常用过程数据校正技术  

Studies on Common-used Process Data Reconciliation Method

作　　者：刘猛[1] 金思毅[1] 陶少辉[1] 

机构地区：[1]青岛科技大学化工学院,山东青岛266042

出　　处：《青岛科技大学学报（自然科学版）》2009年第4期333-336,共4页Journal of Qingdao University of Science and Technology:Natural Science Edition

基　　金：重质油国家重点实验室开放课题基金资助项目(2008-08)

摘　　要：现有的数据校正方法种类繁多,实际应用时难以选择合理的方法。研究了常用的多组分过程数据校正方法,如线性化法、Crowe投影矩阵法和QR正交分解法,对比分析了其原理及方法步骤,并进行了实例数据校正和性能评价。研究结果表明QR分解法精度较高,计算容易实现,因此实际应用时应首选QR分解法。There is a variety of methods available for applying data reconciliation techniques and it is quite difficult to select a proper one in actual process. In order to make the selection easier, several data reconciliation methods are studied and compared based on the theory and methods in this article, including Linearization method, Crowe＇s matrix projection method and QR decomposition method. An example of application is presented and the result shows that the precision of QR decomposition method is higher and the computation process is easier compared with Crowe＇s method and Linearization method. Therefore,QR decomposition method should be selected first for solving the hilinear data reconciliation problem.

关 键 词：数据校正 线性化 投影矩阵 QR正交分解 

分 类 号：TQ015.9[化学工程]