A Simulated System of Battery-Management-System to Test Electric Vehicles Charger Xiangwu Yan, Wei Li, Jiancheng Gu and Xiangning Xiao State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources North China Electric Power University Beijing 102206, China; School of Electrical and Electronic Engineering North China Electric Power University Baoding 071003, Hebei, China.
[email protected],
[email protected] Abstract-This paper presents the implementation of a Simulated System of Battery-Management-System (SS-BMS) based on a computer with CAN communication interface to test the Electric Vehicles (EV) Charger. The SS-BMS includes the mathematical models to simulate the electrical and thermal properties of power battery, provides users with defining a variety of mode to simulate different charging circumstances and real-time information, such as single voltage, single temperature and state of charge (SOC) estimation by taking real-time charging current into account. With the SS-BMS, it is able to make the testing process more controllable and the results more reliable.
designed to meet the charging technical conditions and can provide different operating environments. In detail, the SSBMS has functions that a BMS needs in charging, as battery protection, charge control, thermal management, state of charge (SOC) estimation and CAN communication, and during charging, it can model the electrical and thermal properties of various power batteries. Because of those attributes, the SSBMS can make the testing environment diverse and let the chargers be tested in nearly actual operational situation.
Keywords-component; Simulated System; BMS; Battery Model; Electric Vehicles Charger; CAN
Fig. 1 shows the structure of the SS-BMS. It consists of a man-machine interaction, a central control unit, a measurement block, a CAN communication system and a power battery model.
I.
INTRODUCTION
Recent years, electric vehicle charging station systems are developing in some countries and regions [1, 2], and efficient, low-distortion chargers have been designed or are being designed [3-5]. However, the EV chargers have to satisfy common requirements and cause less negative impact on power quality, such as high harmonics [6, 7], thus the test of EV chargers is indispensable during the design and production of chargers. Especially, in the period that the standards of EV charger have been developing, not only on a global scale, but also on most of developed countries, the high reliability testing work is more important to evaluate the technical solutions of chargers and to carry the perfect global standards off. Therefore, this paper introduces the SS-BMS as a load simulator for the chargers for testing the EV charger, and making the testing platform more universal, intelligent, controllable and multi-functional. At present, the load used to test the EV chargers can be classified into two categories. One is the actual battery pack, usually applied to test certain features or verify the design methods of chargers [4, 8]. The other is the load simulator, mostly used to test the performances of chargers under different operation conditions [9, 10]. But one certain power battery can not match the wide range of testing requirements, and the load simulator can not do the same because of lack of a full-featured simulation system to be coordinated with. According to the standards about EV charger’s technology [11, 12], the SS-BMS is
This research is sponsored by National High-tech R&D Program (863 Program) of China (2011AA11A279) and Key Project of the National Research Program of China (2011BAG02B14).
II.
STRUCTURE OF THE SS-BMS
CAN-bus CAN Communication
EV Charger
Man-machine Interaction Central Control Unit
DC
I Ck & U Ck Measurement Block
Battery Experiment Data Mathematical Model Relationship Chart Thermal Model
Load Power Battery Model
Simulated System of BMS
Figure 1. Structure of the SS-BMS (in the red rectangle)
The man-machine interaction is used to configure the SSBMS and show the real-time information of charging and the status of battery model. According to the configuration, the central control unit can regulate the charging process with the CAN communication interface. In addition, this system using its measurement block to monitor the electrical parameters of charging ensures that they are safe and accurate, and these realtime parameters are the basis for the power battery model to simulate the electrical and thermal properties of the power battery. And as the electric vehicle off-board charger communication protocols in China [12] and the on-board
charger communication protocols of several major manufacturers are embedded in the SS-BMS, this system is compatible with the large part of the charger types in China. The details of the SS-BMS’ working principle are described as following. III.
OPERATION OF THE SS-BMS
Start
Setting the parameters of battery and system operation
SOCk 1 SOCk
1 ek I Ck t 3600CN
Where the subscript k is the discrete-time index, CN (Ah) is the rated capacity, I C (A) is the real-time charge current obtained by measurement block, t (s) is the time interval.
Successful prepare for charging? YES
B. Battery Resistance Calculation Battery internal resistance is calculated using amendment parameters. First, on the base of the experiment database and the relationship between the internal resistance and related state parameters, such as rated capacity, SOC, temperature and charge current of the adoptive battery, determine an initial value of the internal resistance of this kind of battery. After that, according to the real-time state of the battery, the amendment parameters can be estimated by linear interpolation method, and then the value of internal resistance will be obtained.
Computing state variables of battery
Sending battery information
Collecting the charging data from both the charger and battery model
Meet one or more of charging end conditions
A. SOC Estimation Ampere hour counting is more preferable in our conditions. And in order to obtain more accurate result, the equivalent coulombic efficiency e [16] is employed in our method. It is shown as
Connecting and charge configuring
NO
mode or the first step of two-step mode, it is called I-Status. On the other hand, when charging in constant voltage mode or the second step of two-step mode, it is called V-Status. Now, it is assumed that the charger runs in I-Status, and the V-Status will be described later. The methods used to model the responses to charging are listed below:
NO
YES END
Figure 2. Flowchart for the operation of the SS-BMS
The operation of the SS-BMS is shown in Fig. 2. This process can be divided into three segments: setting parameters, preparing charge and simulating charging. First, the parameters of power battery should be set to determine a power battery model, and it is also necessary to choose the charge mode and set limit conditions, such as the maximum permissible charging current, the operating voltage range, the operating temperature range, the desired SOC, and so on. Second, the SS-BMS sends the mode and requirement of charge to the charger to let it prepare for charging. The last and the most important part of the whole process is the simulating charging of SS-BMS. Before discussion, we define two statuses of the charger. When the charger is charging in constant current mode, pulse charge
Moreover, the initial internal resistance of each single cell can be modified, so that the SS-BMS can simulate some abnormal states like the excess temperature or over-voltage of single cell.
C. Temperature Calculation The thermal model used in this paper is a classic model presented by Noboru Sato [17]. He considered total heat generation in the battery reaction Qt (kJ/s), mainly comprised reaction heat value Qr (kJ/s), polarization heat value Q p (kJ/s), which is the energy loss due to electrochemical polarization of the battery, and Joule heat QJ (kJ/s) of the electrical resistance component. It is therefore expressed with (2).
Qt Qr Q p QJ
If the total generated heat in the electrochemical reaction of both electrodes is Q1 (kJ/mol), and battery charge current is I C (A), then the reaction heat value Qr is expressed in (3) for the charging process.
Qr
Q1 I C 1.036 105 Q1 I C F
(U Ck ( K 0 I Ck
K1 K 2 SOCk SOCk
K 3 ln( SOCk ) K 4 ln(1 SOCk ))) / Rink
where, F is Faraday constant. Based on the above and using (2) and (3), charge heat balance per unit time is expressed with (4).
Qt 1.036 105 Q1 I C 1 10 3 I C2 R p 1 103 I C2 Re 1.036 105 Q1 I C 1 103 I C2 Rin
where R p (Ω) and Re (Ω) respectively represent the polarization resistance component and the electrical resistance component. The sum of both can be expressed with the internal resistance Rin (Ω). Accordingly, the temperature can be computed with (5).
Q Tk 1 Tk tk Trk t , Trk f (Tk , Tamb ) C
where C (kJ/K) is the battery heat capacity, Tr (K) is caused by heat emission and can be expressed as a user-defined function f .,. of the battery temperature T (K) and the ambient temperature Tamb (K), and it is different depending on the cooling mode. D. Voltage Calculation A “combined model”, introduced by Gregory L. Plett [1315], has been chosen here. It is shown as
K1 K 2 SOCk SOCk K 3 ln( SOCk ) K 4 ln(1 SOCk )
U c ( k 1) Rink I Ck K 0
where U C (V) is the terminal voltage of cell, K i is fitting constants from experimental data of the relationship between the open voltage and SOC of battery. However, equation (6) is a little different from the one of Gregory L. Plett’s. Here, we used the state parameters in tk to estimate the voltage in next time interval. Though there is a delay of one time interval, it does not influence the whole electrical property of our battery model. When it comes to V-Status, because the mathematical models above use the charge current as the independent variable, it is necessary to convert the charge voltage to the charge current when charged in V-Status. This conversion is as follow:
is the estimate of charge current and replaces where I Ck is calculated by battery insistence in the I Ck in I-Status, Rink last interval added an estimated increment R .
Moreover, using (6) and (7), the charge current of cell can be expressed with (9).
Rin ( k 1) R Rink
K I (U (K 1 K SOC c(k 1) c(k 1) k 0 SOC 2 k K ln(SOC )K ln(1SOC )))/ R 3 k 4 k in(k 1) R ink I Ck R in(k 1)
The charge current of the battery pack can be got by calculating the average of each cell’s charge current. Throughout the whole process of this simulation, the state variables of the battery model are derived from the real-time charge current in I-Status and the real-time charge voltage in V-Status. It is necessary to make the battery’s responses more realistic. In each time interval, furthermore, the central control unit checks the charger’s information received by CAN communication in order to ensure that the charger’s status is normal and its output can meet the requirements sent by SSBMS in last time interval. After that, the central control will check the states of the battery to define the charge requirements of the next time interval, and then send them to the charger along with the state information of the battery. If there are one or more restrictions satisfied, the SS-BMS will send the termination message to the charger and give an alarm signal, or trigger other defined events. During the process of charging, a single running loop will cost one second, which is different from the time interval t . The later is only used to calculate the charge responses of the battery. When t is greater than one second, the simulation time will much less than the charging time of the actual battery. IV.
APPLICATION EXAMPLE
This section describes an example to show several functions of the SS-BMS. The test object, in this example, is a 60kW charger (DC 60kW/500V), which is comprised of six modules in parallel form, the module type is SEVR10000S500 (DC 10kW/500V), which is manufactured by Haowen electronics company of Shenzhen in China. The selected simulation object is a LiFePO4 battery pack including four battery models, and each model consists of 24 single cells in series. The capacity of battery is 200 Ah, and the rated voltage 320V. The other physical parameters refer to the battery information of manufacturers, such as the heat capacity C and the electrochemical reaction heat Q1 . In addition, it is needed
to select the corresponding communication protocol with the charger for test. These are basic configuration. Then several parameters for the operation of the charger test need to be defined. Both the initial and desired values of battery’s SOC, the ambient temperature, and the working conditions of the battery in charging have to be configured to customize the operational environment for the charger test. Additionally, the charge mode and the charge current and voltage are basic settings to test the output of the charger. Above all, by changing the parameters of a single cell, this system can simulate unusual circumstances that may encounter in the actual charge operation. For instance, to set a higher single-voltage or SOC of a cell means an over-voltage protection will be triggered during the charge, and a larger value of the internal resistance means an excess-temperature protection.
likely to be encountered in the real. The SS-BMS is going to be embedded in the software of EV charger testing platform for improving its testing capabilities and performance.
In this example, as setting the internal resistance of No.25 Cell larger than others, the system will simulate battery in an anomalous situation. And the time interval t is set at 60, so that the system can only spend 3 minutes to simulate the charging process which might cost more than three hours. The information of the whole process of charging is shown by Fig.3 to Fig. 5.
Figure 5. Parametric curves of battery
REFERENCES
Figure 3. System operating information [1]
[2]
[3] Figure 4. State parameters of cells (partial)
As Fig.3 to Fig.5 showing, the system simulated the excesstemperature of single cell happened during the test. In this case, not only can we see the output characteristics of the charger in charging, and can also examine the response process of charger in the abnormal environment. V.
CONCLUSIONS
This paper describes the design method of SS-BMS, and provides an example for simulating normal and abnormal charging situations using this system. The advantage of SSBMS is that it permits to set various charging conditions, both normal and abnormal, just setting different running parameters. Its functions are not limited only by the example in this paper. This system can simulate most types of charging objects, which
[4]
[5]
[6]
[7]
[8]
Benzai Deng, and Zhiqiang Wang, “Research on electric-vehicle charging station technologies based on smart grid,” in IEEE 2011 AsiaPacific Power and Energy Engineering Conference, 2011, pp. 1-4. Weidong Xie, and Wei Luan, “Modeling and simulation of public EV charging station with power storage system,” in IEEE 2011 International Conference on Electric Information and Control Engineering, 2011, pp. 2346-2350. A. Bruyère, L. De Sousa, B. Bouchez, P. Sandulescu, X. Kestelyn, and E. Semail, “A multiphase traction/fast-battery-charger drive for electric or Pplug-in hybrid vehicles: solutions for control in traction mode,” in IEEE 2010 Vehicle Power and Propulsion Conference, 2010, pp. 1-7. Y. Lu, K.W.E. Cheng, S.W. Zhao, “Power battery charger for electric vehicles,” Power Electronics on The Institution of Engineering and Technology, vol. 4, Iss. 5, pp. 580-586, 2011. Liangliang Chen, Beibei Zhang, Bin Zhou, Hao Zhang, and Zhigang Ding, “Development of charging module of off-board charger for electric vehicle,” Automation of Electric Power Systems, vol.35, no. 7, pp. 82-85, Apr. 10, 2011 (in Chinese). Paul S. Moses, Mohammad A. S. Masoum, and Keyue M. Smedley, “Harmonic losses and stresses of nonlinear three-phase distribution transformers serving plug-in electric vehicle charging stations,” in IEEE 2011 Innovative Smart Grid Technologies, 2011, pp. 1-6. Na Li, and Mei Huang, “Analysis on harmonics caused by connecting different types of electric vehicle chargers with power network,” Power System Technology, vol.35 no. 1, pp. 170-174, Jan. 2011 (in Chinese). Ying-Chun Chuang, Yu-Lung Ke, and Shun-Yi Chang, “Highlyefficient battery chargers with parallel-loaded resonant converters,” in
[9]
[10]
[11] [12] [13]
IEEE 2009 Industry Applications Society Annual Meeting, 2009, pp. 110. Michael G. Egan, Dara L. O’Sullivan, John G. Hayes, Michael J. Willers, and Christopher P. Henze, “Power-factor-corrected single-stage inductive charger for electric vehicle batteries,” IEEE Transactions on Industrial Electronics, vol. 54, no. 2, pp. 1217-1226, Apr. 2007. Baifen Liu, and Chunjuan Zhan, “Design of high charging power supply for electric vehicle based on BUCK circuit,” in IEEE 2011 International Conference on Electric Information and Control Engineering, 2011, pp. 4904-4908 (in Chinese). SZDB/Z 29.3—2010, Technical specification of electric vehicle charging system Part3: Off-board charger (in Chinese). Q/GDW 235—2009, Electric vehicle off-board charger communication protocols (in Chinese). Gregory, L. P. “Extended Kalman filtering for battery managementsystems of LiPB-based HEV battery packs: Part 1.
[14]
[15]
[16]
[17]
Background,” Journal of Power Sources, vol. 134, iss. 2, pp. 252-261, 12 August 2004. Gregory, L. P. “Extended Kalman filtering for battery managementsystems of LiPB-based HEV battery packs: Part 2. Modeling and identification,” Journal of Power Sources, vol. 134, iss. 2, pp. 262-276, 12 August 2004. Gregory, L. P. “Extended Kalman filtering for battery managementsystems of LiPB-based HEV battery packs: Part 3. State and parameter estimation,” Journal of Power Sources, vol. 134, iss. 2, pp. 277-292, 12 August 2004. Junping Wang, Binggang Cao, Quanshi Chen, and Feng Wang, “Combined state of charge estimator for electric vehicle battery pack,” Control Engineering Practice, vol. 15, iss. 12, pp. 1569-1576, Dec. 2009. Noboru Sato, “Thermal behavior analysis of lithium-ion batteries for electric and hybrid vehicles,” Journal of Power Sources, vol. 99, iss. 1-2, pp. 70-77, August 2001.
