This paper presents an ANN observer that will estimate the position of permanent magnet synchronous motor by a single shunt. The observer with artificial neural network is based on a nonlinear transformation of mathematical model of PMSM in a stationary \alpha -\beta reference frame. By designing appropriate weight value and back propagate learning rule with SVPWM, the permanent magnet synchronous motor will properly get the angular position. The simulation and experiment results show that the scheme has simple structure, satisfactory performance and strong robustness.
Index Terms:
ANN. SVPWM. PMSM. Sensorless control. Observer.
Citation:
Xuepeng Liu, Bin Wang, "ANN Observer of Permanent Magnet Synchronous Motor Based on SVPWM," isda, vol. 1, pp.95-100, Sixth International Conference on Intelligent Systems Design and Applications (ISDA'06) Volume 1, 2006