摘要
根据恒流放电条件下锂离子电池端电压与荷电状态(SOC)映射关系图像的特点,提出了基于BP神经网络和最速下降法的锂离子电池荷电状态(SOC)估计算法。利用matlab软件编写算法程序,结合实验数据对BP神经网络进行了训练,并将经过训练后的BP神经网络应用于SOC预测。实验结果表明,误差在大部分时候低于10%,基本满足动力电池电荷状态估计的精度要求。
According to the characteristics of the mapping relationship between terminal voltage and the SOC of Li-ion battery under constant current discharge, a battery SOC estimation algorithm based on BP neural network and steepest descent method is proposed. By MATLAB software, BP neural network is trained with experimental data, and the trained BP neural network is applied to SOC prediction. The experimental results show that the error is often less than 10%, which basically meets the precision requirements of power battery charge state estimation.
引文
[1]李争,智若东,孙宏旺,等.基于开路电压预测的SOC估算方法[J].河北工业科技,2017,34(1):36-40.
[2]李哲,卢兰光,欧阳明高.提高安时积分法估算电池SOC精度的方法比较[J].清华大学学报(自然科学版),2010,(8):1293-1296.
[3]毕晓君.信息智能处理技术[M].北京:电子工业出版社,2010,118-120.
[4] Charkhgard M,Farrokhi M. State-of-Charge Estimation for Lithium-Ion Batteries Using Neutral Networks and EKF[J].IEEETransactionsonIndustrialElectronics, 2010,57(12):4178-4187.