基于WD-GRU的锂离子电池剩余寿命预测  

Remaining life prediction of lithium ion batteries based on WD-GRU

作　　者：邢子轩 张凡 武明虎[1] 何浩然[1] XING Zixuan;ZHANG Fan;WU Minghu;HE Haoran(School of Electrical and Electronic Engineering,HubeiUniversity of Technology,Wuhan Hubei 430072,China;Hubei Engineering Research Center for SafetyMonitoringofNewEnergy and PowerGrid Equipment,HubeiUniversity of Technology,WuhanHubei 430068,China)

机构地区：[1]湖北工业大学电气与电子工程学院,湖北武汉430072 [2]湖北工业大学新能源及电网装备安全监测湖北省工程研究中心,湖北武汉430068

出　　处：《电源技术》2022年第8期867-871,共5页Chinese Journal of Power Sources

基　　金：中央支持地方专项(Z135050009017);湖北省高等学校优秀中青年科技创新团队计划(T201805)。

摘　　要：为提高锂离子电池剩余寿命预测准确率,提出了一种基于多尺度分解的电池寿命预测方法,可准确捕捉电池退化过程中健康因子的波动。在门控循环单元网络(GRU)预测模型基础上,使用小波分解(WD)方法对健康因子序列分解,提取出序列中波动部分和平滑下降部分分别预测,最后将结果集成输出,即可得到准确的剩余寿命预测结果。最后采用公共数据集对该方法有效性进行验证,并对比了其他几种预测模型,结果表明该方法预测结果均优于其他模型。In order to improve the prediction accuracy of the remaining life of lithium-ion batteries,a battery life prediction method based on multi-scale decomposition was proposed,which could accurately capture the fluctuations of health factors in the process of battery degradation.Based on the gated cycle unit network(GRU)prediction model,the wavelet decomposition(WD)method was used to decompose the health factor sequence,and the fluctuation part and smooth decline part of the sequence were extracted and predicted respectively.Finally,the results were integrated and output,and the accurate residual life prediction results were obtained.The validity of the method was verified by using the public data set,and the results show that the prediction results of this method are better than other models.

关 键 词：锂离子电池 剩余寿命预测 门控循环单元网络 小波分解 

分 类 号：TM912.9[电气工程—电力电子与电力传动]