Sensorless control of PMSM drive using Neural Network Observer

In this paper a Neural Network Observer is designed for sensorless vector control of a PMSM drive. The Neural network observer is designed to give the estimate of speed and position using direct and quadrature current and voltage components. The proposed Neural Network Observer design provides robust estimation of speed. The neural network offers advantage in terms of reduced mathematical computations. Also a sliding mode controller (SMC) controls the speed in the drive. Using the SMC controller, the drive gives an improved performance irrespective of the load variations. Simulation of a PMSM drive is carried out in MATLAB/Simulink environment and presented in this paper subject to system disturbances, variations in speed and load.