A Fault Diagnosis Method for Rolling Bearings Based on SSae and BA-ELM

Abstract

The present invention proposes a rolling bearing fault diagnosis method based on SSAE and BA-ELM, comprising the following steps; Step A1: collecting four types of time-domain signals of rolling bearing normal, rolling element fault, outer ring fault, and inner ring fault; Step A2: Iterative filter decomposition is performed on each sample of each group of time domain signals, and eigenmode components containing different time scales are obtained from each sample; Step A3: Calculate the relative energy and arrangement of the first K eigenmode components of each sample Entropy; Step A4: Calculate the time-domain features of each group of time-domain signals; Step A5: Use the stacked sparse autoencoder SSAE to reduce the dimensionality of the above features, and use the result of dimensionality reduction as the fault feature; Step A6: Use the bat algorithm to optimize the limit The learning machine BA-ELM realizes the identification of the working state and fault type of the rolling bearing; the invention can solve the problems that the commonly used signal processing methods lack sufficient robustness, the data dimensionality reduction method is difficult to retain the most effective features, and the classification accuracy of different fault degrees is low. .

Description

technical field [0001] The invention relates to the technical field of rolling bearing fault diagnosis, in particular to a rolling bearing fault diagnosis method based on SSAE and BA-ELM. Background technique [0002] Rolling bearing is one of the important components in rotating machinery. The harsh working environment such as high temperature, high pressure and alternating load makes rolling bearing one of the most vulnerable parts in mechanical equipment. The health of rolling bearings is closely related to production efficiency. Once a failure occurs, it will inevitably cause certain economic losses, and even endanger personal safety. Therefore, if the working conditions of the rolling bearing can be timely and accurately fed back, the losses will be minimized. [0003] The vibration signal contains rich fault information. The key step in the fault diagnosis of rolling bearing is to extract the bearing state information from the vibration signal. Because the vibration s...

AI Technical Summary

Problems solved by technology

[0006] The present invention proposes a rolling bearing fault diagnosis method based on SSAE and BA-ELM, which can solve the problems of insufficient robustness of commonly used signal processing methods, difficulty in retaining the most effective features in data dimensionality reduction methods, and low classification accuracy of different fault degrees.

Images