2018 IEEE International Conference on Prognostics and Health Management (ICPHM), 11-13 June, 2018, Seattle, WA, USA

Anomaly Detection During Lithium-ion Battery Qualification Testing

Saurabh Saxena1, Myeongsu Kang1, Yinjiao Xing1, and Michael Pecht1,
1Center of Advanced Life Cycle Engineering University of Maryland College Park, USA


Qualification testing of Li-ion batteries usually involves battery capacity fade trend monitoring over a large number of repeated charge/discharge cycles. However, due to manufacturing-induced variations, capacity fade trends of batteries from the same as well different production lots can differ from each other. This paper discusses a real-world problem where the Li-ion batteries from a particular production lot exhibit a different capacity fade trend than that exhibited by healthy batteries from earlier production lots. The paper also outlines the approach to identify these anomalous batteries at the earliest stage of testing and provides a case study that applies one-class support vector machine (SVM) based methods to detect the anomalous capacity fade behavior.

This article is available online here and to CALCE Consortium Members for personal review.

[Home Page] [Articles Page]
Copyright © 2018 by CALCE and the University of Maryland, All Rights Reserved