基于模糊神经网络的定尺剪切线控制系统  

Control system of fixed length cutting line based on fuzzy-neural network

作　　者：崔宝侠[1] 李月明[1] 段勇[1] 

机构地区：[1]沈阳工业大学信息科学与工程学院,沈阳110870

出　　处：《沈阳工业大学学报》2009年第6期676-680,共5页Journal of Shenyang University of Technology

基　　金：国家青年科学基金资助项目(60905054)

摘　　要：为了提高卷钢自动定尺剪切的精度,提出一种模糊神经网络控制系统对卷钢剪的切线位置进行控制.控制系统采用模糊神经网络控制器和神经网络辨识控制器相结合的方式对神经网络的学习算法进行改进,通过对模糊神经网络进行训练学习,优化了网络的连接权值,从而能够很好地控制板材送料位置,使得板材在减速期以理想的减速曲线运行,实现准确停车进行剪切.仿真结果表明:该系统具有响应快、鲁棒性强、控制精度高、控制特性好等优点,能够满足剪切生产的要求.In order to enhance the fixed length cutting accuracy for roll steel, a fuzzy neural network control system for controlling the position of cutting line was proposed. The combined fuzzy neural network controller and neural network identification controller were adopted to improve the learning algorithm of neural network. The connection weight of the network was optimized through performing the learning and training of the network. Thus, the feeding position of steel plate can be well controlled. Steel plate can move according to the ideal slowdown curve, and finally stop at the accurate position. The simulation result shows that the present system has the quick response, strong robustness, high control accuracy and excellent control characteristics, and can satisfy the demand of cutting production

关 键 词：定尺剪切 剪切线 位置控制 模糊神经网络 减速曲线 系统辨识 学习算法 连接权 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]