Study On The Identification And Characterization Of Gas-liquid Two-phase Flow Patterns

Gas-liquid two-phase flow widely exists in various industrial processes andapplications. Parameters like heat and mass transfer rates, momentum loss, andpressure gradient greatly depend on two-phase flow patterns and their dynamicalbehaviors. In this study, the classifier method based on multi-scale signal processingwas employed to identify different gas-liquid two-phase flow patterns, while entropymeasure methods which can be used to uncover the complexity of systems werecombined with multi-scale analysis to characterize flow dynamics of gas-liquidtwo-phase flows.In view of the non-linearity and non-stability of two-phase flow pattern signals,an adaptive analysis method called ensemble empirical mode decomposition (EEMD)was applied to process flow pattern signals. Thus, feature vectors made from intrinsicmode function components on different scales could be obtained as input features ofthe classifier. In this study, feature parameters of typical flow patterns were applied torespectively train the radial basis function (RBF) neural network and continuoushidden Markov models (CHMMs). Finally, flow pattern identification of differentgas-liquid two-phase flow was achieved.Information entropy, permutation entropy and statistical complexity measuremethod were introduced to study the dynamics of two-phase flow patterns.(1)Information entropy distribution of IMF components on different scales was computed.The distributions of entropy values in different flow patterns greatly differ with eachother, and present certain evolutionary characteristics.(2) In view of the fact thatnonlinear parameters on single scale have shortcomings in the reflection of flowdynamics, the permutation entropy method with simple realization and robustness was combined with multi-scale analysis to study the multi-scale entropy and dynamics offlow pattern signals.(3) The statistical complexity measure based on complexityentropy causality plane (CECP) was employed in the study of dynamical complexity ofgas-liquid two-phase flow patterns. The results indicated that multi-scale CECP can beavailable in revealing the nonlinear dynamics of intrinsic flow structure in two-phaseflows.