SYNTHESIS NEURAL NETWORK CONTROLLER FOR TWO-MASS ELECTROMECHANICAL SYSTEM OF LATHE MACHINE’S FEED DRIVE
The analyses of the advantage and disadvantage of the Neural Network Predictive Controller implemented in the Neural Network Toolbox are considered. A mathematical description of two-mass electromechanical drive system for lathe machine feed drive with elastic connections. During synthesis procedure, the neural regulator for stabilizing the speed of linear motion and compensation vibration occurring in the elastic elements of lathe machine’s feed drive is proposed. In this article several algorithms as Moller, Levenberg–Marquardt, Shelb–Ribira, gradient descent for training the neural regulator NARMA-L2 Controller are compared. The parameters of neural regulator NARMA-L2 have significantly affected the quality of control system. The comparative analysis of the quality indicators of the training of the neural network controller under various learning algorithms is given. It is shown that the use of the NARMA-L2 Controller neural network controller makes it possible to improve the quality of the control system (CS) parameters of the lathe machine’s feed drive.
Authors: M. P. Belov, I. S. Nosirov, А. М. Belov
Direction: Electrical Engineering
Keywords: Electromechanical system, lathe machine feed drive with elastic connections, neuroregulator, Levenberg–Marquardt learning algorithm
View full article