Battery state of charge estimation by using a combination of Coulomb Counting and dynamic model with adjusted gain

Abstract

Majority of electric vehicles depend on batteries that require a Battery Management System (BMS). Increasing accuracy of State of charge (SoC) estimation was crucial. The uncomplicated method for SoC estimation is Coulomb Counting (CC), however it has drawback in information of previous SoC initial value and tolerate with error propagation. Dynamic model equipped with integral controller can increase accuracy of the CC method. In this paper adjusted gain of the integral controller is proposed to further improve performance of SOC estimation. The Li-ion battery was modelled using first order RC battery model. Least square method is used to determine parameter of battery model. The simulation results show that the combination of CC and dynamic model with fixed integral gain could increase accuracy of the CC algorithm approximately to 87%, and the mixed algorithm with adjusted integral gain could further increase it to around 89%.