With the large-scale integration of wind power into the grid, the uncertainty of output increases the difficulty of power system dispatch. In this paper, an iterative self-organizing data analysis algorithm is used to segment the wind power prediction value and the corresponding wind power prediction error, and then the prediction error in each wind power interval segment after the adaptive diffusion Gaussian kernel density estimation is used. Then, an optimal scheduling model of the net load prediction error positive and negative rotational reserve capacity probability constraint considering the wind power and load prediction error as a whole is proposed. The probability constraint is converted into an equivalent deterministic constraint by using the opportunity constraint plan. The numerical analysis of three scenarios through the IEEE39 node system shows that after the introduction of the iterative self-organizing data analysis algorithm and adaptive diffusion Gaussian kernel density estimation, the backup cost is reduced by 6.71%, the environmental cost of carbon emissions is reduced by 20.4%, and the total power generation cost is reduced by 2.98%. Finally, the impact of the confidence level on standby capacity and total power generation cost is analyzed. This work is supported by the National Natural Science Foundation of China (No. 62062068 and No. 61761049).