State Estimation of Lithium Battery Degradation Based on Dynamic Complexity Reduction Zonotopic Kalman Filter

To realize the state estimation of lithium batteries under performance degradation, a zonotopic Kalman filter with dynamic complexity reduction is designed. First, a dual polarization equivalent circuit model is constructed, and the zonotopic filter is designed to estimate the state through prediction and update steps, and optimizes zonotope complexity by evaluating variations of the shape generator matrix. Additionally, a dual time-scale framework is introduced for a fast SOC update and slow SOH adjustment. Simulation results show that the proposed algorithm offers a higher accuracy and robustness in capturing battery degradation states.