Research On Fault Detection And Diagnosis Technology Of Rolling Bearing Based On Graph Structure

As an important part of mechanical equipment,especially rotating machinery,rolling bearings helath condition directly affect the safety of the whole production line.Therefore,fault detection and diagnosis is of much importance toreduce maintenance cost and enhance the operation reliability of industrial system,thus becoming a research hotspot in the past decade.This study focusingon the non-lineary and complex rolling bearing signal,explores an unified framework based on the graph model for this problem.The presented work is summarized as follows.First,this study expounds the background and significance of the study on rolling bearing fault detection and diagnosis where the development of the fault diagnosis is introduced and summarized.On the basis of this,the main ideas and reserach focues of this study are described.Second,this study provids an analysis of fault characteristics of rolling bearing.Based on this,a novel graph based model is developed to characterize the dynamical behavior of collection vibration signals.Third,together with the graph-based modeling,an automatic analysis of continous monitoring of vibrations signals by means of entropy,is presented to the early fault detection and diagnosis of rolling bearings during its successive operations.The Martingale-test is used to fault detection;the SVM classifer is used to identify the fault type.Two public data sets are used to validate the proposed method on the fault detection and fault diagnosis,respectively.Finally,we summarize the main content and innovation points,and discuss the shortcomings of this study and present the future work.