In this paper, by measuring the phase voltages and currents of thepermanent magnet synchronous motor (PMSM) drive, a neural-network-basedrotor position and speed estimation method for PMSM is described. Theproposed estimator includes two recurrent neural networks, one is usedto estimate rotor speed and rotor position, and the other is used toestimate stator current. Through using an improved recursive predictionerror algorithm, on-line adaptative estimation is realized. Thesimulation results show that the proposed approach gives a goodestimation of rotor speed and position. Especially, the proposedapproach has low sensitivity to perturbations of the mechanicalparameters and torque disturbances
展开▼