Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Bearing fault diagnosis method based on transfer learning

A technology of transfer learning and fault diagnosis, which is applied in the direction of mechanical bearing testing, etc.

Active Publication Date: 2016-02-03
SOUTHEAST UNIV
View PDF5 Cites 31 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the shortcomings of the existing machine learning bearing fault diagnosis method, the present invention provides a bearing fault diagnosis method based on transfer learning

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Bearing fault diagnosis method based on transfer learning
  • Bearing fault diagnosis method based on transfer learning
  • Bearing fault diagnosis method based on transfer learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0058] The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0059] see figure 1 , a bearing fault diagnosis method based on transfer learning of the present invention comprises the following steps:

[0060] Step A: Use autocorrelation singular value decomposition (SVD) for the target data and auxiliary data to extract the contained fault singular value vectors (ie, fault feature vectors), and perform sample collection according to the selection rules of the training data set T and the test data set S select;

[0061] In step A, the target data source is: the vibration data of the target bearing object in the target environment, and the auxiliary data source is: the vibration data of the non-target environment or non-target object; the target data and auxiliary data of the bearing system are vibration acceleration signals, and the bearing An example of a vibration signal is shown in figure 2 shown;...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a bearing fault diagnosis method based on transfer learning. The method comprises the steps of selection rules of a training data set and a test data set; auxiliary data normalization transferable degree quantification and a threshold selection method; and the algorithm flow of a TrAdaboost algorithm based on weight adjustment used for bearing fault classification. A thinking strategy that auxiliary data in transfer learning assist target data learning is introduced by aiming at problems of lacking of diagnosis generality caused by the conditions of variable working conditions, unavailable direct measurement and new fault types. According to the method, diagnosis precision of conventional machine learning in a small amount of target data can be enhanced, environmental adaptability of bearing fault diagnosis is enhanced, and the method has potential economic value in the field of bearing fault application under variable working condition loads, new faults and indirect measurement.

Description

technical field [0001] The invention relates to a mechanical fault diagnosis method, in particular to a bearing fault diagnosis method based on migration learning. Background technique [0002] Most of the current bearing fault diagnosis is based on the assumption that the training data set and the test data set have the same feature space and data distribution, and most of the models are based on the laboratory environment or ideal environment, ignoring the actual situation of the mechanical system when it is running. Conditions such as variable working conditions, inability to measure directly, and new fault types often lead to less or even unobtainable target fault data in bearing fault diagnosis, and lead to different distribution characteristics of training fault data and target fault data. The traditional machine learning method is based on the premise that the training data and the test data have the same characteristics and the amount of data is sufficient, so it is ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G01M13/04
Inventor 严如强沈飞
Owner SOUTHEAST UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com