A Novel Active Equalization Method for Series-Connected Battery Packs Based on Clustering Analysis With Genetic Algorithm

Abstract

Battery pack performance is the main concern for electric vehicles and energy storage systems. However, charge imbalance is inevitable due to inconsistent manufacturing techniques and environmental conditions. Charge imbalance reduces the power performance and available energy of battery packs. Hence, it is necessary to perform battery equalization. This article proposes an active equalization circuit and a novel equalization strategy to achieve energy redistribution. A bidirectional equalization topology consisting of a forward transformer and switch matrix is proposed first. Then, an innovative equalization strategy based on clustering analysis and genetic algorithm (GA) is developed. Clustering analysis is introduced to identify the target cells to be balanced. To further increase the speed of the equalization process, GA is employed to optimize the classification results. Finally, a series of experiments were conducted to confirm the effectiveness of the proposed topology and the strategy. Both the simulation and experimental results validate that the proposed equalization strategy not only improves the inconsistency but also increases the equalization speed. In practical battery pack experiments, the pack capacity is improved by 16.84% after equalization, and the equalization time is decreased by 23.8% using the proposed method.