Self-supervised learning-based two-phase flow regime identification using ultrasonic sensors in an S-shape riser

被引:7
作者
Kuang, Boyu [1 ]
Nnabuife, Somtochukwu G. [2 ]
Whidborne, James F. [1 ]
Sun, Shuang [3 ]
Zhao, Junjie [1 ]
Jenkins, Karl [1 ]
机构
[1] Cranfield Univ, Sch Aerosp Transport & Mfg, Cranfield MK43 0AL, England
[2] Cranfield Univ, Sch Water Energy & Environm, Cranfield MK43 0AL, England
[3] Civil Aviat Univ China, Coll Aviat Engn, 2898 Jinbei Rd, Tianjin 300300, Peoples R China
关键词
Weak-supervised learning; Multiphase flow; Classification; Feature extraction; Complexity; RECOGNITION; NETWORK;
D O I
10.1016/j.eswa.2023.121414
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Two-phase flow regime identification is an essential transdisciplinary topic that spans digital signal processing, artificial intelligence, chemical engineering, and energy. Multiphase flow systems significantly impact pipeline safety, heat transfer, and pressure drop; therefore, precisely identifying the governing flow regime is crucial for effective modeling and design. However, it is challenging due to the geometrical complexity of flow regimes in multiphase flow. With the advances in sensor measurement and machine learning, applying non-destructive tests and self-supervised learning to practical industrial problems has become technically feasible and costeffective. This study applies a weak-supervised learning-based twophase flow regime identification solution using a nondestructive tests ultrasonic sensor in an S-shape riser experimental bed by proposing a self-supervised feature extraction algorithm. The proposed selfsupervised feature extraction algorithm reduces time/labor consumption and human error in data annotation using SSL, which provides full supervision without manual annotation. The self-supervised feature extraction algorithm uses a bottlenecked neural network and encoder-decoder structure to extract compact features. The self-supervised feature extraction algorithm performance is evaluated using an established convolutional neural network-based classifier. The source data was collected from a 10 x 50 m riser experimental rig. The dataset is made available to the community as part of this study. The performance of the approach is comparable with state-of-the-art methods and is also the first successful attempt to apply selfsupervised learning to multiphase flow regime ultrasonic signal identification. This study achieved 98.84%, 0.000663, 0.00312, and 7.71 x 105 in accuracy, root mean square error, categorical cross-entropy, and model complexity, respectively. The practical experiment justifies the robustness, fairness, and practicability in the practical application environment. The proposed self-supervised feature extraction brings new approaches and inspirations for the feature extraction step in identifying a two-phase flow regime, and it will be beneficial to generalize this study in different riser shapes in the future.
引用
收藏
页数:16
相关论文
共 54 条
[1]  
Abadi M., 2016, TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems. Vol, Vs4-X., DOI DOI 10.1093/LIBRARY/S4-X.3.339
[2]   Non-invasive classification of gas-liquid two-phase horizontal flow regimes using an ultrasonic Doppler sensor and a neural network [J].
Abbagoni, Baba Musa ;
Yeung, Hoi .
MEASUREMENT SCIENCE AND TECHNOLOGY, 2016, 27 (08)
[3]   Host-to-host airborne transmission as a multiphase flow problem for science-based social distance guidelines [J].
Balachandar, S. ;
Zaleski, S. ;
Soldati, A. ;
Ahmadi, G. ;
Bourouiba, L. .
INTERNATIONAL JOURNAL OF MULTIPHASE FLOW, 2020, 132
[4]   Artificial intelligence, cyber-threats and Industry 4.0: challenges and opportunities [J].
Becue, Adrien ;
Praca, Isabel ;
Gama, Joao .
ARTIFICIAL INTELLIGENCE REVIEW, 2021, 54 (05) :3849-3886
[5]   Visualization and modelling of flow pattern transitions in a cross-corrugated plate heat exchanger channel with uniform two-phase distribution [J].
Buscher, Susanne .
INTERNATIONAL JOURNAL OF HEAT AND MASS TRANSFER, 2019, 144
[6]   A practical tutorial on the use of nonparametric statistical tests as a methodology for comparing evolutionary and swarm intelligence algorithms [J].
Derrac, Joaquin ;
Garcia, Salvador ;
Molina, Daniel ;
Herrera, Francisco .
SWARM AND EVOLUTIONARY COMPUTATION, 2011, 1 (01) :3-18
[7]   Artificial neural network based shape optimization of supersonic ejectors in the critical flow regime [J].
Gupta, Pradeep ;
Kumar, Pramod ;
Rao, Srisha M. V. .
APPLIED THERMAL ENGINEERING, 2022, 216
[8]   Internal two-phase flow and spray characteristics of outside-in-liquid twin-fluid atomizers [J].
Hammad, Farid A. ;
Sun, Kai ;
Che, Zhizhao ;
Jedelsky, Jan ;
Wang, Tianyou .
APPLIED THERMAL ENGINEERING, 2021, 187
[9]   Identification of liquid-gas flow regime in a pipeline using gamma-ray absorption technique and computational intelligence methods [J].
Hanus, Robert ;
Zych, Marcin ;
Kusy, Maciej ;
Jaszczu, Marek ;
Petryka, Leszek .
FLOW MEASUREMENT AND INSTRUMENTATION, 2018, 60 :17-23
[10]  
Hashmi K. A., 2022, arXiv