Implementation of Neural Networks for Direction Control of DC-Motor

Abstract

In this paper, a neural network was implemented to control the direction of a DC motor. In this context, attention was paid to the field of artificial intelligence related to neural networks. The network targeted for training is small in size, and the tasks required of it are limited and include the direction of rotation of the DC motor axis, which has two directions (clockwise rotation or counterclockwise rotation). DC motors are used in applications such as robotics and automated manufacturing. It is a rotary actuator or linear actuator that allows control of linear direction. The main purpose of this work is to provide a clear view of the importance of using neural networks in the control field. In this study, the ANN is trained by the back-propagation algorithm, and a multilayer neural network is composed to implement the controller that controls the operation of the model using the back-propagation learning algorithm that regulates the function of the DC motor system. It was noticed that changing some parameters such as weights, learning rates, number of hidden layers, or number of nodes can enhance the output and generate the proper control of the direction.