Evaluation of Active Balancing Algorithms and an Improved Method for a Deployed Active Battery Balancer as Well as Physical Implementation

Abstract

<p><strong>Abstract</strong></p> <p><strong> </strong>Lithium-ion cells have been the workhorse for customer products including laptop computers, cell phones and battery energy storage systems that can store energy from renewable energy sources. The internal resistance of them is very low, resulting in a low amount of wasted energy. Performance of Li-Ion cells is wonderful if they treated well. For this reason, an efficient battery management system (BMS) is essential in order to maximize the battery's capacity, battery means a collection of cells wired in series providing a higher voltage. So by using a BMS, SOC (State Of Charge) levels of cells get closer to each other resulting increased life and capacity of a battery. There are active and passive battery balancers. The energy which is extracted from those cells with higher SOC using passive balancer is wasted in heat, so it is not so efficient in terms of wasted energy in comparison with active balancing algorithms. There are different types of active balancing techniques. Cell to cell technique, cell to battery and battery to cell. Among the above mentioned, cell to battery performs well. However, when push-pull converters (which have high performance) are used, flux imbalance phenomenon ( which is resulted in a asymmetric hysteresis loop of a magnetic core) is unavoidable. In order to prevent this phenomenon, current mode topology is used. Hence, the transistors won't burn out due to this effect. In this thesis I simulated current mode algorithms with push-pull converters with MATLAB SIMULINK. I also physically designed push-pull converters and built it with help of chips ( current mode controllers, optocouplers, transistors, fast recovery diodes, linear and shunt regulators). I did simulate two active balancing methods ( Cell to cell and cell to battery) and compared the results. The results that came from cell to battery indicated a better performance in terms of balancing speed. For three cells balancing time reduced from 3570 seconds in cell to cell method to 518 seconds in cell to battery method.</p>