A fused electrical-mechanical model with extended Kalman filter and adaptive weighting for state-of-charge estimation of lithium iron-phosphate batteries

Accurate state-of-charge (SOC) estimation for lithium iron phosphate (LFP) batteries remains challenging due to their characteristically flat open-circuit voltage (OCV) profile and pronounced hysteresis effects. Though recent advances have explored mechanical signals to improve estimation accuracy, the inherent non-monotonic force–SOC relationship and thermal expansion effects introduce additional complexities that hinder practical deployment. To address these challenges, we propose a SOC estimation framework that fuses an equivalent circuit model with an equivalent force model, explicitly accounting for both intercalation-induced and thermally induced stress. The proposed dual-model structure is integrated via a dual extended Kalman filter with adaptive weighting. This approach outperforms conventional methods under diverse operating conditions and demonstrates robustness against common error sources. Hardware-in-the-loop validation further confirms the real-time applicability of the proposed framework. This work offers a practical and accurate solution for SOC estimation in LFP batteries used in electric vehicles and energy storage systems.