Scholarlink

Lithium-ion power batteries are currently the most widely used energy storage devices in electric vehicles. Rapid and accurate diagnosis of battery faults is crucial for the safe operation of vehicles. This paper proposes a method for diagnosing self-discharge faults in power batteries based on individual adaptive voltage thresholds and Particle Swarm Optimization-Support Vector Machine (PSO-SVM). The research focuses on the voltage signals of power batteries, combines the boxplot method with expert review to label samples of self-discharge faults. Through the sliding window method, 16 features in the time domain and frequency domain are extracted. Principal component analysis is then used to further reduce the dimensionality of voltage features, obtaining the first five principal components with a cumulative variance contribution of 95% as inputs for the PSO-SVM model. The aim of this method is to improve the accuracy of self-discharge fault identification in batteries. The final results show that the proposed method has high recognition accuracy, strong reliability, and potential value in practical electric vehicle applications. It provides theoretical support for enhancing the safety performance of electric vehicles.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

From Time Series to Feature Matrix: A Novel PSO-SVM Framework for Self-discharge Diagnosis of Lithium-Ion Batteries

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Chenghao Liu,
  • Xinyu Zhang,
  • Zhenyu Wang,
  • Yuhao Zhang

摘要

Lithium-ion power batteries are currently the most widely used energy storage devices in electric vehicles. Rapid and accurate diagnosis of battery faults is crucial for the safe operation of vehicles. This paper proposes a method for diagnosing self-discharge faults in power batteries based on individual adaptive voltage thresholds and Particle Swarm Optimization-Support Vector Machine (PSO-SVM). The research focuses on the voltage signals of power batteries, combines the boxplot method with expert review to label samples of self-discharge faults. Through the sliding window method, 16 features in the time domain and frequency domain are extracted. Principal component analysis is then used to further reduce the dimensionality of voltage features, obtaining the first five principal components with a cumulative variance contribution of 95% as inputs for the PSO-SVM model. The aim of this method is to improve the accuracy of self-discharge fault identification in batteries. The final results show that the proposed method has high recognition accuracy, strong reliability, and potential value in practical electric vehicle applications. It provides theoretical support for enhancing the safety performance of electric vehicles.