Fault diagnosis method based on signed. directed graph

Fault Diagnosis Method based on Qualitative Model do not need precise mathematic model of the objects and it has made rapid progress during recent years. SDG model can be used to express the complex cause and effect relations, and has very large capacity of containing process potential information. SDG consisted of nodes and directed branches which can represent the qualitative influence relationships among nodes. In this paper, we perform the research of fault diagnosis model represented by three-pattern SDG instant pattern and we present a modeling method to build SDG fault diagnosis model according to nodes inverse search. principle. After the observed variables are selected as the key nodes, the sub network model is then built for every key node. SDG represents deep knowledge model and normally it is used to build rules in expert system and some aided methods such as scenario is adopted to perform fault diagnosis. We complete real time inference engine used for SDG and realize the real time solving. This algorithm satisfies the requirement of fault diagnosis system and it can be applied in embedded systems after modification. Furthermore, we also discuss a fault diagnosis test system built by semi-object simulation technology and complete SDG fault diagnosis simulation test on this test system. A number of misoperation faults test and real time online fault diagnosis test are also successfully performed.