神经网络模型参考自适应控制及其在带材板形控制系统中的应用

NEURAL NETWORK MODEL REFERENCE ADAPTIVE CONTROL AND ITS APPLICATION IN THE STRIP SHAPE CONTROL SYSTEM

通过对某轧机液压弯辊系统特性的分析，提出了神经网络模型参 考自适应控制策略，兼顾了系统的动态特性和静态特性，成功地解决了未知、不确定、非线 性系统的辨识与控制问题；并将其用于液压弯辊系统仿真，结果表明该系统的性能良好，将 会明显提高产品质量。

A scheme of neural network model reference control is presented by carefully analyzing the characteristic of a hydraulic bending roll system, both the transient and steady state behavior of the system in consideration, the questions of identification and control for unknown, uncertain, non linear system are solved successfully; and application in the hydraulic bending roll system, satisfactory performance of the control system is obtained from the simulation experiments, the quality of products is improved evidently.