基于Elman网络的自适应逆控制电极调节系统  

Adaptive Inverse Control Electrode Regulation System Based on Elman Neural Network

作　　者：郭飞[1] 李华德[1] 

机构地区：[1]北京科技大学信息工程学院,北京100083

出　　处：《煤矿机械》2007年第10期131-133,共3页Coal Mine Machinery

基　　金：北京市重点自然科学基金资助项目(KZ200410005005)

摘　　要：针对交流电弧炉电极调节系统具有非线性、时变、强耦合性的特点,提出一种基于El-man神经网络的自适应逆控制方案。把电极调节系统耦合造成的对系统的影响转化为系统的不确定性扰动问题,并用BP算法在线修正作为控制器和扰动消除器的电极Elman网络逆模型。比PID控制有更好的跟随性、稳定性、抗干扰性、鲁棒性。In view of such characters as nonlinear, parameter- time - varying, strong coupling of electrode regulation system of alternating current electrical arc furnace, an adaptive inverse control scheme is proposed based on Elman neural network to solve the control problem. The effect on the system caused by system coupling is converted into uncertain disturb; and electrode Elman network inverse model as the controller and disturb eliminator is corrected online by BP algorithm. Based on Elman network has better uncoupling control effects on the electrode and has better characteristics of tracking, stability, resisting disturbance, robustness than the PID control.

关 键 词：自适应逆 小生境 遗传算法 ELMAN神经网络 电极 

分 类 号：TP273.3[自动化与计算机技术—检测技术与自动化装置]