Intelligent fault diagnosis using rough set method and evidence theory for NC machine tools

An intelligent fault diagnostic method was presented to satisfy the development requirements of next-generation intelligent NC machine tools. The framework of fault diagnosis unit was established first, which consisted of signal acquisition, diagnosis rule extraction and fault identification mechanism. The technique of diagnosis rule extraction was then studied and an algorithm for acquisition of decision rules was proposed. The algorithm simplified the analysis of core properties and unnecessary properties, and calculated reduction set by the backwards tracking approach. This algorithm reduced complexity in reductions calculation and improved the efficiency of rule extraction. Finally, to process failure data collected by various sensors, a fault identification mechanism using evidence theory was presented. Feasibility and practicability of the proposed method has been verified by the development and the preliminary application of a prototype system.