Improved Coulumb Counting method for state of charge estimation of Lithium-ion battery with temperature consideration

Abstract

In recent decades, Lithium-ion (Li-ion) battery has been dominating the battery technology in wide range of electronics application. A good utilization of battery technology comes with a proper management of its State of Charge (SOC) and state of health. Having a proper management and estimation of SOC will ensure that the life span of battery can be prolong, as well as provide the safety to the users. Therefore, it is very crucial to have an efficient and accurate SOC estimation method. This project proposes an improved Coulumb Counting (CC) method for the SOC estimation of Lithium-ion battery. The conventional CC method suffers from few drawbacks in which it strongly depends on the precision of current sensors, initial value of SOC and the nominal capacity of the battery. A lot of research has been conducted to explore the possibility of improving the CC method. Most of the methods have combined CC method with different techniques of model-based method to overcome the aforementioned drawbacks. However, the research that has been performed only cover one or two limitations at a time. Therefore, this research is conducted to study an improved CC method with the focus on diminishing these three main drawbacks of CC method. The determination of initial SOC and calibration of nominal capacity has been performed by using SOC-OCV data curve which was obtained using the discharge pulse experiment of actual 18650 Li-ion battery. Meanwhile, a temperature compensation circuit is introduced to eliminate the temperature drift of current sensor. This project simulates the effectiveness of this improved CC method using MATLAB Simulink software under wide range operating temperature. An equivalent circuit model of 18650 Li-ion battery cell is used in the simulation to study the effectiveness of proposed method. The simulation result of this improved CC method shows a slight improvement in reducing the accuracy error after implementing each of the method proposed to reduce the drawbacks of CC method. The comparison of the results between conventional CC method and improved CC method implies that the improved CC method able to reduce the disadvantageous of conventional CC method. This method is envisioned to be useful for the future engineer and designer who would like to estimate the SOC of battery using an improved CC method with a higher accuracy requirement.