DocumentCode
437439
Title
A hybrid algorithm of on-line fault diagnosis
Author
Qian, Ma ; Yihan, Yang ; Zheng, Qi
Author_Institution
Dept. of Electr. Eng., North China Electr. Power Univ., Beijing, China
Volume
1
fYear
2004
fDate
21-24 Nov. 2004
Firstpage
991
Abstract
This paper proposes a hybrid algorithm of on-line fault diagnosis, which calls the data of breakers, protections and digital recorders by layer for diagnosis. Breaker tripping information is used first to determine the power failure area that contains fault components. Where after, protection actions information is used to determine the exact fault elements. At last, digital recorder information is used to evaluate protection actions. The analysis on protection information is the main task and character of this paper. There are two steps in processing of protection information. First we find the intersection of regions of acted protections to locate the exact fault elements with rule-based reasoning technique. And then for each element in the suspicious fault components list, the posterior fault probability is calculated as its fault credence degree. Then the elements in the suspicious fault components list can be sequenced by fault credence degree. Theory of evidence is used for calculating the fault credence degree of the suspicious fault components. It is stricter and more reasonable than before. Simulation analyses of the actual systems verify the effectiveness of the algorithm.
Keywords
fault diagnosis; knowledge based systems; power system faults; power system protection; power system simulation; probability; breaker tripping information; digital recorder information; fault credence degree; fault information system; hybrid algorithm; online fault diagnosis; posterior fault probability; power failure area; protection actions information; rule-based reasoning technique; Analytical models; Diagnostic expert systems; Fault diagnosis; Information analysis; Power system faults; Power system interconnection; Power system protection; Power system simulation; Power systems; Probability;
fLanguage
English
Publisher
ieee
Conference_Titel
Power System Technology, 2004. PowerCon 2004. 2004 International Conference on
Print_ISBN
0-7803-8610-8
Type
conf
DOI
10.1109/ICPST.2004.1460138
Filename
1460138
Link To Document