State-of-charge inconsistency estimation for li-ion battery pack using electrochemical model

In order to maximize capacity utilization and guarantee safe operation of Li-ion battery pack, state-of-charge (SOC) inconsistency estimation is essential. And estimating cell electrochemical internal variables is also the requirement of next generation battery management system (BMS). However, it is challenging for the BMS in electric vehicles due to dynamic current conditions and limited computational resources. This paper proposes a novel approach to estimate SOC inconsistency for Li-ion battery pack with reference-plus-difference model where the reference model represents the overall performance of the battery pack. An electrochemical model is used as the reference model for its high-fidelity, low parameter sensitivity, and the capability of estimating cell electrochemical internal variables. Then SOC difference of each battery is achieved by bias correction technique using readily available measurements. The proposed approach is verified by experiments and simulations which show a satisfactory balance between estimation accuracy and computational burden. Based on the SOC inconsistency, battery equalization can be further implemented.