Estimating the remaining useful life of Li-ion batteries with a Bayesian updating model

In this paper, we studied a prediction method for the remaining useful life of Lithium-ion batteries. First, a battery degradation model is obtained based on exponential degradation signal modeling with data collected from second generation 18650-size lithiumion cells from NASA. Using a Bayesian updating procedure, we then obtain the conditional cumulative distribution function (cdf) of the residual life of the battery at various time intervals. Finally, we discuss this method and draw the conclusion that the model is accurate in terms of prediction.