Font Size: a A A

Research On Correlation Flow Measurement Method Based On Tomography

Posted on:2020-08-22Degree:MasterType:Thesis
Country:ChinaCandidate:H GuFull Text:PDF
GTID:2370330578966669Subject:Engineering
Abstract/Summary:PDF Full Text Request
Two-phase flow measurement is common in industrial production,but due to its own complexity and particularity,it is difficult to obtain high measurement accuracy for two-phase flow parameter detection.Two-phase flow measurement includes fluid flow pattern identification and flow parameter detection,and at present,two-phase flow measurement methods are various,and tomography has a wide range of applications in two-phase flow measurement.Tomography has a high flow pattern recognition rate when identifying fluid flow patterns,but it is difficult to obtain accurate measurement results of flow parameters such as fluid velocity of flow and flow rate.Based on the tomography technology(ET)combined with the traditional related flow measurement methods,the accurate measurement of the flow parameters such as phase fraction,flow rate and liquid phase flow of gas-liquid two-phase flow is realized.This paper focuses on the working principle and implementation methods of electrical resistance tomography(ERT)and related flow measurement methods.The image of gas and liquid two-phase flow tomography of A and B cross-section is obtained by ERT imaging system,and the characteristic parameters such as fluid flow pattern and gas-liquid two-phase phase content change and distribution are detected.The gas phase flow rate in the two-phase flow is calculated based on the frequency domain Fourier transform(FFT)combined with the relevant flow measurement principle,and the gas phase flow rate is obtained by combining the phase content ratio;The three-dimensional decomposition of the tomographic image is performed based on the Mallat two-dimensional discrete wavelet to extract the texture features of the image;Based on the gas flow rate and extracted image texture features as input and liquid phase flow as output,a liquid phase flow measurement model based on BP neural network is established.The liquid phase flow measurement model is successfully constructed through 200 sets of sample training and testing,and its performance is analyzed.Through actual experiments,the measurement accuracy of the relevant flow measurement method based on tomography can reach less than 4%,which basically meets the industrial measurement requirements,and proves the feasibility and effectiveness of the flow measurement method.
Keywords/Search Tags:ERT, Flow pattern recognition, Wavelet transform, Texture feature, Related flow measurement, Flow measurement model
PDF Full Text Request
Related items