Track: Maintenance Services
Abstract
This paper discusses the fuzzy expert system for troubleshooting the deodoriser in palm oil refinery plant. Interviews with experienced plant staffs and operators were conducted, and required information regarding the whole process were gathered for identifying the faults, defining the membership functions and providing a sequence of necessary actions. Industrial plant data were used to check the performance of the fuzzy expert system. It was found that the troubleshooting system is capable of predicting the failures and suggesting necessary actions with a similarity value of 80 percent with the system database. In addition, the proposed expert system is useful for operators’ guidance and for field training purposes.