Modelling and Simulation of DC/DC Converter-Based Active Cell Balancing for Battery Management Systems

Abstract

Abstract The transition to sustainable transportation was initiated to mitigate the effects of global warming and decrease CO2 emissions. Electric vehicles are at the forefront of this revolution and rapid technological advancements have been made in the development of electric powertrain’s components, particularly lithium-ion (Li-ion) batteries. These breakthroughs have played a significant role in increasing vehicle performance and range, prompting automotive companies to accelerate their transition to more sustainable vehicles. Li-ion batteries are prominent in their specific energy and specific power while the battery pack comprises low-voltage battery cells connected in series and parallel to meet the voltage and current requirements of the electric vehicles. However, since the cells are limited by voltage and capacity, a serious inconsistency between the cell’s voltage and state of charge (SoC) is generated because of the manufacturing inconsistencies of the individual cells in the battery pack. This leads to growth over time after a number of charging and discharging cycles because of different self-discharge rates, internal resistance, and operating temperature, which will affect the efficiency and life of the entire battery pack. It has become inevitable to keep the cells balanced to achieve the effective usage of energy and to enhance the battery life. This thesis starts with a comprehensive literature study and investigation of different types of cell balancing techniques with a focus on converter-based active balancing. A choice has been made after evaluating the performance and suitability of different converters considering the balancing method, balancing speed, balancing time, and control complexity. In MATLAB/Simulink, a bi-directional buck-boost converter was designed and simulated by integrating with the battery string, performing energy exchange between cells with different SoCs using a state-space modeling approach and cascaded PID control. An easy-to-implement and effective algorithm has been developed, tested, and validated to perform the balancing action in various operating scenarios.