A Fault Diagnosis Method for Rolling Bearings Based on Clustering K-SVD Algorithm

Abstract

The invention discloses a rolling bearing fault diagnosis method based on the clustering K-SVD algorithm. The main steps are as follows: firstly, the original signal is adaptively decomposed by using the time-varying filtering empirical mode decomposition algorithm based on particle swarm optimization, and each original signal is obtained. eigenmode components, and calculate the relative kurtosis index of each component (K cr ) value; then, choose the value with the largest K cr The eigenmode component of the index value is used as the input sample of the clustering K-SVD algorithm for dictionary learning, and an over-complete dictionary D is obtained New ;Finally, using the over-complete dictionary D New , and combined with the orthogonal matching pursuit algorithm to extract the sparse feature of the original signal of the rolling bearing, and perform envelope spectrum analysis on the sparse representation result to extract the fault frequency feature of the rolling bearing. This method effectively solves the problem of low learning accuracy of the classic K-SVD algorithm for the impact characteristics of rolling bearing faults, and is of great significance for the realization of weak fault diagnosis of rolling bearings.

Description

technical field [0001] The invention relates to the technical field of equipment maintenance, in particular to a rolling bearing fault diagnosis method based on a clustering K-SVD algorithm. Background technique [0002] As an important supporting component, rolling bearings have been widely used in rotating machinery. However, due to the harsh operating environment and complex and changeable operating conditions, various damages will inevitably occur to rolling bearings, which will affect the safe and reliable operation of the whole machine. . However, due to the installation location of the sensor, other rotating parts, noisy working environment, electromagnetic interference and other factors, the vibration signal collected by the vibration sensor not only contains periodic pulse components caused by faults, but also has a lot of noise and harmonic interference , so that the effective feature information is submerged, which increases the difficulty of rolling bearing faul...

AI Technical Summary

Problems solved by technology

[0004] However, the vibration signal of rolling bearings not only contains periodic impact components, but also contains a large number of harmonic components and noise interference, so that the learned dictionary will inevitably contain atoms similar to noise or harmonic components, thereby reducing the impact of rolling bearing faults The effective extraction of features leads to the inability to diagnose rolling bearing faults well

Images