Research on State of Charge Estimation Method for Lithium Iron Phosphate Batteries Based on Expansion Force

To enhance the accuracy and robustness of State of Charge (SOC) estimation for lithium iron phosphate (LiFePO₄) batteries and to overcome the limitations of traditional electrical signal-based methods—such as cumulative errors in Coulomb counting and the need for rest periods in open-circuit voltage (OCV) methods—this study proposes a novel SOC fusion estimation algorithm based on mechanical expansion force signals. Addressing the challenge of feature extraction, a model framework integrating the Sparrow Search Algorithm (SSA), Least Squares Support Vector Machine (LSSVM), and Adaptive Extended Kalman Filter (AEKF) is developed. The state equation is constructed via Coulomb counting, while SSA optimizes the LSSVM to establish an observation model centered on expansion force as the input. The AEKF is employed to achieve real-time, precise SOC prediction. Experimental validation under varying temperatures (25°C, 35°C) and dynamic driving cycles (FUDS, UDDS) demonstrate that this fusion algorithm significantly outperforms traditional electrical signal-based methods, with cumulative SOC estimation errors not exceeding 2.2%. The approach exhibits higher accuracy, improved environmental adaptability, and enhanced robustness. This research confirms the feasibility and effectiveness of using expansion force as a non-electrical quantity for SOC estimation, providing a new perspective for high-precision battery state assessment.