LAUSR

Abstract Overcharging is one of the main reasons causing lithium-ion battery thermal abuse, probably leading to vehicle accidents. This paper develops an impedance-based method to characterize the battery heat generation during overcharging process. An electro-thermal model is adopted for better computation efficiency. A series of overcharging experiments at 30 ℃ and 60 ℃ are conducted. Interestingly, three stages can be identified from the results, which are the normal heat-accumulating stage, fast heat-accumulating stage and thermal runaway stage, respectively (Stage I, II and III). During Stage I and II, pulse-relaxation and impedance-measurement methods are developed to parameterize the electro-thermal model, under different state of charge, temperature and charging rate conditions. Results of genetic algorithm with Hybrid Pulse Power Characteristic cycling data are used as benchmark. The simulated surface temperature results during overcharging are validated via experiments, which shows that medium frequency impedance method outputs better equivalent resistance and surface temperature estimation accuracy. The proposed model achieves to reduce the temperature estimation root mean squared error to under 0.9 ℃ in all overcharging situations, with greatly reduced computation complexity.