Real time SOC estimation for Li-ion batteries in Electric vehicles using UKBF with online parameter identification

Abstract

Abstract In the recent era, Lithium ion batteries plays a significant role in EV industry due to their high specific energy density, power density, low self-discharge rate, and prolonged lifespan. Modeling the battery precisely and estimating its State of Charge with great precision is essential to improve the performance of the lithium-ion batteries. Though numerous methods has been proposed for estimating the SOC, accurate estimation approach is not proposed yet since all these approaches consider the discrete-time dynamics of the battery. Hence in this proposed approach, the implementation of Thevenin 2RC battery model in conjunction with the Unscented Kalman Bucy Filter (UKBF) for SOC estimation is suggested. Thevenin 2RC battery model is used to captures the nonlinear relationship between the battery’s voltage, current, and SOC. The UKBF is then used to estimate the SOC by fusing the battery model with noisy measurements of the battery’s voltage and current. The UKBF is able to handle the nonlinearity of the battery model and the noise in the measurements, resulting in a more accurate estimate of the SOC by capturing the continuous-time dynamics of the battery. The model is simulated in Matlab Simulink. With similar covariance noise and measurement noise taken into consideration, the battery’s SOC is estimated using the EKF, UKF, and UKBF. The performance comparison indicate that the UKBF approach provides an accurate estimation of the SOC, with a significantly lower RMSE of 0.003276.

Keywords

• Electric Vehicle
• Lithium ion battery
• State of charge
• Thevenin battery modelling
• Unscented Kalman Bucy Filter