摘要
电池荷电状态(SOC)估计是电动汽车电池管理系统的重要功能。该文首先对现有的SOC估计方法的优缺点进行分析,在此基础上将径向基函数(RBF)网络的思想应用于SOC估算。通过应用RBF逼近电池电压和SOC之间的关系曲线,得到两者之间的关系模型,实现根据电池电压估计SOC。为了提高RBF逼近的精度,定义代价函数,应用梯度下降算法实时修正RBF网络的参数。仿真结果表明,估计误差能够降低到10-4。
The state of charge（SOC）estimation is an important function of electric vehicle battery management system. By analyzing the advantages and disadvantages of the existing SOC estimation methods,this study introduces the radial basis function（RBF）neural network method to SOC estimation. By fitting the curves between the battery voltage and SOC,the input-output model between them is obtained,and then the SOC estimation value can he obtained through the measured voltage value. In order to improve the fitting precision,the performance function is defined and the gradient descent method is used to update the weight parameters of the RBF neural network in real-time. Simulation results demonstrate that the estimation error can be reduced to 10^-4.
出处
《自动化与仪表》
2015年第9期89-92,共4页
Automation & Instrumentation
关键词
电池管理系统
电池荷电状态
径向基函数
代价函数
梯度下降
battery management systems （BMS）
state of charge （SOC）
radial basis function （RBF）
performance function
gradient descent method
