We released an open-source motor failure dataset.
We hope this dataset can benefit your research.
data of Google Drive: (We will release dataset as soon as possible)
This dataset comprises vibration signals and acoustic signals from motors in 6 different health states under 4 distinct operating conditions.
These datasets are publicly available, and anyone can use them to validate diagnosis algorithms.
Publications making use of the HUSTmotor multimodal datasets are requested to cite the following paper.
Chao Zhao, Enrico Zio, Weiming Shen, Domain Generalization for Cross-Domain Fault Diagnosis: an Application-oriented Perspective and a Benchmark Study, Reliability Engineering and System Safety (2024), doi: https://doi.org/10.1016/j.ress.2024.109964.
Fig. 1. (a) Test rig of HUSTbearing dataset.
The Motor fault tests were conducted using a Spectra-Quest Mechanical Fault Simulator, as depicted in Fig. 1
The motors in six health states includes:
(1) healthy----H
(2) bearing fault----BF
(3) bowed rotor----BOW
(4) broken rotor bars----BROKEN
(5) rotor misalignment----MISAL
(6) voltage unbalance----UNBAL
It's important to note that all faults are artificially preset.
A total of 4 different operating conditions were set in experiments. The operating conditions include:
- 5 Hz
- 10 Hz
- 20 Hz
- 30 Hz
The sampling frequency is set to 25.6 kHz. A total of 262144 data points (i.e. 10.2s) are recorded for each sampling.
The raw data file comprises 24 files (6 health states multiplied by 4 working conditions), each in txt format.
For instance, the filename "H_5HZ" indicates a healthy motor under the 5 Hz working condition.
The health states are represented by the following codes:
H: healthy
BF: bearing fault
BOW: bowed rotor
BROKEN: broken rotor bars
MISAL: rotor misalignment
UNBAL: voltage unbalance
If you have any questions or suggestions, do not hesitate to contact:
Mr. Chao Zhao, zhaochao734@hust.edu.cn