汽车工程 ›› 2025, Vol. 47 ›› Issue (6): 1060-1071.doi: 10.19562/j.chinasae.qcgc.2025.06.005

• • 上一篇    

基于TCN-LSTM模型的实车电池健康状态评估方法

胡杰1,2,3(),王浩杰1,2,3,魏敏1,2,3,4,王志红1,2,3,陈琳1,2,3,黄文涛1,2,3,康涵锐1,2,3   

  1. 1.武汉理工大学,现代汽车零部件技术湖北省重点实验室,武汉 430070
    2.武汉理工大学,汽车零部件技术湖北省协同创新中心,武汉 430070
    3.湖北省新能源与智能网联车工程技术研究中心,武汉 430070
    4.上汽通用五菱汽车股份有限公司,柳州 545000
  • 收稿日期:2025-02-07 修回日期:2025-03-20 出版日期:2025-06-25 发布日期:2025-06-20
  • 通讯作者: 胡杰 E-mail:auto_hj@163.com
  • 基金资助:
    第二十七届中国科协年会学术论文。广西科技计划项目(2023AA03009)

A TCN-LSTM Model-Based Approach for Real Vehicle Battery Health State Evaluation

Jie Hu1,2,3(),Haojie Wang1,2,3,Min Wei1,2,3,4,Zhihong Wang1,2,3,Lin Chen1,2,3,Wentao Huang1,2,3,Hanrui Kang1,2,3   

  1. 1.Wuhan University of Technology,Hubei Key Laboratory of Modern Auto Parts Technology,Wuhan 430070
    2.Wuhan University of Technology,Auto Parts Technology Hubei Collaborative Innovation Center,Wuhan 430070
    3.Hubei Technology Research Center of New Energy and Intelligent Connected Vehicle Engineering,Wuhan 430070
    4.SAIC-GM-Wuling Automobile Company Limited,Liuzhou 545000
  • Received:2025-02-07 Revised:2025-03-20 Online:2025-06-25 Published:2025-06-20
  • Contact: Jie Hu E-mail:auto_hj@163.com

摘要:

为解决当前实车数据质量差导致的电池健康状态评估结果精度不足的问题,提出了一种基于TCN-LSTM模型的电池健康状态评估方法。首先提出了基于随机搜索算法的恒流充电电压片段提取方法;然后提出了局部加权回归与3阶多项式回归加权融合的方法来拟合电池容量的整体与局部衰减趋势。接着构建并提取与电池老化相关的特征,创新性地提出了经容量保持率修正的累充容量以及满充电压、电池一致性等特征;最后构建了基于TCN-LSTM的动力电池健康状态评估模型,从多维度来提取特征与电池老化之间的关系。结果表明,TCN-LSTM模型可以准确地评估出实车数据下动力电池复杂的容量衰减变化关系,RMSRE仅为0.002 1。

关键词: 电动汽车, SOH, TCN-LSTM, 电池一致性

Abstract:

For the problem of insufficient accuracy in battery state of health (SOH) evaluation caused by the poor quality of real-world vehicle data, in this study a battery health state evaluation method based on the TCN-LSTM model is proposed. Firstly, the random search algorithm is employed to extract constant-current charging voltage segments. Subsequently, a weighted fusion approach combining locally weighted regression and third-order polynomial regression is used to fit the global degradation trend and the local degradation trend of battery capacity. Then, features related to battery aging, including the capacity retention-corrected cumulative charge capacity, fully charged voltage, and battery consistency, are constructed and extracted. Finally, a TCN-LSTM-based evaluation model for battery health state is constructed to explore the relationship between extracted features and battery aging from multiple dimensions. The results show that the TCN-LSTM model effectively evaluates the complex capacity degradation relationship of power batteries under real-world vehicle data, achieving an RMSRE of only 0.002 1.

Key words: electric vehicles, SOH, TCN-LSTM, battery consistency