At present, wind power has been recognized as a primary renewable energy source, however, its large-scale grid integration poses significant challenges to power system stability. To address these challenges, a hybrid energy storage grouping control optimization strategy was proposed. Firstly, based on grid-connected fluctuation standards, adaptive wavelet packet decomposition and fast Fourier transform were employed to derive the grid-connected reference power, and preliminary power allocation within the hybrid energy storage system was carried out. Subsequently, the state of charge (SOC) of the supercapacitor was dynamically optimized using a fuzzy control algorithm, thereby enabling coordinated correction of its power commands. Finally, to mitigate SOC deviation among battery groups, a dynamic three-battery grouping control strategy was developed by introducing a standby battery unit into the dual-battery control framework. Compare with existing battery grouping control strategies through numerical simulations. The results show that the dynamic grouping strategy of three batteries can reduce the average SOC deviation between battery packs by 9.34% and the maximum SOC deviation by 15.57%. The effectiveness of three battery packs in alleviating charge-discharge imbalance and improving equivalent SOC deviation between battery packs is verified.