A comparative study of kalman filtering based observer and sliding mode observer for state of charge estimation

Abstract

Nowadays, electric mobility is starting to define society and is becoming more and more irreplaceable and essential to daily activities. Safe and durable battery is of a great significance for this type of mobility, hence the increasing interest of research activity oriented to battery studies, in order to assure safe operating mode and to control the battery in case of any abnormal functioning conditions that could damage the battery if not properly managed. Lithium-ion technology is considered the most suitable existing technology for electrical storage, because of their interesting features such as their relatively long cycle life, lighter weight, their high energy density, However, there is a lot of work that is still needed to be done in order to assure safe operating lithium-ion batteries, starting with their internal status monitoring, cell balancing within a battery pack, and thermal management. Tasks that are accomplished by the battery management system (BMS) which uses the state of charge (SOC) as an indicator of the internal charge level of the battery, in order to avoid unpredicted system interruption. Since the state of charge is an inner state of a the battery which cannot be directly measured, a powerful estimation technique is inevitable, in this paper we investigate the performances of tow estimation strategies; kalman filtering based observers and sliding mode observers, both strategies are compared in terms of accuracy, design requirement, and overall performances.