A DESIGN OF AUTOMATIC FAULT DIAGNOSIS SYSTEM FOR ROTATING MACHINERY

摘要

Due to the growing standards in modern industry, the fault diagnosis system increasingly requires auto-inferring and intelligent. Thus, in this paper, several methods are discussed on auto-diagnosis system design: rough set, which effectively simplifies redundant information; rule-regularizing, which helps system realize self-illation and self-study; and fuzzy processing, which makes system more robust and adaptable. Integrated these methods into one system dramatically enhances voluntary process in diagnosis. In this way, an automatic fault diagnosis system for rotating machinery is established, which reduces space-time consumption, weakens human intervention, and accelerates diagnosis speed. The most important achievement of this project is that this system minimizes human-operating during the diagnosis process, which is supported by the experiments in the 3rd part of this paper.