基于ICS-LSSVM的电动汽车锂电池SOC预测  

SOC Prediction of Electric Vehicle Lithium Battery Based on ICS-LSSVM

作　　者：王骏骏 WANG Junjun(State Grid Shiyan Power Supply Company,Shiyan,Hubei 442099,China)

机构地区：[1]国网十堰供电公司,湖北十堰442099

出　　处：《东北电力技术》2022年第12期52-56,共5页Northeast Electric Power Technology

摘　　要：为了提高锂电池的荷电状态(state of charge,SOC)预测精度,将布谷鸟的卵被发现概率和步长控制量进行动态设置,对布谷鸟算法进行改进,采用改进布谷鸟算法(improved cuckoo search,ICS)对最小二乘支持向量机(least squares support vector machine,LSSVM)的惩罚参数和核函数宽度进行优化,建立基于ICS-LSSVM的电动汽车锂电池SOC预测模型,采用磷酸铁锂电池充放电数据进行仿真分析,并与其他模型对比。结果表明,基于ICS-LSSVM的电动汽车锂电池SOC预测模型的均方根误差、平均相对误差和决定系数分别为0.0193、3.14%和0.994,预测效果优于其他模型,验证了模型的正确性和实用性。In order to improve the prediction accuracy of lithium batterystate of charge(SOC),the cuckoo's egg discovery probability and step control quantity are dynamically set,and the cuckoo algorithm is improved.Improved cuckoo search(ICS)algorithm is used to optimize the penalty parameters and kernel function width of least squares support vector machine(LSSVM),and an electric vehicle lithium battery SOC prediction model based on ICS-LSSVM is established.The charge and discharge data of lithium iron phosphate battery are used for simulation analysis,and compared with other models.The results show that the root mean square error of the electric vehicle lithium battery SOC prediction model based on ICS-LSSVM,the average relative error and determination coefficient are 0.0193,3.14%and 0.994 respectively.The prediction effect is better than other models,which verifies the correctness and practicability of the model.

关 键 词：锂电池 荷电状态 改进布谷鸟算法 最小二乘支持向量机 

分 类 号：TM75[电气工程—电力系统及自动化]