CHEMICAL ENGINEERING TRANSACTIONS  
 

VOL. 51, 2016 

A publication of 

 
The Italian Association 

of Chemical Engineering 
Online at www.aidic.it/cet 

Guest Editors: Tichun Wang, Hongyang Zhang, Lei Tian 
Copyright © 2016, AIDIC Servizi S.r.l.,  
ISBN 978-88-95608-43-3; ISSN 2283-9216 

Coordinated Charging Optimization Strategy of Electric 
Vehicles 

Xingping Liu*a,d, Xiangyun Luob, Weidong Lic, Shijun Lid, Haoming Yua 
a College of Electrical & Information Engineering; Hunan Institute of Engineering; Hunan Provincial Key Laboratory of Wind 
Generator and Its Control, Xiangtan 411101, China 
b Hunan Mass Media Vocational Technical college, Changsha 410100, China 
c Powerchina Zhongnan Engineering Co. Ltd., Changsha 410014, China 
d Hunan province Cooperative Innovation Center for Wind Power Equipment and Energy Conversion, Xiangtan 411101, China 
B12090018@hnu.edu.cn  

The large-scale uncoordinated charging threatens the safety operation of power grid. In the paper, two 
coordinated charging modes were proposed: automatic on-off charging mode and the smoothly adjusting 
charging mode. The 2 coordinated charging modes were established based on existing power distribution 
networks in residential areas and the constraint conditions of the distribution transformer and circuit capacity in 
order to transmit maximum electric power to electric vehicles. The established mode formulas were solved with 
stochastic simulation and modified particle swarm algorithm. Taking the charging station in one residential area 
in Shenzhen as an example, we performed the simulation of vehicle charging according to the uncoordinated 
charging mode and the 2 coordinated charging modes. We compared the voltage distribution in the three modes 
respectively. Based on the computation of the profit amount and the load volatility, we proposed the concept of 
cost saving so as to compensate for the expense of involved users. The simulation results showed that the two 
coordinated modes could raise user satisfaction and promote the smooth operation of power grid. Charging 
management system can choose one operation mode as required. 

1. Introduction 

In China, electric vehicles include private cars, buses, taxis and official cars, most of the parked cars in the 
residential areas are private cars. In the coordinated charging optimization mode of electric vehicles, buses, 
taxies, or official cars were not considered. Private cars are often used in the day for work or entertainment and 
parked in residential areas for a short time in the day and a longer time in the night. Private cars allow a slow 
charging speed. Therefore, it is necessary to optimize the coordinated charging mode according to the night 
charging requirements (Rotering and Ilic, 2011). 

2. Analysis of charging modes 

2.1 Automatic on-off charging mode 

In the automatic on-off charging mode, one or several electric cars automatically start or terminate the charging 
process at any time according to the procedure. When the load is approaching the rated capacity of the 
distribution transformer, the charging management system will automatically stop the charging process of one or 
several electric cars; when the common load continues to fall, the management system will automatically 
connect one or several electric cars into the grid network for charging (Cao et.al., 2012). 

2.2 Smoothly adjusting charging mode 
In the smoothly adjusting charging mode, at any time, the charging power of the off-board charger is adjusted 
continuously and automatically. The rated charging power of the battery charger is generally 4 kW , but the 
charging power will continuously vary between 0 and 4 kW to charge the electric vehicles in the smoothly 
adjusting charging mode (Tian et.al., 2013). 

                               
 
 

 

 
   

                                                  
DOI: 10.3303/CET1651205

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Please cite this article as: Liu X.P., Luo X.Y., Li W.D., Li S.J., Yu H.M., 2016, Coordinated charging optimization strategy of electric vehicles, 
Chemical Engineering Transactions, 51, 1225-1230  DOI:10.3303/CET1651205   

1225

mailto:411101,China,3260629090@qq.com


3. Coordinated charging optimization mode 

3.1 Coordinated charging model in the automatic on-off charging mode 
According to the historical average load in the power distribution network in the residential area, the normal load 
curve of a certain day can be predicted with the 96-point daily load curve forecasting method with the interval of 
15 min (Clement et.al., 2010). If the number of the alternating charging points in the residential area is N, 
according to the slow charging method, the battery charger charges the lithium battery under a constant power. 
The model aims to realize the maximum transmission energy to electric vehicles:  

1 1

max ( )
T N

EV i

t i

P x t t
 


       (1) 

Where PEV is the rated charging power of an electric vehicle; in the period t, when the ith charging point is 
disconnected with the electric car or the charging is completed, xi(t)=0, when the ith charging point is connected 
with the electric car and the charging is not completed, xi(t)=1 ; △t is the length of the charging period.

 
3.1.1 Constraint conditions 
3.1.1.1 Constraints of the voltage range of the residential area 

The load of connected electric cars will cause the voltage reduction of the distribution network node, which is 
determined by the access point of electric cars and the charging power. (Sortomme et.al., 2011). The rated 
voltage range of the charging point is:  

min max
 

i
V V V

      (2) 

Where Vi is the voltage of the ith charging point; Vmax and Vmin are the maximum and minimum voltage of the 
power distribution network, respectively.  
3.1.1.2 Constraints of the transformer capacity 

In the period t, the sum of the average load of the residential area and the charging rated load should not bigger 
than the rated output power of the transformer:  

1

( ) ( ) cos
N

EV i N N

i

P x t P t S


   
      (3) 

Where P(t) is the average load of the residential area; SN is the rated capacity of the transformer; cosN is the 
rated power factor of the transformer and generally set as 0.95;  is the load rate of the transformer and it is 
determined by the type and the model of the transformer and assumed to be between 30% and 65%. 
3.1.1.3 Constraints of the charging power 

To avoid the large load fluctuations caused by the sudden access or removal of electric cars, the variation of 
the charging power in the charging station should be restricted in sequential charging periods(Peter et al., 
2012). 

0

1 1

( ) ( 1)
N N

EV i EV i

i i

P x t P x t P
 

    
       (4)

 

Where △P0 is the charging power range between the current charging period and the former charging period 

and defined as 20 kW in this paper. 
3.1.1.4 Constraints of the charging capacity 

,0

1

( ) (1 )
T

EV i i i

t

P x t t SOC


   
       (5) 

Where SOCi,0 is the initial load state of the ith electric car; I i is the battery capacity of the ith electric car. 
3.1.1.5 Constraints of charge state continuity 

, , 1

( )
EV i

i t i t

i

P x t t
SOC SOC




 


      (6) 

1226



where SOCi,t is the charge state of the ith electric car in the end of the charging period t; SOCi,t-1 is the charge 
state of the ith electric car in the end of the charging period (t-1). 
3.1.1.6 Constraints of charging protection 

The management system will monitor the EMS of the electric car in the charging state in the period △t to avoid 
the overcharge phenomenon. 

,
(1 )

EV i t i
P t SOC   

      (7) 

3.1.1.7 Constraints of the circuit heat load 

The heat load of the network element refers to the ratio of the apparent power of the element to its rated power. 
The heat load considered in this paper refers to the heat load of the connecting circuit between transformers:  

maxMC MC
L L

                    (8) 
 

LMC means the heat load of the circuit; LMCmax means the maximum heat load. 
3.2 Coordinated charging model in the smoothly adjusting charging mode 
In order to realize the more balanced distribution of the charging power of electric vehicles we improved the 
objective function mentioned above by introducing SOC factor(H et al., 2012; Su et al., 2013). In the smoothly 
adjusting charging mode, supposing that the charging power can be continuously adjusted, the maximum 
charging energy transmitted to electric vehicles is:  

,

1 1

max (1 ) ( ) ( )
i

T N

i t EV i

t i

SOC P t x t t
 

 
       (9) 

where PMCi(t) is the charging power of the electric vehicle in the ith charging point 

3.2.1 Constraint conditions are respectively expressed below 
The charging power of the electric vehicles should not surpass the rated output charging power 

0 ( )
iEV EV

P t P 
     (10) 

According to the current battery technology, in order to avoid the wide fluctuation of the charging power, the 
constraint of the charging power variation range is:  

0 0
( 1) ( ) ( 1)

i i iEV EV EV
P t p P t P t p     

     (11) 

where △ p0 is the maximum variation range of the charging power of the electric vehicles and the 
access/removal shifting of the charging load of electric vehicle. It is generally defined as 0.5 KW.

  Constraints of the voltage range of the residential area are similar with Eq. (2). 
In the period t, the sum of the average load of the residential area and the charging load should not surpass the 
rated output power of the transformer:  

1

( ) ( ) ( ) cos
i

N

EV i N N

i

P t x t P t S


   
      (12) 

Constraints of charging capacity:  

,0

1

( ) ( ) (1 )
i

T

EV i i i

t

P t x t t SOC


   
     (13) 

Constraints of charge state continuity 

, , 1

( ) ( )
iEV i

i t i t

i

P t x t t
SOC SOC




 

        
(14) 

Because the objective function has introduced the SOC factor, the overcharge phenomenon will not come out 
when charging the high SOC battery with a low charging power. Thus, the overcharge condition should be 
neglected. The computing method of the constraints of the circuit heat load is the same to Eq. (8) 

1227



3.3 Coordinated charging control algorithm 
Combined with the stochastic simulation, the modified particle swarm optimization (PSO) method was adopted 
to solve maximum electric energy transmitted to electric cars in the paper(Zhan. et al., 2012). Taking the 
advantage of randomness and periodicity to generate the initial particle, the search diversity of the algorithm can 
be strengthened. Through the introduction of the algorithm stagnation mutation mechanism, the probability to 
realize global optimization instead of local optimization will be increased, thus avoiding the premature 
convergence of the algorithm(Yang et al., 2013). In addition, the dynamic inertia factor W is introduced to 
balance global search ability and local search ability of the algorithm in the iteration process:  

' max min

max

max

w w
w w t

T


 

      (15) 

where wmax and wmin are the maximum and minimum dynamic inertia factors, respectively. Generally, wmax=0.9, 
wmin =0.9; wmin =0.4; t’ and Tmax are respectively the present iterations and the maximum iterations. 
In this paper, only the charging period of 14 h from 16: 00 to 6: 00 in the next day was considered. 

4. Simulation 

4.1 Power distribution network in residential areas 
Taking a residential area in Shenzhen as an example, the model parameters are provided as follows: 2 
distribution transformers with the capacity of 630 kVA, the no-load voltage ratio of 10/0.4 kV, 2.8-km long 
three-phase line, 1.2km long single cable, and 500 domestic households. The voltage fluctuation range of the 
low-voltage distribution network is -10%~+10% (220 V/380 V). 
It is assumed that there are 60 charging points and 100 electric cars in the residential area. According to the 
normal load distribution, we distributed the charging points into Districts A and B in an unequal way. When all 
the charging points are connected to the electric cars, other electric cars wait for charging. The battery capacity 
of an electric car is 20 kWh. In addition, the car should be fully charged before 6: 00 in the next morning. 

4.2 Comparison of the 3 modes 

4.2.1 Uncoordinated charging mode of electric vehicles 

In the uncoordinated charging mode, the stochastic simulation of the arrival time and SOC of each car was 
performed through the Monte-Carlo method to start the charging process. In the uncoordinated charging 
process, when a car is connected to the charging point, it will be charged at the rated power until the full charging 
process is completed. In the calculating calculation of the average load, the load curve is prone to be gathered in 
the peak time, as shown in Figure 1. In the evening, electric car users will start to charge cars as soon as they 
arrive at the residential area. If the number of charging points is not enough, partial electric cars will wait for 
charging. If one transformer is in an overhaul, the other transformer works alone. In the period (21: 45-22: 30), 
the transformer will be seriously overcharged and the circuit will be overheated. The overheat phenomenon is 
not permitted. 

16:0017:0018:0019:0020:0021:0022:0023:0000:001:00 2:00 3:00 4:00 5:00 6:00
0

100

200

300

400

500

600

700

800

time

lo
a

d
/k

W

 

total load

Charging 

load

User 

load

 

Figure 1: Load curves of the uncoordinated charging mode. 

4.2.2 Automatic on-off charging mode 
Considering the travel habits of users in the residential area, in this paper we only studied the charging time in 
the night from 16: 00 to 6: 00 in the next day. In the automatic on-off charging mode, the management system 
will automatically control the switches of the battery charger to realize the safety operation of the grid power 
according to the charging requirements. 

1228



As shown in Figure 2, the switches of the car will be shut down at the peak time and reconnected at the low-load 
time, thus realizing the purpose of minimizing the charging cost and the gap between peak and valley time by 
peak load shifting. 
 
 

16:00 17:00 18:00 19:00 20:00 21:00 22:00 23:00 00:00 1:00 2:00 3:00 4:00 5:00 6:00
0

100

200

300

400

500

600

Time

L
o

a
d

/k
W

 

  On-off charging mode 
automatically

The charging mode of 

smooth adjustmentTotal 

load

Charging 

load

User load

 

Figure 2: Load curves of the coordinated charging mode. 

4.2.3 Smoothly adjusting charging mode 
In this mode, the charging power of electric cars can be adjusted continuously. We have collected the data every 
15 min. The 12th car starts to charge at 19: 30 and its charging power is rising. However, at the peak time, the 
power decreases to 0. Then, the car is reconnected at 23: 00 and the power rises to the highest power (3.5 kw) 
until 00: 30. The full charging process is completed at 2: 30. 

5. Results and analysis 

Table 1: Parameter settings of energy prices 

Charging periods TOU price (Yuan/kWh) uniform power pricing (Yuan/kWh) 
16: 00-17: 00 0.687 

1.2 
17: 00-21: 00 0.869 
21: 00-00: 00 0.687 
00: 00-06: 00 0.365 

Table 2: Simulation results of coordinated and uncoordinated charging modes 

Charging modes 
Profit 

amount 
Load 

volatility 
Total maximum 

load 
The end time of 

charging 

The 
end charged 

state 
Uncoordinated charging 

mode 
334.7870 9.0413 780 3: 00 100% 

Automatic on-off charging 
mode 

523.1480 4.3014 550 6: 00 100% 

Smoothly adjusting 
charging mode 

576.6970 4.2176 550 6: 00 100% 

50% automatic on-off 
charging mode 

421.6492 4.9947 650 6: 00 100% 

50% smoothly adjusting 
charging mode 

465.1440 6.0316 650 6: 00 100% 

 
Power grid sells the electricity in the form of TOU electric price, and the battery charger management sectors 
develop the uniform power pricing system. The parameter settings of energy prices are provided in Table 1. 
According to the simulating calculation with Monte-Carlo method, total charging electric energy of all the electric 
cars is 1442 kWh. The simulating results of the coordinated charging mode and un-coordinated charging mode 
are provided in Table 2. 

1229



According to the results in Table 2, the coordinated charging mode can increase the profit amount of the 
charging stations in the residential area. However, the charging stations in the residential area are not aimed at 
making profit. After deducting the capital cost of the stations, the rest are assumed as the saving cost, which will 
be paid to the users so as to raise the customer response coefficient. The cost in the two modes mentioned in 
this paper is lower. Therefore, they will attract more users of electric vehicles to participate in the coordinated 
charging mode. 
For the electric power sector, the coordinated charging mode is helpful to reduce the volatility of load curve; for 
the electric car users, the coordinated charging mode is helpful to obtain additional compensation and reduce 
the charging cost. After the electric vehicles are widely used, the charging station with the coordinated charging 
mode will be a mutually beneficial strategy. 

6. Results and analysis 

Taking a residential area in Shenzhen as an example, we compared the two coordinated charging modes with 
the uncoordinated charging mode and found that coordinated charging modes could be directly adopted in the 
original power distribution network without the upgrade or transformation of the power grid infrastructures, thus 
reducing the project cost. 
The smoothly adjusting charging mode has a higher profit potential and the smoother load curve. The automatic 
on-off charging mode is simple and reliable. The two modes can both realize the minimum charging cost and the 
smallest gap between peak and valley time by peak load shifting. If the mode is promoted in a large scale, its 
economic profit will be considerable. 

Reference 

Cao Y., Tang S., Li C., 2012, An optimized EV charging model considering TOU price and SOC curve, IEEE 
Trans. on Smart Grid, 3(1): 388-393. 

Clement N.K., Haesen E., Driesen J., 2010, The impact of charging plug-in hybrid electric vehicles on a 
residential distribution grid, IEEE Trans. on Power Systems, 25(1): 371-380. 

Hu Z.C., Song Y.H., Xu Z.W., 2012, Impacts and utilization of electric vehicles integration into power systems, 
Proceedings of the CSEE, 32(4): 1-8. 

Peter R., Damian F., Andrew K., 2012, Optimal charging of electric vehicles in low-voltage distribution systems, 
IEEE Trans. on Power Systems, 27(1): 268-279. 

Rotering N., Ilic M., 2011, Optimal charge control of plug-in hybrid electric vehicles in deregulated electricity 
markers, IEEE Trans. on Power Systems, 26(3): 1021-1029. 

Sortomme E., Hindi M.M., Macpherson S.D.J., 2011, Coordinated charging of plug-in hybrid electric vehicles to 
minimize distribution system losses, IEEE Trans. on Smart Grid, 2(1): 198-205. Doi: 
10.1109/TSG.2010.2090913 

Sun X M., Wang W., Su S., 2013, Coordinated charging strategy for electric vehicles based on time-of-use price, 
Automation of electric power systems, 37(1): 191-195. 

Tian W.Q., He J.H., Jiang J.C., 2013, Electric vehicle charging load spatial allocation optimization algorithm, 
Transactions of China Electrotechnical Society, 28(3): 269-276. 

Yang B., Wang L.F., Liao C.L., 2013, Research on power-charging demand of large-scale electric vehicles and 
its impacting factors, Transactions of China Electrotechnical Society, 28(2): 22-27. 

Zhan K.Q., Song Y.H., Hu Z.H., 2012, Coordination of electric vehicle charging to minimize active power losses, 
Proceedings of the CSEE, 2012, 32(31): 11-17. 

1230