With the escalating energy and environmental crises, the new energy vehicle (NEV) industry has experienced rapid global expansion. The performance of battery management systems (BMS) and the accuracy of state-of-charge (SOC) estimation are critical for electric vehicle safety. Addressing the limitations of conventional BMS that rely on fixed-parameter equivalent circuit models, this study proposes a novel embedded-oriented battery digital twin framework to dynamically calibrate model parameters and enhance state estimation precision. The research implements parameter identification for the equivalent circuit model and validates SOC estimation using an extended Kalman filter (EKF) algorithm. Experimental results demonstrate the framework’s capability to adapt to dynamic battery characteristics, providing preliminary verification of the proposed digital twin framework’s effectiveness. This approach represents a significant advancement in real-time battery monitoring and management, offering improved reliability for practical BMS applications.