Process Systems Engineering. Fault Diagnosis Using Condition for Cause-Effect Relationship among Stage Variables.

• Takeda Kazuhiro

  Department of Chemical Engineering, Kyushu University

• Tsuge Yoshifumi

  Department of Chemical Engineering, Kyushu University

• Matsuyama Hisayoshi

  Department of Chemical Engineering, Kyushu University

Bibliographic Information

Other Title

• プロセスシステム工学　状態変数間に因果関係が成立するための制約条件を利用する異常診断法
• ジョウタイ ヘンスウカン ニ インガ カンケイ ガ セイリツスル タメ ノ セ

Abstract

It is essential to quantitatively estimate the performance of fault diagnosis. In particular, two quantitative performance criteria are very important : diagnostic reliability and diagnostic accuracy. Each criteria is defined for the design phase and the running phase, respectively. This paper addresses diagnostic reliability for the running phase and diagnostic reliability and accuracy for the design phase.<BR>A fault diagnostic method utilizing signed directed graph is proposed. In this paper, a modified fault diagnosis method which has higher diagnostic accuracy than the aforementioned method's diagnostic accuracy is proposed. The proposed method uses conditions for cause-effect relationship among state variables. Furthermore, the performance of the proposed method and criteria is demonstrated by use of mini-plant.

Journal

• KAGAKU KOGAKU RONBUNSHU

  KAGAKU KOGAKU RONBUNSHU 22 (5), 1009-1017, 1996

  The Society of Chemical Engineers, Japan

Keywords

• Systems Engineering
• Fault Diagnosis
• Graph Theory
• Signed Directed Graph
• Information Content

Details

• CRID: 1390001204508158976
• NII Article ID: 10002669444
• NII Book ID: AN00037234
• DOI: 10.1252/kakoronbunshu.22.1009
• ISSN: 13499203
• ISSN: 0386216X
• NDL BIB ID: 4057608
• Web Site: https://ndlsearch.ndl.go.jp/books/R000000004-I4057608
• Web Site: http://www.jstage.jst.go.jp/article/kakoronbunshu1975/22/5/22_5_1009/_pdf
• Text Lang: ja
• Data Source:
  • JaLC
  • NDL
  • Crossref
  • CiNii Articles
• Abstract License Flag: Disallowed