Predicting pitting severity in gearboxes under unseen operating conditions and fault severities using convolutional neural networks with power spectral density inputs

##plugins.themes.bootstrap3.article.main##

##plugins.themes.bootstrap3.article.sidebar##

Published Oct 26, 2023
Rik Vaerenberg
Douw Marx
Seyed Ali Hosseinli
Fabrizio De Fabritiis
Hao Wen
Rui Zhu
Konstantinos Gryllias

Abstract

The PHM North America 2023 Data Challenge tasked participants to diagnose the pitting fault severity of a gearbox from a three-channel vibration signal.
This work summarizes the authors' proposed diagnostics solution which consists of a convolutional neural network with an ordinal loss criterion, trained on the power spectral density of the signal.
This method is selected based on a rigorous evaluation using three dedicated validation sets, designed to evaluate the model's ability to generalize to unseen operation conditions and fault severities.
Ultimately, the proposed approach achieved a competition validation score of $282.2$ and a test score of $213.3$.

How to Cite

Vaerenberg, R., Marx, D., Hosseinli, S. A., De Fabritiis, F., Wen, H., Zhu, R., & Gryllias, K. . (2023). Predicting pitting severity in gearboxes under unseen operating conditions and fault severities using convolutional neural networks with power spectral density inputs. Annual Conference of the PHM Society, 15(1). https://doi.org/10.36001/phmconf.2023.v15i1.3798
Abstract 188 | PDF Downloads 141

##plugins.themes.bootstrap3.article.details##

Keywords

data challenge, CNN, PSD, Pitting

Section
Data Challenge Papers

Most read articles by the same author(s)