Hierarchical predictive battery thermal management strategy architecture for automotive fast charging

Electric vehicle battery systems exhibit significant thermal hysteresis due to their large heat capacity and limited thermal control capabilities. Thus, it is challenging to adjust the battery system temperature quickly and effectively during dynamic driving scenarios. This paper proposes a hierarchical predictive battery thermal management (BTM) strategy architecture that aims to predict future battery temperature and preemptively precool the battery for fast charging during journey scenario. Strategy architecture includes three layers. The upper layer estimates the future battery system's output power demand based on vehicle navigation and history information. The middle layer predicts the battery temperature changes using battery electrothermal model. The lower layer employs dynamic programming algorithms to determine the optimal inlet temperature trajectory of the battery liquid cooling plate over the entire driving journey, which is then followed by the BTM system. The results demonstrate that, in comparison with rule-based strategies, the proposed strategy can achieve energy-optimal intelligent precooling prior to fast charging, effectively preventing battery temperature from exceeding the threshold during fast charging. Furthermore, this strategy also exhibits notable advantages in vehicle thermal management miniaturization and extreme fast charging thermal management.