Generative adversarial network-based fault diagnosis model for railway point machine in sustainable railway transportation
by Xiao Han; Ning Zhao
International Journal of Sensor Networks (IJSNET), Vol. 43, No. 1, 2023

Abstract: The railway point machine is a significant element of railway signalling equipment that keeps trains running safely. Railway point devices are actuators that move switchblades to reroute trains in different directions. The features are manually designed and used for dimension reduction in conventional methods. This research proposes a generative adversarial network-based fault diagnosis model in the railway point machine based on vibration signals for sustainable railway transportation. It is pre-trained on datasets; the test error can be forecasted without extra hyperparameter tuning in railway points. The system provides a method for obtaining Mel-frequency-cestrum-coefficients from audio information with fewer feature dimensions utilising a gradient-boosted decision tree and the generative adversarial network incorporating a discriminative model in its process for early prediction and classification of abnormalities in railway points. The numerical outcomes indicate that the suggested model enhances the classification accuracy ratio, precision ratio, recall ratio, and fault audio diagnosis performance ratio.

Online publication date: Tue, 03-Oct-2023

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Sensor Networks (IJSNET):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com