Simulink Model for SoC Estimation using Extended Kalman Filter 2021-26-0382

State of Charge (SoC) estimation of battery plays a key role in strategizing the power distribution across the vehicle in Battery Management System. In this paper, a model for SoC estimation using Extended Kalman Filter (EKF) is developed in Simulink. This model uses a 2nd order Resistance-Capacitance (2RC) Equivalent Circuit Model (ECM) of Lithium Ferrous Phosphate (LFP) cell to simulate the cell behaviour. This cell model was developed using the Simscape library in Simulink. The parameter identification experiments were performed on a new and a used LFP cell respectively, to identify two sets of parameters of ECM. The cell model parameters were identified for the range of 0% to 100% SoC at a constant temperature and it was observed that they vary as a function of SoC. Hence, variable resistance and capacitance blocks are used in the cell model so that the cell parameters can vary as a function of SoC. This facilitates the simulation of voltage drop due to internal resistances of the cell at the whole range of SoC, as it is impractical to measure the internal resistances of the cell online. In the EKF algorithm, the Coulomb Counting model equation is used to predict the SoC. In the Measurement Updating step of the algorithm, the predicted SoC is used to calculate the Open Circuit Voltage (OCV) using an SoC-OCV Lookup table. Then using the calculated OCV and the simulated voltage drop due to internal resistance of the cell, the Terminal Voltage of the cell model is calculated and compared with the Terminal Voltage measured by sensors and then the SoC is estimated. The results for SoC estimation using the two different sets of cell parameters were compared with the experimental SoC.