NOVIC+ Motor compound fault dataset (part 1)

https://www.sciencedirect.com/science/article/pii/S0888327025014876?dgcid=coauthor

Please cite above paper when you use this dataset...!

There are part1, part2 and part3. Please download all the dataset.

Part2: https://zenodo.org/records/15743009

Part3: https://zenodo.org/records/15743374

Submitted to Mechanical Systems and Signal Processing on May 9th, 2025

The increasing complexity of rotating machinery and the diversity of operating conditions, such as rotating speed and varying torques, have amplified the challenges in fault diagnosis in scenarios requiring domain adaptation, particularly involving compound faults. This study addresses these challenges by introducing a novel multi-output classification (MOC) framework tailored for domain adaptation in partially labeled (PL) target datasets. Unlike conventional multi-class classification (MCC) approaches, the proposed MOC framework classifies the severity levels of compound faults simultaneously. Furthermore, we explore various single-task and multi-task architectures applicable to the MOC formulation-including shared trunk and cross-talk-based designs-for compound fault diagnosis under PL conditions. Based on this investigation, we propose a novel cross-talk layer structure that enables selective information sharing across diagnostic tasks, effectively enhancing classification performance in compound fault scenarios. In addition, frequency-layer normalization was incorporated to improve domain adaptation performance on motor vibration data. Compound fault conditions were implemented using a motor-based test setup, and the proposed model was evaluated across six domain adaptation scenarios. The experimental results demonstrate its superior macro F1 performance compared to baseline models. We further showed that the proposed mode's structural advantage is more pronounced in compound fault settings through a single-fault comparison. We also found that frequency-layer normalization fits the fault diagnosis task better than conventional methods. Lastly, we discuss that this improvement primarily stems from the model's structural ability to leverage inter-fault classification task interactions, rather than from a simple increase in model parameters.

Please reorganize the file directory as below

dataset_prepared

ㄴ train_data_4s_clf_subsetA.npy

ㄴ train_npy_name_4s_clf_subsetA.npy

...

The subset is divided by the rpm operation condition.

Subset A: Sinusoidal rpm + Manually controlled torque load

Subset B: Triangular rpm + No manually controlled torque load

Subset C: Constant rpm + No manually controlled torque load

Subset E: Dataset composed of all operation conditions written above.

Each dataset consists of two files.

xxx_data_4s_clf_subsetx.npy and xxx_npy_name_4s_clf_subsetx.npy

They are aligned in the same order.

Each file has the following information.

<ex> './processed_data_4s/subsetA/anomaly/inner02_outer02_misalign0_unbalance10034/13_49.npy'

<ex> './processed_data_4s/subsetA/normal/1_14.npy'

1. Subset information. Indicates operating condition.
2. Status of the given folder. The severities of IRF, ORF, misalignment, and unbalance.
3. File index.

The numpy file has below dimensions.

(N, T=102400=25600*4, C=9) Where N is the number of files, T is the time length, and C is the channel.

For T, the sampling rate is 25.6 kHz, and each data has 4 s time length, so the total time index is 102400.

Each channel index indicates below physical meanings.

idx = 0: Vibration data from vibration sensor installed on bearing housing A, with direction perpendicular to the ground.

idx = 1: Vibration data from vibration sensor installed on bearing housing A, with direction parallel to the ground.

idx = 2: Vibration data from vibration sensor installed on bearing housing B, with direction perpendicular to the ground.

idx = 3: Vibration data from vibration sensor installed on bearing housing B, with direction parallel to the ground.

idx = 4: Temperature data from the temperature sensor installed on bearing housing A.

idx = 5: Temperature data from the temperature sensor installed on bearing housing B.

idx = 6: Torque load.

idx = 7: rpm measured before the gearbox.

idx = 8: rpm measured after the gearbox.