Aiming at the problem that the parameters of fault diagnosis model are difficult to be optimized of wind turbine pitch system, a fault diagnosis method of wind turbine pitch system based on multi-class optimal margin distribution machine optimized by the state transition algorithm (mcODM-STA) is proposed. In this method, the wind turbine power output is selected as the main state parameter, and Pearson correlation coefficient is used to analyze the historical operation data of wind turbine in wind power data acquisition and monitoring control system, and the features with low correlation of power output state parameters are eliminated. The remaining features are analyzed twice to reduce the sample features. The data set is divided into training set and test set. The training set is used to train the proposed fault diagnosis model, and the test set is used for testing. The operation data of a domestic wind farm is used for experimental verification. Experimental results show that the proposed method has higher fault diagnosis accuracy and Kappa coefficient than other parameter optimization methods.