A method to diagnose compound fault of rolling bearing with ITD-AF

Abstract

In an engineering practice, the faults of rolling bearing are mostly represented as being compound and hard to diagnose. For that, intrinsic time-scale decomposition (ITD) algorithm was combined with Auto-correlation Function (AF) to extract the characteristics of compound faults of rolling bearing in aviation engine. Firstly, ITD algorithm was used to decompose acceleration signal into multiple rotational and residual trend component; secondly, rotational components were reconstructed to figure out their AF; finally, characteristic frequency of rolling bearing under compound faults mode was extracted by Hilbert spectrum envelope. To validate the effectiveness of the method, a comparative study on sensor installation positions and vibration acceleration signal of different compound faults has been carried out. The result of study shows that the proposed ITD-AF method is capable to extract compound fault characteristics of rolling bearing in an effective and precise manner and the installation positions of sensors, rotation speed and fault type shows insensitivity to extraction.

Fractional Dimensionless Indicator with Random Forest for Bearing Fault Diagnosis under Variable Speed Conditions

“…Rolling bearing is one of the key components among them, which is prone to failure and may cause serious chain reactions [2]. erefore, more and more attention has been paid to the fault diagnosis of aviation bearings [3,4]. In engineering, due to the in uence of starting, stopping, and other working conditions, bearings sometimes run at varying speeds.…”