Remaining Useful Life Prediction of Capacitor Based on Genetic Algorithm and Particle Filter

EasyChair Preprint 5600

Abstract

The failure rate of capacitors is high in the circuit system, and in the system with high requirement for capacitance reliability, it is very important to predict the remaining useful life accurately. In this paper, a particle filter method based on genetic algorithm is proposed to predict the remaining useful life of capacitors. Using the capacitance data set published by NASA, an exponential degradation model is established, and the resampling procedure in traditional particle filter method is optimized by crossover, mutation and optimization in genetic algorithm to increase the particle diversity, and to propel particles move to the high likelihood region. Therefore, the particle depletion problem caused by the resampling step in the traditional particle filter is improved to some extent. The simulation results show that the particle filter method based on genetic algorithm can be used to achieve more accurate prediction of remaining life of electrolyte capacitor.

Keyphrases: Capacitor, Genetic Algorithm, Prognostic, Remaining Useful Life, particle filter

@booklet{EasyChair:5600,
  author    = {Meinan Wang and Wei Niu and Yangyang Zhao},
  title     = {Remaining Useful Life Prediction of Capacitor Based on Genetic Algorithm and Particle Filter},
  howpublished = {EasyChair Preprint 5600},
  year      = {EasyChair, 2021}}