A novel temperature-compensated model for power Li-ion batteries with dual-particle-filter state of charge estimation

The accurate state-of-charge (SOC) estimation of power Li-ion batteries is one of the most important issues for battery management system (BMS) in electric vehicles (EVs). Temperature has brought great impact to the accuracy of the SOC estimation, which greatly depends on appropriate battery models and estimation algorithms. The fact that the model parameters, such as the internal resistance and the open-circuit voltage, are dependent on battery temperature and current detection precision is greatly related to the drift noise in current measurements will lead to errors in SOC estimation. Aiming at this problem, we present a temperature-compensated model with a dual-particle-filter estimator for SOC estimation of power Li-ion batteries in EVs. To overcome the effect of model parameter perturbations caused by temperature, a practical temperature-compensated battery model, in which the temperature and current are taken as model inputs, is presented to study and describe the relationship between the internal resistance, voltage and the temperature comprehensively. Additionally, the drift current is considered as an undetermined static parameter in the battery model to eliminate the effect of the drift current. Then, we build a dual-particle-filter estimator to obtain simultaneous SOC and drift current estimation based on the temperature-compensated model. The experimental and simulation results indicate that the proposed method based on the temperature-compensated model and the dual-particle-filter estimator can realize an accurate and robust SOC estimation.