SuperGraph

Codes are released for paper entitled "SuperGraph: Spatial-temporal graph-based feature extraction for rotating machinery diagnosis", and this paper is revised in IEEE Transactions on Industrial Electronics.

Guide

This project only provides the data and related codes used in the paper that allows anyone to reproduce all results presented in the paper. All algorithms were written by python 3.8, PyTorch 1.7.0, and PyTorch geometric through running on a computer with an Intel Core 298 i7-8700K CPU and a 32G RAM.

Requirements

• Python 3.8
• Numpy 1.16.2
• sklearn 0.21.3
• Scipy 1.2.1
• Pytorch 1.7.0
• Pytroch Geometric

Datasets

• CWRU Bearing Dataset
• Gearbox dataset
• PU Bearing Dataset

Pakages

This repository is organized as:

• code contains the corresonding codes.
• data contains the used data from CWRU dataset, Gearbox dataset, and KAT dataset.

Usage

• Download the datasets
• Modify the file storage address in the code
• Run the corresponding codes

Citation

Paper:

@article{Yang2021,
  title={SuperGraph: Spatial-temporal graph-based feature extraction for rotating machinery diagnosis},
  author={Chaoying Yang, Kaibo Zhou, and Jie Liu},
  journal={IEEE Transactions on Industrial Electronics},
  year={2021},
  doi={10.1109/TIE.2021.3075871},
  url={https://github.com/ChaoyingYang/SuperGraph},
}

Contact

Chaoying Yang - [email protected]

Jie Liu - [email protected]