电动汽车IGBT剩余使用寿命预测  

Remaining useful life prediction of IGBT in electric vehicles

作　　者：杜先君[1,2] 王紫阳 DU Xian-jun;WNAG Zi-yang(College of Electrical and Information Engineering,Lanzhou Univ.of Tech.,Lanzhou 730050,China;Key Laboratory of Gansu Advanced Control for Industrial Processes,Lanzhou Univ.of Tech.,Lanzhou 730050,China)

机构地区：[1]兰州理工大学电气工程与信息工程学院,甘肃兰州730050 [2]兰州理工大学甘肃省工业过程控制重点实验室,甘肃兰州730050

出　　处：《兰州理工大学学报》2024年第2期77-86,共10页Journal of Lanzhou University of Technology

基　　金：国家自然科学基金(61563032);甘肃省教育厅创新基金(2021A-027)。

摘　　要：引入一种基于贝叶斯优化(BOA)的双向长短时记忆网络(Bi-LSTM),同时结合注意力机制,应用于绝缘栅双极型晶体管(IGBT)剩余使用寿命预测,所提方法可有效提高IGBT剩余使用寿命预测的准确性.通过IGBT加速老化试验收集V CE-on,验证了其作为失效特征参数的可行性,并将其作为实验数据集对所提方法进行仿真验证.实验分析结果表明,所提的混合预测模型与经典LSTM及其他预测模型相比,有更低的退化预测误差,具备较高的理论意义和实践价值.As one of the core components of electric vehicles,IGBTs’health monitoring and remaining life prediction play a vital role in proactive maintenance.The Bi-LSTM model based on Bayesian optimization and attention mechanism is proposed to predict the remaining useful life of IGBT in this paper.The proposed method can effectively improve the accuracy of IGBT remaining service life prediction.V CE-on through IGBT accelerated aging test is collected in this study,verifying its feasibility as a failure characteristic parameter.This data is used as an experimental data set to validate the proposed method through simulation.The experimental analysis results show that the proposed hybrid prediction model has lower degradation prediction error than the classical LSTM and other prediction models,demonstrating significant theoretical and practical value.

关 键 词：电动汽车IGBT 剩余寿命预测 贝叶斯优化算法 注意力机制 双向长短时记忆网络 

分 类 号：TN322.8[电子电信—物理电子学]