An Adaptive Battery Capacity Estimation Method Based on Dual-Driven Fusion for Lithium-Ion Batteries in Energy Storage Systems

Accurate estimation of battery capacity is crucial for the safe operation of energy storage power stations. This paper proposes an adaptive battery capacity estimation method based on dual-driven fusion for lithium-ion batteries in energy storage systems. Health features that are highly correlated with the capacity are extracted from historical data as the training dataset for a convolutional neural network, which constructs an online capacity estimation model. Then the discrete Arrhenius degradation model is developed for continuous estimation under time-varying conditions. Finally, a closed-loop Arrhenius model is constructed using the adaptive parameter correction method, based on the difference in capacity estimation between the two models. Experimental results show that the proposed method maintains high prediction accuracy while reducing the burden on battery monitoring equipment.