High-Dimensional State Estimation Using an Adaptive Ensemble Adjustment Kalman Filter for Lithium-Ion Batteries

An efficient model-based state estimation method for lithium-ion (Li-ion) batteries is proposed using an extended electrochemical single particle model enhanced by electrolyte and thermal dynamics. To address the spatially distributed characteristics across the particle radius and the electrode width, an adaptive ensemble adjustment Kalman filter (AEAKF) algorithm is established to effectively reduce the computational burden caused by the high-dimensional nonlinear model for online operation. The uncertainties in the modeling and the measurement error are considered in the proposed algorithm by adaptively adjusting the error covariance matrices in the filtering scheme. The performance of the AEAKF-based distributed battery state estimator is examined via simulation by comparing it with other nonlinear estimation techniques. It is shown that the proposed method exhibits excellent performance in terms of prediction accuracy and computational speed, while the capability to signify the degradation-related internal states can be utilized for advanced health-aware battery management systems designed for various applications such as electric vehicles and smart grids.