Li-Ion battery sizing and dynamic programming for optimal power-split control in a hybrid electric vehicle

This study considers the tradeoff between the weight and the storage capacity among different sizes of LiFePO4 battery packs for optimal powertrain design in a parallel hybrid electric vehicle (HEV). Proper power-split strategy between the engine power and the battery power can improve the performance of the designed powertrain. Dynamic programming (DP) approach is adopted to compute the optimal power-split and to evaluate the performance of each battery storage capacity. The fuel economy of HEV with different battery capacities are evaluated against different driving schedules i.e. Japan 10/15 mode, UDDS, HWFET, UN/ECE, and SFTP. Since the solution to the DP is computed offline (within a few minutes for each profile), some rule extractions are illustrated and can be used to improve the powertrain control strategy onboard.