电动车蓄电池SOC估计的神经网络方法

Neural network based on SOC estimation for electric vehicles

针对蓄电池荷电状态估计问题,将神经网络方法用于电动车蓄电池荷电状态的估计。依据蓄电池荷电状态与可测量参数之间的非线性关系,建立了基于BP神经网络与RBF神经网络的蓄电池荷电状态预测模型。仿真结果表明,经过训练后的预测模型,可以通过蓄电池的端电压、工作电流以及蓄电池的内阻参数预测蓄电池的实时荷电状态。通过比较,RBF预测模型具有较好的泛化能力且稳定性更强,能够更精确的估计出蓄电池的剩余容量。

For the estimation problem of the battery＇s state of charge（SOC）,the neural network method is used to estimate the battery＇s state of charge of the electric vehicles. According to the non-linear relationship between the battery＇s SOC and measurable parameters, it is established that a battery＇ SOC prediction model based on BP neural network and RBF neural network. The simulation results show that the trained prediction model can predict the real-time SOC of the battery by the parameters including the terminal voltage, current and internal resistance of the battery. By comparison,the RBF prediction model has better generalization ability and greater stability, thus more accurately estimating the battery,s state of charge.