State Of Charge estimation using Extended Kalman Filter in Electric Vehicles

Abstract

Because of their inherent safety, fast charging capacity, and extended cycle life, lithium-ion batteries are preferred over other types of batteries in electric vehicle applications. It's critical to be able to determine state factors like state of charge and state of health in order to generate an accurate battery model. The state of charge estimation algorithms for generic Lithium-ion batteries was enhanced using LA92 drive cycle experiment data. To begin, a mathematical model for an analogous circuit battery was created with the goal of accurately imitating the behavior of a lithium-ion battery. The Thevenin model is created by 2 RC branches and identifies the model parameters with the Extended Kalman Filter. The Hybrid Pulse Power Characterization (HPPC) test data obtained at 40°C, 25°C, 10°C, 0°C, and -10°C are used to calculate the SOC 3-dimensional curve as a function of SOC and T. A comparison of the two methods is shown, indicating that the EKF method of battery SOC evaluation is more accurate than the coulomb counting method. The error observed from the EKF results is less than 1% and it shows EKF is reliable for estimating the battery's states. Keywords; Electric Vehicles (EV); Lithium-ion batteries (Li-Ion); State of Charge (SOC); Extended Kalman Filter (EKF)I.