摘要
为了更精确估算车用锂电池荷电状态(SOC)值,采用PID神经网络方法建立电池模型,设定电池电压、放电电流、电池累计放电量和电池电极温度4个变量为模型输入量,电池剩余电量为模型输出量,由此得到了全部神经网络训练数据,并仿真估算出电池SOC值。仿真结果表明,利用该方法对电池SOC进行估算,误差小于3.66%,方法有效。
To estimate state of charge（SOC） of Lithium-ion battery more precisely,a new method using PID neural network was adopted to build the model for Lithium-ion battery.Through setting the battery＇s terminal voltage,discharge current,discharge capacity and temperature as input variables and SOC as output of the model,all training data of the neural network were obtained and the SOC value of the battery was estimated through simulation.Simulation results showed that the maximum estimation error was less than 3.66% and the new method was validated to be effective.
出处
《汽车技术》
北大核心
2012年第10期36-38,43,共4页
Automobile Technology
基金
江苏省科技成果转化项目(BA2010050)
汽车仿真与控制国家重点实验室开放基金(20111110)
江苏省汽车工程重点实验室开放基金(QC200904)
关键词
电动汽车
锂电池
SOC估算
PID神经网络
Electric vehicle
Lithium-ion battery
SOC estimation
PID neural network
