DocumentCode
2518391
Title
Diagnosing novel faults
Author
Bublin, Scott ; Kashyap, R.L.
Author_Institution
Knowledge Based Syst. Lab., Purdue Univ., West Lafayette, IN, USA
fYear
1989
fDate
16-18 Oct 1989
Firstpage
84
Lastpage
93
Abstract
Currently available diagnostic expert systems are limited as they cannot address novel situations and faults. Moreover, they cannot recognize such situations, which often lead to an incorrect diagnosis or at best an exhaustive search through their knowledge bases. The CONSOLIDATE system is a hybrid diagnostic system which integrates experiential and causal-model-based knowledge sources, which retains the efficiency of experiential-based approaches while providing a mechanism to recognize and address previously unobserved situations. A discussion is given the CONSOLIDATE system, issues related to novel fault hypothesis generation, and the approach utilized to address these issues
Keywords
expert systems; knowledge based systems; CONSOLIDATE system; causal-model-based knowledge sources; diagnostic expert systems; experimental knowledge sources; fault hypothesis generation; hybrid diagnostic system; knowledge bases; novel faults; novel situations; unobserved situations; Costs; DC generators; Diagnostic expert systems; Expert systems; Fault diagnosis; Knowledge based systems; Machine intelligence; Machinery; Manufacturing; Problem-solving;
fLanguage
English
Publisher
ieee
Conference_Titel
Data and Knowledge Systems for Manufacturing and Engineering, 1989., Second International Conference on
Conference_Location
Gaithersburg, MD
Print_ISBN
0-8186-1983-X
Type
conf
DOI
10.1109/DKSME.1989.107445
Filename
107445
Link To Document