In this paper, the multi-parameter identification problem of permanent magnet direct-drive wind turbine and the convergence accuracy and slow convergence speed of traditional parameter identification method are proposed. The adaptive space search vector particle swarm optimization (MDPSO) algorithm based on average optimal position variable is proposed. Parameter identification of permanent magnet direct drive wind turbine. Firstly, according to the stator voltage and current model of permanent magnet direct drive wind turbine, the pade approximation and reduced order processing are performed to discretize the generator identification model. Then the adaptive space search vector and the average optimal position variable are introduced to improve the particle swarm optimization algorithm. Finally, Through the MDPSO proposed in this paper, the parameters such as the resistance, inductance and flux linkage of the stator winding of the permanent magnet direct-drive wind turbine are identified. The simulation results show that the proposed identification algorithm has the characteristics of high precision, fast calculation speed and high stability, which verifies the validity of the established direct-drive wind turbine identification model and identification algorithm.