Choosing the best lifetime model for commercial lithium-ion batteries

Abstract

This paper explores three-lifetime models for the commercial Lithium-Ion Batteries, namely, Weibull, Lognormal and Normal distributions. A comparative study is performed on the censored data of an accelerated lifetime test (ALT) to select the best of the proposed models. Although the three models fit the real experimental data and provide similar estimations to the mean time to failure of the batteries, the effect of censoring in estimating the parameters of the distributions is entirely different. The work uses three standard criteria to determine the preferable model. In brief, it is found that the lognormal distribution is the best among the three suggested models.