A Multi-Frequency Fusion-Based Dual-Modal Ultrasonic Tomography Method and System

Abstract

This invention provides a method and system for multi-frequency fusion-based reflection and transmission dual-modal ultrasound tomography. The method includes: sequentially exciting low-frequency ultrasound to each ultrasonic sensor, enabling each sensor to perform reflection and transmission dual-modal tomography at an initial frequency; constructing a low-frequency coefficient matrix and a low-frequency ultrasound image based on the reflected and transmitted signals received by each ultrasonic sensor; determining target frequencies that allow the sound waves from each ultrasonic sensor to penetrate each ultrasound imaging path based on the low-frequency ultrasound image; sequentially exciting high-frequency ultrasound to each ultrasonic sensor, constructing a high-frequency coefficient matrix based on the reflected and transmitted signals received by each ultrasonic sensor; fusing the low-frequency coefficient matrix and the high-frequency coefficient matrix to obtain a fusion coefficient matrix; and reconstructing the image based on the fusion coefficient matrix to obtain a multi-frequency fused two-phase distribution map. This invention improves the imaging effect and accuracy for conditions involving multiple bubbles and bubble clusters.

Description

Technical Field

[0001] This invention relates to the field of liquid metal-bubble two-phase flow imaging technology, and in particular to a reflection-transmission dual-modal ultrasound tomography method and system based on multi-frequency fusion. Background Technology

[0002] Existing ultrasound tomography techniques typically include single-mode imaging and dual-mode imaging. Single-mode imaging usually includes reflection imaging and transmission imaging. However, single-mode imaging can produce many artifacts under certain fluid distributions. When bubbles are close to the ultrasound sensor, the sensor cannot receive the ideal reflected signal. Dual-mode imaging is usually based on a single frequency and is constrained by both the measurement container and resolution. During measurement, in order to obtain a resolvable transmission signal amplitude, an appropriate frequency must be selected according to the required measurement depth. The greater the depth, the lower the frequency needs to be. However, the lower the frequency, the lower the resolution. At the same time, the presence of bubbles in the cross-section will also enhance the attenuation of the ultrasound signal and reduce the imaging effect. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide a dual-modal ultrasonic tomography method and system based on multi-frequency fusion, which can flexibly adapt to different working conditions such as gas content in different cross sections, improve the imaging effect on working conditions with multiple bubbles and bubble groups, and improve the imaging accuracy.

[0004] To achieve the above objectives, the technical solutions adopted in the embodiments of the present invention are as follows:

[0005] In a first aspect, embodiments of the present invention provide a multi-frequency fusion-based dual-modal ultrasound tomography method for reflection and transmission, applied to a multi-frequency fusion-based dual-modal ultrasound tomography system for reflection and transmission. The multi-frequency fusion-based dual-modal ultrasound tomography system for reflection and transmission includes multiple ultrasound sensors, and the multi-frequency fusion-based dual-modal ultrasound tomography method for reflection and transmission includes:

[0006] Each of the ultrasonic sensors is sequentially excited with low-frequency ultrasound, so that each of the ultrasonic sensors performs reflection-transmission dual-modal tomography at an initial frequency. A low-frequency coefficient matrix and a low-frequency ultrasound image are constructed based on the reflected and transmitted signals received by each of the ultrasonic sensors. The initial frequency is the frequency at which the sound waves of each ultrasonic sensor can penetrate each ultrasound imaging path when the cross-sectional gas content of the channel to be detected is the highest predicted cross-sectional gas content.

[0007] Based on the low-frequency ultrasound images, a target frequency is determined that enables the sound waves of each of the ultrasound sensors to penetrate each ultrasound imaging path.

[0008] Each of the ultrasonic sensors is sequentially excited with high-frequency ultrasound, so that each of the ultrasonic sensors performs reflection-transmission dual-modal tomography at the target frequency, and a high-frequency coefficient matrix is ​​constructed based on the reflected and transmitted signals received by each of the ultrasonic sensors.

[0009] The low-frequency coefficient matrix and the high-frequency coefficient matrix are fused to obtain a fusion coefficient matrix. Based on the fusion coefficient matrix, image reconstruction is performed to obtain a two-phase distribution map of multi-frequency fusion.

[0010] Furthermore, embodiments of the present invention provide a first possible implementation of the first aspect, wherein the step of constructing a low-frequency coefficient matrix and a low-frequency ultrasound image based on the reflected and transmitted signals received by each of the ultrasonic sensors includes:

[0011] The low-frequency coefficient matrix and the initial low-frequency ultrasound image are reconstructed using a dual-mode fusion reconstruction algorithm based on the reflected and transmitted signals received by each of the ultrasonic sensors.

[0012] The initial low-frequency ultrasound image is denoised, and bubble sealing is performed on the denoised initial low-frequency ultrasound image to restore the bubble image to complete, thus obtaining the low-frequency ultrasound image.

[0013] Furthermore, this embodiment of the invention provides a second possible implementation of the first aspect, wherein the step of determining the target frequency based on the low-frequency ultrasound image that enables the sound waves of each of the ultrasound sensors to penetrate each ultrasound imaging path includes:

[0014] Based on the low-frequency ultrasound image, determine the frequency corresponding to the maximum ultrasound attenuation on the ultrasound path of each ultrasound sensor, and select the lowest frequency from the frequencies corresponding to the maximum ultrasound attenuation on each ultrasound path.

[0015] Obtain the set frequency corresponding to the set minimum resolvable bubble radius that the ultrasonic sensor can detect;

[0016] The minimum value between the lowest frequency and the set frequency is taken as the target frequency that enables the sound waves of each ultrasonic sensor to penetrate each ultrasonic imaging path.

[0017] Furthermore, this embodiment of the invention provides a third possible implementation of the first aspect, wherein the step of determining the frequency corresponding to the maximum ultrasonic attenuation on the ultrasonic path of each of the ultrasonic sensors based on the low-frequency ultrasonic image includes:

[0018] The lengths of the bubbles and the liquid metal along each ultrasonic path are obtained from the low-frequency ultrasonic images.

[0019] The frequency at which the attenuated sound pressure amplitude on each ultrasonic path is greater than the ambient noise is determined based on the bubble length and liquid metal length on each ultrasonic path.

[0020] Furthermore, this embodiment of the invention provides a fourth possible implementation of the first aspect, wherein the step of constructing a high-frequency coefficient matrix based on the reflected and transmitted signals received by each of the ultrasonic sensors includes:

[0021] The high-frequency coefficient matrix is ​​reconstructed using a dual-mode fusion reconstruction algorithm based on the reflected and transmitted signals received by each ultrasonic sensor.

[0022] Furthermore, this embodiment of the invention provides a fifth possible implementation of the first aspect, wherein the step of fusing the low-frequency coefficient matrix and the high-frequency coefficient matrix to obtain a fused coefficient matrix includes:

[0023] An adaptive gradient weight allocation algorithm is used to adaptively fuse the low-frequency coefficient matrix and the high-frequency coefficient matrix to obtain the fused coefficient matrix.

[0024] Furthermore, this embodiment of the invention provides a sixth possible implementation of the first aspect, wherein the step of obtaining a multi-frequency fused two-phase distribution map by image reconstruction based on the fusion coefficient matrix includes:

[0025] The initial fused image is reconstructed based on the fusion coefficient matrix;

[0026] The initial fused image is denoised, and the denoised initial fused image is bubble-closing processed to restore the bubble image to complete, thus obtaining the two-phase distribution map of the multi-frequency fusion.

[0027] Furthermore, this embodiment of the invention provides a seventh possible implementation of the first aspect, wherein the excitation signal of the ultrasonic sensor is a pulsed step frequency modulation signal;

[0028] In the steps of sequentially exciting low-frequency ultrasound to each of the ultrasonic sensors and sequentially exciting high-frequency ultrasound to each of the ultrasonic sensors, the excitation time interval between each ultrasonic sensor is: ;

[0029] The cross-section of the channel to be detected is circular. The cross-sectional diameter of the channel to be tested is [missing information]. The speed of sound in liquid metal. The speed of sound in air. The gas content of the cross section.

[0030] Secondly, embodiments of the present invention also provide a dual-modal ultrasound tomography system based on multi-frequency fusion, comprising: a controller and multiple ultrasound sensors, wherein the controller includes a processor and a storage device;

[0031] The storage device stores a computer program that, when executed by the processor, performs the method as described in any of the first aspects.

[0032] Furthermore, the present invention provides a first possible implementation of the second aspect, wherein a plurality of ultrasonic sensors are uniformly disposed on the outer surface of the channel to be tested, the cross-section of the channel to be tested is an axisymmetric figure, and the plurality of ultrasonic sensors are distributed in an axisymmetric manner relative to the cross-section of the channel to be tested.

[0033] This invention provides a method and system for dual-modal ultrasound tomography based on multi-frequency fusion, applicable to a dual-modal ultrasound tomography system based on multi-frequency fusion. The system includes multiple ultrasonic sensors. The method comprises: sequentially exciting low-frequency ultrasound to each ultrasonic sensor, enabling each sensor to perform dual-modal ultrasound tomography at an initial frequency; and constructing a low-frequency coefficient matrix and a low-frequency ultrasound image based on the reflected and transmitted signals received by each ultrasonic sensor. The initial frequency is the frequency to be detected. The frequency at which the sound waves from each ultrasonic sensor can penetrate each ultrasonic imaging path is determined when the cross-sectional gas content of the measuring channel is the predicted highest cross-sectional gas content. Based on the low-frequency ultrasonic image, the target frequency at which the sound waves from each ultrasonic sensor can penetrate each ultrasonic imaging path is determined. High-frequency ultrasound is sequentially excited to each ultrasonic sensor, so that each ultrasonic sensor performs reflection-transmission dual-modal tomography at the target frequency. A high-frequency coefficient matrix is ​​constructed based on the reflected and transmitted signals received by each ultrasonic sensor. The low-frequency coefficient matrix and the high-frequency coefficient matrix are fused to obtain a fusion coefficient matrix. Based on the fusion coefficient matrix, the image is reconstructed to obtain a multi-frequency fused two-phase distribution map. This invention first excites low-frequency ultrasound sequentially to each ultrasonic sensor for reflective projection dual-modal tomography. Then, based on the low-frequency ultrasound image obtained from the low-frequency imaging, the target frequency for subsequent high-frequency imaging is determined. By using adaptive frequency for high-frequency imaging, it can ensure that high-frequency sound waves can penetrate the channel under test. By adaptively fusing the low-frequency coefficient matrix and the high-frequency coefficient matrix, the final multi-frequency fused two-phase distribution map can meet the minimum resolution requirements. It can flexibly adapt to different bubble size measurement domains and different cross-sectional gas contents, improving the imaging effect and imaging accuracy for multiple bubbles and bubble groups.

[0034] Other features and advantages of the embodiments of the present invention will be set forth in the following description, or some features and advantages may be inferred from the description or determined without doubt, or may be learned by practicing the techniques described above in the embodiments of the present invention.

[0035] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0036] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0037] Figure 1 The flowchart of a dual-modal ultrasound tomography method based on multi-frequency fusion based on reflection and transmission provided by an embodiment of the present invention is shown.

[0038] Figure 2 The flowchart illustrates an embodiment of the present invention for ultrasonic tomography of a two-phase flow consisting of liquid metal and bubbles. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be described below in conjunction with the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0040] Existing ultrasound tomography techniques typically include single-mode imaging and dual-mode imaging. Single-mode imaging usually employs reflection imaging and transmission imaging.

[0041] The basic principle of dual-modal imaging is as follows:

[0042] When sound waves propagate to the interface between two different media, reflection, transmission, or diffraction occurs due to differences in acoustic properties. Based on the geometric approximation of sector waves, when the incident wavelength... When the sound wave is much smaller than the characteristic size of the obstacle, the diffraction effect can be ignored. The strength of the diffraction effect can be expressed as the product of the wave number and the characteristic size of the obstacle:

[0043] (1)

[0044] in, It is the wave number of the sound wave; These are the frequency, speed, and wavelength of the sound wave, respectively. The characteristic dimension of a sound propagation obstacle is its radius, such as the characteristic dimension of a sphere being its radius.

[0045] Acoustic diffraction effect and dimensionless wavenumber parameter There is a significant correlation: The smaller the value, the stronger the diffraction effect; conversely, a negative correlation trend is shown. Specifically: (1) When When (corresponding to the characteristic scale of the sphere and the order of wavelength), some sound waves undergo diffraction; (2) when When (Rayleigh scattering region), the sound wave produces a significant diffraction effect, and its propagation path is almost unaffected by the sphere; (3) when In the geometric acoustic region, sound waves primarily follow rectilinear propagation, creating a distinct acoustic shadow area behind obstacles. In ultrasonic testing based on reflection and transmission modes, different methods are selected according to the specific needs of the scenario. The value is used to determine the minimum. Generally, in a liquid metal-air system, it is considered that when... When the value is 10, the sound wave will not diffract around the bubble.

[0046] Based on this theory, the smallest bubble size that can be detected in this embodiment is limited by the center frequency of the ultrasonic probe. For example, when =10, when using a 4MHz ultrasonic probe, the smallest resolvable bubble radius in liquid lead-bismuth at 400℃ is approximately:

[0047] (2)

[0048] The strength of reflection and transmission effects at the interface is determined by the acoustic impedance Z:

[0049] (3)

[0050] in, It is the density of the medium. It is the speed of sound in that medium.

[0051] Commonly used transmission coefficient and reflection coefficient The transmission coefficient represents the ratio of the sound pressure of the transmitted and reflected sound waves to the original sound pressure. It is the ratio of the sound pressure of the transmitted and reflected sound waves to the original sound pressure. This is the transmission coefficient when the sound wave propagates from medium 1 to medium 2 without diffraction. and reflection coefficient It can be represented as

[0052] (4)

[0053] (5)

[0054] in, It is the acoustic impedance in medium 1. It is the acoustic impedance in medium 2.

[0055] At room temperature, the acoustic impedance of air is approximately The acoustic impedance in liquid metal is often... On the order of magnitude, such as the acoustic impedance of gallium. This results in a reflectivity of up to 99% when the liquid metal-air interface is incident vertically.

[0056] Building upon this, ultrasonic transmission imaging achieves tomographic imaging by comparing the signal amplitude differences between the background field and the measurement field. Ultrasonic reflection imaging, on the other hand, derives the location of the reflecting interface based on the time elapsed from the transmitter to the receiver.

[0057] Reflection method:

[0058] (6)

[0059] in, d t-r The distance between the boundary point of the discrete phase medium and the axis of the ultrasonic transducer probe is denoted as . TOF t-r This refers to the transit time corresponding to the distance between the boundary points of the discrete phase medium. c w The velocity of sound in the continuous phase medium within the measured field is denoted as .

[0060] Transmission method:

[0061] (7)

[0062] In the formula, P 0 represents the amplitude of the ultrasonic wave before it enters the medium. P This represents the emitted amplitude of the ultrasonic wave after traveling a distance of 1. Let be a function of the ultrasonic attenuation coefficient at any point on the acoustic path. dl For the micro-element on the sound path.

[0063] After taking the logarithm of both sides, let Then equation (7) becomes

[0064] (8)

[0065] For the cross-section of a liquid-solid two-phase flow pipe, the circular field is divided according to an appropriate two-dimensional partitioning rule, and the ultrasonic attenuation coefficient within each partitioned unit is assumed to be constant. This yields a discrete relationship between the projected data and the ultrasonic amplitude attenuation. If we use... Indicates the first j The ultrasonic attenuation coefficient of each subdivided unit, l ii Indicates the first i The sound wave ray in the firsti The acoustic path length in each subdivided unit. Then the discrete form of equation (6) can be expressed as the following system of linear equations:

[0066] (9)

[0067] make , , Then equation (9) is transformed into

[0068] (10)

[0069] Where A is the sound path coefficient matrix, the discrete image reconstruction problem is transformed into the problem of estimating the attenuation coefficient vector x from the given projection data Y.

[0070] The single reflection mode imaging described above provides better information about the bubble boundary and can better depict the overall bubble profile. The single transmission mode, compared to the single reflection mode, provides more information about the bubble's central region and can be used to distinguish the two-phase flow behavior within a bubble group. For example, the gaps filled with liquid metal between multiple bubbles are difficult to distinguish in single reflection mode. However, both transmission and reflection modes have their own drawbacks. Under certain fluid distributions, many artifacts can occur. For instance, when a bubble is close to the sensor, the sensor cannot receive an ideal reflection signal, resulting in poor imaging quality.

[0071] Traditional dual-modal imaging, based on reflection and transmission, is typically limited by a single frequency and constrained by both the measurement container and resolution. These two factors require compromise. During measurement, to obtain a resolvable transmitted signal amplitude, a suitable frequency is selected based on the required measurement depth; the greater the depth, the lower the frequency needs to be. However, a lower frequency results in lower resolution. Furthermore, the presence of air bubbles in the cross-section enhances the attenuation of the ultrasound signal. This also leads to problems such as low image quality and inability to accurately reproduce the original image. Moreover, using a fixed scanning frequency beforehand results in low imaging accuracy.

[0072] To address the aforementioned issues, this invention provides a method and system for dual-modal ultrasound tomography based on multi-frequency fusion of reflection and transmission modes. The embodiments of this invention will be described in detail below.

[0073] This embodiment provides a multi-frequency fusion-based dual-modal ultrasound tomography method, which can be applied to a multi-frequency fusion-based dual-modal ultrasound tomography system. This system includes a controller and multiple ultrasound sensors. (See [link to documentation]). Figure 1 The flowchart shown is for a dual-modal ultrasound tomography method based on multi-frequency fusion, which mainly includes the following steps:

[0074] Step S102: sequentially excite low-frequency ultrasound to each ultrasonic sensor, so that each ultrasonic sensor performs reflection-transmission dual-modal tomography at the initial frequency, and construct a low-frequency coefficient matrix and a low-frequency ultrasound image based on the reflection and transmission signals received by each ultrasonic sensor.

[0075] The aforementioned ultrasonic sensor may be an ultrasonic probe, and multiple ultrasonic sensors are evenly arranged on the outer surface of the cross-section of the channel to be tested. The cross-section of the channel to be tested is an axisymmetric figure, such as a circle, square, or rectangle, and the multiple ultrasonic sensors are axially symmetrically distributed with respect to the cross-section of the channel to be tested, so that the ultrasonic sensors can receive the transmitted signals from the sensors on the symmetrical side.

[0076] The aforementioned initial frequency is the frequency at which the sound waves of each ultrasonic sensor can penetrate each ultrasonic imaging path when the cross-sectional gas content of the channel under test is the predicted highest cross-sectional gas content. The aforementioned predicted highest cross-sectional gas content is the highest cross-sectional gas content expected under all test conditions of the channel under test. For example, the maximum cross-sectional gas content obtained by measurement can be selected from the historical imaging data of the channel under test. For experiments where the range of cross-sectional gas content can be controlled or predicted, the expected highest cross-sectional gas content can also be used, denoted as the predicted highest cross-sectional gas content. The initial frequency is determined based on the frequency corresponding to the attenuated sound pressure amplitude received by the ultrasonic sensor at the predicted highest cross-sectional gas content being higher than the amplitude of the ambient noise, so as to ensure that the ultrasonic sensor can penetrate the ultrasonic imaging path under the current cross-sectional gas content when imaging at low frequencies.

[0077] In one embodiment, the excitation signal of the ultrasonic sensor is a pulsed step frequency modulated signal; the pulsed step signal can be expressed as:

[0078] (11)

[0079] in, For low-frequency signal amplitude, For high-frequency signal amplitude, f 1 represents the initial frequency used to excite low-frequency ultrasound. f 2 is the target frequency used to excite high-frequency ultrasound. and For phase, To the total time of excitation of low-frequency ultrasound (i.e., low-frequency scanning), ~ The time period is the time required to construct the low-frequency coefficient matrix and low-frequency ultrasound image based on the reflected and transmitted signals, and to determine the target frequency for high-frequency imaging based on the low-frequency ultrasound image. and These are the start and end times for exciting high-frequency ultrasound (the start and end times of the high-frequency scan). The total time required to exciting high-frequency ultrasound is the same as the total time required to exciting low-frequency ultrasound. - = ,Right now .

[0080] To ensure phase continuity, the phases of the excitation of low-frequency ultrasound and the excitation of high-frequency ultrasound must satisfy the following:

[0081] (12)

[0082] When exciting low-frequency ultrasound, for a sensor array consisting of multiple ultrasound sensors, multiple ultrasound sensors are excited sequentially. A dual-mode fusion reconstruction algorithm, such as a hybrid binary reconstruction algorithm or a linear reconstruction algorithm, is used to construct a dual-mode coefficient matrix based on the reflected and transmitted signals received by each ultrasound sensor, thereby obtaining the low-frequency coefficient matrix and the low-frequency ultrasound image.

[0083] Step S104: Determine the target frequency that enables the sound waves of each ultrasonic sensor to penetrate each ultrasonic imaging path based on the low-frequency ultrasonic image.

[0084] The lengths of the bubbles and liquid metal on each ultrasonic path are determined based on the low-frequency ultrasonic images. The frequencies corresponding to the maximum allowable ultrasonic attenuation on each ultrasonic imaging path are calculated based on the lengths of the bubbles and liquid metal on each ultrasonic path. The minimum frequency value is selected from the frequencies corresponding to the maximum allowable ultrasonic attenuation on each ultrasonic path and recorded as the minimum frequency. The target frequency that allows the sound waves of each ultrasonic sensor to penetrate each ultrasonic imaging path in high-frequency imaging is determined based on the minimum frequency and the set frequency corresponding to the required maximum resolution.

[0085] Step S106: High-frequency ultrasound is sequentially excited to each ultrasonic sensor, so that each ultrasonic sensor performs reflection-transmission dual-mode tomography at the target frequency, and a high-frequency coefficient matrix is ​​constructed based on the reflection and transmission signals received by each ultrasonic sensor.

[0086] For each ultrasonic sensor array, multiple high-frequency ultrasonic waves are sequentially excited by a pulsed step frequency modulation signal. A hybrid binary reconstruction algorithm is used to construct a dual-mode coefficient matrix based on the reflected and transmitted signals, thereby obtaining the high-frequency coefficient matrix.

[0087] Step S108: The low-frequency coefficient matrix and the high-frequency coefficient matrix are fused to obtain a fused coefficient matrix. Based on the fused coefficient matrix, the image is reconstructed to obtain a two-phase distribution map of multi-frequency fusion.

[0088] A fusion algorithm is used to fuse the low-frequency coefficient matrix obtained from low-frequency imaging with the high-frequency coefficient matrix obtained from high-frequency imaging, resulting in a fused coefficient matrix. This fusion algorithm can be any one or more of weighted fusion, multi-scale fusion, and adaptive fusion. A hybrid binary reconstruction algorithm is then used to reconstruct the fused coefficient matrix into a fused image. The fused image is then denoised and bubble-closing is performed to restore the bubble image, resulting in a multi-frequency fused two-phase distribution map.

[0089] The multi-frequency fusion-based reflection-transmission dual-modal ultrasonic tomography method provided in this embodiment first excites low-frequency ultrasound to each ultrasonic sensor sequentially for reflection-transmission dual-modal tomography, and determines the target frequency for subsequent high-frequency imaging based on the low-frequency ultrasound image obtained from the low-frequency imaging. By using adaptive frequency for high-frequency imaging, it can ensure that high-frequency sound waves can penetrate the channel to be detected. By adaptively fusing the low-frequency coefficient matrix and the high-frequency coefficient matrix, the final multi-frequency fused two-phase distribution map can meet the minimum resolution requirements. It can flexibly adapt to different bubble size measurement domains and different cross-sectional gas content conditions, improving the imaging effect and imaging accuracy for multiple bubbles and bubble groups.

[0090] In one embodiment, this embodiment provides a specific implementation method for constructing a low-frequency coefficient matrix and a low-frequency ultrasound image based on the reflected and transmitted signals received by each ultrasonic sensor:

[0091] The low-frequency coefficient matrix and the initial low-frequency ultrasound image are reconstructed using a dual-mode fusion reconstruction algorithm based on the reflected and transmitted signals received by each ultrasonic sensor.

[0092] The initial low-frequency ultrasound image is denoised, and bubble sealing is performed on the denoised initial low-frequency ultrasound image to restore the bubble image to complete, thus obtaining the low-frequency ultrasound image.

[0093] For a sensor array consisting of N ultrasonic transducers, a total of N×N time-varying ultrasonic signals are received. The measurement field is divided into different pixel blocks according to the two-dimensional partitioning method, and all non-repeating sound wave paths are marked. Based on the reflected and transmitted signals received by each ultrasonic sensor, a dual-mode fusion reconstruction algorithm (such as a hybrid binary reconstruction algorithm or a linear reconstruction algorithm) is used to construct a dual-mode coefficient matrix, denoted as the low-frequency coefficient matrix. The low-frequency coefficient matrix is ​​the weight distribution of the dual-mode reconstructed image, thus obtaining the initial low-frequency ultrasonic image.

[0094] For example, assuming the ultrasonic sensor is 24, the formula for calculating the low-frequency coefficient matrix is:

[0095] (13)

[0096] (14)

[0097] in, The weight distribution of the dual-mode reconstructed image (i.e., the initial low-frequency ultrasound image) is given, where m and n represent the node coordinates of the two-dimensional partition, and the transmission signal matrix is ​​given. and reflected signal matrix It is obtained by analyzing the reflected and transmitted signals received by each ultrasonic sensor. This represents the number of echoes received by the ultrasonic sensor Tx. It is the ratio of the attenuated signal amplitude of the particle passing through the Tx-Ry ultrasound path to the background value. The value ranges from 0 to 1. It is the threshold for signal binarization. For the sensitivity matrix, For unit network weight, The unit weight is for the arc projection.

[0098] The process of constructing the sensitivity matrix is ​​as follows:

[0099] Reading the signal matrix :

[0100] If x = y, proceed to the next step;

[0101] If x≠y and Proceed to the next step;

[0102] If x≠y and Read the reflected signal Based on the minimum value of all TOF data received by the self-transmitting and self-receiving ultrasonic sensor, the positioning arc of the outer contour of the particle system is determined, and the radius of the arc is:

[0103] (15)

[0104] The distance between the ultrasonic sensor and the detection area is greater than For any node, its corresponding It was set to 1.

[0105] All arcs are cumulatively weighted. This is due to the unit grid weighting based on the triangle rule used in the transmission method reconstruction. The value is 1, to balance the weighting effect of the TOF arc received by the ultrasonic sensor itself during superposition, the unit weight of the arc projection is 1. Set as , and These represent the number of effective transmitted signals and the number of reflected signals, respectively.

[0106] The initial low-frequency ultrasound image is binarized and segmented. Then, a filtering algorithm (such as Gaussian filtering or other filtering algorithms) is used to denoise the image. Next, bubble closure is performed. Based on the elliptic kernel, a closing operation is performed. The bubble is first expanded and then eroded to fill the gap inside the bubble contour, connect adjacent contours, and close the bubble edge, thereby restoring the bubble image to its complete state.

[0107] In one embodiment, this embodiment provides an implementation method for determining the target frequency that enables the sound waves of each ultrasonic sensor to penetrate each ultrasonic imaging path based on low-frequency ultrasonic images. Specifically, the following steps can be followed:

[0108] Step (1): Based on the low-frequency ultrasound image, determine the frequency corresponding to the maximum ultrasound attenuation on the ultrasound path of each ultrasound sensor, and select the lowest frequency from the frequencies corresponding to the maximum ultrasound attenuation on each ultrasound path.

[0109] For each ultrasound path corresponding to low-frequency ultrasound imaging, calculate the highest frequency corresponding to the maximum allowable ultrasound attenuation on each ultrasound path, and select the lowest frequency from all ultrasound paths.

[0110] In one specific implementation, the bubble length and liquid metal length on each ultrasonic path are obtained from the low-frequency ultrasonic image; based on the bubble length and liquid metal length on each ultrasonic path, the frequency corresponding to when the attenuated sound pressure amplitude on each ultrasonic path is greater than the ambient noise is determined.

[0111] The calculation methods for attenuation coefficients at different frequencies are as follows:

[0112] Different prediction models can be selected for calculations of different liquid metals, such as the ultrasonic attenuation coefficient below 5 MHz in lead-bismuth liquid metal. The following formula can be used for prediction:

[0113] (16)

[0114] in, For dynamic viscosity, For density, For the speed of sound, Let be the frequency of the i-th ultrasound path.

[0115] Ultrasonic attenuation coefficient of air The following formula can be used for prediction:

[0116] (17)

[0117] in, For air temperature, Reference temperature (25℃). Air pressure. Standard atmospheric pressure (101.325 kPa).

[0118] Based on the low-frequency ultrasound images obtained from the above steps, OpenCV is used to statistically determine the bubble lengths along each ultrasound path i. and liquid metal length In the processed low-frequency ultrasound image, the grayscale value of the bubble's edge and interior is 0, while the grayscale value of the liquid metal is 255. The length of the aforementioned bubble... and liquid metal length The number of pixels below a grayscale threshold (e.g., 127) can be counted. To obtain, that is:

[0119] (18)

[0120] in, This represents the total number of pixels along ultrasound path i. This refers to the pipe diameter.

[0121] Then at the new frequency Below, the attenuated sound pressure amplitude received by the receiving probe of ultrasonic path i is

[0122] (19)

[0123] Select the attenuated sound pressure amplitude For any suitable value higher than the ambient noise, by combining the above formulas (16-19), the frequency corresponding to the maximum ultrasonic attenuation on each ultrasonic path can be calculated. The lowest frequency is selected from the frequencies corresponding to the maximum ultrasonic attenuation of each ultrasonic path. This ensures that the acoustic signals emitted during high-frequency imaging can penetrate all ultrasound paths.

[0124] Step (2): Obtain the set frequency corresponding to the set minimum resolvable bubble radius that the ultrasonic sensor can detect;

[0125] Determine the minimum resolvable bubble radius required, and calculate the set frequency corresponding to the set minimum resolvable bubble radius based on the above formula (1) to improve the resolution so that bubbles of different sizes in the channel to be detected can be imaged.

[0126] Step (3): The minimum value between the lowest frequency and the set frequency is taken as the target frequency that enables the sound waves of each ultrasonic sensor to penetrate each ultrasonic imaging path.

[0127] Based on the requirement of the highest resolution, a suitable ultrasound frequency (i.e., the target frequency) is determined for high-frequency imaging. Target frequency The calculation formula is:

[0128] (20)

[0129] In one embodiment, this embodiment provides a specific implementation method for constructing a high-frequency coefficient matrix based on the reflected and transmitted signals received by each ultrasonic sensor: the high-frequency coefficient matrix is ​​reconstructed using a dual-mode fusion reconstruction algorithm based on the reflected and transmitted signals received by each ultrasonic sensor.

[0130] Based on the reflected and transmitted signals received by each ultrasonic sensor during high-frequency imaging, a dual-mode coefficient matrix is ​​constructed using a dual-mode fusion reconstruction algorithm, denoted as the high-frequency coefficient matrix. That is, the high-frequency coefficient matrix can be calculated using the above formulas (13) to (14).

[0131] In one embodiment, this embodiment provides a specific implementation method for fusing low-frequency coefficient matrices and high-frequency coefficient matrices to obtain a fused coefficient matrix: an adaptive gradient weight allocation algorithm is used to adaptively fuse the low-frequency coefficient matrix and the high-frequency coefficient matrix to obtain a fused coefficient matrix.

[0132] The channel to be tested is divided, and the low-frequency coefficient matrix and the high-frequency coefficient matrix are dynamically weighted and fused using an adaptive gradient weighting method to obtain the fused coefficient matrix.

[0133] Taking the rectangular subdivision method as an example, the space to be measured is subdivided into horizontal and vertical subdivisions. The horizontal gradient is calculated using the Sobel operator for each pixel. and vertical gradient And calculate the gradient magnitude:

[0134] (twenty one)

[0135] Then, dynamic weight allocation is performed, which allocates weights based on the ratio of low-frequency gradient magnitude to high-frequency gradient magnitude at a grid point.

[0136] (twenty two)

[0137] To avoid calculation errors, Take the minimum value, such as , For low-frequency gradient amplitude, For high-frequency gradient amplitude, The weights of the low-frequency coefficient matrix, The weights are the high-frequency coefficient matrix.

[0138] Pixel fusion: Dynamic gradient-weighted fusion is performed on the images corresponding to the low-frequency coefficient matrix and the high-frequency coefficient matrix, and the fusion coefficient matrix is ​​calculated. :

[0139] (twenty three)

[0140] in, It is a low-frequency coefficient matrix. It is a high-frequency coefficient matrix.

[0141] In one embodiment, this embodiment provides a specific implementation method for obtaining a multi-frequency fused two-phase distribution map by image reconstruction based on the fusion coefficient matrix: an initial fused image is reconstructed based on the fusion coefficient matrix; the initial fused image is denoised, and the denoised initial fused image is bubble-closing processed to restore the bubble image to complete, thereby obtaining a multi-frequency fused two-phase distribution map.

[0142] Fusion coefficient matrix To fuse the weight distribution of the images, a dual-mode fusion reconstruction algorithm is used through the fusion coefficient matrix. An initial fused image is obtained through imaging; the initial fused image is binarized and segmented, and then a filtering algorithm (such as Gaussian filtering or other filtering algorithms) is used to denoise the image. Then, bubble closure processing is performed, and a closing operation is performed based on the elliptic kernel. The bubble is first expanded and then eroded to fill the gaps inside the bubble contour, connect adjacent contours, close the bubble edges, and finally obtain a two-phase distribution map of multi-frequency fusion.

[0143] In one implementation, in order to prevent the acoustic signal excited by the previous probe from interfering with the acoustic signal excited by the next probe, when sequentially exciting each sensor, it is necessary to predetermine the time interval between exciting two adjacent ultrasonic sensors.

[0144] In the steps of sequentially exciting low-frequency ultrasound and sequentially exciting high-frequency ultrasound for each ultrasonic sensor, the excitation time interval for each ultrasonic sensor is:

[0145] (twenty four)

[0146] The cross-section of the channel to be tested is circular. Where is the cross-sectional diameter of the channel to be tested. The speed of sound in liquid metal. The speed of sound in air. The gas content of the cross section, This represents the number of ultrasonic sensors.

[0147] For reflection imaging, the system should be designed so that, in the absence of air bubbles, the first echo signal at the farthest distance (i.e., diameter) will not cause interference. For transmission imaging, this interval should be greater than the time it takes for the sensor to receive a single transmission signal. Due to the significant difference in sound velocity between air and liquid metal, the gas content of the cross-section should be considered when considering the transmission method. Therefore, the excitation time interval for each ultrasonic sensor should be the maximum value of the time required to receive the feedback signal in both the reflection and transmission methods.

[0148] For example, for a circular pipe with a diameter of 50 mm, using 24 ultrasonic probes evenly arranged on the circumference of the outer side of the pipe cross-section, with a lead-bismuth-air system at 400°C (the sound velocity of lead-bismuth is approximately 1700 m / s, and the sound velocity of air is approximately 520 m / s), the total measurement time determined by the reflection method is 1.4 ms. However, under extremely high gas content conditions (gas content greater than 60%), the measurement time, constrained by the transmission principle, becomes dominant. This situation is very rare in the reactor field. Therefore, for cases where the bubbles are the dispersed phase, the total time for exciting low-frequency ultrasound is... It can be determined by the following formula:

[0149] (25)

[0150] The multi-frequency fusion-based dual-modal ultrasonic tomography method for reflection and transmission provided in this embodiment, by employing multi-frequency ultrasonic imaging, ensures the capture of discrete bubbles with the smallest resolution size, while also improving the imaging effect on multiple bubbles and bubble swarms. Furthermore, the dual-modal imaging provided in this embodiment does not simply fuse low-frequency and high-frequency imaging results; instead, it uses the low-frequency imaging results to determine the frequency for the next high-frequency imaging step, ensuring that the high-frequency sound waves can penetrate the entire test field and, as far as possible, achieve the minimum resolution requirement, thus improving imaging accuracy. It can be applied to the fusion of more frequencies, such as adding an intermediate frequency to form a three-frequency fused dual-modal ultrasonic tomography for reflection and transmission. Using the intermediate frequency f3 for dual-modal ultrasonic tomography for reflection and transmission yields the coefficient matrix. V mid ( m,n );Will V low ( m,n ), V mid ( m,n )and V high ( m,n The fusion coefficient matrix is ​​obtained by fusion. V fused ( m,n The multi-frequency fused image is then reconstructed.

[0151] Based on the foregoing embodiments, this embodiment provides an example of using the aforementioned multi-frequency fusion-based reflection-transmission dual-modal ultrasonic tomography method to perform ultrasonic tomography on a two-phase flow consisting of liquid metal and bubbles, as shown in the example below. Figure 2 The flowchart shown illustrates the ultrasonic tomography process for a two-phase flow consisting of liquid metal and bubbles. The specific steps are as follows:

[0152] Step 201: Determine the initial frequency for low-frequency imaging based on the predicted highest cross-sectional gas content. f 1. Determine the set frequency based on the set maximum resolution. f 0;

[0153] At the predicted highest cross-sectional gas content, the frequency corresponding to the attenuated sound pressure amplitude received by the ultrasonic sensor being higher than the amplitude of the ambient noise is determined as the initial frequency.

[0154] The minimum resolvable bubble radius is determined based on the set maximum resolution, and the set frequency corresponding to the set minimum resolvable bubble radius is calculated based on the above formula (1).

[0155] Step 202: Sequentially excite low-frequency ultrasound to each ultrasonic sensor, and perform reflection-projection dual-mode tomography at the initial frequency to obtain the low-frequency coefficient matrix and the initial low-frequency ultrasound image.

[0156] Step 203: Binarize and block bubble process the initial low-frequency ultrasound image to obtain a low-frequency ultrasound image (i.e., a low-frequency two-phase distribution map).

[0157] The initial low-frequency ultrasound image is binarized and the interior of the bubble is sealed to obtain the low-frequency ultrasound image. Specifically, a Python script calls OpenCV for optimal binarization segmentation, Gaussian filtering is used to remove noise, and a closing operation is performed based on the elliptic kernel. Dilation followed by erosion is then used to fill the gap inside the contour, connect adjacent contours, and close the edges.

[0158] Step 204: Calculate the frequencies at which the sound waves from each ultrasonic sensor can just penetrate each ultrasonic imaging path based on the attenuation effect, and select the lowest frequency from the frequencies corresponding to the maximum ultrasonic attenuation of each ultrasonic path. ;

[0159] Step 205: Modulate to the target frequency (i.e., the new frequency) using a step frequency modulation signal. ;

[0160] Step 206, at the target frequency A reflection-projection dual-modal tomography was performed to obtain the high-frequency coefficient matrix;

[0161] Step 207: Adaptively fuse the low-frequency coefficient matrix and the high-frequency coefficient matrix to obtain the fusion coefficient matrix and reconstruct the multi-frequency fusion image;

[0162] Step 208: Binarize and close the bubble in the multi-frequency fusion image to obtain the two-phase distribution map of the multi-frequency fusion.

[0163] The multi-frequency fusion method provided in this embodiment improves upon existing dual-modal imaging methods to enhance imaging accuracy. Furthermore, by employing adaptive frequency, it can flexibly adapt to different measurement domain sizes and varying gas content across different cross-sections. This method is applicable to all liquid metal-gas systems, not just the lead-bismuth-air system described above. The method is applicable to any symmetrically arranged ultrasonic tomography system, not limited to circular pipes, but also applicable to rectangular channels.

[0164] Corresponding to the multi-frequency fusion-based dual-modal ultrasound tomography method provided in the above embodiments, this invention provides a multi-frequency fusion-based dual-modal ultrasound tomography system, including: a controller and multiple ultrasound sensors, the controller including a processor and a storage device; the storage device stores a computer program, and the computer program executes the multi-frequency fusion-based dual-modal ultrasound tomography method provided in the above embodiments when run by the processor.

[0165] In one embodiment, the plurality of ultrasonic sensors are uniformly disposed on the outer surface of the channel to be tested, the cross-section of the channel to be tested is an axisymmetric figure, and the plurality of ultrasonic sensors are distributed in an axisymmetric manner with respect to the cross-section of the channel to be tested.

[0166] The system provided in this embodiment has the same implementation principle and technical effects as the aforementioned embodiments. For the sake of brevity, any parts not mentioned in the system embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0167] This invention provides a computer-readable medium storing computer-executable instructions. When these computer-executable instructions are invoked and executed by a processor, they cause the processor to implement the methods described in the above embodiments.

[0168] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system described above can be referred to the corresponding process in the foregoing embodiments, and will not be repeated here.

[0169] Furthermore, in the description of the embodiments of the present invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention based on the specific circumstances.

[0170] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0171] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0172] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A dual-modal ultrasound tomography method based on multi-frequency fusion of reflection and transmission modes, characterized in that, An application is made to a multi-frequency fusion-based dual-modal ultrasound tomography system, wherein the multi-frequency fusion-based dual-modal ultrasound tomography system includes multiple ultrasound sensors, and the multi-frequency fusion-based dual-modal ultrasound tomography method includes: Each of the ultrasonic sensors is sequentially excited with low-frequency ultrasound, so that each of the ultrasonic sensors performs reflection-transmission dual-modal tomography at an initial frequency. A low-frequency coefficient matrix and a low-frequency ultrasound image are constructed based on the reflected and transmitted signals received by each of the ultrasonic sensors. The initial frequency is the frequency at which the sound waves of each ultrasonic sensor can penetrate each ultrasound imaging path when the cross-sectional gas content of the channel to be detected is the highest predicted cross-sectional gas content. Based on the low-frequency ultrasound images, a target frequency is determined that enables the sound waves of each of the ultrasound sensors to penetrate each ultrasound imaging path. Each of the ultrasonic sensors is sequentially excited with high-frequency ultrasound, so that each of the ultrasonic sensors performs reflection-transmission dual-modal tomography at the target frequency, and a high-frequency coefficient matrix is ​​constructed based on the reflected and transmitted signals received by each of the ultrasonic sensors. The low-frequency coefficient matrix and the high-frequency coefficient matrix are fused to obtain a fusion coefficient matrix. Based on the fusion coefficient matrix, image reconstruction is performed to obtain a two-phase distribution map of multi-frequency fusion. The step of determining the target frequency that enables the sound waves of each ultrasonic sensor to penetrate each ultrasonic imaging path based on the low-frequency ultrasonic image includes: Based on the low-frequency ultrasound image, determine the frequency corresponding to the maximum ultrasound attenuation on the ultrasound path of each ultrasound sensor, and select the lowest frequency from the frequencies corresponding to the maximum ultrasound attenuation on each ultrasound path; obtain the set frequency corresponding to the set minimum resolvable bubble radius that the ultrasound sensor can detect; and use the minimum value between the lowest frequency and the set frequency as the target frequency that allows the sound waves of each ultrasound sensor to penetrate each ultrasound imaging path. The step of fusing the low-frequency coefficient matrix and the high-frequency coefficient matrix to obtain a fused coefficient matrix includes: An adaptive gradient weight allocation algorithm is used to adaptively fuse the low-frequency coefficient matrix and the high-frequency coefficient matrix to obtain the fused coefficient matrix.

2. The method according to claim 1, characterized in that, The step of constructing a low-frequency coefficient matrix and a low-frequency ultrasound image based on the reflected and transmitted signals received by each of the ultrasonic sensors includes: The low-frequency coefficient matrix and the initial low-frequency ultrasound image are reconstructed using a dual-mode fusion reconstruction algorithm based on the reflected and transmitted signals received by each of the ultrasonic sensors. The initial low-frequency ultrasound image is denoised, and bubble sealing is performed on the denoised initial low-frequency ultrasound image to restore the bubble image to complete, thus obtaining the low-frequency ultrasound image.

3. The method according to claim 1, characterized in that, The step of determining the frequency corresponding to the maximum ultrasonic attenuation on the ultrasonic path of each ultrasonic sensor based on the low-frequency ultrasonic image includes: The lengths of the bubbles and the liquid metal along each ultrasonic path are obtained from the low-frequency ultrasonic images. The frequency at which the attenuated sound pressure amplitude on each ultrasonic path is greater than the ambient noise is determined based on the bubble length and liquid metal length on each ultrasonic path.

4. The method according to claim 1, characterized in that, The step of constructing a high-frequency coefficient matrix based on the reflected and transmitted signals received by each of the ultrasonic sensors includes: The high-frequency coefficient matrix is ​​reconstructed using a dual-mode fusion reconstruction algorithm based on the reflected and transmitted signals received by each ultrasonic sensor.

5. The method according to claim 1, characterized in that, The step of reconstructing the image based on the fusion coefficient matrix to obtain a multi-frequency fused two-phase distribution map includes: The initial fused image is reconstructed based on the fusion coefficient matrix; The initial fused image is denoised, and the denoised initial fused image is bubble-closing processed to restore the bubble image to complete, thus obtaining the two-phase distribution map of the multi-frequency fusion.

6. The method according to claim 1, characterized in that, The excitation signal of the ultrasonic sensor is a pulsed step frequency modulation signal; In the steps of sequentially exciting low-frequency ultrasound to each of the ultrasonic sensors and sequentially exciting high-frequency ultrasound to each of the ultrasonic sensors, the excitation time interval between each ultrasonic sensor is: ; The cross-section of the channel to be detected is circular. The cross-sectional diameter of the channel to be tested is [missing information]. The speed of sound in liquid metal. The speed of sound in air. The gas content of the cross section.

7. A dual-modal ultrasound tomography system based on multi-frequency fusion, characterized in that, include: A controller and multiple ultrasonic sensors, the controller including a processor and a storage device; The storage device stores a computer program that, when executed by the processor, performs the method as described in any one of claims 1 to 6.

8. The system according to claim 7, characterized in that, Multiple ultrasonic sensors are evenly disposed on the outer surface of the channel to be tested. The cross-section of the channel to be tested is an axisymmetric figure, and the multiple ultrasonic sensors are distributed in an axisymmetric manner with respect to the cross-section of the channel to be tested.

Images