NMPC-Based Performance Evaluation of Active Balancing Networks of Li-ion Batteries for Overnight EV Charging

A battery pack plays a vital role in an Electric Vehicle (EV), for it stores the energy and releases it to provide a reliable energy output. However, because of variations and imbalance in individual cell characteristics, the cells connected in a pack do not contain the same State of Charge (SoC) level—leading to poor EV performance and battery health deterioration. Therefore, in this work, active cell balancing is performed during overnight charging to ensure the same charge level among the cells by the end of the charging cycle. For this purpose, a high-fidelity mean current model is presented for a switched capacitor- (SC) based architecture comprising <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">N</i> cells for a generalized battery pack. Furthermore, this model is incorporated into a feedback control system utilizing Non-linear Model Predictive Control (NMPC). Since balancing losses are a primary factor during overnight charging, the SC-based NMPC is also compared with a buck-boost (BB) converter architecture under the same conditions. For each framework, extensive simulations are conducted by varying the weights of different terms in the cost function of the NMPC problem. The simulation results demonstrate that both SC- and BB-based architectures have been able to achieve balancing, wherein the average balancing time of SC is 186 % more than that of BB. However, in terms of average power losses and energy dissipation, SC incurred 288 % and 31 % fewer losses than BB, respectively.