Model Parameter Identification of State of Charge Based on Three Battery Modelling using Kalman Filter.

State of Charge (SOC) is the ratio of current versus total capacity of the battery. In the context of Battery Management System (BMS), the SOC is estimated by using a battery model. In this research, three battery models were presented, including (1) Thevenin battery model, (2) modified Thevenin battery model, and (3) simple battery model. Then, the SOC of those battery models was estimated using Coulomb Counting, Open Circuit Voltage (OCV), and Kalman Filter method. The simulation evaluated the performance of the SOC estimation methods, including the correctability of SOC initialization error. The simulation results showed that the proposed battery models could accurately estimate SOC. In terms of SOC initialization error, the Coulomb Counting, OCV Model 1, and OCV Model 2 could not correct the initialization error of SOC. However, the application of OCV Model 3 and Kalman Filter could provide an accurate SOC estimation with excellent correction of SOC initialization error. Compared to OCV model 3, the error correction in the Kalman Filter method was performed 25 minutes faster. Therefore, this finding suggests that Kalman Filter is the most suitable estimation method for BMS due to the high accuracy of SOC estimation (RMSE = 0.0014) and fast correction of SOC initialization error (time < 20 seconds). [ABSTRACT FROM AUTHOR]