In recent years, permanent magnet synchronous motor drive systems have gradually become a hot topic in high-performance industrial applications. As a new type of motor, PMSM has broad application prospects in various fields. There are many existing PMSM sensorless control technologies, but currently PMSM The main problems of motor control methods include slow response, inability to resist interference, low estimation accuracy, and large errors. Considering the instability of parameters during actual motor operation, coupled with the presence of various interference factors, PMSM control will be further complicated. The difficulty cannot be adjusted accurately. This article proposes an extended Kalman filter method and designs the implementation steps of the algorithm in detail. The extended Kalman filter algorithm adds the fuzzy module function and does not linearize the nonlinear function, so it will not lead to an increase in errors and can be effectively. To improve control performance and accuracy, use MATLAB/Simulink tools to build a PMSM control system model based on EKF, and conduct relevant debugging and control of control functions and parameters. Analyze parameters such as position angle response, speed response, etc., and obtain waveform diagrams. With the simulation diagram, the simulation results, response curves, and error curves were analyzed and summarized, verifying the rationality and effectiveness of the extended Kalman filter algorithm in estimating the PMSM speed.