Design of neural network controller for VRM inverter drive
F. Ismail, S. Wahsh, Amal Z. Mohamed · 2002
In this paper, the design of a dynamic neurocontroller with good robustness properties is presented for a variable reluctance motor (VRM) drive system. The goal of this controller is to determine the optimal excitation parameters that can be fed to the inverter which forces the drive system to operate at the required operating conditions of torque and speed. The neurocontroller is generated by a multilayer feedforward neural network which is trained with a back propagation algorithm.>