Journal of Power Sources, Volume 239, pp. 253264, October 1, 2013

Prognostics of Lithium-Ion Batteries Based on Relevance Vectors and a Conditional Three-Parameter Capacity Degradation Model

Dong Wang, Qiang Miao, Michael Pecht

Department of Systems Engineering & Engineering Management, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong, China
School of Mechanical, Electronic and Industrial Engineering, University of Electronic Science and Technology of China, Chengdu, Sichuan 611731, China
Center for Advanced Life Cycle Engineering (CALCE), University of Maryland, College Park, MD 20742, USA

Abstract:

Lithium-ion batteries are widely used as power sources in commercial products, such as laptops, electric vehicles (EVs) and unmanned aerial vehicles (UAVs). In order to ensure a continuous power supply, the functionality and reliability of lithium-ion batteries have received considerable attention. In this paper, a battery capacity prognostic method is developed to estimate the remaining useful life of lithium-ion batteries. This capacity prognostic method consists of a relevance vector machine and a conditional three-parameter capacity degradation model. The relevance vector machine is used to derive the relevance vectors that can be used to find the representative training vectors containing the cycles of the relevance vectors and the predictive values at the cycles of the relevance vectors. The conditional three parameter capacity degradation model is developed to fit the predictive values at the cycles of the relevance vectors. Extrapolation of the conditional three-parameter capacity degradation model to a failure threshold is used to estimate the remaining useful life of lithium-ion batteries. Three instance studies were conducted to validate the developed method. The results show that the developed method is able to predict the future health condition of lithium-ion batteries.

Complete article is available from the publisher and to CALCE Consortium Members.



[Home Page] [Articles Page]
Copyright © 2015 by CALCE and the University of Maryland, All Rights Reserved