Online Model Identification for State of Charge Estimation for Lithium-ion Batteries with Missing Data

Online model identification is critical for determining the accurate state of charge (SOC) for a battery based on a model, which depends on the use of complete and reliable measurement data. To ensure precision for parameter identification and reliability for SOC estimation in practical applications, the occurrence of data loss and noise interference must be considered. In this paper, a first-order resistor- capacitor equivalent circuit model is developed to simulate the "black box" system of a battery. A recursive least square method based on a variable interval auxiliary model is proposed to compensate for the missing data in an unreliable actual environment. Meanwhile, a forgetting factor is introduced to prevent the influence of historical data in parameter identification. To further reduce the noise effects on SOC estimation, the extended Kalman filter (EKF) is applied to the algorithm. The proposed method is verified using CALCE experimental data. The experimental results show that the proposed method can be used to realize accurate model parameter identification and reliable online SOC estimation under conditions of data loss and noise interference.