Health and safety enhancement of lithium-ion batteries for fast-charging

Abstract

With the ever-increasing demand for battery electric vehicles, the expectation of safety, range, and fast charging has become a major focus for industry players and researchers. The research presented herein sought to improve the safety of Lithium-ion (Li-ion) batteries and the health of the cells while fast charging. The work presented has introduced recurrent neural network models and other innovative processes for the estimation of the core temperature in Li-ion cells, and the detection of thermal and voltage faults that could cause thermal runaway. A novel health-conscious fast charging (HCFC) strategy for Li-ion battery packs validated through experimentation was also developed to minimize the charge time while maximizing the longevity and useable capacity of the battery pack. Results from this research verified the reliability of recurrent neural networks for battery signal estimation and fault detection while also demonstrating the performance of the HCFC strategy compared to the standard approach.