轧钢加热炉钢坯加热质量的检测研究(II)

Measurements of Quality Variables for an Industrial Rolling Mill Reheating Furnace (II)

轧钢加热炉中钢坯温度的检测是一类典型的工业过程质量参数难测量问题。本文首先分析了该问题的产生原因及对生产过程的不良影响 ,然后运用多元统计投影原理建立了钢坯温度变量和过程变量之间的主元回归软测量模型和偏最小二乘软测量模型 ,最后基于工业拖偶试验数据对两类模型的参数矩阵进行了求取。根据工业实际生产数据进行的模型校验和误差分析表明 ,模型预测误差满足工业应用的精度要求 ,且较作者前期研究的结果更精确。全文分为两部分 ,这是第二部分。

The on-line slab temperature measurement for an industrial rolling mill reheating furnace is one class of typical hard-measuring problem in industrial process. After discussing the difficulty of on-line measuring for slab temperature and the influence on process operation, this paper developed two slab temperature soft-sensing models, the principal component regression model and partial least square model, based upon multivariable statistical projection technique. As a key step of this work, model verification is presented by using real operating data gathered from industrial experiments. Research results show that the models’ precision is good enough to satisfy the engineering demands. This paper is divided into two parts, and this is the second part.