This is Python code for our paper. The CWRU dataset is too large, please download it yourself if you are interested！
For JNU dataset (data_jnu): The ReadData_JNU_5000.py is for sampling. The CNN_J.py is the model for jnu dataset. The mmd.py is used for domain alignment（MK-MMD） and the TCTL_jnu.py is for test model.
For CWRU dataset (data_cwru): The ReadData200.py is for sampling. The CNN_C.py is the model for cwru dataset. The mmd.py is used for domain alignment（MK-MMD） and the TCTL_cwru.py is for test model.
If you want to use this code, please
- download datasets (cwru and jnu) in the code dir, and rename their folder name to
data_cwruanddata_jnu. - if you want to run model in cwru, run
python TCTL_cwru.py --src 0 --tar 1, where--srccan be chosen in 0, 1, 2, 3, and--tarcan be chosen in 0, 1, 2, 3. Notice that--src!=--tar! - if you want to run model in jnu, run
python TCTL_jnu.py --src 600 --tar 800, where--srccan be chosen in 600, 800, 1000, and--tarcan be chosen in 600, 800, 1000. Notice that--src!=--tar!
I hope my code proves helpful to you. Please feel free to reach out if you have any questions!
The citation format:
Zheng J, Jiang B, Yang C. Proportional periodic sampling for cross-load bearing fault diagnosis[J]. International Journal of Machine Learning and Cybernetics, 2024: 1-13.
@article{zheng2024proportional, title={Proportional periodic sampling for cross-load bearing fault diagnosis}, author={Zheng, Jianbo and Jiang, Bin and Yang, Chao}, journal={International Journal of Machine Learning and Cybernetics}, pages={1--13}, year={2024}, publisher={Springer} }
Postscript：The reference work associated with this work is our previous MDPS at https://github.com/IWantBe/MDPS