Speed Sensorless Control Based Artificial Neural Network for Direct Current Motor

Abstract

In this paper, A robust adaptive speed sensorless by a neural network approach of a separately excited (DC) motor without the use of speed sensors is presented. The simulation carried out by Matlab backage. The actual speed is estimated using a feed-forward neural network, the neural estimator have three inputs which are armature current, armature voltage, and the speed in the previous step of time; and one output which represents the estimated speed. The obtained results are presented to show the validity and efficiency of sensorless control scheme, which in turn obtains high performances.