Research On The Method Of System Fault Diagnosis Based Ontology And Rule

Fault diagnosis can tell why and where fault occurs in the system using variety methods. Fault tree is an inverted logic graph consisting of event symbols, logic gate symbols and transition symbols. Fault tree analysis is one of the main methods used in fault diagnosis. However, because the knowledge resources of establishing the fault tree are scattered as well as inequality and lack of accurate semantics information, it may lead to wrong diagnosis results and difficult maintenance work brings great difficulties to system analysis.It is an urgent problem of introducing accurate semantic information into knowledge of fault tree at the base of being unrelated to fault diagnosis. This paper uses an ontology description language OWL to build a general shared fault tree ontology based on the analysis of knowledge of fault tree domain. Then, it gives the method of transforming seven different logic gates into SWRL rules. The main work is as follows:(1)Analyzing the fault tree knowledge, this paper extracts the important terms of fault tree and carries the classification about concept level. Utilizing the Protégé ontology tool, this paper constructs a fault tree domain ontology.(2)Based on the proposed fault tree domain ontology, this paper studies how to transform the and gate, or gate, priority and gate, voting gate, XOR gate and forbidden gate into SWRL rules, and gives out the transforming method. It provides the rule basis for fault diagnosis.(3)Based on the above theory and method, this paper designs and carries out a system fault diagnosis tool based on ontology and rules. The tool is utilized to analyze the phenomenon of system faults and finds out all reasons of the trigger of the system faults. An example shows the method is correct and effective.