A rest-time-based prognostic model for remaining useful life prediction of lithium-ion battery

Research output: Journal Publications and ReviewsRGC 21 - Publication in refereed journalpeer-review

31 Scopus Citations
View graph of relations

Author(s)

Detail(s)

Original languageEnglish
Pages (from-to)2035–2046
Number of pages12
Journal / PublicationNeural Computing and Applications
Volume33
Issue number6
Online published21 Jun 2020
Publication statusPublished - Mar 2021

Abstract

This paper proposes a novel empirical model for the remaining useful life prediction of lithium-ion battery. The proposed model is capable of modeling both the global degradation and local degradation of lithium-ion battery aging process. The global degradation process is regarded as the ideal aging profile without any interference by regeneration phenomenon. However, the regeneration phenomenon inevitably occurs in practical usage of lithium-ion battery and affects the local degradation significantly. Therefore, we separate the local degradation process from the global degradation process to represent the local battery aging process affected by regeneration phenomenon. We unify the modeling method for the global and local degradation process by exponential functions with cleverly designing of the corresponding cycles. The particle filter framework is employed to estimate the model parameters with measurement data. The future capacity is predicted after the identification, and the remaining useful life is extracted by calculating the difference between the predicted capacity and failure threshold. Model comparisons on benchmark battery datasets have been conducted. The results demonstrate that our proposed method is capable of capturing the degradation and regeneration phenomena, and the remaining useful life prediction performance of our proposed model is better than state-of-the-art modeling methods.

Research Area(s)

  • Capacity degradation, Capacity regeneration, Lithium-ion battery, Particle filter, Remaining useful life prediction