| Gas-liquid two-phase widely exists in petroleum industry,chemical industry,metallurgical industry,power industry,energy industry and so on.Two-phase flow parameters measurement is very important to the national economy and production safety.Flow pattern,gas fraction and flowrate are important parameters in gas-liquid two-phase.Flow pattern not only affects the flow characteristics,but also affects the accurate measurement of other parameters in gas-liquid two-phase flow.Gas fraction can provide the basis for calculating the average density of the two-phase flow mixture,measuring the flowrate and the pressure gradient,and analyzing the flow condition in pipe.Flowrate is the precondition to ensure the balance of energy and material.The accurate measurement of flowrate affects the safe operation of equipment.The flow characteristics of gas-liquid two-phase flow are complicated,which makes it difficult to measure the parameters.According to previous experience,the differential pressure signal of gas-liquid two-phase flow contains abundant fluid flow information.The dissertation researches the flow pattern recognition and the measurement methods of gas fraction and flowrate based on the differential pressure fluctuation signal of Venturi by means of Extreme-point Symmetric Mode Decomposition(ESMD)and fuzzy pattern recognition.The main work of this dissertation includes:(1)ESMD is introduced into the trend term extraction,denoising and pseudo component analysis of the gas-liquid two-phase flow differential pressure signals,and the results are compared with the results of Hilbert Huang Transform(HHT).In view of the shortcomings of HHT,the trend term is too rough and the defect of time-frequency analysis,ESMD makes up for these defects.Because the noise component of the differential pressure signal has the characteristics of high frequency and low energy,the frequency and energy analysis of the Intrinsic Mode Function(IMF)is carried out to denoise.The correlation coefficient method is used to determine whether the low-frequency IMF is a pseudo component.The experimental results show that the IMF1 of differential pressure signal decomposed by ESMD is the noise component,which meets the characteristics of high frequency and low energy.Compared with HHT,the number of pseudo components decomposed by ESMD is less,and the influence of the pseudo components energy on the total signal energy is also less.(2)The flow pattern characteristic parameters are extracted and the fuzzy pattern recognition is used to identify the flow pattern.The relative variance,skewness and kurtosis are extracted as the flow pattern characteristic parameters since there are differences between the time-domain characteristics of differential pressure signals under different flow pattern.In addition,the differential pressure signal can be considered as the signal source.According to the information theory,the Shannon entropy of the differential pressure signal is further selected as the characteristic parameter.The experimental results show that these four characteristic parameters can effectively identify different flow patterns.(3)The fluctuation coefficient of differential pressure signal is extracted and the correlation equations between the fluctuation coefficient and gas volume fraction are established.According to the results of ESMD of differential pressure signal,the trend term and fluctuation component are extracted.It is found that the trend term is mainly affected by the water flowrate,and when the water flowrate is constant,the fluctuation component is affected by the air flowrate.On this basis,the fluctuation coefficient related to the gas volume fraction is proposed.According to the acceleration pressure drop theory of the gas-liquid two-phase flow,the relationship between the fluctuation coefficient and gas volume fraction is established,and the effects of the flow pattern,flowrate and density on the relationship are analyzed.The modified fluctuation coefficient is obtained by eliminating the effects of these factors,and the correlation equations of the modified fluctuation coefficient and the gas volume fraction are established under bubble flow,plug flow and slug flow respectively.The experimental results show that the proposed method can effectively estimate the value of gas volume fraction,and the root mean square error(RMSE)is less than 4%.(4)In the study of the gas-liquid two-phase flowrate measurement,the flowrate measurement accuracy of the homogeneous flow model,the James model,the separated flow model,the Murdock model and the lin-zonghu model are analyzed and compared.The appropriate flow measurement models are modified according to different flow patterns and a new flow measurement method is put forward.The experimental results show that James model is suitable for the bubble flow and the RMSE is 16.5%.The Murdock model can give a good measurement result for the plug flow and the slug flow and the RMSEs are 22.67% and 12.09% respectively.The James model under bubble flow and the Murdock model under plug flow and slug flow are modified,and the RMSEs are both less than 5%.In addition,the mass quality can be obtained by the estimated gas volume fraction,and the combination of mass quality and the total mass flowrate gives the split mass flowrate.The experimental results show that the error of the liquid mass flowrate is less affected by the mass quality,and the measurement results of the liquid mass flowrate are similar to those of the total mass flowrate,and the root mean square error of the three flow patterns is less than 5%.The gas mass flowrate is greatly affected by the mass quality,and the root mean square error is larger than that of the total mass flowrate,and the root mean square error of the three flow patterns is between 5% and 5.5%. |
PDF Full Text Request