The Performance analysis of artificial neural network for speed control of DC motors

m ABSTRACT MSc.Thesis THE PERFORMANCE ANALYSIS OF ARTIFICIAL NEURAL NETWORK FOR SPEED CONTROL OF DC MOTORS Cafer BAL Fırat University Graduate School of Natural and Applied Sciences Department of Electronics and Computer Science 2002, Page: 82 This thesis describes the speed control of DC motors using adaptive activation function neural networks. Comparative performance of the adaptive sigmoid, polynomial and spline activation functions used in the hidden layer of the multi-layer neural networks are tested for the control of DC motor including a nonlinear actuator. This comparison aims to determine the most suitable activation function, which reduces the training time of the artificial neural network and increases the control performance of the motor, under variable load conditions. Simulation results showing the effective of the adaptive activation function neural networks for speed control of DC motors, are given. Keywords : Artificial neural networks, Adaptive activation functions, DC motors