Fault Diagnostic Expert System Of Turbine Generator Sets Based On Rule Reasoning And Case Reasoning

Turbine generator sets are very important equipments in electric power production and their performance will affect directly on the stability of electric power system and the whole national economy. Therefore, diagnosing and eliminating faults accurately and efficiently have great significance for turbine generator sets to keep good working condition.The main contents of this paper are as follows:(1) Knowledge base of turbine generator sets is established. The knowledge base should be of validity, availability and integrity. The table of turbine generator sets fault based on the analysis about the relationships between the turbine generator set fault and the fault feature is founded. A knowledge representation method of the rule based on confidence-based and production-rule-base is presented and the method of object-oriented is adopted in case representation in this paper.(2) Case Retrieval Strategy. The ratio similarity algorithm and the hierarchical retrieval strategy are adopted in this paper. When the diagnosing case is inputted, the system would isolate automatically the diagnosing case in its corresponding case class (sub-case base) based on fault features data of diagnosing cases.Then similarity degree between diagnosing cases and historical cases is calculated, according to the match criteria.(3) The framework design of fault diagnosis expert system for turbine generator sets based on rule based reasoning and cases based reasoning is studied. The diagnosis system is developed by using CLIPS, C#.NET and SQL, full advantages of the three kinds of software are developed to improve the efficiency of system development. Last, it is verified by the instance tests, the results show that the expert system can diagnose quickly and accurately.