Model-Based Quantitative Analysis of Power Losses-Aware Active Cell Balancing Networks With Load

Cell balancing has a paramount significance in a battery management system (BMS) for improving battery performance and safety. Performance characterization of different active cell balancing networks (ACBNs) is important to understand their utilization. In this work, switched capacitor (SC), buck-boost (BB), and flyback (FB) ACBNs with static and dynamic parameters are quantitatively analyzed during charging/discharging and static conditions. For this purpose, a high-fidelity mean current modeling approach is employed, and an urban dynamometer driving schedule (UDDS) input current profile is used. The analysis of each ACBN is verified by incorporating a practical scenario in Simscape toolbox of SIMULINK. A state-dependent Kalman filter (SDKF) is also designed to estimate the state of charge (SoC), as precise estimation of the SoC is critical for determining the performance of any balancing architecture. The sensitivity analysis manifests that consideration of static and dynamic parameters results in a percentage change of 26.75%, 80%, and 108% in the cell balancing speed of SC, BB, and FB, respectively. Furthermore, it is observed that the trade-off exhibited in the efficiency and cell balancing speed of balancing architectures. It is worth observing that the SC network is efficient but slower than BB and FB cell balancing techniques. While the BB is slower than FB but almost similar in terms of efficiency.