Multi-physics-informed fast charging strategy of lithium-ion batteries for aging suppression and thermal safety

Fast charging strategy of lithium-ion batteries (LIBs) attract widespread attention in the context of electric vehicles range anxiety. However, most existing fast charging strategies lead to battery abnormal accelerated degradation or thermal safety issues. To alleviate battery abnormal accelerated degradation and increase thermal safety during rapid charging of LIBs, this study proposed a novel fast charging strategy by considering multi-physics-informed data-driven model. More specifically, an electro-thermal-aging coupled LIB model (ETACM) is developed to comprehensively characterize electrical properties, thermal features and aging dynamics of batteries. Next, to improve charging speed while ensuring thermal safety and battery lifetime, a new multi-constrained optimization problem is formulated. Then, a fast charging optimized framework incorporating improved soft actor-critic (SAC) and long short-term memory network is developed, which can provide new real-time charging strategy for LIBs. Two experimental studies are conducted to validate the effectiveness of the proposed multi-physics-informed fast charging strategy, which is better than existing fast charging methods including constant current-constant voltage (CCCV) charging protocol, SAC-based strategy, model predictive control (MPC)-based strategy. Results demonstrate that the proposed method effectively achieved improvements in charging efficiency, thermal safety and capacity degradation suppression, resulting in at least 11.4% faster charging than the comparative methods without accelerating battery degradation.