The Research on Intelligent Management System of Li-ion Power Battery String of Electric Vehicle

Zhengwei Zhu
School of Information Science and Engineering Changzhou University Changzhou, China Zhuzw888@126.com
Abstractentering the 21st century, the energy shortage seriously restricts the development of human society. Lithium battery with high monomer voltage and high energy density has become the first choice of future power battery. In order to provide electric cars with safe and reliable power sources, this paper put forward an overall design scheme of the battery energy management system. Firstly, according to the requirement of battery management system, the hardware and software of the PC and the monitoring module ware designed. Secondly, in order to meet the needs of battery equilibrium, the charge balance control strategy was proposed. Finally, the battery group was introduced, and a viable battery grouping scheme in engineering was presented. Through practical test, the battery energy management system could effectively manage the power of lithium-ion battery. Keywords- electric cars; BMS; Lithium battery

Bo yang
School of Information Science and Engineering Changzhou University Changzhou, China 779740038@qq.com efficiency, thus increasing the driving mileage, prolonging the service life, reducing the operating cost and improving the reliability of battery electric vehicles. BMS has become the indispensable core component of electric cars. At present, to vigorously develop the battery application technology, especially to develop the battery management system becomes ithe main direction of development of electric vehicles. II. DESIGN OF THE BATTERY MANAGEMENT SYSTEM

I.

INTRODUCTION

In pace with the rapid development of China's economy and the high speed growth of cars ownership, peoples green awareness is growing unceasingly. It becomes an extremely urgent task to reduce vehicle exhaust pollution. As the price of oil soaring, the cost of the traditional car increase quickly. If the number of vehicles continues in accordance with the present growth rate, it will cause energy shortages, thus will restrict the country's economic development. Therefore, pollution-free electric vehicles will be a breakthrough in transportation industry [1]. The main obstacles of popularizing electric vehicles are rechargeable mileage and initial price. Power is the bottleneck in the development of electric vehicles, if power supply problems are properly resolved, electric automobiles socialization will be realized in the near future. The history of Lithium-ion battery can be traced to the early 90s; it basically solved the two technical problems of battery that are poor safety and short service life of the charging and discharging. At the same time, the lithium-ion battery with high battery monomer voltage, high specific energy and energy density. With its industry development, the cost of lithium-ion batteries is steadily falling. For its incomparable superiority, lithium-ion batteries will be the best choice of future power battery [2]. The battery management system (BMS) is used to monitor the safety of storage battery and to improve the battery's use

The battery contains 13 battery packs, each battery pack is connected in series, and each battery pack composed of 10 single cells. On the one hand, in order to ensure the basic balance of each cell in the course of using, the overall control scheme uses a distributed control system [3] [4]: Configures each cell with a monitoring module ICB. The lower computer, which is responsible for the collection of single cell's temperature and voltage, the equalization of current control and the sending/ receiving signals of PC, is composed of 13 ICBs, current and voltage sampling controllers and a balancing module controller; Using ARM9, and connect its touch screen with the lower computer through 485 buses, then summarize the simulation quantities of 120 batteries' voltage from 13 battery packs and 13 temperature values. On the other hand, use IPM charge/discharge controller for communication, collecting data and sending control commands. The touch screen supports USB and CAN bus what can be used to control the charger charging process and send commands to the motor controller to control the operation of the motor. The structural design principle diagram of control system is shown as Figure 1.

Fig.1 Control structure of the battery management system

978-1-4577-0321-8/11/$26.00 2011 IEEE

2370

A. design of hardware system The whole system can be divided into three parts: PC, the battery monitoring module and the charging equalization module. The monitoring module collects the single cell's voltage and temperature, also transmits the dates to PC. The PC collects dates from the lower position machine, and makes a corresponding decision after a comparative analysis. At the same time, PC collects the total current and voltage of the battery, then provides the vehicle's CAN network with the highest battery voltage and No., the minimum battery voltage and No., total current, total voltage at intervals of one second, and outputs over-temperature alarm signal, overvoltage and undervoltage alarm signal, communication interrupt alarm signal randomly. PC communicates with the monitoring modules of single cells through the RS485. Through the CAN bus interface, PC can realize data sharing with other systems. The monitoring module mainly consists of 14 pieces of American TI Company's MSP430 MCU which contains 12 channels 12 AD and 4 serial communications [5], and the model No. is MSP430F5418. There are 13 MSP430F5418s that be used to collected the voltage and temperature (ten battery voltage and the one temperature value) of the 13 battery packs. The remaining one is used as a current, voltage sampling controller, it collects the total voltage and total current or other physical quantities, and carries ou over voltage protection, over current protection and low voltage protection. Monitoring module hardware design diagram is shown as Figure 2.

So there the absolute error of single cell voltage is less than 0.025V. Formula 1 show that the main factor that influences the voltage measurement accuracy is the precision of voltage reference that is the zener diode's precision. This paper chooses LE45CD zener diode. If need higher voltage measurement accuracy, can choose voltage reference chips with higher accuracy. Temperature acquisition: In consideration of the battery temperature range is not large, this paper use a single bus temperature sensor (1-Wire) device DS18B20 integrated with a temperature sensor chip. DS18B20 can realize two-way communication only needs one IO port when connected with the microprocessor. The temperature range of DSl8820 is 55 ~ 125, and the inherent temperature resolution is 0.5. The conversion time of DSl8820 is less than 750ms. As DS18B20 sends a serial of digital signal, the circuit needs no A\D conversion. The ten bits encode lines of the single cells correspond to the RB1 ~ RB10 I/O interfaces in MCU. The encoding circuit using 10K 9 ports for weak pull-up. By coding the ten bits coded address, the number of each single cell could be obtained. The 485 communication port of the monitoring module corresponds to pin17 and pin18 in MCU, and uses the photoelectric isolation device PC817 for isolation. The model no. of 485 communication control chip is MAX1487, its power comes from a 4-core communication cables. The upper machine BMS consists of five parts: data receiving and calculation, BMS status display, BMS mode control, CAN device control and data record. B. design of software system Software design uses the top-down modular method. The software system can be divided into the initialization module, multiple switch gate control module, A/D transformation control module and an ISR (interrupt service routine, ISR) module. Initialization module completes the initialization of the registers of the MCU system and the definition of working mode. Multiple switch gate control module choose the channel of multiple switch circularly, and send the input analog signal after isolation to the A/D transformation intelligent control module circularly. A/D transformation intelligent control module mainly transforms the analog quantity to digital quantity. ISR receives instruction from PC and transmits the conversion results to PC. The main program flow of software system is shown as figure 3.

Fig.2 Schematic diagram of the monitor module Voltage acquisition: In general, the voltage range of single cell is 2.5-4.25V. After the series dividers R3and R4, the voltage of single cell is inputted to the SCM's A0 port, and then does AD sampling. The zener diode D1 provides the control chip with an 2.5V voltage for AD sampling. Voltage range of single battery is 2.5-4.25V in general, after R3 and R4 , it was inputted to the SCM's AN0 port, both R3 and R4 are 1K Omega and the accuracy is 0.1%.The accuracy of AD sampling is about 0.025% with of 12 digits. Thus controlled the precision of AD sampling of single cell voltage at 0.6%, the calculate method is as following: n = 1 - (1-0.5%) * (1-0.1%) * (1-0.025%) = 0.6%. (1)

Fig.3 Control structure of the battery management system

2371

III.

STATE ESTIMATION OF THE BATTERY

Fig.4 SOC principle diagram Firstly, estimate the SOC value at the moment k+1 based on the SOC value at k and the current value at present. Secondly, calculate the estimation voltage value of the battery at k+1 according to the battery model, and compare the value with the measured voltage value, thereby; the voltage estimation error can be obtained. Thirdly, modify the filter gain by the estimation error, then using the modified gain to modify the estimation SOC, then obtain the SOC value at k+1 after filtering. Finally, using the SOC value obtained above as the initial SOC value of the next time, and then starts the next estimation process. IV. CONCLUSIONS

One of the important functions of battery management system is to estimate the SOC accurately. The common methods include the Ampere-Hour counting method and the open-circuit voltage method. Because of the errors of the current sampling and charge/discharge efficiency, the AmpereHour counting method will cause the SOC's accumulated error larger and larger. So the Ampere-Hour counting method is not suitable for long time working. The open-circuit voltage method is suitable for the battery open circuit and constant current charging/discharging work situation, but the amplitude range of current changes is wide and quickly when electric vehicles are working. So the open-circuit voltage method is not applicable [6] [7]. In this paper, we use Kalman optimal filtering theory based SOC estimation method. Regard the power battery as a system, the charged state SOC and the polarization voltage of battery as the state variables of the system, the currentas the systems input variable, the voltage of the battery as the system's observed quantity, then the battery SOC estimation problem can be transformed into the estimation problem of system state inspired by the current input. The basic principle of the method is shown as Figure 4.

The battery management system is the key guarantee of electric cars. Experiments show that the battery management system can realize the fast electric vehicle energy supply for lithium-ion. As the discharge current changes randomly when the electric cars are in operation, it is difficult to accurately measure the battery voltage and capacity. So, the residual capacity of battery group, and the remaining mileage estimation problems still need to strengthen research. REFERENCES
[1] [2] [3] qingquan Cheng, green traffic tools--electric cars in 2l century, Tsinghua university press, 2000. sizhong Cheng, the key technology and development prospects of the electric car, Beijing university of science and technology press, 2002 Chengning Zhang, Design on management system for traction battery packs in electric transmission vehicle [J], Computer Engineering and Application, 2006. fangwei Yuan, Research of electric vehicle battery management system [J],Automobile Research&Development,2003. MSP430x5xx FamilyUserpsGuide. www. ticom. cn, 2011. chengtao Lin, The electric car SOC estimation principle and application [J] , Power technologies, 2004. zili Deng, Kalman filtering and wiener filtering, Harbin industrial university press, 2001.

[4] [5] [6] [7]

2372