摘要
针对轧机位置内环 (APC)中存在着许多非线性不确定因素和慢时变参数的现象 ,提出一种神经网络鲁棒控制方法 .这种控制方法采用线性模型与神经网络在线逼近非线性相结合共同辨识实际模型 ,在鲁棒控制的基础上加入神经网络逼近误差对系统进行控制 ,从而能有效的消除不确定性引起的控制误差 .仿真研究表明 ,神经网络鲁棒控制器应用于APC的有效性及控制器的动态性能良好 .
A controller of neural network and robust is formulated considering the nonlinear and parameter changing.The object is identified by the neural network and linear model.The error of robust is compensated by using the approaching error of neural network.The control error which caused by uncertainty is been avoided.The result proves that the controller is effective and the arithmetic is stable.
出处
《河北大学学报（自然科学版）》
CAS
2002年第1期1-4,共4页
Journal of Hebei University(Natural Science Edition)
