ELECTRIC SERVO DRIVE OF TURNING LATHE CONTROL SYSTEM USING NEURAL GENERALIZED PREDICTIVE CONTROL
This paper deals with mathematical description of the algorithm for neural generalized Predictive Control (NGPC) with prediction and the Newton-Raphson optimization algorithm. The Newton–Raphson algorithm allows minimize the Cost Function Minimization (CFM) which determines the input signal necessary to obtain the desired behavior quality control. Identification procedure is considered that allows build neural network plant. The synthesis of neural network control is considered to ensure high static and dynamic accuracy when reproducing contour-positional displacements. A model control system for electric drive of a lathe with using Neural Network Predictive Control (NNPC) is developed in Matlab/Simulink environment. Several variants of the NNPC were synthesized with different values to establish its optimal parameters. The parameters of NNPC regulator prediction are determined that they significantly effect to the quality servo control system.
Authors: M. P. Belov, I. S. Nosirov, T. H. Phuong
Direction: Electrical Engineering
Keywords: Electric drive of turning of the lathe, neural generalized Predictive Control, synthesis of the neuron controller, Newton-Raphson algorithm
View full article