DESIGN OF A FUZZY-LOGIC BASED DIAGNOSTIC MODEL FOR TECHNICAL PROCESSES

The design and implementation of the heuristic knowledge base for a fuzzy logic-based technical diagnostic problem solving strategy integrated as a heuristic level within a computer assisted fault diagnosis system is considered. Based on knowledge acquisition via expert knowledge (linguistically expressed as fuzzy conditional statements); simulation-based observations of the diagnosed process (as a feature extraction technique for technical diagnostic relevant information); process history and fault statistics, a fuzzy diagnostic model is developed as a fuzzy symptoms-faults mapping and implemented within the heuristic knowledge base. Diagnostic reasoning by MATCHING the proposed fuzzy diagnostic model with the current values of the symptoms (either analytically generated or subjectively observed by the operator) is presented as well as the validation of the design method. first by illustrating it on a very simple simulated process (a low-pass filter) and then by applying it to a simulated DC motor.