Modality transition-based network from multivariate time series for characterizing horizontal oil–water flow patterns

Abstract

The simultaneous flow of oil and water through a horizontal pipe is a common occurrence during petroleum industrial processes. Characterizing the flow behavior underlying horizontal oil–water flows is a challenging problem of significant importance. In order to solve this problem, we carry out experiment to measure multivariate signals from different flow patterns and then propose a novel modality transition-based network to analyze the multivariate signals. The results suggest that the local betweenness centrality and weighted shortest path of the constructed network can characterize the transitions of flow conditions and further allow quantitatively distinguishing and uncovering the dynamic flow behavior underlying different horizontal oil–water flow patterns.