An Improved Convolutional Neural Network for Pipe Leakage Identification Based on Acoustic Emission DOI Creative Commons
Weidong Xu, Jiwei Huang,

Lianghui Sun

et al.

Journal of Marine Science and Engineering, Journal Year: 2024, Volume and Issue: 12(10), P. 1720 - 1720

Published: Sept. 30, 2024

Oil and gas pipelines are the lifelines of energy market, but due to long-term use environmental factors, these prone corrosion leaks. Offshore oil pipeline leaks, in particular, can lead severe consequences such as platform fires explosions. Therefore, it is crucial accurately swiftly identify leaks on offshore platforms. This significant importance for improving early warning systems, enhancing maintenance efficiency, reducing economic losses. Currently, efficiency identifying still needs improvement. To address this, present study first established an experimental simulate a marine environment. Laboratory leakage signal data were collected, on-site noise gathered from “Liwan 3-1” platform. By integrating signals with data, this aimed closely mimic real-world application scenarios. Subsequently, several neural network-based identification methods applied integrated dataset, including probabilistic network (PNN) combined time-domain feature extraction, Backpropagation Neural Network (BPNN) optimized simulated annealing particle swarm optimization, Long Short-Term Memory (LSTM) Mel-Frequency Cepstral Coefficients (MFCC). Corresponding models constructed, effectiveness leak detection was validated using test sets. Additionally, paper proposes improved convolutional (CNN) technology named SART-1DCNN. optimizes architecture by introducing attention mechanisms, transformer modules, residual blocks, combining them Dropout optimization algorithms, which significantly enhances recognition accuracy. It achieves high accuracy rate 99.44% dataset. work capable detecting

Language: Английский

Milling Machine Fault Diagnosis Using Acoustic Emission and Hybrid Deep Learning with Feature Optimization DOI Creative Commons

Muhammad Umar,

Muhammad Siddique,

N. Ullah

et al.

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(22), P. 10404 - 10404

Published: Nov. 12, 2024

This paper presents a fault diagnosis technique for milling machines based on acoustic emission (AE) signals and hybrid deep learning model optimized with genetic algorithm. Mechanical failures in machines, particularly critical components like cutting tools, gears, bearings, account significant portion of operational breakdowns, leading to unplanned downtime financial losses. To address this issue, the proposed method first acquires AE from machine. signals, capturing dynamic responses machine components, are transformed into continuous wavelet transform (CWT) scalograms further analysis. Gaussian filtering is applied enhance clarity these scalograms, effectively reducing noise while maintaining essential features. A convolutional neural network (CNN) VGG16 architecture utilized spatial feature extraction, followed by bidirectional long short-term memory (BiLSTM) capture temporal dependencies scalograms. The algorithm (GA) used optimize selection ensure most relevant features improve model’s performance. finally fed fully connected (FC) layer classification. achieves an accuracy 99.6%, significantly outperforming traditional approaches. offers highly accurate efficient solution detection allowing more reliable predictive maintenance efficiency industrial settings.

Language: Английский

Citations

9
Critical challenges and advances in vibration signal processing for non-stationary condition monitoring DOI
Anil Kumar, Agnieszka Wyłomańska, Radosław Zimroz

et al.

Advanced Engineering Informatics, Journal Year: 2025, Volume and Issue: 65, P. 103290 - 103290

Published: April 4, 2025

Language: Английский

Citations

0
Enhanced semi-supervised model for acoustic leak detection in water distribution networks DOI
Changjiang Wang, Wei Qian, Shuanglin Shen

et al.

Automation in Construction, Journal Year: 2025, Volume and Issue: 175, P. 106228 - 106228

Published: April 25, 2025

Language: Английский

Citations

0
Stuck Pipe Detection in Oil and Gas Drilling Operations Using Deep Learning Autoencoder for Anomaly Diagnosis DOI Creative Commons

Hasan N. Al-Mamoori,

Jialin Tian, Haifeng Ma

et al.

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(9), P. 5042 - 5042

Published: May 1, 2025

Stuck pipe events remain a critical challenge in oil and gas drilling operations, leading to increased non-productive time substantial financial losses. Traditional detection methods rely on manual monitoring expert judgment, which are prone delays human error. This study proposes deep learning autoencoder-based anomaly diagnosis approach enhance the of stuck incidents. Using high-resolution series data from Volve field, autoencoder model was trained exclusively normal conditions learn operational patterns detect deviations indicative events. The proposed leverages reconstruction error as an metric, effectively distinguishing between cases. results demonstrate that achieves accuracy 99.06%, with area under receiver operating characteristic curve (AUC) 0.958. Additionally, attained precision 97.12%, recall 91.34%, F1-score 94.15%, significantly reducing false positives negatives. findings highlight potential learning-based approaches improving real-time detection, offering scalable cost-effective solution for mitigating disruptions. research contributes advancing intelligent systems industry, risks, enhancing efficiency.

Language: Английский

Citations

0
An Improved Convolutional Neural Network for Pipe Leakage Identification Based on Acoustic Emission DOI Creative Commons
Weidong Xu, Jiwei Huang,

Lianghui Sun

et al.

Journal of Marine Science and Engineering, Journal Year: 2024, Volume and Issue: 12(10), P. 1720 - 1720

Published: Sept. 30, 2024

Oil and gas pipelines are the lifelines of energy market, but due to long-term use environmental factors, these prone corrosion leaks. Offshore oil pipeline leaks, in particular, can lead severe consequences such as platform fires explosions. Therefore, it is crucial accurately swiftly identify leaks on offshore platforms. This significant importance for improving early warning systems, enhancing maintenance efficiency, reducing economic losses. Currently, efficiency identifying still needs improvement. To address this, present study first established an experimental simulate a marine environment. Laboratory leakage signal data were collected, on-site noise gathered from “Liwan 3-1” platform. By integrating signals with data, this aimed closely mimic real-world application scenarios. Subsequently, several neural network-based identification methods applied integrated dataset, including probabilistic network (PNN) combined time-domain feature extraction, Backpropagation Neural Network (BPNN) optimized simulated annealing particle swarm optimization, Long Short-Term Memory (LSTM) Mel-Frequency Cepstral Coefficients (MFCC). Corresponding models constructed, effectiveness leak detection was validated using test sets. Additionally, paper proposes improved convolutional (CNN) technology named SART-1DCNN. optimizes architecture by introducing attention mechanisms, transformer modules, residual blocks, combining them Dropout optimization algorithms, which significantly enhances recognition accuracy. It achieves high accuracy rate 99.44% dataset. work capable detecting

Language: Английский

Citations

3