State of Charge Estimation of Lithium Ferro Phosphate Battery Using Extended Kalman Filter

Abstract

Rapidly evolving technology of smart grid and Electrical Vehicle (EV) requires efficient energy storage device. So it is important to improve performance of Battery Management System (BMS). One of the challenges in BMS is accurate estimation of State of Charge (SOC). SOC is important in determining range of Electrical Vehicle (EV), since range anxiety is one of the biggest obstacles in getting EV cost competitive and popular. An accurate SOC estimation can enhance the performance of the battery and ensures safe operation of the battery. SOC estimation algorithm is expected to be accurate and easy to implement. In this work Thevenin’s 3rd order battery model is considered, Open Circuit Voltage (𝑉𝑂𝐶) is a function of SOC, nonlinear relationship between 𝑉𝑂𝐶 and SOC is described by curve fitting of 5th order polynomial equation. Resistances and Capacitances in battery model are function of SOC which is measured at 0%,25%,50%,75%,100% SOC by Electrochemical Impedance Spectroscopy (EIS) and relationship between these two is established using 5th order polynomial. SOC of Lithium Ferro Phosphate Battery is estimated using Extended Kalman Filter (EKF) algorithm. The algorithm is tested using Battery Management Unit Simulator (BMUS) and PIC32 Microcontroller. Mathematical results of SOC estimation using Coulomb count are compared with the experimental results to validate the EKF algorithm for SOC estimation.
Keywords: Battery models, state of charge, Kalman filter, open circuit voltage