Refining Furnace Electrode Based On Neural Networks To Adjust The Intelligent Control System

The older L.A.F control system is typical control one.The controlling of three-phase electrode is absolute.The current of three-phase electrode is unbalance. Its running effect is not satisfied enough. It is very important to control the electrode of ladle furnace. Considering LF is a strong nonlinear system with the charcteristic of the three-phase unblance ,strong couping ,time variation,we hope to find an appropriate method of intelligent control applied in electrode current of ladle furnace. Neural network can approach nonlinear function randomly and it can have multi-input and multi-output. It can be implemented with computer also. The neural network can learn by itself. So it can be used for controlling the electrode. This paper made a study on intelligent control of electrode of ladle furnace based on neural network.The author construct a neural network modle for it based on the data from the practical system,design a BP neural network inverse modle controller using the primary idea of decouping control and theory of directiy inverse control.Then the model of electrode is trained and simulated by the BP models of neural network. The simulation shows that it,s a very effectual control method and can yield a satisfied control quality.