电弧炉电极调节系统双模控制器设计及应用  

Design and Application of an Arc Furnace Electrode Control System Based on Dual-mode Control

作　　者：张世峰[1] 张绍德[1] 

机构地区：[1]安徽工业大学电气信息学院,安徽马鞍山市243002

出　　处：《重庆大学学报（自然科学版）》2008年第1期38-42,共5页Journal of Chongqing University

基　　金：安徽省"十五"攻关资助项目(01012053);安徽省教育厅自然科学基金资助项目(2004KJ059)

摘　　要：针对三相交流电弧炉电极调节系统具有非线性时变、多变量、强耦合及存在随机干扰的特点,提出一种基于RBF神经网络实时在线辨识的模型参考自适应解耦控制(MRAC)与比例微分(PD)控制相结合的双模控制策略。该策略通过在线建立电弧炉电极系统的精确模型并用于控制,使得系统具有快速适应对象和过程变化的能力以及解耦特性,稳定性分析和应用结果证实了双模控制器提高了系统的动态跟踪性能,增加了抗干扰能力,具有较强的鲁棒性。This paper presents a control strategy based on a Mode reference Adaptive Decoupling controller combined with a PD controller. The controller employs RBF（ Radial Basis Function） neural network real-time identification on line and neural network decoupling. This work is in accordance with characteristics of electrode system controllers in arc furnaces such as highly nonlinear, time-variant, uncertain, and stochastic disturbances as well as multivariable input and output coupling. The accurate arc furnace model can be identified on-line. This control strategy can process Changes and the decoupling characteristic frequently. The control strategy is quite robust and improves dynamic tracking performance and system disturbance resistance. The validity is verified by stability analyses and the result of application.

关 键 词：电弧炉 电极控制 神经网络 解耦 双模控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]