The development of a used-oil analysis expert system for triboloigcal condition monitroing and fault diagnosis

摘要

A used-oil analysis expert system for triboloigcal condition montroring and fault diagnosis- Oil-CMES is develoepd.In the system,fuzzy sets are used to process uncertain monitoring data and to express triboloigcal conditions.An object entity with fuzzy state distributions is defined to construct the dianosis knowledge units.Fuzzy resoning netwroks are built to support the inuction procedures in the diangois.Operating in MS Windows 95,Oil-CMES can deal with the data from AES,Traditional used oil tests,Ferrography and Infrared analysis and classify wear debris separated by analystical ferrography.It can also evaluate the lubricant and wear conditions of a machine,diagnose faults and generate ordinary monitoring reports.In this paper,the construction of Oil-CMES is introduced.Some fuzzy data processing methods and reasoning techniques utilized by the system are discussed.