Multilayer Network from Multivariate Time Series for Characterizing Nonlinear Flow Behavior.

Zhong-Ke Gao,Shan-Shan Zhang,Wei-Dong Dang,Shan Li,Qing Cai
DOI: https://doi.org/10.1142/s0218127417500596
IF: 2.45
2017-01-01
International Journal of Bifurcation and Chaos
Abstract:The exploration of two-phase flows, as a multidisciplinary subject, has drawn a great deal of attention on account of its significance. The dynamical flow behaviors underlying the transitions of oil–water bubbly flows are still elusive. We carry out oil–water two-phase flow experiments and capture multichannel flow information. Then we propose a novel methodology for inferring multilayer network from multivariate time series, which enables to fuse multichannel flow information at different frequency bands. We employ macro-scale, meso-scale and micro-scale network measures to characterize the generated multilayer networks, and the results suggest that our analysis allows uncovering the nonlinear flow behaviors underlying the transitions of oil-in-water bubbly flows.
What problem does this paper attempt to address?