PMSM mechanical sensor-less vector control based on extended Kalman filter

The rotor position and speed of a permanent magnet synchronous motor (PMSM) are estimated for mechanical sensor-less vector control by an extended Kalman filter (EKF) algorithm suitable for solving the nonlinear PMSM equations. The equations use rotating coordinates synchronous with oriented by the rotor flux where the parameters such as inductances are constants, so the observer can be applied to both salient and non-salient motors. The EKF computation is simplified for practical applications. With the load torque as a state variable, the observer can also observe the load torque which can be used for feed-forward compensation to improve the dynamic performance. The system performance is verified by experimental results.