A Novel Deep Learning Framework for Industrial Multiphase Flow Characterization

被引:51
作者
Dang, Weidong [1 ]
Gao, Zhongke [1 ,2 ]
Hou, Linhua [1 ]
Lv, Dongmei [1 ]
Qiu, Shuming [3 ]
Chen, Guanrong [4 ]
机构
[1] Tianjin Univ, Sch Elect & Informat Engn, Tianjin 300072, Peoples R China
[2] Tianjin Univ, Key Lab Efficient Utilizat Low & Medium Grade Ene, Minist Educ, Tianjin 300350, Peoples R China
[3] Elect Sci Co Ltd, Tianjin Res Inst, Tianjin 300180, Peoples R China
[4] City Univ Hong Kong, Dept Elect Engn, Hong Kong, Peoples R China
基金
中国国家自然科学基金;
关键词
Impedance measurement; Electrodes; Impedance; Deep learning; Oils; Sensitivity; Solid modeling; flow behavior; impedance sensor measurement system; oil-water flow; CONVOLUTIONAL NEURAL-NETWORKS; WATER HOLDUP; TIME-SERIES; PREDICTION; SENSOR;
D O I
10.1109/TII.2019.2908211
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Due to the inherent disturbances associated with flow structures, measurement of the complicated flow parameters in multiphase flows remains a challenging problem of significant importance. The flow dynamical behaviors are still elusive. In this paper, a multichannel complex impedance measurement system is designed to cope with this difficult issue. First, the geometry of the distributed multielectrode impedance sensor is optimized and a matched hardware measurement system is developed. After performance evaluation, a convolutional neural network and long short-term memory based measurement model is formulated for measuring flow parameters with high accuracy. The mean absolute error is only 0.36 for water cut and 0.77 for total flow velocity. Further, from the perspective of Lempel-Ziv complexity and mutual information, the relationship between the diverse flow structures and spatial flow behaviors is explored, leading to a deeper understanding of oil-water flows. All the experimental and analytical results demonstrate that the combination of deep learning and the designed impedance sensor measurement system allows measuring the complicated flow parameters, thereby characterizing the flow structures and behaviors. This opens up a new venue for exploring industrial multiphase flows and serving for an efficient oilfield exploitation as well.
引用
收藏
页码:5954 / 5962
页数:9
相关论文
共 50 条
[21]   Deep Air Quality Forecasting Using Hybrid Deep Learning Framework [J].
Du, Shengdong ;
Li, Tianrui ;
Yang, Yan ;
Horng, Shi-Jinn .
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING, 2021, 33 (06) :2412-2424
[22]   Deep learning models for improved accuracy of a multiphase flowmeter [J].
Manami, Mohammadreza ;
Seddighi, Sadegh ;
Orlu, Ramis .
MEASUREMENT, 2023, 206
[23]   A Novel Forecasting Model for Solar Power Generation by a Deep Learning Framework with Data Preprocessing and Post processing [J].
Quoc-Thang Phan ;
Wu, Yuan-Kang ;
Quoc-Dung Phan ;
Lo, Hsin-Yen .
2022 IEEE/IAS 58TH INDUSTRIAL AND COMMERCIAL POWER SYSTEMS TECHNICAL CONFERENCE (I&CPS), 2022,
[24]   Efficient deep learning techniques for multiphase flow simulation in heterogeneous porousc media [J].
Wang, Yating ;
Lin, Guang .
JOURNAL OF COMPUTATIONAL PHYSICS, 2020, 401
[25]   A novel sequential-based hybrid approach incorporating physical modeling and deep learning for multiphase subsurface flow simulation [J].
Du, Shuyi ;
Zhang, Jingyan ;
Yue, Ming ;
Xie, Chiyu ;
Wang, Yuhe ;
Song, Hongqing .
GAS SCIENCE AND ENGINEERING, 2023, 118
[26]   Hybrid deep learning framework for human activity recognition [J].
Pushpalatha, S. ;
Math, Shrishail .
INTERNATIONAL JOURNAL OF NONLINEAR ANALYSIS AND APPLICATIONS, 2022, 13 (01) :1225-1237
[27]   A novel passenger flow prediction model using deep learning methods [J].
Liu, Lijuan ;
Chen, Rung-Ching .
TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES, 2017, 84 :74-91
[28]   XcelNet14: A Novel Deep Learning Framework for Aerial Scene Classification [J].
Ahmed, Bilal ;
Akram, Tallha ;
Naqvi, Syed Rameez ;
Alsuhaibani, Anas ;
Attique Khan, Muhammad ;
Kraiem, Naoufel .
IEEE ACCESS, 2024, 12 :196266-196281
[29]   A Novel Deep Learning Framework on Brain Functional Networks for Early MCI Diagnosis [J].
Kam, Tae-Eui ;
Zhang, Han ;
Shen, Dinggang .
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION, PT III, 2018, 11072 :293-301
[30]   Intelligent fault diagnosis for triboelectric nanogenerators via a novel deep learning framework [J].
Wu, Hao ;
Xu, Xing'ang ;
Xin, Chuanfu ;
Liu, Yichen ;
Rao, Runze ;
Li, Zhongjie ;
Zhang, Dan ;
Wu, Yongxi ;
Han, Senzhe .
EXPERT SYSTEMS WITH APPLICATIONS, 2023, 226