function curves of
predicted maximum output of wind power and photovoltaic power P 2 . (b) Fuzzy membership function curves of SOC b . (c) Fuzzy
generation in this area is shown in Fig. 8. Assume that the capacity membership function curves of P RESs
of ES is 5000kWh, the number of electric vehicles in this area is
500 and EV battery capacity is 32 kWh, the maximum discharge
Table III. Fuzzy control rules of guiding output power of RESs
power and maximum charging power allowed in charging pile
are 6 kW and 12 kW. According to the EV trip chain simulation P 2
based on Monte Carlo method proposed in Ref. [21], the charging
requirements of EVs are obtained. SOC b NB NM M NS ZE PS S PM PB
Z RP H SM M L SL ZE ZE ZE
5.2. Analysis of system operation results The VS RP H SM M L SL ZE ZE ZE
regional base load, the distribution load curve in EV instant S RP H M L SL ZE ZE ZE ZE
charging mode and strategy proposed in this paper are shown M RP SM M SL ZE ZE ZE ZE ZE
in Fig. 9. It can be seen from Fig. 9 that the power grid is at B SM M L SL ZE ZE ZE ZE ZE
the peak of power consumption between 8:00 and 10:00, 19:00 F SM M SL ZE ZE ZE ZE ZE ZE
and 22:00. At this time, if the EV charging is not controlled
properly, access of a large number of EVs will further increase
the load peak of the power grid, the distributed energy storage Figure 10 shows the power curve of renewable energy con-
characteristics of EVs are failed to utilize, and is not conducive to sumption in different situation. Table IV shows the peak-to-valley
the stable operation of the power grid. While the control strategy difference and the renewable energy consumption rate in differ-
proposed in this paper can meet the charging needs of EV users, ent strategies, and the maximum power generation of regional
and effectively delay the charging demand of EVs connected renewable energy is 3838.6 kWh. From Fig. 10, it can be seen
during peak load periods, even supply energy to power grid. It that the power grid can only consume RESs during peak period
can not only avoid the increase of load spikes, but also effectively between 8:00–10:00 and 19:00–22:00 without EV charging, and
shift the charging demand to low load period, which successfully the renewable energy consumption rate is extremely low. Accord-
realizes the function of assisting power grid to cut peak and fill ing to Fig. 10 and Table IV, the consumption rate of renewable
valley and makes the load curve more stable. energy has been improved a little with strategic EV charging, while

1335 IEEJ Trans 17: 1330–1338 (2022)

 19314981, 2022, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tee.23623 by Cochrane Thailand, Wiley Online Library on [27/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Q. LIU ET AL.

1.6 350
without EV
1.4 300 Unstrategic EV charging
strategy In this paper
1.2
Base Load/MW

250 Maximum output of RESs

Active power/MW
1
200
0.8
150
0.6
100
0.4

0.2 50
6:00 9:00 12:00 15:00 18:00 21:00 24:00 3:00 6:00
t/h 0
6:00 9:00 12:00 15:00 18:00 21:00 24:00 3:00 6:00
t/h
Fig. 7. Base load curve of residential area
Fig. 10. Curve of consumed renewable energy power
1
Wind power
Photovoltaic power Table IV. Comparison of peak to valley difference and RESs
0.8 Total RESs
consumption rate
Active power/MW

0.6
Base Unstrategic Strategy in
load EV charging this paper
0.4
Peak valley difference 1.238 1.742 0.922
0.2 (MW)
consumed renewable 721.4 1746.6 2854.9
0 energy (kWh)
6:00 9:00 12:00 15:00 18:00 21:00 24:00 3:00 6:00
t/h
Renewable Energy 18.79% 45.50% 74.37%
Consumption Rate
Fig. 8. Predicted maximum output power of RESs

2.2 1.4
The 7th EV,Arrived at 07:58,leave at 14:07
2
The 221th EV,Arrived at 16:42,leave at 17:52
1.8 1.2 The 286th EV,Arrived at 18:18,leave at 06:43

1.6

1.4 1
Load/MW

1.2
0.8
1
SOC

0.8
0.6
0.6
strategy In this paper
0.4
Unstrategy EV charging
Base Load 0.4
0.2
0
6:00 9:00 12:00 15:00 18:00 21:00 24:00 3:00 6:00 0.2
t/h

Fig. 9. Load curve in different control strategy 0

6:00 9:00 12:00 15:00 18:00 21:00 24:00 3:00 6:00
t/(h)
in the three-layer architecture proposed in this paper, the consump-
Fig. 11. SOC curves of typical EVs
tion rate of renewable energy has been greatly improved. It can
be seen from Figs. 9 and 10 that the control strategy proposed in
this paper can not only get RESs fully consumed during the peak electric energy into the power grid at peak load, make full use
period, but also can effectively improve the consumption of RESs of the characteristics of distributed energy storage, and help the
during valley period by shifting a large number of EV loads to power grid to cut peak and fill valley.
valley period. In addition, the simulation model based on the strategy in
Figure 11 shows some EVs’ SOC curve in coordinated charging this paper is fast, and does not need to solve complex planning
strategy proposed in this paper. It can be seen from Fig. 11 models. The overall running time is only 156 s, corresponding to
that charging plans arranged by V2G management system are 288 time periods of 1 day, the average time of per time interval
different due to the different conditions in power grid, charging is only 0.542 s, which means the running time of this three-layer
requirements, and residence time of the EVs, but they can complete energy management process is reasonable for the practical
the charging task within the prescribed residence time, supply works.

1336 IEEJ Trans 17: 1330–1338 (2022)

 19314981, 2022, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tee.23623 by Cochrane Thailand, Wiley Online Library on [27/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
THREE-LAYER MANAGEMENT STRATEGY OF V2G, RENEWABLE ENERGY AND ENERGY STORAGE BASED ON FUZZY CONTROL

6. Conclusion (10) Zhang Y, Liu Q, Hong C, Tang G. Charging and discharging dispatch
strategy of regional V2G based on fuzzy control. Electric Power
At present, the coordinated optimization of V2G and renewable Automation Equipment 2019; 39:147–153.
energy is a research hotspot. This paper constructs a three-layer (11) Yu D, Song S, Zhang B, Han X. Impacts of electric vehicles on power
structure of V2G, renewable energy and energy storage firstly, systems as well as the associated dispatching and control problem.
and makes the reasonable scheduling between V2G and renewable Automation of Electric Power Systems 2011; 35:24–29.
energy by fuzzy control method, then the final simulation result (12) Jiang X, Wang J, Han Y, Zhao Q. Coordination dispatch of electric
shows that: vehicles charging/discharging and renewable energy resources power
in microgrid. Procedia Computer Science 2017; 107:157–163.
(13) Cheng S, Li Z. Multi-objective network reconfiguration considering
1) The V2G management system based on the fuzzy control
V2G of electric vehicles in distribution system with renewable
algorithm proposed in this paper has high resource utiliza- energy. Energy Procedia 2019; 158:278–283.
tion rate, fast optimized calculation speed and high operat- (14) Huang G, Lei X, Yang Y, Wang Y, Chen X. Two-layer Smart charge-
ing efficiency, which is suitable for situation of accessing discharge strategies for electric vehicles considering wing generation
large-scale electric vehicles. ang Users’ satisfaction. Transactions of China Electrotechnical Soci-
2) The V2G management system can effectively realize the ety 2015; 30:85–97.
function of assisting the power grid to cut peaks and fill (15) Pan C, Qin B. Research on power model of large-scale wind
valley under the premise of meeting the needs of EV users. farm based on measured data. Power System Technology 2017;
3) The three-layer architecture control strategy adopted in this 41:2625–2631.
(16) Li L, Wen S, Tseng M, Wang C. Renewable energy prediction:
paper can arrange the charging and discharging of EVs rea-
A novel short-term prediction model of photovoltaic output power.
sonably, improve the consumption rate of renewable energy
Journal of Cleaner Production 2019; 228:359–375.
source effectively, and increase the overall economic ben- (17) Peng C, Chen N, Gao B. Ultra-short-term WindPower forecasting
efits of system operation. method combining multiple clustering and hierarchical clustering.
Automation of Electric Power Systems 2019; 43:1–8.
Acknowledgment (18) Lucas A, Chondrogiannis S. Smart grid energy storage controller
for frequency regulation and peak shaving, using a vanadium redox
This work is supported by State Grid Co., Ltd project ‘Research on flow battery. International Journal of Electrical Power 2016; 80:
the technology of lightweight power conversion and power Flexibility 26–36.
Coordination control for offshore DC wind farm’, (5100-202199543A-0- (19) Jayamaha C, Costabeber A, Williams A, Sumner M. An indepen-
5-ZN). dently controlled energy storage to support short term frequency
fluctuations in weak electrical grids. International Journal of Electri-
References cal Power 2018; 103:562–576.
(20) Tang ZX, Lim YS, Morris S, Yi JL, Lyons PF, Taylor PC.
A comprehensive work package for energy storage systems as
(1) Clement-Nyns K, Haesen E, Driesen J. The impact of charging plug- a means of frequency regulation with increased penetration of
in hybrid electric vehicles on a residential distribution grid. IEEE photovoltaic systems. International Journal of Electrical Power
Transactions on Power Systems 2010; 25:371–380. 2019; 110:197–207.
(2) Richardson P, Flynn D, Keane A. Impact assessment of varying (21) Wen J, Tao S, Xiao X, Luo C, Liao K. Analysis on charging demand
penetrations of electric vehicles on low voltage distribution systems. of EV based on stochastic simulation of trip chain. Power System
IEEE 2010:1–6. Technology 2015; 39:1477–1484.
(3) Sinsel SR, Riemke RL, Hoffmann VH. Challenges and solution tech-
nologies for the integration of variable renewable energy sources—A
review. Renewable Energy 2020; 145:2271–2285.
Qihui Liu (Non-Member), male, doctor, associate professor,
(4) Weng G, Huang F, Zhang Y, Xie L, Qi J. V2G strategy for energy
dispatch of Island microgrid with EBBG. Electric Power Automation
main research directions: New Energy Power
Equipment 2016; 36:31–37. Generation and Grid-connected Control.
(5) Liu D, Zhong Q, Wang Y, Liang Y. A Synchronverter-based V2G
Smart charging and discharging control strategy for electric vehicles.
Proceedings of the CSEE 2017; 37:544–557.
(6) Chen K, Niu Y. Research on optimal scheduling strategy of electric
vehicle charging and discharging based on V2G. Power System
Protection and Control 2019; 47:1–9.
(7) Zhang C, Dong M, Luan TH, Ota K. Battery maintenance of Pedelec
sharing system: Big data based usage prediction and replenishment
Ruohan Tian (Non-Member), female, postgraduate, main research
scheduling. IEEE Transactions on Network Science and Engineering directions: New Energy Power Generation
2020; 7:127–138. https://doi.org/10.1109/TNSE.2019.2901833. and Grid-connected Control.
(8) Thirugnanam K, Joy TPER, Singh M, Kumar P. Modeling and control
of contactless based smart charging station in V2G scenario. IEEE
Transactions on Smart Grid 2014; 5:337–348.
(9) Singh M, Kumar P, Kar I. Implementation of vehicle to grid
infrastructure using fuzzy logic controller. IEEE Transactions on
Smart Grid 2012; 3:565–577.

1337 IEEJ Trans 17: 1330–1338 (2022)

19314981, 2022, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tee.23623 by Cochrane Thailand, Wiley Online Library on [27/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

IEEJ Trans 17: 1330–1338 (2022)

Simin Pang (Non-Member), female, postgraduate, main research
directions: New Energy Power Generation
and Grid-connected Control.
Q. LIU ET AL.

1338
Chenwei Hong (Non-Member), male, postgraduate, main research
directions: New Energy Power Generation
and Grid-connected Control.