State feedback using artificial neural network for speed control of DC motor

This paper introduces the idea of using artificial neural networks for the speed control of a DC motor whose parameters are not constant. Motor armature current and speed are taken as state variables. The control is achieved through state feedback and output controllers whose parameters are adjusted by the neural network by observing current and speed history. The scheme is illustrated by its application to a particular DC motor. The responses of conventional controllers and those obtained with the proposed scheme are compared.