EVALUATION OF SCHEMES FOR DETECTING AND IDENTIFYING GROSS ERRORS IN PROCESS DATA.

被引：0

作者：

Rosenberg, Joseph ^{[1
]}

Mah, Richard S.H. ^{[1
]}

Iordache, Corneliu ^{[1
]}

机构：

[1] Northwestern Univ, Evanston, IL, USA, Northwestern Univ, Evanston, IL, USA

来源：

Industrial and Engineering Chemistry Research | 1987年 / 26卷 / 03期

关键词：

MEASUREMENT ERRORS;

D O I：

暂无

中图分类号：

学科分类号：

摘要：

The authors propose two new composite tests, DMT and EMT, which make use of more than one statistical test to reduce mispredictions and to account for bound violation. These tests were evaluated along with global and measurement tests (GT and MT). Although the composite tests are not restricted to linear constraints and a single gross error, the evaluation was carried out for mass flow networks with at most one gross error present in the data. The effects of various factors on the performance of the tests are summarized in the form of basic rules and guidelines. Guidelines for choosing an appropriate detection scheme are also developed. The composite tests are found to have superior overall performance.

引用

页码：555 / 564