Online estimation of negative electrode overpotential and detection of lithium plating of batteries using electrochemistry-driven Kalman filter closed-loop framework

Diagnosing lithium plating in batteries can be achieved by monitoring the polarity of the negative electrode overpotential (NEO), which cannot be directly measured for commercial batteries. The electrochemical models have the ability to calculate the NEO, however, existing methods cannot guarantee estimation accuracy and computational efficiency simultaneously under complex working conditions, limiting their practicality. To address the problem, this work proposed a novel closed-loop NEO estimation framework that combines a simplified electrochemical model with a filtering algorithm. First, a parameter-corrected SP+ model with practical applicability under high C-rate conditions is built. Following that, a closed-loop NEO estimation framework involving the extended Kalman filtering (EKF) is constructed, in that, NEO is treated as the only state variable and adjusted by terminal voltage observations. Then the proposed method is validated experimentally by measuring NEO using a 50-Ah LFP battery with a referenced electrode. Results showed that the accuracy, efficiency, convergence, and robustness can all be guaranteed benefiting from the simplicity of the electrochemical model and the adaptiveness of the EKF, and the proposed method is well-suited for online monitoring lithium plating in battery energy storage applications.