TMDoctor: A Fuzzy Rule- and Case-Based Expert System for Turbomachinery Diagnosis

Abstract This paper presents an expert system that combines fuzzy logic and certainty factors for turbomachinery diagnosis. Given symptoms associated with a vibration problem, the system can identify and rank possible causes by performing incremental forward chaining. This system also incorporates a simple case-based reasoning component to allow past experience to be applied for providing more accurate diagnoses. The ability of the expert system to identify causes of typical vibration problems in rotating machinery is supported with tests on real cases.