Remaining useful life prediction of lithium-ion battery with unscented particle filter technique

Abstract

Accurate prediction of the remaining useful life of a faulty component is important to the prognosis and health management of a system. It gives operators information about when the component should be replaced. In recent years, a lot of research has been conducted on battery reliability and prognosis, especially the remaining useful life prediction of the lithium-ion batteries. Particle filter (PF) is an effective method for sequential signal processing. It has been used in many areas, including computer vision, target tracking, and robotics. However, the accuracy of the PF is not high. This paper introduces an improved PF algorithm-unscented particle filter (UPF) into the battery remaining useful life prediction. First, PF algorithm and UPF algorithm are described separately. Then, a degradation model is built based on the understanding of lithium-ion batteries. Finally, the prediction results can be obtained using the degradation model and the UPF algorithms. According to the analysis results, it can be seen that UPF can predict the actual RUL with an error less than 5%.