Highly dynamic gas-liquid two-phase flow measurement method based on correlation coefficient constrained cross-correlation

Abstract

The present invention provides a high-dynamic gas-liquid two-phase flow measurement method based on correlation coefficient-constrained cross-correlation. The method includes: generating a simulation sample based on actual conditions, and screening the measurement capacitance in combination with a selection method based on contribution analysis; installing a dual-plane ECT sensor and a temperature and pressure sensor in a pipeline to measure the measurement capacitance and temperature and pressure data; using the temperature and pressure data to compensate for the screening and normalize the measurement capacitance, perform visualization, and obtain the liquid holdup; using an adaptive cross-correlation algorithm with correlation coefficient constraints combined with the dual-layer ECT sensor data to obtain the transit time and flow rate; using the temperature and pressure data to compensate for the gas-liquid two-phase density, and combining the two-phase density, liquid holdup, and flow rate to ultimately obtain the mass flow rate. The present invention provides a high-dynamic gas-liquid two-phase flow measurement method based on correlation coefficient-constrained cross-correlation, which can obtain better imaging and more accurate measurement of liquid holdup, and realize dynamic real-time measurement of the mass flow rate of high-dynamic gas-liquid two-phase flow.

Description

Technical Field

[0001] The present invention relates to the field of electromagnetic measurement, and in particular to a high-dynamic gas-liquid two-phase flow measurement method based on correlation coefficient-constrained cross-correlation. Background Art

[0002] In recent years, the demand for real-time dynamic measurement of highly dynamic multiphase fluids has been growing in fields such as scientific research, biomedicine, energy and power, and environmental monitoring. For example, during the coolant injection process for laser surgery cooling protection, the refrigerant injection rate from a small nozzle must be monitored in real time to achieve precise control and prevent safety accidents. Similarly, in liquefied gas propulsion systems, the propellant flow rate within a small pipeline must be monitored in real time to achieve efficient, reliable control, and rational planning. Therefore, a method for real-time measurement of the mass flow rate of highly dynamic gas-liquid two-phase fluids is urgently needed. Mass flow can be measured using a single-phase flowmeter, a direct mass flowmeter, or a method that combines flow rate and liquid holdup. Single-phase flowmeters cannot function properly when the two-phase flow fluctuates significantly. Direct mass flowmeters are complex, bulky, and expensive. Therefore, a method that combines flow rate and liquid holdup is preferred. Highly dynamic gas-liquid two-phase fluids often exhibit rapid flow pattern changes, necessitating a flow pattern-insensitive measurement method. Electrical capacitance tomography (ECT) reconstructs the dielectric constant distribution by measuring boundary capacitance, obtaining the two-phase fluid distribution and, in turn, the liquid holdup necessary for flow measurement, with minimal impact from flow pattern. Using a dual-plane ECT sensor, cross-correlation is then used to measure the flow velocity, yielding the mass flow rate. However, achieving such measurements faces the following difficulties: Highly dynamic gas-liquid two-phase flows experience rapid dynamic changes in shape, which do not fully conform to the frozen flow assumption. This results in poor correlation between the measurement signals between upstream and downstream ECT sensors, making it difficult to stably obtain transit times. Highly dynamic gas-liquid two-phase flows require high flow channel strength, often resulting in thicker pipe walls, which produce capacitance that contributes little to image reconstruction, severely degrading ECT image quality and, consequently, reducing the accuracy of subsequent liquid holdup measurements. Highly dynamic gas-liquid two-phase flows are often accompanied by dramatic temperature and pressure fluctuations, which can also cause changes in the dielectric constant of the liquid propellant, and thus the measured capacitance. Therefore, it is essential to design a high-dynamic gas-liquid two-phase flow measurement method based on correlation coefficient-constrained cross-correlation. Summary of the Invention

[0003] The purpose of the present invention is to provide a high-dynamic gas-liquid two-phase flow measurement method based on correlation coefficient constrained cross-correlation, which can obtain better imaging and more accurate measurement of liquid holdup, offset the adverse effects of temperature and pressure changes on the measurement results, and realize dynamic real-time measurement of high-dynamic gas-liquid two-phase flow mass flow.

[0004] To achieve the above object, the present invention provides the following solutions:

[0005] A high-dynamic gas-liquid two-phase flow measurement method based on correlation coefficient constrained cross-correlation includes the following steps:

[0006] Step 1: Generate simulation samples based on actual application conditions and use a selection method based on contribution analysis to screen the measured capacitance.

[0007] Step 2: Install a dual-plane ECT sensor and a temperature and pressure sensor on the pipeline near the tank to measure capacitance and temperature and pressure data;

[0008] Step 3: Use temperature and pressure data to compensate for and filter the normalized measured capacitance, visualize it, and obtain the liquid holdup;

[0009] Step 4: Use the adaptive cross-correlation algorithm with correlation coefficient constraint to combine the double-layer ECT sensing data to obtain the transit time and then the flow velocity;

[0010] Step 5: Use temperature and pressure data to compensate for the gas-liquid two-phase density, and finally obtain the mass flow rate by combining the two-phase density, liquid holdup, and flow rate.

[0011] Optionally, based on actual application conditions, simulated samples are generated and combined with a selection method based on contribution analysis to screen out the measured capacitance with the main contribution, specifically:

[0012] Setting the holdup β as a function of the normalized capacitance λ yields:

[0013] β=f β (λ)

[0014] Considering the first-order components and their interactions, the above formula can be rewritten as:

[0015]

[0016] Where k is the coefficient, M is the number of measured capacitors, and λ(i) is the normalized capacitance measurement. The coefficient k is obtained by regression combined with simulated samples. The normalized contribution E is calculated as follows:

[0017]

[0018] The normalized contribution E of each component is arranged in descending order, and the cumulative contribution is calculated. When the cumulative contribution reaches 50% of the total, the capacitance associated with the component involved in the cumulative contribution is selected for imaging.

[0019] Optionally, a dual-plane ECT sensor and a temperature and pressure sensor can be installed near the tank in the pipeline to measure capacitance and temperature and pressure data. Specifically:

[0020] The dual-plane ECT sensor is arranged near the tank port or in the tank, the temperature and pressure sensor is arranged adjacent to the dual-plane ECT sensor, and the dual-plane ECT sensor includes an upstream ECT sensor and a downstream ECT sensor with a distance L therebetween.

[0021] Optionally, the measured capacitance can be normalized after compensation and screening using temperature and pressure data, and visualized to obtain the liquid holdup, specifically:

[0022] The capacitance measured in a biplanar ECT sensor is represented by two wall capacitances C w1 、C w2 and internal capacitor C X The series connection of C w1 and C w2 Considered as a single capacitor C w , get the measured capacitance C M :

[0023]

[0024] When the tube is filled with a dielectric constant of ε ref When the medium is a homogeneous medium, the measured capacitance can be expressed as C X =ε ref C0, where C0 is the capacitance inside the pipe wall in the empty field, in addition to the measured empty calibration capacitance C L In addition, another dielectric constant ε ref The uniform reference medium is filled in the pipe, and the mutual capacitance between the electrodes is measured to obtain C ref , then C W and C0 can be calculated as:

[0025]

[0026] where ε L is the dielectric constant of the empty field, which is generally taken as 1, and the full calibration capacitance C is given by the following formula H :

[0027]

[0028] Among them, ε liquid It can be obtained by combining the measured dielectric constant data of the liquid phase with the following fitting formula:

[0029]

[0030] Where P and T are the temperature and pressure of the measured capacitor, respectively. b0, b1, b2, and b3 are the coefficients to be fitted. During measurement, the full-scale capacitance of the filtered capacitance measurement is updated in real time based on the temperature and pressure data. This means that parallel normalization is used in real time to obtain the compensated normalized measured capacitance:

[0031]

[0032] The reconstructed image can be obtained using the linear back projection algorithm:

[0033]

[0034] Where S is the sensitive field matrix, which is obtained based on the simulation model. After obtaining the reconstructed image, the liquid holdup can be obtained:

[0035]

[0036] Where K is the number of pixels in the reconstructed image, A i and A are the area of ​​the ith pixel and the cross-sectional area of ​​the pipe, respectively.

[0037] Optionally, an adaptive cross-correlation algorithm with correlation coefficient constraints is used in combination with the double-layer ECT sensing data to obtain the transit time and thus the flow velocity, specifically:

[0038] Get the velocity-dependent transition time:

[0039] The traditional cross-correlation algorithm formula is as follows:

[0040]

[0041] Where x(t) and y(t) are the collected upstream and downstream ECT signals, t0 is the start time of sampling, and t s is the sampling time. At the same time, the normalized cross-correlation result is often used for speed measurement. The normalization formula is as follows:

[0042]

[0043] R xy,coeff(τ) The peak value is recorded as corr xy , occurs when the similarity between the upstream ECT signal and the downstream ECT signal reaches its maximum value, and the corresponding time τ0 represents the conversion time of the flow propagating from the upstream ECT sensor to the downstream ECT sensor:

[0044] corr xy =max[R xy,coeff (τ)]

[0045] R xy,coeff (τ0)=max[R xy,coeff (τ)]

[0046] In order to reduce the poor correlation between upstream and downstream signals, an adaptive cross-correlation algorithm is used, which is expressed as:

[0047]

[0048] τ c =τ i +τ r , stR xy (τ r )=max[R xy (τ)]

[0049] Among them, τ i is the estimated initial flight time, τ r is the remaining transit time, τ c is the corrected flight time, τ0 is obtained by conventional cross-correlation technique during initialization, and then the corrected flight time τ is used to obtain the corrected flight time τ c To update,

[0050] τ i =τ c

[0051] By fine-tuning τ i Get the optimal correlation, take the peak point near the origin to get the remaining transition time, when the correlation coefficient is lower than the threshold corr th When the transition time is calculated without updating the transit time, the result τ of the last calculation is continued to be used. i-1 , then we can get:

[0052] τ c =τ i +τ r

[0053] stτ r =min{|τ||R xy (τ)=peak[R xy (τ)]}

[0054]

[0055] If the weak correlation persists, i.e., in five consecutive calculations corr xy Less than corr th , the traditional cross-correlation will be reapplied to reinitialize τ i To prevent errors in the cross-correlation procedure, after obtaining the transmission time, combined with the distance between the two planes of the ECT sensor, the average flow velocity can be calculated as follows:

[0056]

[0057] Optionally, the gas-liquid two-phase density can be compensated using temperature and pressure data, and the mass flow rate can be finally obtained by combining the two-phase density, liquid holdup, and flow rate. Specifically,

[0058] The gas-liquid two-phase densities are calculated by the ideal gas state equation and the fitting formula as follows:

[0059]

[0060] Among them, c0, c1, c2, c3, c4, c5 are the parameters to be fitted, ρ ref The dielectric constant is ε ref Density of the homogeneous reference medium, T ref The dielectric constant is ε ref The temperature of the capacitor is measured after the uniform reference medium is formed. The fitting formula parameters are obtained by combining the measured dielectric constant data of the liquid phase fluid. The mass flow rate of each phase of the high dynamic gas-liquid two-phase flow is calculated by the measured flow rate and liquid holdup as follows:

[0061]

[0062] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects: the present invention provides a high-dynamic gas-liquid two-phase flow measurement method based on correlation coefficient constrained cross-correlation, proposes an adaptive cross-correlation algorithm with correlation coefficient constraints, and can dynamically and stably measure flow rate under drastic flow changes; adopts a measurement capacitor selection method based on contribution analysis to screen out low-contribution measurement capacitors, effectively alleviating the negative impact of thick pipe walls on ECT imaging; a compensation method based on temperature and pressure measurement effectively compensates for measurement errors caused by temperature and pressure changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0064] Figure 1 This is a schematic diagram of the real-time dynamic measurement process of the mass flow rate of a highly dynamic gas-liquid two-phase flow;

[0065] Figure 2 Schematic diagram of the simulation sample with variable liquid holdup;

[0066] Figure 3 Schematic diagram of a typical dual-plane ECT sensor;

[0067] Figure 4 It is a schematic diagram of the ECT equivalent capacitance model;

[0068] Figure 5 Flowchart of the adaptive cross-correlation algorithm with correlation coefficient constraints.

[0069] Reference numerals: 1, annular flow; 2, laminar flow; 3, upstream ECT sensor; 4, downstream ECT sensor. DETAILED DESCRIPTION

[0070] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0071] The purpose of the present invention is to provide a high-dynamic gas-liquid two-phase flow measurement method based on correlation coefficient constrained cross-correlation, which can obtain better imaging and more accurate measurement of liquid holdup, offset the adverse effects of temperature and pressure changes on the measurement results, and realize dynamic real-time measurement of high-dynamic gas-liquid two-phase flow mass flow.

[0072] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0073] like Figure 1 As shown, a high dynamic gas-liquid two-phase flow measurement method based on correlation coefficient constrained cross-correlation includes the following steps:

[0074] Step 1: Generate simulation samples based on actual application conditions and use a selection method based on contribution analysis to select the measurement capacitors with the main contributions.

[0075] Step 2: Install a dual-plane ECT sensor and a temperature and pressure sensor on the pipeline near the tank to measure capacitance and temperature and pressure data;

[0076] Step 3: Use temperature and pressure data to compensate for and filter the normalized measured capacitance, visualize it, and obtain the liquid holdup;

[0077] Step 4: Use the adaptive cross-correlation algorithm with correlation coefficient constraint to combine the double-layer ECT sensing data to obtain the transit time and then the flow velocity;

[0078] Step 5: Use temperature and pressure data to compensate for the gas-liquid two-phase density, and finally obtain the mass flow rate by combining the two-phase density, liquid holdup, and flow rate.

[0079] Based on actual application situations, simulation samples are generated and the selection method based on contribution analysis is combined to screen out the measurement capacitors with major contributions, specifically:

[0080] The simulation flow pattern is set to variable liquid holdup annular flow 1 and laminar flow 2, as shown in Figure 2As shown in FIG, noise is added to the sample according to the actual situation of the data acquisition system, and 40dB white noise can usually be added.

[0081] Thicker tube walls (relative to the tube diameter) can severely degrade ECT image quality, reducing the accuracy of subsequent holdup measurements. This is due to the adverse effects of capacitance, which contributes little to image reconstruction. However, the impact of each capacitance measurement on image reconstruction is difficult to quantitatively assess. Therefore, its contribution to the measured holdup, which contains global information about the reconstructed image, is calculated to filter out capacitance measurements with significant contributions.

[0082] In order to select the most valuable capacitance measurement value from all the measured capacitances of the ECT sensor, their contribution to the holdup liquid measurement is analyzed. The contribution of the individual measured capacitances and their interactions are considered and obtained as follows. The holdup factor β is set as a function of the normalized capacitance vector λ, and β = f β (λ),

[0083] Considering the first-order components and their interactions, β=f β (λ) can be rewritten as

[0084]

[0085] Where k is the coefficient, M is the number of measured capacitors, λ(i) is the normalized capacitance measurement, the coefficient k is obtained by regression combined with simulated samples, and the normalized contribution E is calculated by the following formula

[0086]

[0087] It indicates how much change the holdup can reflect when the capacitance or its interaction undergoes a normalized change. The E are arranged in descending order, and then their cumulative contributions are calculated. When the cumulative contribution reaches 50% of the total, it can be considered that the components involved in the accumulation have contained most of the effective knowledge related to the holdup, and only the capacitances related to these selected components are selected for imaging.

[0088] Install a dual-plane ECT sensor and a temperature and pressure sensor on the pipeline near the tank to measure capacitance and temperature and pressure data. Specifically:

[0089] Near the tank, the high dynamic gas-liquid two-phase flow usually has a small slip between the gas phase and the liquid phase. Therefore, the dual-plane ECT sensor needs to be placed near the tank port. A typical dual-plane ECT sensor design is as follows: Figure 3 As shown, it consists of an upstream ECT sensor 3 and a downstream ECT sensor 4 with a spacing of . The temperature and pressure sensor is arranged adjacent to the dual-plane ECT sensor. When the dual-plane ECT sensor and the temperature and pressure sensor are difficult to install on the pipeline, they are installed in the tank.

[0090] The measured capacitance is normalized after temperature and pressure data compensation and screening, and then visualized to obtain the liquid holdup, specifically:

[0091] The main purpose of using temperature and pressure data to compensate for the capacitance of the dual-plane ECT sensor after screening is to compensate for its full-scale capacitance, and then compensate for the normalized result of the measured capacitance. When the tube is filled with liquid propellant, the full-scale capacitance can be measured. However, as the temperature and pressure change, the dielectric constant of the liquid propellant and therefore the measured capacitance will also change. Figure 4 As shown, the capacitance measured in the dual-plane ECT sensor can be expressed as two wall capacitances C w1 、C w2 and internal capacitor C X The series connection of C w1 and C w2 Considered as a single capacitor C w , get the measured capacitance C M :

[0092]

[0093] When the tube is filled with a dielectric constant of ε ref When the medium is a homogeneous medium, the measured capacitance can be expressed as C X =ε ref C0, where C0 is the capacitance inside the pipe wall in the empty field, in addition to the measured empty calibration capacitance C L In addition, another dielectric constant ε ref The uniform reference medium is filled in the pipe, and the mutual capacitance between the electrodes is measured to obtain C ref , then C W and C0 can be calculated as:

[0094]

[0095] where ε L is the dielectric constant of the empty field, which is generally taken as 1, and the full calibration capacitance C is given by the following formula H :

[0096]

[0097] Among them, ε liquid It can be obtained by combining the measured dielectric constant data of the liquid phase with the following fitting formula:

[0098]

[0099] Where P and T are the temperature and pressure of the measured capacitor, respectively. b0, b1, b2, and b3 are the coefficients to be fitted. Then, during measurement, the full-scale capacitance of the filtered capacitance measurement is updated in real time based on the temperature and pressure data. Parallel normalization can be used in real time to obtain the compensated normalized measured capacitance:

[0100]

[0101] The reconstructed image can be obtained using the linear back projection algorithm:

[0102]

[0103] Where S is the sensitive field matrix, which can be obtained based on the simulation model. After obtaining the reconstructed image, the liquid holdup can be obtained:

[0104]

[0105] where K is the number of pixels in the reconstructed image, A i and A are the area of ​​the ith pixel and the cross-sectional area of ​​the pipe, respectively.

[0106] like Figure 5 As shown in Figure 1, the adaptive cross-correlation algorithm with correlation coefficient constraint is combined with the double-layer ECT sensing data to obtain the transit time and then the flow velocity, specifically:

[0107] Combined with the cross-correlation technique, the average flow velocity of the propellant two-phase flow can be obtained using the dual-plane ECT sensor. According to the freezing assumption, the upstream ECT sensor 3 and the downstream ECT sensor 4 have the same characteristic flow noise. The cross-correlation is used to obtain the velocity-related transition time, which is defined as follows:

[0108]

[0109] Where x(t) and y(t) are the collected upstream and downstream ECT signals, t0 is the start time of sampling, and t s is the sampling time. At the same time, the normalized cross-correlation result is often used for speed measurement. The normalization formula is as follows:

[0110]

[0111] Among them, R xy，coeff(τ) is the normalized cross-correlation result, R xy，coeff(τ) The peak value is recorded as corr xy , occurs when the similarity between the upstream ECT signal and the downstream ECT signal reaches its maximum value, and the corresponding time τ0 represents the conversion time of the flow propagating from the upstream ECT sensor 3 to the downstream ECT sensor 4:

[0112] corrxy =max[R xy，coeff (τ)]

[0113] R xy，coeff (τ0)=max[R xy，coeff (τ)]

[0114] Since the upstream and downstream signals are usually collected synchronously, the length of the weak correlation part between them varies greatly with the flow rate, which will greatly reduce the correlation coefficient value at the peak and lead to errors in the identification of the peak position in many cases. In order to reduce the poor correlation part of the signal between the upstream and downstream, an adaptive cross-correlation algorithm is used, which is expressed as:

[0115]

[0116] τ c =τ i +τ r , stR xy (τ r )=max[R xy (τ)]

[0117] Among them, τ i is the estimated initial flight time, τ r is the remaining transit time, τ c is the corrected flight time. Initialization τ0 can be obtained by conventional cross-correlation technology, and then the corrected flight time τ is used. c To update,

[0118] τ i =τ c

[0119] By fine-tuning τ i The optimal correlation can be obtained. However, since the highly dynamic gas-liquid two-phase flow does not fully conform to the freezing assumption and the correlation between the upstream and downstream signals is weak, false peaks in the correlation graph are mistakenly identified, resulting in large measurement errors or even outliers. Since the flow rate cannot change suddenly, the remaining transit time τ r should be close to zero, so the peak point near the origin is taken to obtain the remaining transition time; in addition, a constraint is imposed on the correlation coefficient. When the correlation coefficient is lower than the threshold, the transition time is calculated in a more conservative way. When the correlation coefficient is lower than the threshold corr th When the transition time is calculated without updating the transit time, the result τ of the last calculation is continued to be used. i-1 , then we can get:

[0120] τ c =τ i +τ r

[0121] stτ r =min{|τ||R xy (τ)=peak[R xy (τ)]}

[0122]

[0123] When corrxy is greater than corrth, the two signals should have a strong correlation. Therefore, corrth can be set based on specific application experience. To further stabilize the solution, if the weak correlation persists, that is, corrxy is less than corrth in five consecutive calculations, the traditional cross-correlation will be re-applied to reinitialize τi to prevent the cross-correlation procedure from erroring. After obtaining the transmission time, combined with the distance between the two planes of the ECT sensor, the average flow velocity can be calculated by the following formula

[0124]

[0125] The gas-liquid two-phase density is compensated using temperature and pressure data, and the mass flow rate is finally obtained by combining the two-phase density, liquid holdup and flow rate. Specifically:

[0126] The gas-liquid two-phase density can be calculated by the ideal gas state equation and the fitting formula as follows:

[0127]

[0128] Among them, c0, c1, c2, c3, c4, c5 are the parameters to be fitted, ρ ref The dielectric constant is ε ref Density of the homogeneous reference medium, T ref The dielectric constant is ε ref The temperature of the capacitor is measured after the uniform reference medium is added. The fitting formula parameters can be obtained by combining the measured dielectric constant data of the liquid phase fluid. The mass flow rate of each phase of the highly dynamic gas-liquid two-phase flow can be calculated by the measured flow rate and liquid holdup as follows:

[0129]

[0130] The present invention discloses the following technical effects: the present invention provides a high-dynamic gas-liquid two-phase flow measurement method based on correlation coefficient constrained cross-correlation, proposes an adaptive cross-correlation algorithm with correlation coefficient constraints, and can dynamically and stably measure flow rate under drastic flow changes; adopts a measurement capacitor selection method based on contribution analysis to screen out low-contribution measurement capacitors, effectively alleviating the negative impact of thick pipe walls on ECT imaging; a compensation method based on temperature and pressure measurement effectively compensates for measurement errors caused by temperature and pressure changes.

[0131] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A high dynamic gas-liquid two-phase flow measurement method based on correlation coefficient constrained cross-correlation, characterized in that: The steps include: Step 1: Generate simulation samples based on actual application conditions and use a selection method based on contribution analysis to screen the measured capacitance. Step 2: Install a dual-plane ECT sensor and a temperature and pressure sensor on the pipeline near the tank to measure capacitance and temperature and pressure data; Step 3: Use temperature and pressure data to compensate for and filter the normalized measured capacitance, visualize it, and obtain the liquid holdup; Step 4: Use the adaptive cross-correlation algorithm with correlation coefficient constraint to combine the double-layer ECT sensing data to obtain the transit time and then the flow velocity; Step 5: Use temperature and pressure data to compensate for the gas-liquid two-phase density, and finally obtain the mass flow rate by combining the two-phase density, liquid holdup, and flow rate; The adaptive cross-correlation algorithm with correlation coefficient constraint is combined with the double-layer ECT sensor data to obtain the transit time and then the flow velocity, specifically: Get the velocity-dependent transition time: The traditional cross-correlation algorithm formula is as follows: ； Where x(t) and y(t) are the collected upstream and downstream ECT signals, t0 is the start time of sampling, and t s is the sampling time. At the same time, the normalized cross-correlation result is used for velocity measurement. The normalization formula is as follows: ； R xy,coeff(τ) The peak value is recorded as corr xy , occurs when the similarity between the upstream ECT signal and the downstream ECT signal reaches its maximum value, and the corresponding time τ0 represents the conversion time of the flow propagating from the upstream ECT sensor to the downstream ECT sensor: ； ； In order to reduce the poor correlation between upstream and downstream signals, an adaptive cross-correlation algorithm is used, which is expressed as: ； ； Among them, τ i is the estimated initial flight time, τ r is the remaining transit time, τ c is the corrected flight time, τ0 is obtained by conventional cross-correlation technique during initialization, and then the corrected flight time τ is used to obtain the corrected flight time τ c To update, ； By fine-tuning τ i Get the optimal correlation, take the peak point near the origin to get the remaining transition time, when the correlation coefficient is lower than the threshold corr th When the transition time is calculated without updating the transit time, the result τ of the last calculation is continued to be used. i-1 , then we can get: ； ； ； If the weak correlation persists, i.e., in five consecutive calculations corr xy Less than corr th , the traditional cross-correlation will be reapplied to reinitialize τ i To prevent errors in the cross-correlation procedure, after obtaining the transmission time, combined with the distance between the two planes of the ECT sensor, the average flow velocity is calculated by the following formula: 。 2. The high dynamic gas-liquid two-phase flow measurement method based on correlation coefficient constrained cross-correlation according to claim 1 is characterized in that: Based on actual application situations, simulation samples are generated and the selection method based on contribution analysis is combined to screen out the measurement capacitors with major contributions, specifically: Setting the holdup β as a function of the normalized capacitance λ yields: ； Considering the first-order components and their interactions, the above formula can be rewritten as: ； Where k is the coefficient, M is the number of measured capacitors, λ(i) is the normalized capacitance measurement, the coefficient k is obtained by regression combined with simulated samples, and the normalized contribution E is calculated as follows: ； The normalized contribution E of each component is arranged in descending order, and the cumulative contribution is calculated. When the cumulative contribution reaches 50% of the total, the capacitance associated with the component involved in the cumulative contribution is selected for imaging.

3. The high dynamic gas-liquid two-phase flow measurement method based on correlation coefficient constrained cross-correlation according to claim 2, characterized in that: Install a dual-plane ECT sensor and a temperature and pressure sensor on the pipeline near the tank to measure capacitance and temperature and pressure data. Specifically: The dual-plane ECT sensor is arranged near the tank port or in the tank, the temperature and pressure sensor is arranged adjacent to the dual-plane ECT sensor, and the dual-plane ECT sensor includes an upstream ECT sensor and a downstream ECT sensor with a distance L therebetween.

4. The high dynamic gas-liquid two-phase flow measurement method based on correlation coefficient constrained cross-correlation according to claim 3, characterized in that: The measured capacitance is normalized after temperature and pressure data compensation and screening, and then visualized to obtain the liquid holdup, specifically: The capacitance measured in an ECT sensor is represented by two wall capacitances C w1 、C w2 and internal capacitor C X The series connection of C w1 and C w2 Considered as a single capacitor C w , get the measured capacitance C M : ； When the tube is filled with a dielectric constant of ε ref When the medium is a homogeneous medium, the measured capacitance can be expressed as C X =ε ref C0, where C0 is the capacitance inside the pipe wall in the empty field, in addition to the measured empty calibration capacitance C L In addition, another dielectric constant ε ref The uniform reference medium is filled in the pipe, and the mutual capacitance between the electrodes is measured to obtain C ref , then C W and C0 can be calculated as: ； in is the dielectric constant of the empty field, which is taken as 1, and the full calibration capacitance C is given by H : ； Among them, ε liquid It can be obtained by combining the measured dielectric constant data of the liquid phase with the following fitting formula: ； Where P and T are the temperature and pressure of the measured capacitance, respectively. b0, b1, b2, and b3 are the coefficients to be fitted. Subsequently, during measurement, the full calibration capacitance of the filtered capacitance measurement is updated in real time based on the temperature and pressure data. This means that parallel normalization is used in real time to obtain the compensated normalized measured capacitance: ； The reconstructed image can be obtained using the linear back projection algorithm: ； Where S is the sensitive field matrix, which is obtained based on the simulation model. After obtaining the reconstructed image, the liquid holdup can be obtained: ； Where K is the number of pixels in the reconstructed image, A i and A are the area of ​​the ith pixel and the cross-sectional area of ​​the pipe, respectively.

5. The high dynamic gas-liquid two-phase flow measurement method based on correlation coefficient constrained cross-correlation according to claim 4, characterized in that: The gas-liquid two-phase density is compensated using temperature and pressure data, and the mass flow rate is finally obtained by combining the two-phase density, liquid holdup and flow rate. Specifically: The gas-liquid two-phase densities are calculated by the ideal gas state equation and the fitting formula as follows: ； Among them, c0, c1, c2, c3, c4, c5 are the parameters to be fitted. The dielectric constant is ε ref The density of the homogeneous reference medium, The dielectric constant is ε ref The temperature of the capacitor is measured after the uniform reference medium is formed. The fitting formula parameters are obtained by combining the measured dielectric constant data of the liquid phase fluid. The mass flow rate of each phase of the high dynamic gas-liquid two-phase flow is calculated by the measured flow rate and liquid holdup as follows: 。

Images