Research Of Battery Management System Based On TMS320F2812 For Electric Vehicle

With the consuming of non-renewable energy such as coal and petroleum, energy is increasingly becoming a bottleneck, which restricts the development of international society and economy. Automotive exhaust contains lots of harmful gases and causes environmental pollution seriously. More and more countries begin to implement the green energy source plan, in order to solve environmental problems. Therefore, research on electric vehicle and its related technologies have become a pop field. What's more, the research on battery management system (BMS), which is one of the core and key electric vehicle technologies, has important theoretical value and practical significance.At present, some enterprises and research institutions have empoldered some battery management systems, which have been applied into electric vehicles. These battery management systems have some basic functions, such as the estimation for state of charge (SOC) and temperature measurement, but they can not enter the commercial application. The problem has become one of main factors which baffle the marketization of electric vehicle's application.Firstly, the main functions of BMS at home and abroad are commented. The shortages in existing BMS are summarized. The composition of Li-ion batteries and their principle are introduced. The concept of SOC is proposed and the advantages and shortages of different estimation methods of SOC are analyzed. In the case of the thevenin model combined with unscented Kalman filtering (UKF), a new method to estimate SOC of li-ion battery is proposed and a simulation model based on hardware-in-loop is established. Field test shows the simulation model on SOC runs reliably.To configuration of TMS320F2812+LTC6802, a new solution of BMS is presented in this paper. General design idea of BMS is described. The hierarchical structure of control mainboard and measurement sub-panels are demonstrated. The process of hardware design is given. The design takes the mainboard as the core. The overall hardware design scheme is performed, such as sampling module, communication module, the control circuit, measurement and safety protection module etc. On CCS3.3 development platform, the method of overall planning for modularization design is discussed. Particular program design flow of key parts is given, such as main program, data acquisition, data processing, protection, CAN, data storage, thermal management, balance management with software anti-interference etc.In the end, hardware-in-the-loop simulation and actual measurement are progressed. The simulation result shows a new method in this paper is superior to the extended Kalman filtering (EKF), especially in the case that the SOC is below 50%. The test result shows that the SOC estimation design in this paper is stable. The error of monomer balance is less than 1%.and the temperature can be controlled between 30℃and 50℃.The whole design conforms to design requirements.