A MSTUKF-based Technique for SOC Estimation of Li-ion Batteries for Electric Vehicles

Lithium-ion batteries have received widespread use by electric vehicles because of their high energy density, and so forth. Accurate estimation of state of charge (SOC) is very important for the high efficiency use and lifetime prolong of batteries. In this work, a strong tracking unscented Kalman filter method using multiple sub-optimal fading factors (MSTUKF) is employed to estimate lithium ion battery SOC. This method introduces multiple sub-optimal fading factors to traditional unscented Kalman filter algorithm, which effectively solves the robustness problem of the unscented Kalman filter method facing with the state mutation, and the model mismatch for battery degradation. Simulation results prove that this MSTUKF based method has a smaller prediction error rate than the traditional unscented Kalman filter method. The results also showed that this MSTUKF based method can be used for lithium-ion battery management of electric vehicles.