A novel model-driven method for lithium-ion battery remaining useful life prediction

Accurate prediction of lithium-ion battery remaining useful life (RUL) is a critical technology for battery management system. It can help maximize the battery performance and avoid accidents. By comprehensive comparison, model-driven method is an effective method for revealing the mechanism of ageing process and compensating the effects of temperature on capacity. This paper presents a novel model-driven method based on capacity fading micro model (CFM) and temperature-related capacity compensation model (TRCC). These two models are obtained based on the state transition of lithium ions in the micro level. CFM has the same form as dual-exponential empirical model. By mutual authenticating, they strengthen the effectiveness and credibility. TRCC is a parabolic model. It can compensate for the capacity related to temperature. The novel RUL prediction method consists of compensation and Particle Filter process. An experiment system is built to verify the models and RUL prediction method. According to the experiment results, it can be seen that this novel method can improve the accuracy of RUL prediction.