A Novel Method for SOC Estimation of Dynamic Reconfigurable Battery Networks

With the increasing integration of renewable energy sources into power systems, Battery Energy Storage Systems (BESSs) play a critical role in mitigating power fluctuations. Dynamic Reconfigurable Battery Networks (DRBN) enhance operational efficiency by dynamically balancing battery units, but accurate State of Charge (SOC) estimation remains challenging due to pulsed currents and parameter uncertainties. This paper proposes a novel SOC estimation method tailored for DRBN systems. First, a hybrid parameter identification approach combines offline and online techniques to address the dynamic operating conditions of DRBN. Offline methods leverage pulsebased responses, while online methods employ Recursive Least Squares (RLS) for real-time parameter updates. Second, an Adaptive Extended Kalman Filter (AEKF) is integrated to iteratively refine SOC estimates, improving robustness against noise and initial value errors. Experimental validation on a 3P14S DRBN setup demonstrates the method's effectiveness under noisy signals and sparse data, outperforming traditional coulomb counting. The results highlight enhanced accuracy and adaptability, making the method suitable for real-world DRBN applications.