Abstract: In order to solve the problem of the peak pattern of the wind speed distribution mismatching with the fitting function,a wind speed probability distribution model based on peak pattern identification is proposed. Before fitting the wind speed distribution,the fitting function is selected on the basis of peak pattern identification,and the probability density function of the wind speed distribution is estimated. Aad then,the wind power density is estimated by using the parameters of the mixed double Weibull distribution function.The calculation results show that the wind speed probability distribution model based on peak is pattern identification is accurate in identifying various peak pattern of measured wind speed distribution,and well matched with the measured wind speed distribution of various peak pattern. The calculation accuracy of wind power density estimation by using the parameters of mixed double Weibull distribution function is high.