運用簡易高斯過程粒子濾波器預測剩餘壽命

Abstract

因應目前全球面臨的能源危機和環保問題，質子交換膜燃料電池被視為一個相當重要的替代能源。然而目前燃料電池的應用還受限於壽命不足的問題上，因此管理電池運作時的剩餘壽命是一個相當重要的課題。而粒子濾波器在預測燃料電池的剩餘壽命上被認為是一個相當成功的方法。在本篇論文中，我們提出了簡易高斯過程粒子濾波器來預測電池的剩餘壽命，它比傳統的粒子濾波器更加地具有彈性，我們使用此方法達到更準確的壽命預測。

In response to the worldwide energy crisis and environmental issues, proton exchange membrane fuel cell is regarded as a very important alternative energy sources. However, the current fuel cell applications are also limited by the lack of life issues. Therefore, the remaining useful life of the battery management is significant. The particle filter is considered to be successful in the prediction of the remaining life of the fuel cell. In this study, we propose a Simple Gaussian Process Particle Filter to predict the remaining life of the battery, and it is more flexible than traditional particle filter. We use this method to achieve a more accurate life prediction.