Sliding Observer Approach for Sensorless Operation of PMSM Drive

To eliminate the mechanical sensors of the permanent magnet synchronous motor (PMSM) drive, a nonlinear RBF network sliding mode speed/position observer was proposed for sensorless control of PMSM. The sliding surface was defined based on the errors between actual and estimated currents. Sliding mode variable is the RBF neural network input, network output is the observer control strategy. The estimated back electromotive force (EMF) was observed to calculate the position of the rotor, and the speed of the rotor. Convergence of the estimated parameters was analyzed based on the Lyapunov function. The proposed algorithm is easy to design and implement. Experimental result s showed that the proposed method can precisely estimate t he speed and position of the rotor. The system has good dynamic and static performances.