Experimental Investigations of Distributed Fiber Optic Sensors for Water Pipeline Monitoring DOI Creative Commons
Manuel Bertulessi, Daniele Fabrizio Bignami, Ilaria Boschini

et al.

Sensors, Journal Year: 2023, Volume and Issue: 23(13), P. 6205 - 6205

Published: July 6, 2023

Water Loss (WL) is a global issue. In Italy, for instance, WL reached 36.2% of the total fresh water conveyed in 2020. The maintenance supply system strategic task that requires huge amount investment every year. this work, we focused on use Distributed Fiber Optic Sensors (DFOS) based Stimulated Brillouin Scattering (SBS) technology monitoring pipeline networks. We worked High-Density Polyethylene (HDPE) pipes, today most widely used creating pipelines. By winding and fixing optic fiber cable pipe’s external surface, verified ability to detect strain related pressure anomalies along pipeline, e.g., those caused by leakage. performed two experimental phases. first one, assessed sensibility sensor layout an HDPE solicited with static pressure. investigated viscoelastic rheology material calibrating validating parameters Burger model, which Maxwell Kelvin-Voigt models are connected series. second phase, instead, detection anomaly produced leakage circuit set up running moved pump. theoretical studies returned overall positive feedback DFOS Future developments will be more detailed solution industrial production “natively smart” pipes cables integrated into surface during extrusion process.

Language: Английский

Performance analysis of various machine learning algorithms for CO2 leak prediction and characterization in geo-sequestration injection wells DOI Creative Commons
Saeed Harati, Sina Rezaei Gomari, Mohammad Azizur Rahman

et al.

Process Safety and Environmental Protection, Journal Year: 2024, Volume and Issue: 183, P. 99 - 110

Published: Jan. 4, 2024

The effective detection and prevention of CO2 leakage in active injection wells are paramount for safe carbon capture storage (CCS) initiatives. This study assesses five fundamental machine learning algorithms, namely, Support Vector Regression (SVR), K-Nearest Neighbor (KNNR), Decision Tree (DTR), Random Forest (RFR), Artificial Neural Network (ANN), use developing a robust data-driven model to predict potential incidents wells. Leveraging wellhead bottom-hole pressure temperature data, the models aim simultaneously location size leaks. A representative dataset simulating various leak scenarios saline aquifer reservoir was utilized. findings reveal crucial insights into relationships between variables considered characteristics. With its positive linear correlation with depth leak, could be pivotal indicator location, while negative relationship well demonstrated strongest association size. Among predictive examined, highest prediction accuracy achieved by KNNR both localization sizing. displayed exceptional sensitivity size, able identify magnitudes representing as little 0.0158% total main flow relatively high levels accuracy. Nonetheless, underscored that accurate sizing posed greater challenge compared localization. Overall, obtained can provide valuable development efficient well-bore systems.

Language: Английский

Citations

17
A Hybrid Deep Learning Approach: Integrating Short-Time Fourier Transform and Continuous Wavelet Transform for Improved Pipeline Leak Detection DOI Creative Commons
Muhammad Siddique, Zahoor Ahmad, Niamat Ullah

et al.

Sensors, Journal Year: 2023, Volume and Issue: 23(19), P. 8079 - 8079

Published: Sept. 25, 2023

A hybrid deep learning approach was designed that combines with enhanced short-time Fourier transform (STFT) spectrograms and continuous wavelet (CWT) scalograms for pipeline leak detection. Such detection plays a crucial role in ensuring the safety integrity of fluid transportation systems. The proposed model leverages power STFT CWT to enhance capabilities. pipeline's acoustic emission signals during normal operating conditions undergo transformation using CWT, creating representing energy variations across time-frequency scales. To improve signal quality eliminate noise, Sobel denoising filters are applied scalograms. These filtered then fed into convolutional neural networks, extracting informative features harness distinct characteristics captured by both CWT. For computational efficiency discriminatory power, principal component analysis is employed reduce feature space dimensionality. Subsequently, leaks accurately detected classified categorizing reduced dimensional t-distributed stochastic neighbor embedding artificial networks. achieves high accuracy reliability detection, demonstrating its effectiveness capturing spectral temporal details. This research significantly contributes monitoring maintenance offers promising solution real-time diverse industrial applications.

Language: Английский

Citations

35
Pipeline Leak Detection: A Comprehensive Deep Learning Model Using CWT Image Analysis and an Optimized DBN-GA-LSSVM Framework DOI Creative Commons
Muhammad Siddique, Zahoor Ahmad,

N. Ullah

et al.

Sensors, Journal Year: 2024, Volume and Issue: 24(12), P. 4009 - 4009

Published: June 20, 2024

Detecting pipeline leaks is an essential factor in maintaining the integrity of fluid transport systems. This paper introduces advanced deep learning framework that uses continuous wavelet transform (CWT) images for precise detection such leaks. Transforming acoustic signals from pipelines under various conditions into CWT scalograms, followed by signal processing non-local means and adaptive histogram equalization, results new enhanced leak-induced scalograms (ELIS) capture detailed energy fluctuations across time-frequency scales. The fundamental approach takes advantage a belief network (DBN) fine-tuned with genetic algorithm (GA) unified least squares support vector machine (LSSVM) to improve feature extraction classification accuracy. DBN-GA precisely extracts informative features, while LSSVM classifier distinguishes between leaky non-leak conditions. By concentrating solely on capabilities ELIS processed through optimized DBN-GA-LSSVM model, this research achieves high accuracy reliability, making significant contribution monitoring maintenance. innovative capturing complex patterns can be applied real-time leak critical infrastructure safety several industrial applications.

Language: Английский

Citations

15
Advancing deep learning-based acoustic leak detection methods towards application for water distribution systems from a data-centric perspective DOI
Yipeng Wu,

Xingke Ma,

Guancheng Guo

et al.

Water Research, Journal Year: 2024, Volume and Issue: 261, P. 121999 - 121999

Published: June 24, 2024

Language: Английский

Citations

14
Pipeline Leak Detection System for a Smart City: Leveraging Acoustic Emission Sensing and Sequential Deep Learning DOI Creative Commons

N. Ullah,

Muhammad Siddique, Saif Ullah

et al.

Smart Cities, Journal Year: 2024, Volume and Issue: 7(4), P. 2318 - 2338

Published: Aug. 20, 2024

This study explores a novel approach utilizing acoustic emission (AE) signaling technology for pipeline leakage detection and analysis. Pipeline leaks are significant concern in the liquids gases industries, prompting development of innovative methods. Unlike conventional methods, which often require contact visual inspection with surface, proposed time-series-based deep learning offers real-time higher safety efficiency. In this study, we propose an automatic system efficient transportation liquid (water) gas across city, considering smart city approach. We AE-based framework combined time-series algorithms to detect through features. The AE signal module is designed capture subtle changes state caused by leaks. Sequential models, including long short-term memory (LSTM), bi-directional LSTM (Bi-LSTM), gated recurrent units (GRUs), used classify response into normal from minor seepage, moderate leakage, major ruptures pipeline. Three sensors installed at different configurations on pipeline, data acquired 1 MHz sample/sec, decimated 4K sample/second efficiently constraints remote system. performance these models evaluated using metrics, namely accuracy, precision, recall, F1 score, convergence, demonstrating classification accuracies up 99.78%. An accuracy comparison shows that BiLSTM performed better mostly all hyperparameter settings. research contributes advancement technology, offering improved reliability identifying addressing integrity issues.

Language: Английский

Citations

10
Microleakage Acoustic Emission Monitoring of Pipeline Weld Cracks under Complex Noise Interference: A Feasible Framework DOI
Zhifen Zhang, Jing Huang,

Yan-Long Yu

et al.

Journal of Sound and Vibration, Journal Year: 2025, Volume and Issue: unknown, P. 118980 - 118980

Published: Jan. 1, 2025

Language: Английский

Citations

1
Application of machine learning to leakage detection of fluid pipelines in recent years: A review and prospect DOI

Jianwu Chen,

Xiao Wu, Zhibo Jiang

et al.

Measurement, Journal Year: 2025, Volume and Issue: unknown, P. 116857 - 116857

Published: Jan. 1, 2025

Language: Английский

Citations

1
Advanced transformer model for simultaneous leakage aperture recognition and localization in gas pipelines DOI
Pengyu Li, Xiufang Wang, Chunlei Jiang

et al.

Reliability Engineering & System Safety, Journal Year: 2023, Volume and Issue: 241, P. 109685 - 109685

Published: Sept. 24, 2023

Language: Английский

Citations

18
Machine Learning Model for Leak Detection Using Water Pipeline Vibration Sensor DOI Creative Commons
Suan Lee,

B. Kim

Sensors, Journal Year: 2023, Volume and Issue: 23(21), P. 8935 - 8935

Published: Nov. 2, 2023

Water leakage from aging water and wastewater pipes is a persistent problem, necessitating the improvement of existing leak detection response methods. In this study, we conducted an analysis essential features based on data collected sensors installed at meter boxes outlets pipelines. The pipeline through vibration sensor were preprocessed by converting it into tabular form frequency band applied to various machine learning models. characteristics each model analyzed, XGBoost was selected as most suitable with high accuracy 99.79%. These systems can effectively reduce time, minimize waste, economic losses. Additionally, technology be fields that utilize pipes, making widely applicable.

Language: Английский

Citations

17
A review of leak detection methods based on pressure waves in gas pipelines DOI

Linkun Zhao,

Zheng Cao, Jianqiang Deng

et al.

Measurement, Journal Year: 2024, Volume and Issue: 236, P. 115062 - 115062

Published: June 6, 2024

Language: Английский

Citations

8