Wanqian Yang

journal · 2025

Self-supervised condition monitoring for wind turbine gearbox based on adaptive feature selection and contrastive residual graph neural network

Authors: Wanqian Yang, Coauthor A, Coauthor B

Venue: Manuscript

Status: Under review

Wind turbineGearboxSCADAGraph neural networkSelf-supervised learning

This work develops a self-supervised condition monitoring framework for wind turbine gearboxes using SCADA data.

The framework combines adaptive feature selection, graph construction based on feature relationships, contrastive residual graph representation learning, and health-index-based early warning.

You can replace this placeholder description with the final abstract, contribution list, figures, BibTeX, and paper links.