摘要
锂离子电池在运行过程中内部电化学反应复杂,为使电池稳定运行,实时监测电池的健康状态是电池管理系统的核心任务之一。在实际运行中,锂离子电池的健康状态存在衰退现象,且衰退过程呈非线性和波动性。针对传统建模方法实现锂离子电池健康状态准确估计较为困难、耗时长等问题,构建了基于改进的鲸鱼群算法优化的反向传播(Back Propagation,BP)神经网络,利用鲸鱼群算法(Whale Optimization Algorithm,WOA)进行网络初值参数优化,可有效避免网络易陷入局部极值,并引入非线性收敛因子,加快算法局部寻优能力,提高了算法收敛速度。通过对比传统BP算法、粒子群优化的BP算法,在马里兰大学不同电池组老化数据上的验证结果表明,该算法对于锂离子电池健康状态预测具有更小的估计误差以及较强的学习性能,可以满足估算要求。
The internal electrochemical reaction of lithium⁃ion battery is complex during operation.In order to ensure the stable operation of the battery,real⁃time monitoring the health state of the battery is one of the core tasks of the battery management system.In actual operation,the health state of the lithium⁃ion battery will gradually decrease,and the reduction process is nonlinear and fluctuating.A BP(Back Propagation)neural network optimized based on the improved whale swarm algorithm was constructed to solve the problems that it is difficult and time⁃consuming to accurately estimate the health state of lithium⁃ion battery by traditional modeling methods.The whale optimization algorithm(WOA)was used to optimize the initial parameters of the network,avoiding falling into local extremum.The nonlinear convergence factor was introduced to accelerate the local optimization ability of the algorithm and improve the convergence speed of the algorithm.Using the aging data of different battery packs from the University of Maryland,the traditional BP algorithm and WOA⁃BP algorithm were compared.The results show that the algorithm has less estimation error and strong learning performance for the health state prediction of lithium⁃ion battery,and can meet the estimation requirements.
作者
薛太林
张超
闫来清
吴杰
靳玉祥
XUE Tailin;ZHANG Chao;YAN Laiqing;WU Jie;JIN Yuxiang(School of Electrical Power,Civic Engineering and Architecture,Shanxi University,Taiyuan 237016,China;State Grid Shanxi Electric Power Company Yangquan Power Supply Company,Yangquan 045000,China;Shanxi Guojin Coal Power Corporation,Jiaocheng 030500,China)
出处
《山东电力技术》
2022年第10期16-22,37,共8页
Shandong Electric Power
基金
中央高校基本科研基金资助项目(2018QN097)。
