锂离子电池剩余容量与剩余寿命预测  

Prediction of Li-ion battery's remaining capacity and remaining useful life

作　　者：谢建刚 李其仲 黄妙华[1,2] 王树坤 XIE Jian-gang;LI Qi-zhong;HUANG Miao-hua;WANG Shu-kun(Hubei Key Laboratory of Advanced Technology for Automotive Components(Wuhan University of Technology),Wuhan Hubei 430070,China;Hubei Collaborative Innovation Center for Automotive Components Technology,Wuhan Hubei 430070,China)

机构地区：[1]现代汽车零部件技术湖北省重点实验室(武汉理工大学),湖北武汉430070 [2]汽车零部件技术湖北省协同创新中心,湖北武汉430070

出　　处：《电源技术》2018年第10期1438-1440,共3页Chinese Journal of Power Sources

基　　金：国家科技支撑计划(2015BAG08B02)

摘　　要：锂离子电池具有优异的性能,在电动汽车中得到广泛应用。剩余容量和剩余寿命预测是电池健康管理的关键所在。支持向量回归机(support vector regression,SVR)作为一种具有良好的非线性、泛化性的预测算法,能有效提高锂离子电池剩余容量和剩余寿命的预测精度。在分析SVR算法原理的基础上,提出了一种基于蚁群算法(ant colony optimization,ACO)的参数优化方法,增强了SVR关键参数全局最优搜索能力,改善了SVR算法的预测能力。与基于网格搜索的SVR算法预测结果比较,仿真结果表明:改进ACO_SVR算法有更好的预测精度,能为电池管理系统提供可靠的数据。The lithium-ion battery is widely used in electric vehicles,due to its excellent performance.The estimations of remaining capacity and remaining useful life of a faulty battery are at the center of battery health management.Thanks to its good nonlinear function approximation,generalization ability and stability,the prediction accuracy of residual capacity could be improved by SVR algorithm.Based on the analyze of the fundamental principle of SVR algorithm,the SVR algorithm based on Ant colony optimization(ACO)was proposed.Then the ability in searching for the global optimal solution was enhanced,and the prediction accuracy was improved.Compared with the results of SVR based on grid search algorithm,the simulation results show that ACO_SVR algorithm has higher prediction accuracy and can provide reliable data for battery management system.

关 键 词：电动汽车 支持向量回归机 剩余容量 蚁群算法 

分 类 号：TM912[电气工程—电力电子与电力传动]