Research on Pipeline Leak Signal Reconstruction and Multi-Aperture Classification Method Based on Multi-Task 1D U-Net
Bibliographic record
Abstract
Pipeline leak detection is a key technology to ensure the safe operation of oil-gas and industrial pipelines. However, actually collected monitoring signals are often affected by noise interference, data loss, and difficult feature extraction, which seriously restrict detection accuracy and reliability. To address these challenges, this paper proposes a multi-task 1D U-Net based pipeline leak detection model, realizing collaborative optimization of signal reconstruction, denoising, and aperture classification. The model adopts a shared encoder and task-specific decoders architecture, integrates channel and spatial attention mechanisms to focus on key features, and uses uncertainty learning to dynamically adjust task loss weights, improving training stability and reducing manual tuning costs. Experimental verification on a 4,778-sample multi-channel dataset (each sample is a 5-channel,720-length time-series signal) shows significant performance: denoising task MSE drops to 0.0013(training set) and 0.0016(validation set) with over 98% reduction, reconstruction MSE improves by 44.8%, and aperture classification training set accuracy reaches 78.98%(F1-score 75.52%). SNR improvement rates are 5.08% and 4.83%. Compared with traditional single-task models, the method achieves feature sharing and complementarity, reducing training costs and improving performance. Despite slight validation fluctuations in classification and reconstruction, the architecture shows strong potential in denoising and feature extraction. Future work will integrate Transformer and self-supervised learning to enhance generalization and engineering value.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 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 source (direct Gemma or distilled Codex), 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".