Summary: | 碩士 === 國立臺灣科技大學 === 電機工程系 === 98 === This thesis presents an architecture of conditioning based maintenance management systems by integrating IEDs (Intelligent Electrical Device) and FMEA (Failure Mode and Effect Analysis) Fuzzy rules. IEDs acquire analog signals and send messages which are parameters for estimating conditions of electrical equipments to remote servers by Ethernet. This thesis proposes risk priority number to analyse and optimize the current maintenance systems which inspect conditions of electrical equipments based on time period. Furthermore, the experts can establish fine fuzzy rules with the analysis results. The conditioning based maintenance management systems integrate data distributed in different departments and information acquired by IEDs to infer the possible failure rate of components of electrical equipments with fuzzy rules. Therefore, operators can look for the best chance to maintain the facilities.
This thesis establishes fuzzy rules to forecast failure rates of iron cores, copper windings, and tap changers of power transformers. Eight examples demonstrate the inference process of failure rates with different index inputs related to the conditions of power transformers. The final results show the fuzzy rule can judge the condition of the power transformer.
|