Characteristic Analysis Of Gas-liquid Two-phase Flow Patterns And Implementation Of Software System

Two-phase and multiphase flow is generally referred to as "refractory fluid".Now whether at home or abroad,there is no perfect method of the two-phase flow pattern characteristics.But because it has an extremely important role in industrial production,so in recent years it has become the subject which the relevant areas of scholars would actively explore.In this paper,the flow characteristics of gas-liquid two-phase flow are analyzed by the method of multi-scale energy characteristics.And equiprobable symbolization sample entropy(ESSE)and fuzzy entropy(FuzzyEn)are introduced into the two-phase flow signal analysis field,providing some ways of thinking for the dynamic characterization of the flow.In order to overcome the disadvantages of nonlinear and non-stationary characteristics of gas-liquid two-phase flow pattern signals,more deeply reveal the characteristics of two-phase flow pattern signals.We use the dual tree complex wavelet decomposition method,and the gas-liquid two-phase flow conductance fluctuation signals are decomposed into 7 scales,further analysize the two-phase flow characteristics of different frequency band signals.We extract root mean square(RMS)value of each scale and multi-scale RMS curve linear fitting slope.The RMS values and multi-scale RMS curve linear fitting slope are taken as characteristic parameters.In order to make it more rational and scientific,the method of multi-scale energy feature is used to identify the flow pattern of two-phase flow,and the recognition effect is achieved.Basing on the theory of ESSE and FuzzyEn,we studied the dynamic characteristics of different flow patterns.Based on the analysis of typical chaotic signal and noise signal,the ability of ESSE method to distinguish different types of chaotic signals and the ability to resist noise are verified.The FuzzyEn algorithm isused to compare the results of the two phase flow signal analysis with the traditional sample entropy(SE)and approximate entropy(ApEn)analysis.Finally,using LabVIEW and MATLAB mixed programming,we design and implement a LabVIEW software system which can complete the data acquisition,filtering and analysis.