Digital Twin Driven Reconfiguration of Li-ion Batteries with Capacity Delivery Maximization

Traditional fixed-topology battery equalization circuits are typically difficult to deliver maximal capacity of battery pack due to the state-of-health (SoH) unbalance of cells. To address this challenge, a dynamical reconfigurable battery management method is proposed based on the digital twin model of batteries. First, the digital twin model is developed to estimate the SoH distribution of batteries in the pack. Then, a capacity delivery maximization reconfiguration problem is constructed by utilizing the SoH distribution of batteries. A dynamic programming method is used to solve this problem by changing the topology of reconfigurable batteries. Finally, the validity of the reconstruction strategy is verified by extensive experimental data of Li-ion batteries. Experiment results show that capacity utilization of the battery pack is significantly improved when compared with existing methods.