文献类型：期刊文献

中文题名：分馏塔产品质量估计MIMO软测量

英文题名：MIMO Soft Measurement for Quality Estimation for the Products of Hydrogen Cracking Fractional Column

作者：仲蔚[1];刘爱伦[1];俞金寿[1]

机构：[1]华东理工大学自动化所,上海200237

年份：2000

卷号：21

期号：3

起止页码：8

中文期刊名：自动化仪表

外文期刊名：Process Automation Instrumentation

收录：CSTPCD;;北大核心:【北大核心1996】；

基金：国家教委博士点基金资助项目

语种：中文

中文关键词：加氢裂化;分馏塔;软测量;产品质量估计

外文关键词：Soft measurement MIMO system Partial least square method

摘要：针对加氢裂化分馏塔多个产品的质量指标需同时预报的一类ＭＩＭＯ软测量建模问题，研究了适用于ＭＩＭＯ软测量建模的一种改进的ＰＬＳ算法。通过采用实际过程数据进行建模，指出该算法具有所需训练数据少、允许数据相关、算法收敛快等优点，是处理多变量系统软测量建模的一种有效方法。
An improved PLS algorithm which is suitable for modeling in MIMO soft measurement is investigated. MIMO soft measurement is adapted in simultaneous estimation for multiple products of hydrogen cracking fractional column. From the experience of establishing the model with practical process data it is stated that by using this method, less training data are needed, correlative data are allowable, and fast convergence is obtained. This is an effective method for establishing the model for soft measurement of multiple variables systems.