Aiming at the problem that the control parameters of traditional fuzzy PI controller are fixed and the control performance is poor and the adaptive ability of the system is decreased, this paper proposed a parameter optimization method of fuzzy PI controller based on improved particle swarm optimization. The Arctan function adaptive inertia weight is introduced to optimize the global characteristics of the particle swarm optimization algorithm, and the quantization factor and scale factor of the fuzzy control process are optimized to make the system achieve better control effect. The simulation model of PMSM vector control speed regulation is built under MATLAB/Simulink, and the effectiveness of the proposed control method is verified by two working conditions. The simulation results show that compared with the traditional fuzzy PI control and particle swarm fuzzy PI control, the adjustment time, overshoot and steady-state error of the fuzzy PI control method based on improved particle swarm optimization algorithm are reduced by 52.1%, 98.9%, 67.9% and 13.9%, 76.9%, 40.1% respectively in the first working condition. The adjustment time, overshoot and steady-state error decreased by 60.4%, 59.9%, 33.8% and 40.2%, 57.3%, 27.2% respectively. This method improves the dynamic response speed of the permanent magnet synchronous motor control system, reduces the overshoot and fluctuation, and makes the system achieve better control effect.