FAULT DIAGNOSIS SYSTEM AND AUTOMATED DESIGN SYSTEM THEREFOR

摘要

PROBLEM TO BE SOLVED: To automatically design a fuzzy diagnostic rule, using the measured vibration/sound signal measured on an industrial rotating machine, for the fault diagnosis of the rotating machine.;SOLUTION: An optimizing module 10 takes the fuzzy symptom calculated from the vibration/sound signal measured on the industrial rotating machine as input data, selects, on the basis of the fuzzy rough sets theory, the fuzzy symptom having important information on the fault diagnosis of the rotating machine 1 from among fuzzy symptoms, and, depending on the learning by knowledge-base networks(KBN), automatically creates new fuzzy rules (fault detection rule, fault isolation rule, fault magnitude identification rule) of which each selected fuzzy symptom is the input and the status of the rotating machine is the output.;COPYRIGHT: (C)2002,JPO