Prediction of battery-SOC of pure electric vehicle

The accurate value of battery SOC is one of the prerequisites of a pure electric vehicle energy management to achieve optimal control. However, the battery is a highly nonlinear system and its charge-discharge process is difficult to establish accurate mathematical model. Taking into account nonlinear characteristics of BP neural network, and it has parallel structure and ability to learn, so it is suitable for online estimation of battery SOC value. Train BP neural network through Matlab programming, and predict the performance of the battery based on the built neural network model to get the battery SOC prediction. Simulation results show that the the established BP neural network has good adaptability, and is able to predict the mapping relationship between the battery voltage and SOC. This method is access to battery SOC prediction quickly and easily, the maximum error is less than 1% and the results meet the accuracy requirements.