Moisture flow measurement method fusing vortex street frequency and vortex street cross-correlation speed measurement

Abstract

The invention relates to a moisture flow measurement method fusing vortex street frequency and vortex street cross-correlation speed measurement, a measurement pipeline is a vertically downward pipeline, and a used sensor comprises a front piezoelectric probe and a rear piezoelectric probe which are located behind a vortex street generator; the vortex street convection velocity is obtained through cross-correlation calculation of signals of the upstream and downstream piezoelectric probes; performing fast Fourier transform based on time sequence data of the piezoelectric probe to obtain vortex street frequency; a fusion measurement model is established by using vortex street convection characteristics and vortex street frequency characteristics, and the moisture flow is solved based on Newton iteration.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of fluid flow measurement. Specifically, it relates to a method for measuring wet gas flow. BACKGROUND

[0002] In the aspect of measuring wet gas phase separation flow (gas phase and liquid phase) by using vortex flowmeter, the main problem to be solved is to correct the over-reading in two-phase vortex measurement. There are currently three methods: one is to directly measure the liquid content information and establish the relationship between over-reading and liquid content, so as to correct the over-reading. Traditional measurement methods include microwave method, ray method, equal-speed sampling method, ultrasonic method, optical method, etc. These measurement methods are limited by cost, measurement environment, complexity, etc., and are difficult to popularize and use. For example, the microwave method and the ray method have safety hazards, the equal-speed sampling method cannot realize online measurement, the ultrasonic method has a complex signal processing process, and the optical method generally requires transparent pipeline and low pressure, and the optical element has high cost. In addition, the liquid content in wet gas is small, and the liquid measurement accuracy is difficult to guarantee, which may cause over-correction and cause greater measurement error.

[0003] The second method is to correct the vortex over-reading by using vortex additional information (such as instantaneous ridge frequency fluctuation, vortex amplitude). For example, patent ZL201910469779.4 proposes a vortex wet gas over-reading compensation and flow measurement method combined with wavelet ridge frequency fluctuation, wherein the instantaneous frequency needs to be assisted by time-frequency processing method, which has high complexity and high requirement for hardware operation capacity. Patent ZL201910474178.2 proposes a wet gas over-reading compensation and flow measurement method combined with vortex amplitude characteristics, wherein the non-dimensionalization of signal amplitude involves 7 parameters, and the calibration process is complex. In addition, when the liquid content is high, the amplitude curve shows gradual saturation characteristics, at this time, the sensitivity of signal amplitude to liquid change will decrease. In the extreme case, the liquid content change no longer causes the amplitude change, at this time, the iterative process may not converge to the true value, and the algorithm will fail, so the applicable range is limited.

[0004] The third method is to use the conductivity-vortex street dual mode detection method, to establish a liquid film parameter model and a vortex street over-reading model, to solve the two equations together and combine an iterative algorithm to establish a wet gas phase separation measurement model. The patents CN202110129932.6 and CN202110129883.6 use liquid film thickness information for vortex street over-reading correction and phase separation flow measurement, and the modeling and prediction accuracy depends on the liquid film thickness measurement accuracy, and the calibration of the liquid film thickness is seriously dependent on the medium conductivity. Once the wet gas conductivity changes, the wet gas prediction accuracy will be greatly reduced. Therefore, the method based on liquid film thickness measurement is limited in application and has high requirements for medium conductivity. The patent 202310242114.6 proposes to use disturbance wave frequency information for vortex street over-reading correction, without the need for liquid film thickness calibration, greatly reducing the requirement for medium conductivity. However, the above-mentioned method based on conductivity-vortex street dual mode measurement increases the cost due to the use of conductivity sensors, and the collection of liquid film signals requires the use of high-speed AD-DA modules and FPGAs for excitation and signal collection, with high signal complexity and difficult to ensure real-time measurement. The patent CN202010300026.3 proposes a segmented wet gas flow measurement method combining vortex frequency and convection characteristics, wherein the determination method of the critical value is not explicitly given, only the calibration results under a typical carrier gas working condition (gas phase pressure and flow) are given, and it cannot be used for actual variable carrier gas working condition wet gas flow measurement.

[0005] The present application proposes a new method of wet gas flow measurement combining vortex frequency and vortex cross-correlation velocity measurement. The method uses a double-probe vortex cross-correlation velocity measurement system to accurately obtain the vortex convection velocity parameter and extract the vortex frequency parameter. By establishing a vortex convection coefficient model and a vortex over-reading factor model, the frequency characteristics and convection characteristics of the vortex flowmeter are effectively utilized, the measurement error caused by relying on a single characteristic is avoided, the measurement accuracy of the wet gas flow is greatly improved, the liquid phase flow measurement range is expanded, and the measurement robustness and applicability are better.

[0006] REFERENCES

[0007] [1] J. X. Li, C. Wang, H. B. Ding, Z. X, Zhang and H. J. Sun, “EMD and spectrum-centrobaric-correction-based analysis of vortex street characteristics in annular mist flow of wet gas”, IEEE Trans. Instrum. Meas., vol. 37, no. 5, pp. 1150-1160, May 2018. SUMMARY

[0008] This invention addresses the overreading problem of vortex flowmeters in moist two-phase flow by providing a method for measuring moist gas flow that integrates vortex frequency and vortex cross-correlation velocities, thereby improving the accuracy of vortex flowmeter measurements in moist two-phase flow. The technical solution is as follows:

[0009] A method for measuring moist gas flow rate by integrating vortex shedding frequency and vortex shedding cross-correlation velocity measurements is disclosed. The measurement pipe is a vertically downward pipe, and the sensors used include two piezoelectric probes located behind the vortex generator. The vortex shedding velocity is calculated by cross-correlation of the upstream and downstream piezoelectric probe signals. The vortex shedding frequency is obtained by performing a fast Fourier transform based on the time-series data from the piezoelectric probes. A fusion measurement model is established using the vortex shedding convection characteristics and vortex shedding frequency characteristics, and the moist gas flow rate is solved based on Newton's iteration. The steps are as follows:

[0010] 1) Collect the pressure P, temperature T, and the raw vortex street signals from the upstream and downstream piezoelectric probes, s1(t) and s2(t);

[0011] 2) Calculate the gas density ρ g Liquid density ρ l And the surface tension of the liquid phase σ; extract the vortex shedding frequency f vs and vortex street convection velocity U c ;where U c =L / τ0, where L is the distance between the two piezoelectric probe sensors, and τ0 is the transit time obtained by cross-correlation calculation using the vortex shear signals from the upstream and downstream piezoelectric probes;

[0012] 3) Calculate the initial value of gas phase flow rate Q g,apparent =3600f vs / K v , where: K v The instrument coefficient of the vortex flow meter is obtained through a dry gas calibration experiment;

[0013] 4) Combine the overread factor correlation equations OR = 1 + bφ p Correlation between K and vortex sheave convection coefficient c =c1We g +c2+c3φ p Constructing a gas phase volumetric flow rate Q g Functions with parameters and its derivative

[0014] in:

[0015] ·Vortex overread factor OR = Q g,apparent / Q g Q g,apparent Q represents the indicated gas volumetric flow rate in a moist two-phase flow, i.e., the uncorrected vortex flowmeter reading. g This represents the actual gas phase volumetric flow rate in a wet two-phase flow;

[0016] • vortex convection coefficient K c = U c / U, U is the average velocity of the flow;

[0017] • droplet loading φ p = m p / m g , m p represents the mass of droplets flowing through the cross section per unit time, m g represents the mass of gas flowing through the cross section per unit time;

[0018] • gas Weber number σ represents the surface tension coefficient of the liquid phase, ρ g represents the gas phase density, ρ l represents the liquid phase density, U sg is the apparent velocity of the gas phase, U sg = Q g / (0.25πD 2 ), D is the nominal diameter of the pipeline;

[0019] • constant coefficients b, c1, c2, c3 are obtained by least squares fitting;

[0020] • η = c1K / A 2 , Cross-sectional area A = πD 2 / 4;

[0021] • γ = (c2-c3 / b), ε = (c3Q g,apparent / b)-U c A;

[0022] 5) The uncorrected vortex flowmeter indication Q g,apparent is taken as the initial value of the iteration of the gas phase volumetric flow rate Q g , that is, Q g,0 = Q g,apparent = 3600f vs / K v , subscript n = 0 represents the initial value;

[0023] 6) The Newton iteration formula is brought in to iteratively solve, and the gas phase volumetric flow rate of the n+1 measurement is obtained:

[0024]

[0025] Where: Q g,n represents the last iteration result, Q g,n+1 represents the current iteration result;

[0026] 7) Set the convergence threshold, compare the gas phase volumetric flow rates obtained by the previous and next iterations, and judge whether the convergence condition is met, that is, |Qg,n+1 -Q g,n | / Q g,n Is it less than the convergence threshold? If the convergence condition is met, the iteration ends; if the convergence condition is not met, let n = n + 1 and jump to step 6) to continue iterative solution until convergence is achieved or the maximum number of iterations is reached.

[0027] 8) Output the volumetric flow rate Q of the gas phase in the humid gas. g .

[0028] Furthermore, the upstream and downstream piezoelectric probes are installed facing each other. Based on the estimated vortex shedding wavelength, the installation distance between the two piezoelectric probes is greater than one wavelength.

[0029] Furthermore, the vortex shedding velocity U c Estimation process:

[0030] The raw signals collected by the two piezoelectric probes upstream and downstream are preprocessed to obtain the preprocessed raw signals;

[0031] The vortex shear signal acquired by the downstream piezoelectric probe is reversed, and a cross-correlation function based on moving average filtering is introduced to perform moving average filtering on the preprocessed original signal. Then, cross-correlation calculation is performed to obtain the cross-correlation function.

[0032] Find the second peak point (k, ρ) in the cross-correlation function curve of the moving average filter. k ), at point (k,ρ k If a second interpolation is performed near the location of the transit point, the transit time is estimated as follows:

[0033]

[0034] Where, ρ k-1 ,ρ k and ρ k+1 f represents the cross-correlation values ​​corresponding to the (k-1), k, and k+1 delay points, respectively. s The sampling frequency.

[0035] Calculate the vortex sheave velocity U c =L / τ0, where L is the distance between the two piezoelectric probes.

[0036] Furthermore, the raw signals collected by the upstream and downstream piezoelectric probes are preprocessed to obtain the preprocessed raw signals. The method is as follows: remove outliers according to the Pauta criterion, eliminate outliers and retain the piezoelectric force fluctuation information to obtain the preprocessed raw signals.

[0037] The present invention has the following advantages due to the adoption of the above technical solutions:

[0038] (1) The wet gas flow measurement method fusing vortex street frequency and vortex street cross-correlation velocity measurement can effectively improve the measurement accuracy of wet gas two-phase flow, expand the liquid phase flow measurement range, and has better measurement robustness and applicability;

[0039] (2) The measurement method provided by the application applies Newton iteration method to solve the gas phase flow calculation equation, and the iteration converges quickly, which is beneficial to obtaining accurate gas phase flow results; BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 Vortex street piezoelectric double probe detection system

[0041] 1-pressure sensor 2-vortex flowmeter 3-vortex piezoelectric front probe 4-temperature sensor

[0042] 5-vortex street generator 6-vortex piezoelectric rear probe

[0043] Figure 2 Signal acquisition flowchart

[0044] Figure 3 Vortex over-reading modeling result

[0045] Figure 4 Vortex convection velocity estimation flowchart

[0046] Figure 5 Vortex convection coefficient and droplet loading amount diagram

[0047] Figure 6 Vortex convection coefficient modeling result based on droplet loading amount considering liquid phase influence

[0048] Figure 7 Wet gas flow measurement flowchart fusing vortex street frequency and vortex street cross-correlation velocity measurement

[0049] Figure 8 Wet gas gas phase volume flow prediction error distribution based on Weber number

[0050] Figure 9 Wet gas gas phase volume flow prediction error distribution based on droplet loading amount DETAILED DESCRIPTION

[0051] The application will be further described in combination with the drawings and implementation.

[0052] The application designs a double-piezoelectric-probe vortex cross-correlation measurement system, establishes a convection coefficient model based on vortex convection velocity measurement, establishes a vortex over-reading factor model based on vortex frequency, proposes a wet gas flow measurement method fusing vortex convection and frequency characteristics, and solves and calculates based on Newton iteration method.

[0053] This example demonstrates a specific implementation of a method for measuring moist gas flow rate that integrates vortex shedding frequency and vortex shedding cross-correlation velocimetry. The moist gas operating conditions are: pressure P = (100–400) kPa, apparent gas velocity U... sg = (12~38) m / s, the medium is air and water. The nominal inner diameter of the pipe is D = 0.015 m.

[0054] Measuring device as attached Figure 1 As shown, it mainly consists of pressure sensor 1, vortex flow meter 2, upstream piezoelectric probe 3, temperature sensor 4, downstream piezoelectric probe 6, and 5 is the vortex generator. The signal acquisition flowchart is attached. Figure 2 As shown: The system collects operating pressure P, operating temperature T, and vortex shedding signals s1(t) and s2(t). s1(t) and s2(t) are measured by a piezoelectric sensor. The probe converts the flow signal into an electrical signal. Hardware circuitry amplifies the original signal through charge and voltage amplification, followed by bandpass filtering (200–2500 Hz). Data is then acquired by an NI-USB data acquisition card and displayed and stored using LabVIEW software. The sensor sampling frequency is f. s =60kHz, sampling time is 7s.

[0055] A vortex flow meter is a velocity-type flow meter that measures the frequency f of vortex shedding. vs The volumetric flow rate can then be obtained. When a vortex flow meter is used for moisture measurement, the two-phase Strouhal number will be higher due to the presence of a liquid phase in the fluid, leading to an overestimation of the gas phase volume and causing instrument overreading. To correct for vortex overreading, a dimensionless overreading factor OR is introduced, denoted as:

[0056]

[0057] Among them, Q g,apparent The apparent gas volumetric flow rate is expressed as...

[0058]

[0059] Among them, K v The instrument coefficient of the vortex flowmeter in single-phase gas is obtained through calibration experiments in dry gas. In this example, the instrument coefficient is calibrated as K. v =3.2506×10 5 (m -3 ).

[0060] According to the conclusion of reference [1], the droplet loading amount φ p The main factor affecting the overread factor OR is the droplet mass loading parameter φ. p

[0061]

[0062] where m p and m LF represent the mass flow rate of droplets and liquid film, respectively. l and m g represent the mass flow rate of liquid and gas phase, respectively.

[0063] To accurately measure the gas phase flow rate in wet gas two-phase flow, the present patent proposes a wet gas flow rate measurement method that combines vortex shedding frequency and vortex shedding cross-correlation velocimetry. First, the over-reading factor OR and the vortex shedding convection coefficient K c are modeled, and then the fusion method based on the vortex shedding over-reading factor and the vortex shedding convection coefficient is modeled to establish a gas phase flow rate prediction model based on the Newton iteration algorithm.

[0064] The over-reading factor OR changes with the dimensionless parameter droplet loading φ p , as shown in the accompanying Figure 3 , which is modeled as

[0065] OR = 1 + bφ p #(4)

[0066] The coefficients in the model correlation expression are calibrated, and the constant coefficient b is fitted using the least squares method to obtain the coefficient b = 0.883, the determination coefficient R 2 = 0.882, and the relative root mean square error rRMSE = 0.42%. The relative error is within ±1.0% error range.

[0067] The flowchart of the vortex shedding convection velocity U c estimation process is shown in the accompanying Figure 4 . The upstream and downstream piezoelectric probe sensors in the accompanying Figure 1 are installed in opposite positions. The downstream signal of the sensor is taken as negative due to the introduction of an additional 180° phase lag, s1(t) = s1(t), s2(t) = -s2(t), and then a moving average filter is used to preprocess the signals collected by the two sensors. According to the Pauta criterion, data outliers can be effectively excluded while maintaining pressure dynamic wave information, i.e.

[0068]

[0069] where X i represents the original pressure sequence s1(t) or s2(t), and σ0 is the standard deviation of the sequence.

[0070] The cross-correlation function based on the moving average filter is introduced, and the moving average filter is used to preprocess the original signal

[0071]

[0072] where X is the pre-processed signal, i is the index of the signal sequence X(t), f s is the sampling frequency, and M is the length of the moving average filter, M=10 in this example.

[0073] In this example, the wavelength of the vortex street is estimated to be 12.25-13.65 mm, and the sensor installation distance is greater than one wavelength, so most of the maximum transit time should occur at the second peak of the cross-correlation function. The second peak point (k, p k ) is found in the moving average filtered cross-correlation function curve, and then a quadratic interpolation such as p k =a1x xy +a2x+a3 is performed near the point (k, p 2 ), and the transit time can be estimated as:

[0074]

[0075] where p k-1 , p k , and p k+1 are the cross-correlation values corresponding to the k-1, k, and k+1 delay points, respectively, and f s is the sampling frequency.

[0076] Finally, the vortex street convection velocity U c =L / τ0 can be obtained, where L is the distance between the two piezoelectric probes.

[0077] In order to obtain a more objective vortex street convection velocity U c , the analysis result is divided by the inlet (inlet) velocity U to obtain the dimensionless vortex street convection coefficient K c , that is

[0078]

[0079] The vortex street convection coefficient K c as a function of the droplet loading φ p is shown in the attached Figure 5 figure, and it can be seen that the vortex street convection coefficient K c increases with the increase of the droplet loading φ p , and the slope of the relationship curve is basically a constant under different working conditions, and the intercept changes with the working condition: the greater the gas volume flow rate and gas density, the greater the curve intercept. The gas volume flow rate and gas density parameters can be characterized by the dimensionless gas Weber number We g :

[0080]

[0081] where U sgρ represents the apparent velocity of the gas phase, σ represents the surface tension coefficient of the liquid phase, and ρ represents the apparent velocity of the gas phase. g ρ represents the gas phase density. l The value represents the liquid density, and D is the nominal diameter of the pipe.

[0082] Based on the above description and appendix Figure 5 The vortex shedding coefficient K is shown. c We in different gas phase Weber numbers g The amount of liquid added with the droplet φ p The change graph is modeled as

[0083] K c =c1We g +c2+c3φ p #(10)

[0084] The coefficients in the model correlation equation were calibrated, and the constant coefficients c1, c2, and c3 were fitted using the least squares method, yielding coefficient c1 = -3.747 × 10⁻⁶. -5 c2 = 0.80, c3 = 0.649, coefficient of determination R 2 = 0.9602. The error distribution of the vortex convection coefficient prediction considering the influence of the liquid phase is shown in the attached figure. Figure 6 As shown, there are no systematic errors as the droplet loading increases, and the prediction error is within ±1.0%, indicating good prediction performance.

[0085] A fusion moisture measurement model is established using the overreading factor correlation and vortex convection coefficient correlation described above. The overreading factor correlation and vortex convection coefficient correlation are as follows:

[0086]

[0087] For ease of calculation, let η = c1K / A 2 γ=(c2-c3 / b), ε=(c3Q) g,apparent / b)-U c A. Among them Cross-sectional area A = πD 2 / 4. Then establish the gas phase volumetric flow rate Q g The equation with parameters is

[0088]

[0089] Establish with Q g A function f(Q) with unknowns g )for

[0090]

[0091] Construct the Newton iteration format as follows

[0092]

[0093] f(Q g ) is obtained from equation (13) g The derivative f'(Q g ) of f(Q g ) with respect to Q g is

[0094]

[0095] Substitute equation (13) and equation (15) into equation (14), we get

[0096]

[0097] The uncorrected vortex shedding flowmeter indication Q g,apparent is taken as the initial value of iteration of the gas phase volumetric flow rate Q g , i.e. Q g,0 = Q g,apparent = 3600f vs / K v , subscript n=0 represents the initial value, and is substituted into the Newton iteration format for calculation, and the iteration calculation process is shown in the attached figure. Figure 7

[0098] According to the principle of Newton iteration algorithm, the Newton iteration format equation (16) established in the present example satisfies the following two conditions, i.e. equation (16) is convergent:

[0099] 1. The second order derivative of the iteration function f(Q g ) is not 0, i.e. f''(Q g )≠0;

[0100] 2. The initial value Q g,apparent is near Q g , i.e. OR=Q g,apparent / Q g <2;

[0101] In the process of iteration calculation, when the iteration stop condition is met:

[0102]

[0103] the iteration is considered to be convergent, and in order to avoid the special case of iteration not converging, the calculation will also be stopped when the iteration number reaches the maximum iteration number 20. After the iteration is completed, the volumetric flow rate Q g of the gas in the wet gas can be obtained.

[0104] According to the above modeling and calibration results, the wet gas flow measurement of fusion vortex street frequency and vortex mutual correlation velocity measurement is carried out, as follows:

[0105] 1) Collect pressure P, temperature T and vortex street signals s1(t), s2(t);

[0106] 2) Calculate the gas density ρ g Liquid density ρ l And the surface tension of the liquid phase σ; extract the vortex shedding frequency f vs and vortex street convection velocity U c ;

[0107] 3) Based on the vortex shedding frequency f vs The apparent gas volumetric flow rate Q without overread correction is calculated according to equation (2). g,apparent ;

[0108] 4) By combining the overreading factor correlation (4) and the vortex shedding coefficient correlation (10), a gas phase volumetric flow rate Q is constructed. g The function (13) with parameters and its derivative (15);

[0109] 5) The uncorrected vortex flowmeter reading Q g,apparent As the gas phase volumetric flow rate Q g The initial value for iteration, i.e., Q g,0 =Q g,apparent The subscript n = 0 represents the initial value;

[0110] 6) Substitute into Newton's iteration formula (14) and iterate to obtain the gas phase volumetric flow rate (16) measured for the (n+1)th time;

[0111] 7) Set a convergence threshold δ, compare the gas phase volumetric flow rates obtained from the two iterations, and determine whether the convergence condition is met according to equation (17), i.e., |Q g,n+1 -Q g,n | / Q g,n Is it less than the convergence threshold? If the convergence condition is met, the iteration ends and the next operation is performed. If the convergence condition is not met, let n = n + 1 and determine whether the maximum number of iterations 20 has been reached. If not, jump to step 6) to continue iterative solution until convergence is achieved or the maximum number of iterations 20 is reached.

[0112] 8) Output the volumetric flow rate Q of the gas phase in the humid gas. g .

[0113] In this example, the distribution of prediction errors for gas phase volumetric flow rate under different carrier gas pressures and flow conditions is shown in the attached figure. Figure 8 Appendix Figure 9 As shown, where We g For the gas phase Weber number, φ pFor droplet loading, the error PE(%)=(predicted value-real value) / real value*100. In the vapor measurement, the iteration results based on the fusion method are that the vapor phase prediction error is 95.65% within ±3.0%, the mean absolute error MPE=0.122%, the mean absolute percentage error MAPE=1.239%, and the uncertainty=1.15%. There is no obvious systematic error with the vapor Weber number and the liquid phase content, and theoretically, the liquid phase flow measurement range can be expanded, and the measurement robustness and applicability are better.

Claims

1. A method for measuring moist gas flow rate by integrating vortex frequency and vortex cross-correlation velocity measurements, wherein the measurement pipe is a vertically downward pipe, and the sensors used include two piezoelectric probes located behind the vortex generator; the vortex convection velocity is calculated by cross-correlation of the upstream and downstream piezoelectric probe signals; the vortex frequency is obtained by performing a fast Fourier transform based on the time-series data of the piezoelectric probes; a fusion measurement model is established using the vortex convection characteristics and vortex frequency characteristics, and the moist gas flow rate is solved based on Newton's iteration, the steps of which are as follows: 1) Collect the pressure P, temperature T, and the raw vortex street signals from the upstream and downstream piezoelectric probes, s1(t) and s2(t); 2) Calculate the gas density ρ g Liquid density ρ l And the surface tension of the liquid phase σ; extract the vortex shedding frequency f vs and vortex street convection velocity U c ;where U c =L / τ0, where L is the distance between the two piezoelectric probe sensors, and τ0 is the transit time obtained by cross-correlation calculation using the vortex shear signals from the upstream and downstream piezoelectric probes; 3) Calculate the initial value of gas phase flow rate Q g,apparent =3600f vs / K v , where: K v The instrument coefficient of the vortex flow meter is obtained through a dry gas calibration experiment; 4) Combine the overread factor correlation equations OR = 1 + bφ p Correlation between K and vortex sheave convection coefficient c =c1We g +c2+c3φ p Constructing a gas phase volumetric flow rate Q g Functions with parameters and its derivative in: Vortex overread factor OR = Q g,apparent / Q g Q g,apparent Q represents the indicated gas volumetric flow rate in a moist two-phase flow, i.e., the uncorrected vortex flowmeter reading. g This represents the actual gas phase volumetric flow rate in a wet two-phase flow; Vortex convection coefficient K c =U c / U, where U is the average velocity of the incoming flow; Droplet loading amount φ p =m p / m g m p The mass of a droplet flowing through a cross-section per unit time, m g It represents the mass of gas phase flowing through the cross-section per unit time; Vascular Weber Numbers σ represents the surface tension coefficient of the liquid phase, ρ g ρ represents the gas phase density. l U represents the density of the liquid phase. sg U is the apparent velocity in the gas phase. sg =Q g / (0.25πD 2 D is the nominal diameter of the pipe; The constant coefficients b, c1, c2, and c3 were obtained by least squares fitting. Cross-sectional area A = πD 2 / 4; γ=(c2-c3 / b),ε=(c3Q g,apparent / b)-U c A; 5) The uncorrected vortex flowmeter reading Q g,apparent As the gas phase volumetric flow rate Q g The initial value for iteration, i.e., Q g,0 =Q g,apparent =3600f vs / K v The subscript n = 0 represents the initial value; 6) Substitute the values ​​into Newton's iteration formula and solve iteratively to obtain the gas phase volumetric flow rate measured in the (n+1)th time: Among them: Q g,n Q represents the result of the previous iteration. g,n+1 This represents the result of this iteration; 7) Set a convergence threshold, compare the gas phase volumetric flow rates obtained from the two iterations, and determine whether the convergence condition, i.e., |Q|, is met. g,n+1 -Q g,n | / Q g,n Is it less than the convergence threshold? If the convergence condition is met, the iteration ends; if the convergence condition is not met, let n = n + 1 and jump to step 6) to continue iterative solution until convergence is achieved or the maximum number of iterations is reached. 8) Output the volumetric flow rate Q of the gas phase in the humid gas. g .

2. The method for measuring moist gas flow rate by integrating vortex street frequency and vortex street cross-correlation velocimetry according to claim 1, characterized in that, The upstream and downstream piezoelectric probes are installed facing each other. Based on the estimated vortex shedding wavelength, the installation distance between the two piezoelectric probes is greater than one wavelength.

3. The method for measuring moist gas flow rate by integrating vortex street frequency and vortex street cross-correlation velocimetry according to claim 2, characterized in that, Vortex convection velocity U c Estimation process: The raw signals collected by the two piezoelectric probes upstream and downstream are preprocessed to obtain the preprocessed raw signals; The vortex shear signal acquired by the downstream piezoelectric probe is reversed, and a cross-correlation function based on moving average filtering is introduced to perform moving average filtering on the preprocessed original signal. Then, cross-correlation calculation is performed to obtain the cross-correlation function. Find the second peak point (k, ρ) in the cross-correlation function curve of the moving average filter. k ), at point (k,ρ k If a second interpolation is performed near the location of the transit point, the transit time is estimated as follows: Where, ρ k-1 ,ρ k and ρ k+1 f represents the cross-correlation values ​​corresponding to the (k-1), k, and k+1 delay points, respectively. s The sampling frequency; Calculate the vortex sheave velocity U c =L / τ0, where L is the distance between the two piezoelectric probes.

4. The method for measuring moist gas flow rate by integrating vortex street frequency and vortex street cross-correlation velocities according to claim 3, characterized in that, The method for preprocessing the raw signals collected by the upstream and downstream piezoelectric probes to obtain the preprocessed raw signals is as follows: remove outliers according to the Pauta criterion, eliminate outliers and retain the piezoelectric force fluctuation information to obtain the preprocessed raw signals.

Images