dataset: The folder where the dataset is stored. Open the Readme file in the dataset directory to download the public dataset of this paper.
main: outputs directory stores the best weight parameter file of each model, and log file in main directory is the experimental log file (log generated by running code). Main.py contains the code of most models in this experiment. main_cwt.py stores the code transformed by CWT and the model experiment (only CNN), main_bls.py stores the code of the width learning network, and main_ml.py stores the experimental code of the machine learning method on this dataset.
Models: contains FCNN, CNN, BLS, LSTM, QCNN, WDCNN, RNN, SAC, SACD model framework code.
utils: data_process_cwt.py in the utils directory is the code of the continuous wavelet transform, dataset.py stores each implementation class of the loaded dataset, and see.py is the fixed seed file. The seed is 2023, that is, the year of the experiment.
If you have any questions about the code file, please contact the author at lil_ken@163.com.
ps: The first public code experiment, there may be a lot of inconsiderate places, but also please understand the researchers.
lf you find this repo helpful, please cite the following paper:
@inproceedings{zeng-etal-2024,
title = "Fault Diagnosis of Rotating Equipment Unbalance Problem Based on Denoising Stacked Autoencoders",
author = "Zeng, Peijian and Lin Nankai and Lin Jianghao and Yang Aimin and Hou Liheng and Lu Maohua",
booktitle = "International Conference on Intelligent Computing: ICIC 2024",
year = "2024",
publisher = "Lecture Notes in Artificial Intelligence (LNAI)" }