Toward dynamic programming-based management in reconfigurable battery packs

Abstract

Well-designed battery energy management algorithms are integral and important parts of battery management and maintenance in various applications ranging from smart grid backup systems to Electric and Hybrid Electric Vehicles (EV/HEV). Management in smart reconfigurable battery systems tend to be more complicated, flexible, and sophisticated since the systematic ability allows access to individual cells for monitoring and control purposes in a real-time fashion. Therefore, in this study, based on our previous work on adaptive reconfigurable battery network, a dynamic programming based management strategy is proposed and validated to fully utilize systematic ability and to optimize energy efficiency while keeping cell to cell states balanced and ensuring safety. To validate the proposed algorithm, battery model is first set up using experimental data on 26650 lithium ion batteries with the help of Arbin test-bed. Simulations are then conducted using the established battery models to figure out the gain in terms of energy efficiency of the proposed algorithm compared with traditional fixed battery system design. The comparison indicates the effeteness of the algorithm. Furthermore, the proposed method is applicable to battery packs with different types of battery cells in terms of capacity and electrochemistry, which is the key idea of software defined battery.