Parameter Characterization & Health Estimation With Unscented Kalman Filter Approach For Lithium-Ion Battery Pack In Electric Vehicle Application

In the electric vehicle application, battery models must be accurate and robust for proper operation. Nevertheless, the process of parametrically identifying these models necessitates the utilization of comprehensive and advanced techniques in order to attain sufficient levels of accuracy. In this study, the Hybrid Pulse Power Characterization (HPPC) Test is used to parametrize the lithium-ion battery pack using the 2RC equivalent circuit model and its parameters are identified with the available Optimization Toolbox in the MATLAB. It is widely recognized for its ability to capture the dynamic behavior of batteries under realistic driving conditions. The test foundation is established by utilizing voltage/current profile that closely resembles practical driving conditions of electric vehicles. The performance of the battery pack is evaluated for the UDDS (Urban Dynamometer Drive Schedule) and compared with the original cell characteristics. Furthermore, the Unscented Kalman Filter (UKF)-based state space model is employed to estimate the battery's state of charge and state of health. This estimation is crucial for ensuring reliability, safety, and improved battery life as it is a key indicator for assessing the capacity of lithium-ion (Li-ion) batteries to store and release energy during the stable operation of electric vehicles. The simulation results show the efficacy of approach to provide accurate prognosis of SoC and SoH.