Remaining Useful Life Prediction for Lithium-Ion Batteries Based on Exponential Model and Particle Filter

As the secondary widely used battery, lithium-ion batteries (LIBs) have become the core component of the energy supply for most devices. Accurately predicting the current cycle time of LIBs is of great importance to ensure the reliability and safety of the equipment. In this paper, considering the nonlinear and non-Gaussian capacity degradation characteristics of LIBs, a remaining useful life (RUL) prediction method based on the exponential model and the particle filter is proposed. The cycle life test data of LIBs published by prognostics center of excellence in national aeronautics and space administration were exponentially experiencing the rule of degradation. And then the extrapolation method was used to get the quantitative expression of the uncertainty of life expectancy of LIBs, i.e. the prediction mean and the probability distribution histogram. The prognostic horizon index and the new specific accuracy index were applied to evaluate the prediction performance. Moreover, the prediction error under different prediction starting points is given. Compared with other methods such as the auto-regressive integrated moving average model, the fusion nonlinear degradation autoregressive model and the regularized particle filter algorithm, the proposed algorithm has a better prediction performance. According to the accuracy index, the proposed prediction method has better prediction accuracy and convergence. The RUL prediction for LIBs can provide a better decision support for the maintenance and support systems to optimize maintenance strategies, and reduce maintenance costs.