Permanent magnet synchronous motor (PMSM) was widely used in the field of ship electric propulsion because of its excellent torque characteristics and wide speed range. The sensorless control was an important guarantee for the reliable operation of the system. However, the parameters of the motor would change with the operating conditions due to the temperature change, the magnetic saturation effect and the cross coupling of the magnetic circuit. Therefore, realtime control of operating parameters of PMSM was an important guarantee for the control quality of the system. In view of the above problems, the model reference adaptive system was used to identify the parameters of PMSM. The full rank adjustable model was established by using the Runge-Kutta method. The adaptive law was derived based on the Popov super stability theorem. Finally, the experimental test platform was used to verify the algorithm. The simulation and experimental results showed that the online parameter identification algorithm could identify the motor parameters accurately in real time.