30 / 2025-03-26 18:26:15
Remaining Useful Life Prediction Based on PSO-PF and Empirical Degradation Model
RUL, PSO-PF, empirical degradation model, dynamic estimation
全文待审
燕山 李 / 中国核动力研究设计院
常华 聂 / 中国核动力研究设计院
The remaining useful life (RUL) of complex systems and critical components is a key metric for evaluating their reliability and maintainability. Accurate RUL prediction enhances maintenance efficiency, reduces costs, and improves overall operational safety. Traditional physics-based models for RUL prediction typically rely on fixed parameters, which limit their adaptability and real-time effectiveness. Additionally, parameter estimation in these models is often based on complete degradation data. While this approach captures the overall degradation trend, it fails to account for fluctuations throughout the degradation process. To address this limitation, [14] introduced an RUL prediction method that integrates particle filtering (PF) with degradation modeling, allowing for dynamic updates of model parameters. However, this method does not account for the issue of particle weight degradation during the resampling process in PF, which can compromise prediction accuracy. To mitigate this issue, this paper proposes replacing the standard PF algorithm with a particle swarm optimized particle filtering (PSO-PF) algorithm, aiming to enhance prediction accuracy. The effectiveness of the proposed approach is validated using a publicly available lithium battery dataset.
重要日期
  • 会议日期

    08月22日

    2025

    08月24日

    2025

  • 04月25日 2025

    初稿截稿日期

主办单位
中国自动化学会技术过程的故障诊断与安全性专业委员会
承办单位
新疆大学
新疆自动化学会
移动端
在手机上打开
小程序
打开微信小程序
客服
扫码或点此咨询