电弧炉神经网络预测补偿控制系统设计与实现

The Design of the Control System Compensated by Neural Network Predicting for Electric Arc Furnace

利用神经网络预测技术设计实现了一套电弧炉电极智能控制系统,介绍了系统的整体结构和实现技术。系统采用神经网络预测电弧炉电弧下一时刻的状态,然后利用此预测结果对PID控制器的输出进行调整,使系统较好地适应负载和外扰的变化,从而获得满意的控制质量。

The paper presents a new Electric Arc Furnace （E. A. E） control method compensated by neural network predicting, and introduces the entire structure and realizing technique of the system. The neural network is used here to predict the state ofE. A. F., then to control an actual furnace by compensating the outputting of the PID controller. The system can adjust related parameters to suit the variations in load and external interference, which has yielded satisfied control quality.