Wavelet Convolutional Neural Network with Multilabel Classifier: A Compound Fault Diagnosis Framework and Its Interpretability Analysis
Bibliographic record
Abstract
With the development of intelligent sensing, measurement, and data analytics technologies, exploring intelligent fault diagnosis methods has attracted enormous attention in the past decade, in which many successful attempts have been achieved for compound fault diagnosis of industrial equipment. However, the traditional intelligent compound fault diagnosis methods are pure deep learning algorithms, resulting in these methods being “black box” solutions and cannot be interpreted. To explore a solution for the above problem, an intelligent and interpretable compound fault diagnosis method, named wavelet convolutional neural network with multilabel classifier (WavCNN-MLC), is proposed for industrial equipment. First, the WavCNN-MLC is constructed with the following steps: 1) a wavelet convolutional layer is introduced to substitute the first layer of the traditional convolutional neural network (CNN), which aims to learn features with interpretable meaning from vibration signals; 2) a multilabel classifier is employed to substitute the classifier of traditional CNN, which endows the diagnosis model with the ability to decouple the compound fault intelligently. Second, the WavCNN-MLC is trained and optimized with the single and compound fault samples in a supervised way. Finally, a variant of Gradient-weighted Class Activation Mapping, called Grad-CAM++, is applied to analyze the relationship between the target class and the key activation areas of inputs, which can provide visual explanations for the classification results of the WavCNN-MLC. The experimental results on a gearbox dataset show that the proposed method can effectively decouple the compound faults and demonstrate that the WavCNN-MLC is interpretable through the visualization of the saliency map.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".