Accurate State of Charge Estimation in Lithium‐Ion Batteries by Second‐Order Sliding Mode Observer

ABSTRACT Accurate state‐of‐charge (SoC) estimation in lithium‐ion batteries is crucial for efficient energy management, safe operation, and extended battery lifespan. Although sliding mode observers (SMOs) are widely used for this purpose, conventional first‐order designs often suffer from chattering and slow convergence, resulting in noisy and less reliable estimation signals. This paper proposes a finite‐time second‐order sliding mode observer (SO‐SMO) for accurate SoC estimation based on an equivalent circuit model of the battery. The proposed observer analytically derives a closed‐form expression for the finite convergence time, enabling predictable estimation dynamics. Moreover, it eliminates chattering and significantly improves estimation smoothness and robustness against modeling uncertainties and measurement noise. A comparative analysis with the Extended Kalman Filter and traditional SMO demonstrates that the proposed method achieves higher estimation accuracy and faster convergence while maintaining lower computational complexity, making it well‐suited for real‐time applications. Theoretical analysis and simulation results confirm that the SO‐SMO offers a superior balance between accuracy, robustness, and efficiency, establishing its potential for next‐generation battery management systems in electric and hybrid vehicles.