| Coriolis flowmeters have very high mass flow measurement accuracy,and can provide multi-parameter measurement,such as density,temperature,volume flow rate,etc.,which are widely used in various industries.The main problem faced by Coriolis flowmeters is complex flow-phase conditions,such as gas-liquid two-phase flow.The interaction of gas-liquid twophase leads to complex flow patterns and volatility,resulting in a decline in the measurement performance of Coriolis flowmeters and a sharp increase in mass flow errors.The measurement error is complex and non-linear,which means modeling is difficult.Especially for gas-liquid two-phase flow conditions with a Gas Volume Fraction(GVF)of 0-30%,the model is difficult to have universality and low accuracy.When a Coriolis flowmeter is used in two-phase flow measurement,the differential pressure provides information of the GVF and flow patterns,which can effectively reduce mass flow measurement error.However,the external differential pressure transmitter brings the problems of inconvenience to installation and space limit.In order to simplify the measurement method,an external differential pressure transmitter can be integrated to the measuring tube,or the signals from the Coriolis flowmeter itself can be studied.Using a combination of simulation,experiment,and machine learning,following research work was carried out in the thesis for two-phase flow measurement model establishment and integrated measurement.(1)In order to integrate the differential pressure transmitter to the measuring tube,the principle and simulation method for selecting the position of the pressure ports were proposed.The feasibility of the simulation method were verified by experiments.The pressure uniformity at different tube sections and the effect of flow rate on the pressure uniformity studied for different pressure ports.The best pressure port position was chosen.(2)In order to establish a two-phase flow measurement model,a DN10 Coriolis flowmeter with twin V shaped bent tube was tested under gas-liquid two-phase flow conditions with liquid phase flow of 200kg/h-880kg/h and GVF of 0-30%.Experiments provide 107 data sets under different two-phase flow conditions.The reasons for the difference of measurement errors from a same GVF at different flow rates were explained by two-phase flow simulation using Fluent software.The relationships between different parameters,the GVF,and liquid flow rate of the twophase flow were analyzed.It was found that the trend of the differential pressure variation is roughly linear with the increase of the gas flow rate as the water flow rate is constant.The linear relationship between the GVF and the flow charge coefficient is fitted for eliminate the influence of the pressure.(3)Research on the signal output by the Coriolis flowmeter itself,mainly analyze the dynamic signals of differential pressure and mass flow in the time domain and frequency domain with obvious regularity with GVF.It was found that the trend of the relative fluctuation of differential pressure is similar to that of the mass flow error.(4)A two-phase flow measurement model based on PSO-SVM(Particle Swarm Support Vector Machine)for mass flow error correction and GVF prediction was designed.The relative error of the model for mass flow correction is within ± 1.5%,and the absolute error of the GVF prediction is within ± 3%.Moreover,a measurement model based on the information from Coriolis flowmeter was designed.The relative error of mass flow is within ± 2%,and the absolute error of the GVF prediction is within ± 3%. |