DocumentCode
2468100
Title
A method for searching and evaluating diagnosable sequence fault sets of a dependency matrix
Author
Shi, Junyou ; Lin, Xiegui ; Lv, Kaiyue
Author_Institution
Sch. of Reliability & Syst. Eng., Beihang Univ., Beijing, China
fYear
2012
fDate
23-25 May 2012
Firstpage
1
Lastpage
7
Abstract
Based on the single-fault diagnosis theory and the evaluation requirements of residual diagnosis information in a dependency matrix, a new method for searching and evaluating diagnosable sequence fault sets of a dependency matrix is proposed. Definitions of diagnosable sequence fault set and dependency sub-matrix are given. The generation algorithm of a dependency sub-matrix which is based on original faults is studied, with the exit condition of this generation algorithm established. Based on the dependency sub-matrix, the searching flow of full diagnosable sequence fault sets is given, and the parameters of residual diagnosis information evaluation for the sets are determined. These parameters include maximum depth, minimum depth, average depth and fault unique isolation percentage. A case of an avionic subsystem with a dependency matrix is given, and the above method is applied. The result shows that all diagnosable sequence fault sets of a dependency matrix can be found, and this method can be used to evaluate the residual diagnosis information in a dependency matrix.
Keywords
fault diagnosis; matrix algebra; average depth; avionic subsystem; dependency matrix; dependency submatrix; diagnosable sequence fault set; evaluation requirements; exit condition; fault unique isolation percentage; generation algorithm; maximum depth; minimum depth; residual diagnosis information evaluation; sequence fault sets; single-fault diagnosis theory; Computer aided software engineering; Dependency Sub-Matrix; Diagnosable Sequence Fault Sets; Residual Diagnosis Information; Sequence Fault;
fLanguage
English
Publisher
ieee
Conference_Titel
Prognostics and System Health Management (PHM), 2012 IEEE Conference on
Conference_Location
Beijing
ISSN
2166-563X
Print_ISBN
978-1-4577-1909-7
Electronic_ISBN
2166-563X
Type
conf
DOI
10.1109/PHM.2012.6228789
Filename
6228789
Link To Document