Synthesis of a neural network controller for an automated system controlling a cascade of centrifugal extractors
Download article in PDF format
Authors: Emelyanov A. M., Kozlova Yu. M., Nadezhdin I. S.
Annotation: The paper considers the synthesis of an automated control sys-tem for a cascade of centrifugal extractors with a neural network controller trained by the method of error back propagation through a direct neuroemmulator with a reference model. The identification of the control object was carried out, the given standard deviation is 1.63%. The quality indicators of the con-trol system with a thermostat are superior to the control system with a PID controller. The developed automated control system for the extraction unit based on a neural network controller has been tested for rudeness.
Keywords: neural network controller, artificial neural network, extraction, automated control system