For the scenario of wind power fluctuation smoothing, taking the wind and energy storage combined power-generation system as the research object, a FIR filter combined with smooth strategy of wind power fluctuation of amplitude reduction link is proposed, and then the optimization dispatching of hybrid energy storage is realized through adaptive reconstruction based on EMD of filtered fluctuation components. In this strategy, a FIR filter that can dynamically adjust the combination of filtering order and cut-off frequency is selected for filtering, and FRA amplitude reduction link is introduced to address high-amplitude data points that FIR filters cannot smooth well, so as to smooth wind power fluctuations. The smoothed and filtered wind power fluctuations are used as hybrid energy storage power for EMD decomposition, and the IMF component is adaptively reconstructed into two groups of high-low frequency power allocated to super capacitors-batteries respectively by an algorithm, so as to obtain the optimal energy storage optimal dispatching parameters. Through the numerical example verification analysis, the FIR filtering added with FRA can adapt to the power characteristics of different frequencies and amplitude, effectively reduce the power fluctuations of 1 min and 10 min scales, and avoid the problem of increasing the energy storage burden caused by excessive smoothing. The adaptive reconstruction of IMF components after EMD decomposition can obtain the power of batteries and super capacitors to compensate for different charge and discharge characteristics, significantly reducing the energy storage output burden and improve dispatching economy.