Intelligent Battery Management Algorithm for Second-Life Battery Systems Using Voltage Based Active Cell Balancing

Battery Energy Storage Systems (BESS) are widely used in the present for domestic and commercial applications. Energy generated from renewable and non-renewable sources will be stored in BESS for utilization on demand. Lithium-ion battery cells are mostly used in present BESS due to the main advantages of high energy density and longer cycle life. Lithium-ion battery cells need strict control of cell voltage and cell temperature within operating ranges using Battery Management System (BMS) for safety and longevity. With the rapidly growing Electric Vehicle (EV) usage in the world, Lithium-ion battery cells used in EVs are used for BESS for second-life usage. Unlike first-life battery cells, Second-Life Battery (SLB) cells tend to develop cell capacity mismatches due to cycle life and non-homogenous temperature variations during first-life usage. The cell capacity mismatch causes cell voltage imbalance which leads to usable capacity reduction and accelerated degradation. The traditional passive BMSs are not sufficient to overcome cell capacity mismatch in SLB cells. This paper presents Dynamic Time Warping (DTW) based relative cell capacity estimation with Reinforced Learning (RL) based active cell balancing algorithm. A simulated virtual model of a SLB system is used for the analysis which replicates the actual battery cell parameters. The performance of the introduced algorithm is evaluated in comparison with a conventional passive BMS algorithm. Compared with the conventional approaches, the proposed method eliminates the exposure of the low-capacity cells to extended highest and lowest cell voltage conditions improving the overall performance of SLB systems.