Condition-based monitoring system for rolling element bearing using a generic multi-layer perceptron

Rolling element bearings are critical mechanical components in rotating machinery and fault detection in the early stages of damage is important to prevent their malfunctioning and failure. Vibration monitoring is the most widely used and cost-effective monitoring technique to detect, locate and dis...

Full description

Saved in:
Bibliographic Details
Published in:Journal of vibration and control 2015-12, Vol.21 (16), p.3456-3464
Main Authors: de Almeida, Luis F, Bizarria, José WP, Bizarria, Francisco CP, Mathias, Mauro H
Format: Article
Language:English
Subjects:
Bearings
Fault diagnosis
Monitoring systems
Motion control
Rolling
Vibration
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Rolling element bearings are critical mechanical components in rotating machinery and fault detection in the early stages of damage is important to prevent their malfunctioning and failure. Vibration monitoring is the most widely used and cost-effective monitoring technique to detect, locate and distinguish faults in rolling element bearings. This paper purposes single hidden layer architecture for fault diagnosis of rolling element bearings. The particular of this proposed architecture is its ability to generalize for solving both basic classification and fault identification. The network uses the features of time-domain vibration signals with normal and defective bearings. The Multi Layer Perceptron (MLP) was trained and tested with a set of experimental data obtained from previous experiments developed by FEG, CWRU and RANDALL laboratories. The results show the effectiveness of the MLP to diagnose the machine condition for the various data used.
ISSN:1077-5463
1741-2986
DOI:10.1177/1077546314524260