The Study For Charging State Of Electic Vehicle Power Battery Which Based On The UKF Algorithm

In recent years, with the haze weather gradually increased, people’s health has been affected, so the fuel vehicles are often limited to the line .The electric vehicle which has the advantage of low power consumption and zero emission is becoming popular for people. For the research of pure electric vehicle, the most important part is the power battery. And the core of Power battery is the battery management system. It can realize many functions, for example the battery state detection、information transmit、 security alarm and protection. When the electric vehicle is running, these functions ensured the energy security, reasonable and efficient utilization. So the research on it is very important.According to the battery charge state which is the core of BMS estimation is analyzed from the inside to the outside. On the basis of the mechanism and characteristics of lithium iron phosphate batteries, this article built a model of the battery. This model which is increased the Battery fade module, temperature module, self discharge module, and two order the venin equivalent circuit module is the basic model of power battery. According to experimental data,we got the relationship between SOC and the open circuit voltage. And the nonlinear correction factor has been fitted from the experimental data. At the same time, we simulated and verified the model in Matlab/Simulink.For the estimation of SOC, this paper compared several common SOC estimation methods,and analyzed their advantages and disadvantages. According to the experimental results and the actual situation, this paper proposes a method which combined with the unscented Kalman filter(UKF) algorithm, Ampere-hour integral method and open circuit voltage method. This algorithm has also added a correction factor to correct the results of estimation. In doing so, the system can be stable reflect the actual working conditions through SOC in the complex environment. According to the simulation experiments, we can find that this method estimation error can be controlled in the range of 5% under the certain conditions.The hardware circuit design is used the ARM STM32F103 for the main control board, and used the LTC6802 for the data acquisition board. This hardware can realize data acquisition,data transmission, and SOC data analysis. Through the transplant algorithm and simulation, this paper verified the feasibility of algorithm in the practical application.