基于BP神经网络的动力电池SOC估算

Estimation of power battery SOC based on BP neural network

电池荷电状态(SOC)的预测是影响电动汽车发展的关键技术之一,采用经典BP神经网络控制算法完成了动力电池的SOC估算研究。通过设计工况实验,在Matlab中对该算法进行了仿真验证,结果表明该算法能够很好地拟合动力电池充放电特性,误差可以减小到5%以内。

The estimation of power battery state of charge（SOC） is one of the key technologies for electric vehicles＇ development. The research of battery SOC estimation was completed based on classic BP neural network control algorithm. It was tested and simulated in Matlab though designing working conditions. The results show that this algorithm could perfectly fit the charging and discharging characteristics of power battery.