论文部分内容阅读
现代控制理论应用于冷连轧过程是一个新的研究课题。本文讨论了MIMO(多输入多输出)线性系统的辨识问题,针对钢铁工业中的多机架冷连轧机实际生产系统,利用实际系统辨识方法和计算机技术,对数学模型的建立问题进行了研究。讨论了建立实用的冷连轧机系统数学模型的方法,建立了两种描述稳态冷连轧过程的动态方程。并用多变量最小二乘法和静态逐步回归方法,分别对描述系统的多变量典范差分方程和扩充的差分方程进行了辨识,得到了描述冷连轧机系统实用的一阶模型。经拟合仿真验证,表明效果较好。以此为基础,建立了一个小型线性系统辨识软件包,为一般系统的辨识研究提供了有效的工具。
The application of modern control theory to cold tandem rolling process is a new research topic. In this paper, the identification of MIMO (Multiple Input Multiple Output) linear systems is discussed. Aiming at the actual production system of multi-stand tandem cold rolling mill in steel industry, the establishment of mathematical model is studied by using the actual system identification method and computer technology. The method of establishing mathematical model of cold rolling mill system is discussed. Two dynamic equations describing the process of steady cold rolling are established. The Multivariable Canonical Difference Equation and the Extended Difference Equation describing the system were respectively identified by multivariate least square method and static stepwise regression method. The practical first-order model describing the cold tandem mill system was obtained. The fitting simulation shows that the effect is better. Based on this, a small linear system identification software package is established, which provides an effective tool for the general system identification research.