QUANTIFYING SIGNED DIRECTED-GRAPHS WITH THE FUZZY SET FOR FAULT-DIAGNOSIS RESOLUTION IMPROVEMENT

NCU Institutional Repository > 工學院 > 化學工程與材料工程研究所 > 期刊論文 > Item 987654321/26686

jsp.display-item.identifier=請使用永久網址來引用或連結此文件: http://ir.lib.ncu.edu.tw/handle/987654321/26686

题名:	QUANTIFYING SIGNED DIRECTED-GRAPHS WITH THE FUZZY SET FOR FAULT-DIAGNOSIS RESOLUTION IMPROVEMENT
作者:	HAN,CC;SHIH,RF;LEE,LS
贡献者:	化學工程與材料工程學系
关键词:	CHEMICAL PROCESS;EXPERT SYSTEM;QUALITATIVE SIMULATION;KNOWLEDGE;ALGORITHM;FAILURES;PLANTS;STATE
日期:	1994
上传时间:	2010-06-29 17:34:16 (UTC+8)
出版者:	中央大學
摘要:	The fault diagnosis method of signed directed graphs (SDGs) has difficulties verifying the seriousness of an abnormal process variable and providing an accurate resolution of fault origin, when the abnormal variable is in the neighborhood of the designed threshold. In this paper, the fuzzy set theory was incorporated into SDGs to overcome these obstacles. This proposed algorithm begins with an SDG representing the qualitative cause-effect relations among the process variables. The primarily formulated SDG is then simplified by edge consistency and the combination of unmeasured variables, an acyclic graph consisting of strongly connected components was formed, and then the technique of the depth-first search determined the specific strong component where the possible fault origin was located. Finally, the quantitative fuzzy set manipulation was introduced, the degree of membership of the process variables was determined, and then the variables were sequentially arranged by their degree of membership to determine the possible fault origins. A process consisting of a jacketed CSTR and a vaporizer with a designated plugged product valve was diagnosed by the proposed approach. The result shows that the proposed approach improves the accuracy of resolution and provides more valuable information on the arrangement of fault origin candidates by the seriousness of the abnormality of the variables.
關聯:	INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH
显示于类别:	[化學工程與材料工程研究所] 期刊論文

文件中的档案:

档案	描述	大小	格式	浏览次数
index.html		0Kb	HTML	493	检视/开启

在NCUIR中所有的数据项都受到原著作权保护.

社群 sharing

数据加载中.....