Design of Modular Battery Management System with Point-to-point SoC Estimation Algorithm

Abstract

Lithium-ion (Li-ion) batteries are proven to be the best type of battery in the market especially for electric vehicle (EV) applications. However, Li-ion battery must be monitored properly to prevent operation outside the safe operating area (SOA). In using the Li-ion battery, the voltage, current and SoC must be monitored properly to avoid overvoltage, undervoltage, overcurrent, undercurrent, overcharge and overdischarge scenarios which could be harmful to both the Li-ion battery and to the users. Especially in dynamic systems such as EV in which the load supplied by the battery is not constant and could vary drastically. Thus, battery management systems (BMS) are often designed to perform such monitoring functions. BMS must be designed well to fit a specific application. In this study we design a battery management system that is capable of battery voltage and current monitoring, and SoC estimation. For SoC estimation we propose a point-to-point estimation technique which utilizes machine learning algorithm to estimate the SoC consumed as the EV is driven from one point to another.