scholarly journals A novel complex network-based deep learning method for characterizing gas–liquid two-phase flow

2020 ◽  
Author(s):  
Zhong-Ke Gao ◽  
Ming-Xu Liu ◽  
Wei-Dong Dang ◽  
Qing Cai
Keyword(s):  
Deep Learning ◽  
Complex Network ◽  
Two Phase Flow ◽  
Phase Flow ◽  
Learning Method ◽  
Two Phase ◽  
Novel Complex

scholarly journals Prediction of two-phase flow patterns in upward inclined pipes via deep learning

Energy ◽  
2020 ◽  
Vol 210 ◽  
pp. 118541 ◽  
Author(s):  
Zi Lin ◽  
Xiaolei Liu ◽  
Liyun Lao ◽  
Hengxu Liu
Keyword(s):  
Deep Learning ◽  
Two Phase Flow ◽  
Flow Patterns ◽  
Phase Flow ◽  
Two Phase ◽  
Inclined Pipes

Complex network analysis in inclined oil–water two-phase flow

2009 ◽  
Vol 18 (12) ◽  
pp. 5249-5258 ◽  
Author(s):  
Gao Zhong-Ke ◽  
Jin Ning-De
Keyword(s):  
Network Analysis ◽  
Complex Network ◽  
Two Phase Flow ◽  
Phase Flow ◽  
Two Phase ◽  
Complex Network Analysis ◽  
Oil Water

CHARACTERIZATION OF HORIZONTAL GAS–LIQUID TWO-PHASE FLOW USING MARKOV MODEL-BASED COMPLEX NETWORK

2013 ◽  
Vol 24 (05) ◽  
pp. 1350028 ◽  
Author(s):  
LI-DAN HU ◽  
NING-DE JIN ◽  
ZHONG-KE GAO
Keyword(s):  
Complex Networks ◽  
Complex Network ◽  
Chemical Engineering ◽  
Two Phase Flow ◽  
Transition Probability ◽  
Flow Patterns ◽  
Phase Flow ◽  
Two Phase ◽  
Markov Transition ◽  
Liquid Flows

Horizontal gas–liquid two-phase flow widely exists in many physical systems and chemical engineering processes. Compared with vertical upward gas–liquid two-phase flow, investigations on dynamic behavior underlying horizontal gas–liquid flows are quite limited. Complex network provides a powerful framework for time series analysis of complex dynamical systems. We use a network generation method based on Markov transition probability to infer directed weighted complex networks from signals measured from horizontal gas–liquid two-phase flow experiment and find that the networks corresponding to different flow patterns exhibit different network structure. To investigate the dynamic characteristics of horizontal gas–liquid flows, we construct a number of complex networks under different flow conditions, and explore the network indices for each constructed network. In addition, we investigate the sample entropy of different flow patterns. Our results suggest that the network statistic can well represent the complexity in the transition among different flow patterns and further allows characterizing the interface fluctuation behavior in horizontal gas–liquid two-phase flow.

Sign in / Sign up

Export Citation Format

Share Document