Study on Sampling Control in Hydraulic AGC Based on Neural Network

This paper introduces a hydraulic automatic gauge control system for Hengtong 1270 mm cold mill in Tangshan, China. Identification and control are achieved based on RBF (radial basis function) and BP (back propagation) NN ( neural network) in APC (automatic position control) system. Sampling control is applied to solve the dead-time control problem because of the gauge's location. Considering the inertia of the hydraulic servo-system, we propose a new method of selecting sample period. The simulation result and the application show that the control effect is satisfied, and this method can be used widely.