Design and Testing Neuro Networks Control System for Direct Current Machine

Abstract

The paper presents an adaptive intelligent control method to overcome effects of some indeterminate and undealt factors that a DC. drive is suffered. In the speed loop, we use a three-layer neural networks through a backpropagation (BP) algorithm out of line learning to realize the fuzzy-control tactics. We use unit neuron through Hebb algorithm on-line dynamic learning to realize adaptive mechanism. The simulation is based on a MATLAB neural networks toolbox with simulink. The results of the simulation show that adaptive intelligent control method enables the system to have good dynamic and stability performance. The proposed method develops the use of simulink in the field of electrical drive of adaptive intelligent control.