Applied Energy 190 (2017) 591–599

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

An optimal dispatching strategy for V2G aggregator participating in

supplementary frequency regulation considering EV driving demand and
aggregator’s benefits q
Chao Peng, Jianxiao Zou ⇑, Lian Lian, Liying Li
School of Automation Engineering, University of Electronic Science and Technology of China, Chengdu, Sichuan 611731, China

h i g h l i g h t s

A dispatching strategy for V2G aggregator participating in frequency regulation.

Satisfying EV driving demand and maximizing benefits of aggregator simultaneously.

A driving demand calculation module is used to obtain the required EV battery SOC.

An optimal regulation power calculation model is designed to optimize aggregator profits.

A regulation power allocation model is built to allocate regulation power to EVs.

a r t i c l e i n f o a b s t r a c t

Article history: With the development of Vehicle-to-Grid (V2G) technology and increasing number of electric vehicles
Received 1 July 2016 (EVs) integrating in power grid, supplementary frequency regulation service provided by V2G aggregator
Received in revised form 11 December 2016 has been seen as the most promising grid ancillary service provided by the integrated EVs. In this paper,
Accepted 12 December 2016
an optimal dispatching strategy of V2G aggregator is proposed to satisfy the driving demand of EV own-
ers and maximize the economic benefits of aggregator simultaneously when it participates in supple-
mentary frequency regulation. A judgment module is designed to determine EVs in aggregator
Keywords:
whether participating in frequency regulation according to EV battery SOC for EVs’ driving demand,
V2G aggregator
Dispatching strategy
which is calculated by EVs’ daily driving distance. An optimal regulation power calculation model is built
Supplementary frequency regulation to optimize profits of aggregator and tracking performance of frequency load control signal from grid
EV driving demand operator. A fair regulation power allocation module is designed to avoid over-discharging of EVs in aggre-
Economic benefits gator. Finally, the proposed strategy is implemented in the simulation experiments to demonstrate its
effectiveness.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction
Since 90s of last century, the environment pollution problems
caused by traditional vehicle exhaust have attracted extensive
public attentions. Especially, in recent years, the continuous
increasing of vehicle makes these problems more and more seri-
ous. With the rapid development of technology of electric vehicles
(EV), using EV instead of traditional vehicle has been considered as
one of the most effective solutions [1]. Meanwhile, the amount of
global EVs has increase dramatically year by year.
q
This work was supported by National Natural Science Foundation of China
under Grant No. 61201010.
⇑ Corresponding author.
E-mail address: jxzou@uestc.edu.cn (J. Zou).
With the rapid increasing of EVs, the integrating of such large
scale of EVs would bring great challenges to power grid operation.
To promote the coordinated development of power grid and EVs,
the concept of vehicle-to-grid (V2G) was proposed in 1997 [2]. In
the past near twenty years, V2G technology has been received
extensive researches and has become mature. By using V2G tech-
nology, EV chargers could be operated in bidirectional mode, i.e.,
EV batteries can be charged when they are plugged in power grid,
the energy stored in the EV batteries can also be delivered back to
the power grid [3].
Now, the EVs integrating in grid not only are charging loads, but
also could be used as power generation and energy storage units in
power grid [4]. They could be utilized to participate in grid ancil-
lary services to maintain power grid’s stability and security, such
as frequency regulation [5], voltage regulation [6], spinning reserve

http://dx.doi.org/10.1016/j.apenergy.2016.12.065
0306-2619/Ó 2016 Elsevier Ltd. All rights reserved.
592 C. Peng et al. / Applied Energy 190 (2017) 591–599
Nomenclature
Ropt ðtÞ optimal regulation power of aggregator at time t
Rnopt ðtÞ optimal regulation power of nth EV at time t
Preg ðtÞ frequency regulation price at time t
Pem ðtÞ market electricity price at time t
N the total number of EVs managed by EV aggregator
SOC n ðtÞ battery SOC of nth EV at time t
SOC nexp battery SOC of driving demand of nth EV owner
C nmax rated capacity of nth EV battery
[7], load peak shifting [8] and so on. However, the capacity of single
EV is too little to participate in these grid ancillary services. Thus, a
concept of V2G aggregator is proposed, which introduces a central-
ized control system to control a large scale of EVs to provide grid
ancillary services [9]. Considering the communication delay
between grid operator and V2G aggregator, communication delay
between V2G aggregator and each EVs controlled by aggregator,
V2G aggregator commonly is used to participate in the supplemen-
tary frequency regulation service, which requires V2G aggregator
to follow a load frequency control (LFC) signal [10–12], which is
within the achievable power capacity of V2G aggregator. The dis-
patching strategy is the key for aggregator, which regulates the
charging and discharging power of each EV to participate in grid
frequency regulation. The dispatching strategy determines the fea-
sibility and economics of V2G aggregator participating in supple-
mentary frequency regulation.
Currently, the dispatching strategy for V2G aggregator partici-
pating in supplementary frequency regulation has become a
research hotspot. Some dispatching strategies are proposed consid-
ering on the problem of satisfying EV’s charging demand or driving
demand when EV aggregator participating in supplementary fre-
quency regulation [13–17]. To satisfy charging demand of EVs in
aggregator, a dispatching strategy considering charging demands
is proposed in [14]. To reduce the EV owner’s cost and obtain the
desired EV battery SOC when EV departs, a model predictive con-
trol based dispatching strategy is proposed in [15]. A dispatching
strategy considering the operational status and mobility needs is
proposed in [16]. In [17], a dispatching strategy is proposed to
maximize the profit for the EV owner and satisfy EV charging
demand during the whole parking time.
Meanwhile, many dispatching strategies are proposed to opti-
mize the aggregator’s profit when it participating in supplemen-
tary frequency regulation [17–22]. A dispatching strategy is
proposed to maximize the revenue obtained from aggregator par-
ticipating in the supplementary frequency regulation in [18]. To
optimize aggregator’s profit under electricity price uncertainty,
the dispatching strategies considering the electricity price is pro-
posed in [19,20]. In [21], a dispatching strategy is proposed to min-
imize the operating cost of aggregator when it participating in
frequency regulation. In [22], a real-time welfare-maximizing reg-
ulation allocation based dispatching strategy is proposed to fairly
allocate the regulation power capacity among the EVs. In [23], a
dispatching strategy is proposed to optimize the aggregator’s profit
considering the fair allocation of charging or discharging power of
EVs. A stochastic dynamic programming based dispatching strat-
egy is proposed to optimize the charging and frequency regulation
capacity bids of EVs in [24]. In [25], a robust frequency regulation
scheduling algorithm is proposed to maximize the revenue of
aggregator under the frequency regulation performance-based
compensation scheme.
According to the mentioned above, the satisfying driving
demand of EV and optimization of benefits of aggregator are the
two main problems faced by EV aggregator dispatching strategy
Rich max battery maximum charging power limitation of nth EV
Ridisch max battery maximum discharging power limitation of nth
EV
SðtÞ frequency regulation command signal at time t
participating in supplementary frequency regulation. The first
problem refers to the satisfaction of EV owners, the second prob-
lem refers to the attraction for V2G aggregator participating in fre-
quency regulation. The above existing dispatching strategies have
not considered on both of them comprehensively.
The contribution of this paper is the development of a novel
optimal dispatching strategy for V2G aggregator participating in
supplementary frequency regulation, to (I) optimizes the tracking
performance of load frequency control signal, (II) satisfy EV own-
er’s driving demand and maximize the economic benefits of V2G
aggregator as much as possible in same time. The SOC for EV own-
er’s driving demand is calculated by the daily driving distance
information. A judge module is designed to determine which EVs
in aggregator participating in frequency regulation. An optimal
regulation power participating in frequency regulation calculation
module is built for aggregator to optimize the profits of aggregator
and tracking performance of frequency regulation command sig-
nal. A regulation power allocation module is designed to fairly allo-
cate the charging or discharging power of each EV participating in
frequency regulation based on their capacity.
The rest of this paper is organized as follows. In Section 2, the
structure of V2G aggregator participating in supplementary fre-
quency regulation is described. In Section 3, the principle and
design objective of the proposed dispatching strategy are pro-
posed. The detailed design of the proposed strategy is presented
in Section 4. The results of simulation experiments are discussed
in Section 5. Finally, the conclusion is presented in Section 6.
2. V2G aggregator participating in supplementary frequency
regulation
V2G aggregator is a control center of EVs, which manages the
charging and discharging behavior of each EV in aggregator. The
system structure of V2G aggregator participating in supplementary
frequency regulation is shown in Fig. 1.
As seen in Fig. 1, when aggregator participates in supplemen-
tary frequency regulation, aggregator receives load frequency con-
trol signal from grid operator and sends its participating power
capacity information back to grid operator. Meanwhile, aggregator
dispatches the charging or discharging power command to each EV
according to the load frequency control signal, information of EV
owner’s requirements and state information of each EV. The aggre-
gator plays a connector role between EVs and power grid.
3. The proposed dispatching strategy
3.1. Principle
The proposed dispatching strategy consists of driving demand
calculation module, EVs participating frequency regulation judge
module, optimal regulation power participating in frequency
 C. Peng et al. / Applied Energy 190 (2017) 591–599 593

Fig. 1. System structure of V2G aggregator participating in supplementary frequency regulation.

regulation calculation module and regulation power allocation
module. Its scheme is shown in Fig. 2.
The driving demand calculation module calculates the EV bat-
tery SOC for EV owner driving demand according to each EV daily
driving distance. The EVs participating frequency regulation judg-
ment module determines which EVs in aggregator participating
in frequency regulation. The optimal regulation power calculation
module calculates the regulation power of aggregator participating
in frequency regulation according to the SOC for driving demand of
EV owners and the current SOC state of EV battery. The regulation
power allocation module dispatches the charging and discharging
power of each EV according to the regulation power of aggregator
participating in frequency regulation and SOC state of each EV.
3.2. Objectives
When aggregator participates in supplementary frequency reg-
ulation, grid operator would ask aggregator to track the load fre-
quency control signal. However, charging service provided
aggregator is prior to participating in supplementary frequency

Fig. 2. Scheme of the proposed dispatching strategy.

594 C. Peng et al. / Applied Energy 190 (2017) 591–599

regulation for EV owners. Their EVs in aggregator should be ticipating in supplementary frequency regulation and optimization
charged to meet their driving demand when they departure. Mean- of the tracking performance of load frequency control signal. They
while, the attraction for aggregator is its economy benefits are given as follows,
obtained from participating in supplementary frequency regula- 8
tion. Thus, the objectives of the proposed dispatching strategy of < max Ireg ðtÞ
Ropt ðtÞ
V2G aggregator could be given as follows, ð4Þ
: min Dsig ðtÞ
Ropt ðtÞ

(1) The SOC of each EV in aggregator should satisfy the driving

demand of EV owners.
(2) Optimizing the economy profits of aggregator and tracking
performance of load frequency signal from grid operator as
much as possible.
4. Design of the proposed dispatching strategy
4.1. Driving demand calculation module
The calculation model for driving demand calculation module is
built based on the EV daily driving distance, EV battery energy con-
sumption and additional energy for uncertainty of driving and bat-
tery safety. It is written as follow,
DE ¼ M d  Em
DE 1 t respectively, Dt is the time interval of load frequency control sig-
SOC exp ¼ SOC g  20% þ þ Em 
S S
where DE is the energy for driving demand. Md is the daily driving
distance of EV. Em is the energy consumption per kilometer. SOCgexp
is the battery SOC for driving demand, SOCg is upper bound SOC of
EV battery. S is the maximum capacity of battery. Considering on
the battery working temperature, current limits, the SOC range of
EV battery always are selected between 20% and 95% in the
calculation.
where Ropt ðtÞ is the optimal regulation power of aggregator partici-
pating in supplementary frequency regulation, Ireg ðtÞ is the total
economy benefits obtained from participating in frequency regula-
tion, Dsig ðtÞ is the tracking error between load frequency control sig-
nal and actual regulation power participating in supplementary
frequency regulation.
Considering on the income of aggregator participating in sup-
plementary frequency regulation, charging income and discharging
cost of EVs in the process of participating in frequency regulation,
economy benefits Ireg ðtÞ could be written as follow,
Ireg ðtÞ ¼ Preg ðtÞRðtÞDt þ ðsgnðSðtÞÞÞPem ðtÞRðtÞDt ð5Þ
where RðtÞ is the regulation power of aggregator participating in
ð1Þ supplementary frequency regulation, Preg ðtÞ and Pcha ðtÞ are fre-
quency regulation energy price and market electricity price at time
ð2Þ nal from grid operator.
The indices of tracking error square could be used to measure
the tracking performance of load frequency control signal from
grid. Thus, Dsig ðtÞ could be written as follow,
Dsig ðtÞ ¼ ðRðtÞ  SðtÞÞ2 ð6Þ
To simplify the calculation, the multi-objective in Eq. (4) could
be transformed into the following single objective,

4.2. EVs participating frequency regulation judgment module max UðtÞ ¼ aP reg ðtÞRopt ðtÞDt þ bðsgnðRopt ðtÞÞÞP em ðtÞRopt ðtÞDt
Ropt ðtÞ

The EVs participating frequency regulation judgment module
compares the current SOC of each EVs in aggregator with their
SOC for driving demand of EV owners. It would select which EVs
in V2G aggregator to participate in supplementary frequency reg-
ulation. Its selection condition is whether current SOC of nth EV
is above its SOC for driving demand, which is given as follows,
( charging in the process of frequency regulation, c is the weighting
n
1; ðSOC n ðtÞ > SOC nexp Þ
F ¼ ; ðn ¼ 1; 2; . . . NÞ
0; ðSOC n ðtÞ 6 SOC nexp Þ
 cP reg ðtÞðRopt ðtÞ  SðtÞÞ2 ð7Þ
where UðtÞ is the total benefits of EVs participating in supplemen-
tary frequency regulation, a is the weighting factor for income of
aggregator participating in supplementary frequency regulation, b
is the weighting factor for income of EVs charging or cost of EVs dis-
factor for tracking performance of load frequency signal from grid
ð3Þ operator.

where Fn is a flag to represent whether nth EV in aggregator partic- 4.3.2. Constraints

ipates in frequency regulation. Fn = 1 represents nth EV would par- There are two constraints in optimal regulation power calcula-
ticipate in frequency regulation, Fn = 0 represents nth EV would not tion model.
participate in frequency regulation.
(1) The satisfaction of driving demand of EV owners. When
4.3. Optimal regulation power participating frequency regulation aggregator participates in up frequency regulation, the EVs
calculation module participating in frequency would be discharged. However,
their SOC should not be below the SOC for driving demand.
Optimal regulation power participating frequency regulation
calculation module calculates the regulation power of aggregator X
N X
N
F n SOC n ðtÞ  Ropt ðtÞDt P F n SOC nexp ð8Þ
participating in supplementary frequency regulation according to
n¼1 n¼1
the SOC for driving demand of EV in aggregator, load frequency
control signal from grid operator, electricity price and frequency
regulation price. An optimal regulation power calculation model (2) The limitation of max SOC. When aggregator participates in
is used in this module and its optimization objectives and con- down frequency regulation, EVs would be charged. However,
straints are given bellow. considering on battery safety, their SOC should not be
charged above the maximum of SOC
4.3.1. Optimization objectives X
N X
N
F n SOC n ðtÞ  Ropt ðtÞDt 6 F n SOC nmax ð9Þ
The optimization objectives of optimal regulation power calcu- n¼1 n¼1
lation model are maximizing economy benefits obtained from par-
 C. Peng et al. / Applied Energy 190 (2017) 591–599 595

To sum up, the optimal regulation power calculation for aggre- (4) The limitation of EV battery charging/discharging rates, i.e.,
gator participating in frequency regulation could be given as the regulation power of each EV would be limited in its nor-
follow, mal range.
max UðtÞ ¼ aPreg ðtÞRopt ðtÞDt þ bðsgnðRopt ðtÞÞÞPem ðtÞRopt ðtÞDt F n ðSOC n ðtÞ þ Rn ðtÞDtÞ 6 F n SOC nmax ð17Þ
Ropt ðtÞ

 cðRopt ðtÞ  SðtÞÞ 2

ð10Þ To sum up, the allocated regulation power calculation model
could be given as follow,
s.t.
X
N  n 
 ðtÞSOC n ðt1Þ
X
N X
N minJðtÞ ¼ F n SOC
SOC n SOC n

F n SOC n ðtÞ  Ropt ðtÞDt P F n SOC nexp
n
ð11Þ R ðtÞ
n¼1
max exp

! ð18Þ
n¼1 n¼1 X
N X
N
jSOC n ðtÞ  SOC n ðt  DtÞj ¼ Rn ðtÞDt
X
N X
N n¼1 n¼1
F n SOC n ðtÞ  Ropt ðtÞDt 6 F n SOC nmax ð12Þ
n¼1 n¼1 s.t.

X
N
4.4. Regulation power allocation module Rn ðtÞ ¼ Ropt ðtÞ ð19Þ
n¼1
Regulation power allocation module allocates the optimal regu-
lation power to each EV in aggregator. To allocate the optimal reg- SOC nexp 6 SOC n ðtÞ  Rn ðtÞDt ð20Þ
ulation power fairly and reduce the over-discharging of one EV
battery, a calculation model based on the available capacity of EV
battery is used. Its objective function and constraints are given SOC n ðtÞ þ Rn ðtÞDt 6 SOC nmax ð21Þ
bellow.
Rnmin 6 Rn ðtÞ 6 Rnmax ð22Þ
4.4.1. Objective function
The objective of regulation power allocation calculation model
is calculating the regulation power of each EV participating in sup- 5. Simulation and performance analysis
plementary frequency regulation according to their available SOC
range and make the SOC change of EVs small as much as possible 5.1. Simulation experiments
when they participate in frequency regulation. The objective func-
tion could be given as follows, In this section, the proposed dispatching strategy will be imple-
  mented in an EV aggregator participating in supplementary fre-
XN  n n 
n SOC ðtÞ  SOC ðt  1Þ quency regulation simulation experiment.
min JðtÞ ¼ F   ð13Þ
Rn ðtÞ
n¼1
 SOC max  SOC exp 
n n
An EV aggregator with 10 EVs of three different types will be
used in the simulation experiment. The types of EVs are Mini-E,
where SOC n ðtÞ  SOC n ðt  DtÞ ¼ Rn ðtÞDt, Rn ðtÞ is the allocated regu- Nissan Leaf and i-MiEV, their battery capacities are 35 kWh [26],
lation power of nth EV participating in frequency regulation, J(t) 24 kWh [27], 16 kWh [28] respectively. The initial SOC of EV bat-
represents the SOC change of EVs participating in frequency teries are random. The expected SOCs of driving demand are calcu-
regulation. lated based on daily travel distances information of EVs from
statistical data in [29]. The detail parameters of the EVs in aggrega-
tor are shown in Table 1.
4.4.2. Constraints
In the simulation experiment, the frequency regulation price,
There are four constraints in regulation power allocation calcu-
electricity price and load frequency control signal are selected from
lation model.
New York Independent System Operator (NYISO) market in March
11, 2012. The time interval of frequency regulation DT is 15 min.
(1) The sum of regulation power of EVs participating in fre-
The frequency regulation price and electricity price are shown in
quency regulation must equal the calculated optimal regula-
Fig. 3.
tion power.
To illustrate the effectiveness of the proposed dispatching strat-
X
N egy, two dispatching strategies are compared in the simulation
F n Rn ðtÞDt ¼ Ropt ðtÞ ð14Þ experiment: (a) the proposed dispatching strategy, (b) the tradi-
n¼1 tional economic benefits based dispatching strategy [18].

(2) The satisfaction of driving demand, i.e., the battery SOC of
each EV participating in frequency regulation would not be
bellow the SOC of driving demand of EV owner.
F n
SOC nexp n n n
6 F ðSOC ðtÞ  R ðtÞDtÞ
(3) The limitation of EV battery capacity, i.e., the battery SOC of
each EV participating in frequency regulation would not be
above its maximum SOC limitation.
F n ðSOC n ðtÞ þ Rn ðtÞDtÞ 6 F n SOC nmax
5.2. Simulation results discussions
5.2.1. Load frequency control signal tracking performance
Power grid operator sends the load frequency signal control sig-
ð15Þ
nal to V2G aggregator at every 15 min. The tracking results of load
frequency signal are shown in Fig. 4. As seen in Fig. 4, dispatching
strategy (b) only considers on maximizing economic benefits of
aggregator, which leads it could not track load frequency control
signal well. The tracking error of the proposed strategy in this
paper, i.e., dispatching strategy (a) is within ±3%, which is much
ð16Þ
smaller than that of dispatching strategy (b).
596 C. Peng et al. / Applied Energy 190 (2017) 591–599

Table 1
Parameters of EVs.

EV No Capacity/kWh SOCinit SOCexp Power of charge/kW Power of regulation/kW

EV1 35 0.342 0.372 0–12 14 to 14
EV2 35 0.632 0.532 0–12 14 to 14
EV3 35 0.098 0.284 0–12 14 to 14
EV4 24 0.279 0.504 0–10 12 to 12
EV5 24 0.547 0.578 0–10 12 to 12
EV6 24 0.838 0.550 0–10 12 to 12
EV7 24 0.965 0.526 0–10 12 to 12
EV8 16 0.158 0.517 0–8 10 to 10
EV9 16 0.533 0.627 0–8 10 to 10
EV10 16 0.957 0.606 0–8 10 to 10

35
market electricity price
30 Frequency regulation price

25
Price ($/MWh)

20

15

10

0
0 1 2 3 4 5 6 7 8 9 10 11 12
Time (h)

Fig. 3. Regulation capacity price and energy price from NYISO.

40
load frequency signal
30 dispatching strategy a)
dispatching strategy b)
Regulation Power (kW)

20

10

-10

-20

-30
0 1 2 3 4 5 6 7 8 9 10 11 12
Time (h)

Fig. 4. Load frequency control signal tracking results.

5.2.2. Economic benefits
The economy benefits of V2G aggregator obtained from partic-
ipating in supplementary frequency regulation could be calculated
by following equation, which considers the benefits from partici-
pating supplementary frequency regulation, the charging benefits
from EV owners, the aggregator penalty imposed by grid operator
for load frequency control signal tracking violation [30].
X
M
In ¼ ðP reg ðt þ kDTÞRðt þ kDTÞDT þ ðsgnðRðt þ kDTÞÞPem ðt þ kDTÞ
k¼1
 Rðt þ kDTÞDT  kPreg ðt þ kDTÞjRðt þ kDTÞ  Sðt þ kDTÞjÞ
where k is the penalty coefficient of load frequency control signal
tracking violation, is selected as 0.45.
According to the load frequency signal tracking results in Fig. 4
and calculation of income in Eq. (23), the income by using the pro-
posed strategy, i.e., dispatching strategy (a) is $45.03, which is 7%
more than that obtained by dispatching strategy (b). This shows
that the economic benefits of aggregator participating in supple-
mentary frequency obtained by using the proposed dispatching
strategy is better than that obtained by using dispatching strategy
(b), when considering the frequency load control signal tracking
ð23Þ violation penalty imposed by grid operation.
 C. Peng et al. / Applied Energy 190 (2017) 591–599 597

0.8

SOC
0.6

0.4

0.2
0 1 2 3 4 5 6 7 8 9 10 11 12
Time (h)
1

0.8
SOC

0.6

0.4

0.2
0 1 2 3 4 5 6 7 8 9 10 11 12
Time (h)
1

0.8
SOC

0.6

0.4

0.2
0 1 2 3 4 5 6 7 8 9 10 11 12
Time (h)
1

0.8
SOC

0.6

0.4

0.2
0 2 4 6 8 10 12
Time (h)

Fig. 5. The simulation experiment results of EVs’ SOC curve (‘O-’ is the SOC change obtained by dispatching strategy a), ‘⁄-’ is the SOC change obtained by dispatching strategy
b), ‘+-’ is SOC of EV owner driving demand).

5.2.3. Battery SOC and driving demand of EVs obvious differences in SOC maximums of EVs by using these two
The maximum, minimum and change range of EVs’ battery dispatching strategies. The SOC minimums of all EVs are above
SOCs during aggregator participating in supplementary frequency 50% by using the proposed dispatching strategy. By using dispatch-
regulation are shown in Table 2. As seen in Table 2, there is no ing strategy (b), the SOC minimums of EVs are between 26.37% and
598 C. Peng et al. / Applied Energy 190 (2017) 591–599

Table 2
Comparison of EVs’ SOC change.

EV No SOC maximum SOC minimum SOC change range

(a) (b) (a) (b) (a) (b)
EV1 93.63% 91.35% 55.32% 29.16% 38.31% 62.19%
EV2 95.00% 93.50% 53.18% 26.37% 41.80% 67.13%
EV3 90.60% 91.28% 56.33% 35.82% 34.27% 55.46%
EV4 86.55% 89.64% 51.79% 42.86% 34.76% 46.78%
EV5 91.52%. 90.92% 52.48% 38.90% 39.04% 52.02%
EV6 93.28% 88.49% 50.77% 33.65% 42.51% 54.84%
EV7 94.52% 94.70% 52.57% 26.61% 41.95% 68.09%
EV8 92.47% 91.31% 50.39%. 28.25%. 42.08% 63.06%
EV9 88.65% 83.42% 58.63% 39.49% 30.20% 43.93%
EV10 94.28% 94.24% 60.55% 26.58% 33.73% 67.66%

42.86%. The SOC change ranges by using the proposed dispatching
strategy are between 41.80% and 30.20%. It is obvious that the SOC
change ranges by using the proposed dispatching strategy are less
than the those by using dispatching strategy (b) during the process
of aggregator participating in supplementary frequency regulation.
The proposed dispatching strategy could reduce the SOC change
ranges of EVs and avoid over-discharging when aggregator partic-
ipating in supplementary frequency regulation.
The simulation experiment results of EV battery SOC change
curve of four randomly selected EVs are shown in Fig. 5. As seen
in Fig. 5, the SOCs of four EVs by using strategy (a) are above their
expected SOC for driving demand in most time. However, by using
strategy (b), the SOCs are below the expected SOC for driving
demand in some hours. Especially, at times of 2nd to 5th hour
and 8th hour to 11th hour, when aggregator participates in up fre-
quency regulation.
To sum up, by using strategy (a), the SOCs of EVs could satisfy
the driving demand of EV owner during the most time of process
of aggregator participating in frequency regulation. By using strat-
egy (b), the SOCs of EVs could not satisfy the driving demand of EV
owner during the near half time of process of aggregator partici-
pating in frequency regulation.
6. Conclusion
To satisfy the driving demand of EV owners and maximize the
economic benefits of EV aggregator obtained from participating
in frequency regulation service simultaneously, a novel dispatch-
ing strategy for V2G aggregator is proposed in this paper. In the
proposed strategy, a driving demand calculation module is
designed to calculate the battery SOC for EV owner’s driving
demand. An optimal regulation power participating frequency reg-
ulation calculation module and a regulation power allocation mod-
ule are built to optimize the economic benefits of aggregator and
aggregator’s tracking performance of load frequency control signal
from grid operator. Finally, the simulation experiment is conducted
to demonstrate the effectiveness of the proposed dispatching strat-
egy. The simulation results show that the proposed dispatching
strategy not only could optimize the economy benefits of aggrega-
tor participating in supplementary frequency regulation, but also
make the battery SOC of EVs in aggregator satisfying the driving
demand of EV owners and avoid over-discharging of EVs.
References
[1] Siang FT, Chee WT. A review of energy source and energy management system
in electric vehicles. Renew Sustain Energy Rev 2013;2:82–102.
[2] Richardson DB. Electric vehicles and the electric grid: a review of modeling
approaches, impacts and renewable energy integration. Renew Sustain Energy
Rev 2014;19:247–54.
[3] Tan KM, Ramachandaramurthy VK, Yong JY. Integration of electric vehicles in
smart grid: a review on vehicle to grid technologies and optimization
techniques. Renew Sustain Energy Rev 2016;55:720–32.
[4] Jin Z, Lina H, et al. Coordinated control for large-scale EV charging facilities and
energy storage devices participating in frequency regulation. Appl Energy
2014;123(15):253–62.
[5] Fan H, Jiang L, Zhang CK, et al. Frequency regulation of multi-area power
systems with plug-in electric vehicles considering communication delays. IET
Gener Transm Distrib 2016;10(14):3481–91.
[6] Rogersk M, Kiump R, Khurana H, et al. An authenticated control framework for
distributed voltage support on the smart grid. IEEE Trans Smart Grids 2010;1
(1):40–7.
[7] Pavic I, Capuder T, Kuzle I. Value of flexible electric vehicles in providing
spinning reserve services, vol. 157; 2015. p. 60–74.
[8] Aziz M, Oda T, Mitani T, Watanabe Y, Kashiwagi T. Utilization of electric
vehicles and their used batteries for peak-load shifting. Energies 2015;8
(5):3720–38.
[9] Lam AYS, Leung KC, Li OK. Capacity estimation for vehicle-to-grid frequency
regulation services with smart charging mechanism. IEEE Trans Smart Grid
2016;7(1):156–66.
[10] Fan H, Jiang L, Zhang CK, et al. Frequency regulation of multi-area power
systems with plug-in electric vehicles considering communication delays. IET
Gener Transm Distrib 2016:1–11.
[11] Masuta T, Yokoyama A. Supplementary load frequency control by use of a
number of both electric vehicles and heat pump water heaters. IEEE Trans
Smart Grid 2012;3(3):1253–62.
[12] Ko K, Han S, Sung DK. A new mileage payment for EV aggregators with varying
delays in frequency regulation service. IEEE Trans Smart Grid 2016;PP
(99):1–8.
[13] Wu C, Mohsenian-Rad H, Huang J. Vehicle-to-aggregator interaction game.
IEEE Trans Smart Grid 2012;3(1):434–42.
[14] Liu H, Hu Z, Song Y, et al. Vehicle-to-Grid control for supplementary frequency
regulation considering charging demands. IEEE Trans Power Syst 2015;30
(6):3110–9.
[15] Wang R, Li Y, Wang P, Niyato D. Design of a V2G aggregator to optimize PHEV
charging and frequency regulation control. In: IEEE international conference
on smart grid communications, Vancouver, Canada; 2013. p. 127–32.
[16] Karfopoulos EL, Panourgias KA, Hatziargyriou ND. Distributed coordination of
electric vehicles providing V2G regulation services. IEEE Trans Power Syst
2016;31(4):2834–46.
[17] Shi W, Wong VWS. Real-time vehicle-to-grid control algorithm under price
uncertainty. In: IEEE international conference on smart grid communication,
Brussels, Belgium; 2011. p. 261–6.
[18] Sekyung H, Soohee H. Development of an optimal vehicle-to-grid aggregator
for frequency regulation. IEEE Trans Smart Grid 2010;1(1):65–72.
[19] Lee J, Young M, Kim Y, et al. Optimal coordination of charging and frequency
regulation for electric vehicle aggregator using least square Monte-Carlo
(LSMC) with modeling of electricity price uncertainty. J Electr Eng Technol
2013;8(6):1269–75.
[20] Sortomme E, El-sharkawi MA. Optimal scheduling of vehicle-to-grid energy
and ancillary services. IEEE Trans Smart Grid 2012;3(1):351–9.
[21] Mojdehi MN, Ghosh P. An approach to engage EVs for frequency regulation:
aggregator perspective. In: IEEE conference on technologies for sustainability,
Ogden, USA; 2015. p. 229–33.
[22] Sun S, Dong M, Liang B. Real-time welfare-maximizing regulation allocation in
dynamic aggregator-EVs system. IEEE Trans Smart Grid 2014;5(3):1397–409.
[23] Escudero-Garzas JJ, Garcia-Armada A, Seco-Granados G. Fair design of plug-in
electric vehicles aggregator for V2G regulation. IEEE Trans Veh Technol
2012;61(8):3406–19.
[24] Donadee J, Ilic M. Stochastic optimization of grid to vehicle frequency
regulation capacity bids. IEEE Trans Smart Grid 2014;5(2):1061–9.
[25] Yao E, Wong VWS, Schober R. Robust frequency regulation capacity scheduling
algorithm for electric vehicles. IEEE Trans Smart Grid 2016;pp(99):1–14.
[26] Berdichevsky G, Kelty K, Straubel JB, Toomre E. Tesla motors, tesla roadster
battery system; August 2006. Available: <http://www.teslamotors.com/
display_data/TeslaRoadsterBatterySystem.pdf>.
 C. Peng et al. / Applied Energy 190 (2017) 591–599 599

[27] Nissan Zero Emission Website. Leaf specs; May 1, 2010. Available: <http:// Liying Li received Ph.D. degree in Automation from
www.nissan-zeroemission.com/EN/LEAF/specs.html>. University of Electronic Science and Technology of
[28] Mitsubishi Motors. Mitsubishi motors to bring new-generation EV i-MiEV to China in 2010, respectively. Currently, she is an associ-
market; June 2009. Available: <http://media.mit-subishimotors.com/ ate professor with school of automation engineering in
pressrelease/e/products/detail1940.html>. Electronic Science and Technology of China. Her
[29] National Statistics. Department for Transport, Transport Statistics Bulletin- research interests are smart grid and its communication
National Travel Survey: 2008; April 9, 2009. technologies.
[30] Han S, Han S, Sezaki K. Estimation of achievable power capacity from plug-in
electric vehicles for V2G frequency regulation: case studies for market
participation. IEEE Trans Smart Grid 2011;2(4):632–41.

Chao Peng received the M.S. degree and Ph.D. degree in

Automation from University of Electronic Science and
Technology of China in 2007 and 2012, respectively.
Currently, he is a lecturer with school of automation
engineering in Electronic Science and Technology of
China. His research interests are smart grid, Intelligent
information processing and control technologies.
Lian Lian received the B.S. degree in Automatic Control
from Southeast University, China, in 2012. She is cur-
rently work toward the M.S. degree in school of
automation engineering in Electronic Science and
Technology of China. Her research interests are smart
grid and charging and discharging technology for elec-
tric vehicle.

Jianxiao Zou received the M.S. degree and Ph.D. degree

in Automation from University of Electronic Science and
Technology of China in 2003 and 2009, respectively.
Currently, he is a professor with school of automation
engineering in Electronic Science and Technology of
China. His research interests are smart grid and
renewable power generation technologies.