Ultrasonic guided wave mode tracking method and apparatus, electronic device, and storage medium

Abstract

The present disclosure relates to an ultrasonic guided wave mode tracking method and device, electronic equipment and storage medium. The method comprises: performing finite element meshing on a waveguide to form a plurality of nodes; obtaining ultrasonic wave characteristic data of each node in the plurality of nodes under different guided wave modes and different frequencies; determining target ultrasonic wave characteristic data of all eigenvalues in a target frequency range based on the ultrasonic wave characteristic data of each node under different guided wave modes and different frequencies; determining a first cosine distance between any two eigenvalues based on the target ultrasonic wave characteristic data of each eigenvalue; obtaining a similarity matrix based on all first cosine distances and a nearest neighbor propagation algorithm; performing clustering processing on all eigenvalues based on the similarity matrix to obtain at least one class; and each class corresponds to an ultrasonic guided wave mode. The technical solution provided by the present disclosure can improve the tracking accuracy of different ultrasonic guided wave modes in a complex waveguide.

Description

Technical Field

[0001] This disclosure relates to the field of nondestructive testing technology, and in particular to an ultrasonic guided wave mode tracking method, apparatus, electronic device, and storage medium. Background Technology

[0002] Non-destructive testing technology based on ultrasonic guided waves has attracted widespread attention in industry due to its excellent characteristics such as economy, high efficiency, and sensitivity to defects. Using guided waves for structural health inspection / monitoring requires a thorough understanding of the dispersion relationships of different ultrasonic guided wave modes within the structure. The dispersion relationship refers to the relationship between frequency f and wavenumber k. This can be achieved through the phase velocity calculation formula. The attenuation calculation formula α = Im(k) yields the phase velocity spectrum and attenuation coefficient spectrum of the same ultrasonic guided wave mode at different frequencies. This information on wave velocity variations and attenuation helps determine the required ultrasonic frequency range and identify the most suitable ultrasonic guided wave mode for material testing. Therefore, obtaining the dispersion relationships of different ultrasonic guided wave modes is a crucial foundation for structural testing.

[0003] Taking a flat plate structure as an example, when it is subjected to ultrasonic excitation, the waveguide types inside the flat plate structure are A, S, and SH. Figure 1 This is a dispersion curve obtained from an ultrasonically excited flat plate structure. Figure 1 As can be seen, the dispersion relationships of different waveguide modes vary significantly. Some waveguide modes, such as the SH0 mode, do not exhibit dispersion; while others, such as the SH1 mode, show significant dispersion in specific frequency ranges. These dispersion curves provide important guidance for selecting specific excitation waveguides.

[0004] Currently, apart from simple structures like flat plates and circular tubes where analytical solutions for different ultrasonic guided wave modes can be calculated, complex structures (such as railway tracks, welds, or irregularly shaped beams) cannot have their ultrasonic guided wave mode dispersion relationships obtained analytically. Therefore, finite element method (FEM) approaches are widely used to calculate the guided wave modes and dispersion relationships of complex structures. There are currently two main categories of FEM-based methods for identifying ultrasonic guided wave modes and their dispersion relationships. One is manual clustering and identification of dispersion relationships; the other is algorithmic analysis using the curvature of the dispersion relationship for point-by-point searching.

[0005] Figure 2 This is a true dispersion curve of an ultrasonically excited object with a complex structure. From... Figure 2 It can be seen that the dispersion relation of objects with complex structures may fall into at least one of the following three categories: First, such as... Figure 2 In the case defined by A, there is overlap between different waveguide modes; secondly, as... Figure 2In the case defined by B, the data points overlap between different waveguide modes. Thirdly, as... Figure 2 In the case defined by C, some guided wave modes are discontinuous, meaning that some guided wave modes only appear within a specific frequency range.

[0006] Since both manual clustering and identification of dispersion relationships and algorithmic search based on the curvature of dispersion relationships are based on superficial analysis, their accuracy in identifying the above three situations is relatively low. Summary of the Invention

[0007] To solve the above-mentioned technical problems, or at least partially solve them, this disclosure provides an ultrasonic guided wave mode tracking method, apparatus, electronic device, and storage medium.

[0008] In a first aspect, this disclosure provides an ultrasonic guided wave mode tracking method, including:

[0009] The waveguide is divided into finite element meshes to form multiple nodes;

[0010] Acquire ultrasonic dynamic characteristic data of each of the plurality of nodes under different waveguide modes and frequencies;

[0011] Based on the ultrasonic dynamic characteristic data of each node at different frequencies under different guided wave modes, the target ultrasonic dynamic characteristic data of all intrinsic solutions within the target frequency band are determined; the guided wave modes and / or frequencies of different intrinsic solutions are different.

[0012] Based on the target ultrasonic dynamic characteristic data of each of the intrinsic solutions, determine the first cosine distance between any two of the intrinsic solutions;

[0013] Based on all the first cosine distances and the nearest neighbor propagation algorithm, a similarity matrix is ​​obtained;

[0014] Based on the similarity matrix, all intrinsic solutions are clustered to obtain at least one class; each class corresponds to an ultrasonic guided wave mode.

[0015] Secondly, this disclosure also provides an ultrasonic guided wave mode tracking device, comprising:

[0016] The node determination module is used to divide the waveguide into finite element meshes to form multiple nodes;

[0017] An ultrasonic dynamic characteristic data determination module is used to acquire ultrasonic dynamic characteristic data of each of the plurality of nodes under different guided wave modes and frequencies.

[0018] The target ultrasonic dynamic characteristic data determination module is used to determine the target ultrasonic dynamic characteristic data of all intrinsic solutions within the target frequency band based on the ultrasonic dynamic characteristic data of each node at different guided wave modes and frequencies; the guided wave modes and / or frequencies of different intrinsic solutions are different.

[0019] The cosine distance determination module is used to determine the first cosine distance between any two of the intrinsic solutions based on the target ultrasonic dynamic feature data of each of the intrinsic solutions.

[0020] The nearest neighbor propagation module is used to obtain a similarity matrix based on all the first cosine distances and the nearest neighbor propagation algorithm;

[0021] The clustering module is used to cluster all intrinsic solutions based on the similarity matrix to obtain at least one class; each class corresponds to an ultrasonic guided wave mode.

[0022] Thirdly, this disclosure also provides an electronic device, including: a processor and a memory;

[0023] The processor executes the steps of any of the above methods by calling programs or instructions stored in memory.

[0024] Fourthly, this disclosure also provides a computer-readable storage medium that stores a program or instructions that cause a computer to perform the steps of any of the above methods.

[0025] The technical solution provided in this disclosure has the following advantages compared with the prior art:

[0026] The technical solution provided in this disclosure involves setting up ultrasonic dynamic characteristic data of each node at different frequencies under different guided wave modes to determine the target ultrasonic dynamic characteristic data of all intrinsic solutions within the target frequency band; different intrinsic solutions have different guided wave modes and / or frequencies; based on the target ultrasonic dynamic characteristic data of each intrinsic solution, the first cosine distance between any two intrinsic solutions is determined; based on all first cosine distances and the nearest neighbor propagation algorithm, a similarity matrix is ​​obtained; based on the similarity matrix, all intrinsic solutions are clustered to obtain at least one class; each class corresponds to one ultrasonic guided wave mode. Essentially, this involves a deep analysis of the mode shape characteristics of any two intrinsic solutions among multiple intrinsic solutions to determine their correlation. This can solve the problem of low clustering accuracy caused by the overlap between different guided wave modes, data point coincidence, and discontinuity in some guided wave modes, which leads to clustering multiple intrinsic solutions based on appearance alone. This achieves the goal of improving the clustering accuracy of intrinsic solutions. Attached Figure Description

[0027] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0028] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, those skilled in the art can obtain other drawings based on these drawings without creative effort.

[0029] Figure 1 The dispersion curve obtained by ultrasonically excited plate structure;

[0030] Figure 2 This is a true dispersion curve of an ultrasonically excited object with a complex structure.

[0031] Figure 3 A flowchart of an ultrasonic guided wave mode tracking method provided in this embodiment of the disclosure;

[0032] Figure 4 This is a schematic diagram of a waveguide after being divided according to an embodiment of the present disclosure;

[0033] Figure 5 This is a schematic diagram illustrating the decomposition of a particle's displacement, angular momentum, and average energy flux density at a frequency in a guided wave mode.

[0034] Figure 6 A schematic diagram illustrating the principle of the nearest neighbor relationship determination method provided in this embodiment of the disclosure;

[0035] Figure 7 The dispersion curve of the railway track obtained by using the ultrasonic guided wave mode tracking method provided in the embodiments of this disclosure;

[0036] Figure 8 This is a schematic diagram of the structure of an ultrasonic guided wave mode tracking device according to an embodiment of the present disclosure;

[0037] Figure 9 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this disclosure. Detailed Implementation

[0038] To better understand the above-mentioned objectives, features, and advantages of this disclosure, the solutions disclosed herein will be further described below. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0039] Numerous specific details are set forth in the following description in order to provide a full understanding of this disclosure, but this disclosure may also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only some, and not all, of the embodiments of this disclosure.

[0040] Figure 3 A flowchart illustrating an ultrasonic guided wave mode tracking method provided in this embodiment of the disclosure. See also... Figure 3 The ultrasonic guided wave mode tracking method includes:

[0041] S110. Divide the waveguide into a finite element mesh to form multiple nodes.

[0042] A waveguide is a medium for wave propagation. In this application, the waveguide can be an object requiring ultrasonic excitation. In some embodiments, the waveguide is an object to be inspected using nondestructive testing.

[0043] For example, in one scenario, it is desirable to use ultrasonic excitation to investigate whether there are morphological defects such as slag inclusions, porosity, cracks, incomplete fusion / penetration, pits, undercut, and weld beads in a metal welded component, where the metal welded component is a waveguide.

[0044] The essence of "dividing the waveguide into a finite element mesh to form multiple nodes" is to select a large number of nodes at different locations in the waveguide and use the wave equation solutions of these nodes to characterize the wave characteristics of ultrasonic waves propagating in the waveguide.

[0045] There are various ways to implement this step, and this application does not limit this one. For example, the implementation method of this step includes: using a mesh generation method in topological geometry to perform finite element mesh generation on the waveguide, obtaining multiple meshes, and using the intersection of two adjacent meshes as nodes. Optionally, the shape of each mesh can be triangular, quadrilateral, pentagonal, or hexagonal, etc.

[0046] This step can also include: dividing the target cross-section of the waveguide into a finite element mesh to form multiple nodes. This application does not restrict the specific location of the target cross-section within the waveguide. In practice, the specific location of the target cross-section can be determined based on the research objective. The purpose of this setup is to analyze the ultrasonic guided wave mode using the target cross-section as a representative example.

[0047] Figure 4 This is a schematic diagram of a waveguide target cross section after finite element meshing, provided as an embodiment of this disclosure. Figure 4 In this design, the waveguide is represented by a railway track. The cross-section of the waveguide target is divided into multiple triangles, and the intersection of any two adjacent triangles is a node. Figure 4 In the image, image E is an enlarged view of the area enclosed by the white frame.

[0048] S120. Obtain ultrasonic dynamic characteristic data of each of the multiple nodes under different guided wave modes and frequencies.

[0049] This application does not limit the specific content of the ultrasonic dynamic characteristic data of each node at different frequencies in different waveguide modes. For example, the ultrasonic dynamic characteristic data of each node at each frequency in each waveguide mode includes: particle displacement, particle vibration velocity, and average energy flux density. Here, "particle displacement" should be understood as the displacement of the particle corresponding to the node, and "particle vibration velocity" should be understood as the vibration velocity of the particle corresponding to the node. The particles are those that constitute the waveguide.

[0050] There are various ways to implement this step, and this application does not limit this one. For example, the implementation method of this step includes: calculating the ultrasonic dynamic characteristic data of each of the multiple nodes at different frequencies in different guided wave modes using finite element numerical methods. Finite element numerical methods include boundary element numerical methods and / or semi-analytical finite element numerical methods.

[0051] It is important to emphasize that, for each node, the finite element method can yield multiple particle displacements, multiple particle velocities, and multiple average energy flux densities. Furthermore, it can determine which particle displacement, particle velocity, and average energy flux density have a corresponding relationship, and these corresponding particle displacements, velocities, and average energy flux densities correspond to the same mode and frequency. However, it cannot determine the specific mode corresponding to each particle displacement, particle velocity, or average energy flux density, but it can determine the specific frequency corresponding to each particle displacement, particle velocity, or average energy flux density.

[0052] S130. Based on the ultrasonic dynamic characteristic data of each node at different frequencies under different guided wave modes, determine the target ultrasonic dynamic characteristic data of all intrinsic solutions within the target frequency band; the guided wave modes and / or frequencies of different intrinsic solutions are different.

[0053] The eigenvalue, also known as the guided wave eigenvalue, corresponds to only one frequency-mode point in the phase velocity spectrum. In other words, the eigenvalue can be used as a data point to be plotted in a dispersion relation diagram, serving as the data point for constructing the dispersion relation curve.

[0054] It should be noted that each eigensol has a unique guided wave mode and a unique frequency. However, when performing this step using the technical solution provided in this application, the specific guided wave mode of each eigensol is unknown.

[0055] The target frequency band is the frequency band that researchers are interested in, and this application does not limit the specific frequency range that the target frequency band refers to.

[0056] The target ultrasonic dynamic characteristic data of the intrinsic solution is the result of reorganizing the ultrasonic dynamic characteristic data of all nodes at different frequencies under different guided wave modes according to the guided wave mode and frequency. For example, ultrasonic dynamic characteristic data with the same guided wave mode and frequency are grouped together as the target ultrasonic dynamic characteristic data of the intrinsic solution corresponding to that guided wave mode and frequency.

[0057] There are various ways to implement this step, and this application does not limit them. In one embodiment, the ultrasonic dynamic characteristic data of each node at each frequency in each guided wave mode includes: particle displacement, particle vibration velocity, and average energy flux density. The implementation method of this step includes: constructing a rectangular coordinate system, with the waveguide located in the rectangular coordinate system; converting the particle vibration velocity of each node at each frequency in each guided wave mode into the angular momentum of each node relative to the origin of the rectangular coordinate system; decomposing the particle displacement, particle angular momentum, and average energy flux density of each node at each frequency in each guided wave mode to obtain the displacement components, angular momentum components, and average energy flux density components of each node at each frequency in each guided wave mode on each coordinate axis of the rectangular coordinate system; summarizing the displacement components, angular momentum components, and average energy flux density components of each node at the same frequency in the same guided wave mode to obtain the target ultrasonic dynamic characteristic data of each eigensol.

[0058] For example, suppose that N nodes are determined in S110. After exciting the waveguide with ultrasound, for each node, M particle displacements, M particle vibration velocities, and M average energy flux densities can be obtained. These M particle displacements, M particle vibration velocities, and M average energy flux densities have M corresponding relationships. Each correspondence includes one particle displacement, one particle vibration velocity, and one average energy flux density. The corresponding particle displacements, particle vibration velocities, and average energy flux densities correspond to the same guided wave mode and the same frequency. In other words, there are a total of M eigensols. In S120, a data matrix of size M×N×3 will be obtained.

[0059] Alternatively, a formula can be used. The particle vibration velocity of each node at each frequency in each waveguide mode is converted into the angular momentum of each node relative to the origin of the Cartesian coordinate system. Angular momentum; The position vector of a node is the vector pointing from the origin of the rectangular coordinate system to the node. denoted as , where is the particle vibration velocity at the node.

[0060] Assume there are N nodes and M eigensols. Since each eigensol corresponds to only one frequency-mode point in the phase velocity spectrum, there are a total of M frequency-mode points in the phase velocity spectrum. Number each eigensol, and let the particle displacement of the j-th node under the m-th eigensol be U. m,j Angular momentum L m,j and average energy flux density P m,j . Figure 5 This is a schematic diagram illustrating the decomposition of a particle's displacement, angular momentum, and average energy flux density at a frequency in a guided wave mode. (See also...) Figure 5 After constructing a Cartesian coordinate system in the space where the waveguide is located, U can be further... m,j Decompose the components onto the coordinate axes of a rectangular coordinate system to obtain... as well as L m,j Decompose the components onto the coordinate axes of a rectangular coordinate system to obtain... as well as P m,j Decompose the components onto the coordinate axes of a rectangular coordinate system to obtain... as well as By summing the displacement components, angular momentum components, and average energy flux density components of each node at the m-th guided wave mode-frequency, the target ultrasonic dynamic characteristic data of the m-th eigensol are obtained. The target ultrasonic dynamic characteristic data of the m-th eigensol can be expressed as:

[0061]

[0062] in,

[0063] as well as Both are N×1 arrays.

[0064] It should be noted that this application selects three characteristic quantities—particle displacement, angular momentum, and average energy flux density—after fully considering the physical background of the ultrasonic guided wave problem. This selection method is beneficial for a deeper analysis of the mode shape characteristics of the two intrinsic solutions, thereby achieving the goal of improving the clustering accuracy of the intrinsic solutions.

[0065] S140. Based on the target ultrasonic dynamic characteristic data of each eigensol, determine the first cosine distance between any two eigensoles.

[0066] There are various ways to implement this step, and this application does not limit this one. For example, the implementation method of this step includes: calculating the second cosine distance between the target displacement components, the second cosine distance between the target angular momentum components, and the second cosine distance between the target average energy flux density components of any two intrinsic solutions under each coordinate axis, based on the target ultrasonic dynamic characteristic data of any two intrinsic solutions; determining the weight values ​​of the target displacement components, the target angular momentum components, and the target average energy flux density components of any two intrinsic solutions under each coordinate axis, based on the target ultrasonic dynamic characteristic data of any two intrinsic solutions; and determining the first cosine distance between any two intrinsic solutions based on the second cosine distance between the target displacement components, the target angular momentum components, and the target average energy flux density components of any two intrinsic solutions under each coordinate axis, as well as the weight values ​​of the target displacement components, the target angular momentum components, and the target average energy flux density components of any two intrinsic solutions under each coordinate axis.

[0067] There are multiple methods for "calculating the second cosine distance of the target displacement component, the second cosine distance of the target angular momentum component, and the second cosine distance of the target average energy flux density component based on the target ultrasonic dynamic characteristic data of any two intrinsic solutions under each coordinate axis," and this application does not limit this method. For example, a specific implementation method may include: calculating the third cosine distance of the target displacement component, the third cosine distance of the target angular momentum component, and the third cosine distance of the target average energy flux density component based on the target ultrasonic dynamic characteristic data of any two intrinsic solutions and the definition of cosine distance under each coordinate axis; if the third cosine distance of the target displacement component of two intrinsic solutions under a coordinate axis is greater than or equal to 0, and less than or equal to... The third cosine distance is taken as the second cosine distance between the target displacement components of two eigensoles under the coordinate axis; if the third cosine distance between the target displacement components of any two eigensoles under a coordinate axis is greater than... Calculate the difference between π and the third cosine distance, and use this difference as the second cosine distance between the target displacement components of the two eigensoles on the coordinate axis; if the third cosine distance between the target angular momentum components of the two eigensoles on a coordinate axis is greater than or equal to 0, and less than or equal to... The third cosine distance is taken as the second cosine distance between the target angular momentum components of two eigensoles under the coordinate axis; if the third cosine distance between the target angular momentum components of any two eigensoles under a coordinate axis is greater than... Calculate the difference between π and the third cosine distance, and use this difference as the second cosine distance between the target angular momentum components of the two eigensoles under the coordinate axis; if the third cosine distance between the target average energy flux density components of the two eigensoles under a coordinate axis is greater than or equal to 0, and less than or equal to... The third cosine distance is taken as the second cosine distance between the target average energy flux density components of two eigensoles under the coordinate axis; if the third cosine distance between the target average energy flux density components of any two eigensoles under a coordinate axis is greater than... Calculate the difference between π and the third cosine distance, and use the difference as the second cosine distance between the target average energy flux density components of the two eigensols under the coordinate axis.

[0068] The definition of cosine distance specifically includes: obtaining two vectors through the inverse cosine function. and The angle between them, and this dimensionless angle value is defined as distance, its mathematical expression is:

[0069]

[0070] In the definition of cosine distance, θ AB The range of values ​​for (i.e., the third cosine distance) is [0, π]. From the definition of cosine distance, we know that θ... AB The larger the value, the worse the linear correlation between the two vectors.

[0071] If the positive direction is not specified beforehand before executing S120, when determining the third cosine distance between two eigensoles, if Equation (1) is used directly to calculate the third cosine distance between the two eigensoles, the third cosine distance between the two eigensoles that are essentially of the same mode may be -π. This will result in the final determination that the correlation between the two eigensoles is small, while in fact the two eigensoles belong to the same mode and have a large correlation.

[0072] Therefore, formula (2) can be used to calculate the second cosine distance between the target displacement components of the m-th eigensol and the n-th eigensol under the x-axis:

[0073]

[0074] in, Let m be the target displacement component of the m-th eigensol on the x-axis. Let m be the target displacement component on the x-axis of the nth eigensol. Both m and n are positive integers, and both m and n are less than or equal to M. M is the total number of eigensoles. For array The length of the mold, For array The length of the mold,

[0075] Similarly, formula (3) can be used to calculate the second cosine distance between the target displacement components of the m-th eigensol and the n-th eigensol on the y-axis:

[0076]

[0077] in, Let m be the target displacement component of the m-th eigensol on the y-axis. Let m be the target displacement component on the y-axis of the nth eigensol. Both m and n are positive integers, and both m and n are less than or equal to M. M is the total number of eigensoles. For array model The length of the mold,

[0078] Similarly, formula (4) can be used to calculate the second cosine distance between the target displacement components of the m-th eigensol and the n-th eigensol on the z-axis:

[0079]

[0080] in, Let m be the target displacement component of the m-th eigensol on the z-axis. Let m be the target displacement component on the z-axis of the nth eigensol. Both m and n are positive integers, and both m and n are less than or equal to M. M is the total number of eigensoles. For array The length of the mold, For array The length of the mold,

[0081] Formula (5) can be used to calculate the second cosine distance between the target angular momentum components of the m-th eigensol and the n-th eigensol on the x-axis:

[0082]

[0083] in, Let m be the target angular momentum component of the m-th eigensol on the x-axis. Let m be the target angular momentum component of the nth eigensol on the x-axis. Both m and n are positive integers, and both m and n are less than or equal to M. M is the total number of eigensoles. For array The length of the mold, For array The length of the mold,

[0084] Formula (6) can be used to calculate the second cosine distance between the target angular momentum components of the m-th eigensol and the n-th eigensol on the y-axis:

[0085]

[0086] in, Let m be the target angular momentum component of the m-th eigensol on the y-axis. Let m be the target angular momentum component of the nth eigensol on the y-axis. Both m and n are positive integers, and both m and n are less than or equal to M. M is the total number of eigensoles. For array The length of the mold, For array The length of the mold,

[0087] Formula (7) can be used to calculate the second cosine distance between the target angular momentum components of the m-th eigensol and the n-th eigensol under the z-axis:

[0088]

[0089] in, Let be the target angular momentum component of the m-th eigensol on the z-axis. Let m be the target angular momentum component of the nth eigensol on the z-axis. Both m and n are positive integers, and both m and n are less than or equal to M. M is the total number of eigensoles. For array The length of the mold, For array The length of the mold,

[0090] Formula (8) can be used to calculate the second cosine distance between the target average energy flux density components of the m-th eigensol and the n-th eigensol on the x-axis:

[0091]

[0092] in, Let m be the target average energy flux density component of the m-th eigensol on the x-axis. Let m be the target average energy flux density component of the nth eigensol on the x-axis. Both m and n are positive integers, and both m and n are less than or equal to M. M is the total number of eigensoles. For array The length of the mold, For array The length of the mold,

[0093] Formula (9) can be used to calculate the second cosine distance between the target average energy flux density components of the m-th eigensol and the n-th eigensol under the y-axis:

[0094]

[0095] in, Let m be the target average energy flux density component of the m-th eigensol on the y-axis. Let m be the target average energy flux density component of the nth eigensol on the y-axis. Both m and n are positive integers, and both m and n are less than or equal to M. M is the total number of eigensoles. For array The length of the mold, For array The length of the mold,

[0096] Formula (10) can be used to calculate the second cosine distance between the target average energy flux density components of the m-th eigensol and the n-th eigensol under the z-axis:

[0097]

[0098] in, Let m be the target average energy flux density component of the m-th eigensol on the z-axis. Let m be the target average energy flux density component of the nth eigensol on the z-axis. m and n are both positive integers, and both m and n are less than or equal to M. M is the total number of eigensoles. For array The length of the mold, For array The length of the mold,

[0099] Using formulas (2) to (10) above, the range of the second cosine distance can be limited to [0, π / 2]. Under this limitation, the smaller the second cosine distance, the closer the modes of the two eigenfunctions are. In subsequent clustering, the two eigenfunctions with smaller distances will be assigned to the same class.

[0100] Optionally, the following formulas (11), (12), and (13) can be used to determine the weight values ​​of the target displacement components of the m-th and n-th eigensols under the x-axis, respectively. Determine the weight values ​​of the target displacement components of the m-th and n-th eigensols along the y-axis. Determine the weight values ​​of the target displacement components of the m-th and n-th eigensols along the z-axis.

[0101]

[0102]

[0103]

[0104] In the above formula (11)-formula (13),

[0105] Alternatively, the weights of the target angular momentum components of the m-th and n-th eigensols along the x-axis can be determined using the following formulas (14), (15), and (16). Determine the weight values ​​of the target angular momentum components of the m-th and n-th eigensols along the y-axis. Determine the weight values ​​of the target angular momentum components of the m-th and n-th eigensols along the z-axis.

[0106]

[0107]

[0108]

[0109] In the above formula (14)-formula (16),

[0110] Alternatively, the weights of the target average energy flux density components of the m-th and n-th eigensols on the x-axis can be determined using the following formulas (17), (18), and (19). Determine the weight values ​​of the target average energy flux density components of the m-th and n-th eigensols along the y-axis. Determine the weight values ​​of the target average energy flux density components of the m-th and n-th eigensols along the z-axis.

[0111]

[0112]

[0113]

[0114] In the above formula (17)-formula (19),

[0115]

[0116] Furthermore, regarding as well as Normalization is performed to make right as well as Normalization is performed to make right as well as Normalization is performed to make

[0117] Optionally, the first cosine distance between the m-th eigensol and the n-th eigensol can be expressed as follows:

[0118]

[0119] Those skilled in the art will understand that the second cosine distance measures the difference in mode shape among different eigensoles, without considering the amplitude information of the mode shapes. Directly summing the second cosine distances treats the mode shape amplitudes in all directions as being of the same order of magnitude, which is inconsistent with the physical reality. This is because, in some guided wave modes, there is a concept of a dominant direction. In the dominant direction, nodes have large particle displacement / angular momentum / average energy flux density, while in non-dominant directions, the particle displacement / angular momentum / average energy flux density of nodes is very small. In this case, if the sum of all second cosine distances is directly used as the first cosine distance, the calculated particle displacement / angular momentum / average energy flux density results for directions with small particle displacement / angular momentum / average energy flux density introduce a large numerical error, equivalent to introducing significant noise, which will cause distortion in the measurement of the first cosine distance.

[0120] If appropriate weight values ​​are determined for each target displacement component, each target angular momentum component, and each target average energy flux density component, and the first cosine distance between two eigensoles is calculated using formula (20), it can be achieved that if the particle displacement / angular momentum / average energy flux density in one direction of the mode shape corresponding to a certain eigensol is very small, then its corresponding weight value is small; if the particle displacement / angular momentum / average energy flux density in one direction of a certain eigensol is very large, then its corresponding weight value is large. In this way, the effectiveness of the "distance" metric can be improved.

[0121] S150. Based on all first cosine distances and the nearest neighbor propagation algorithm, the similarity matrix is ​​obtained.

[0122] The similarity matrix can be viewed as a set, in which the elements reflect the similarity between two different eigenvalues.

[0123] There are multiple ways to implement this step, and this application does not limit this one. For example, the implementation method of this step may include: determining the maximum neighborhood radius corresponding to each eigensol; if the first cosine distance θ between the m-th eigensol and the n-th eigensol... mn If the radius of the largest neighborhood of the m-th eigensol is greater than that of the m-th eigensol, but the q-th eigensol and the m-th eigensol are neighbors, and the n-th eigensol and the q-th eigensol are neighbors, then the first cosine distance θ between the m-th and n-th eigensols is determined. mn The first cosine distance θ between the m-th eigensol and the q-th eigensol. mq and the first cosine distance θ between the q-th eigensol and the n-th eigensol.qn minimum value θ min-mqn ; set the minimum value θ min-mqn The opposite of the value is used as an element in the similarity matrix; if the first cosine distance θ between the k-th eigensol and the f-th eigensol is... kf The first cosine distance θ between the k-th and f-th eigensols is less than the maximum neighborhood radius of the m-th eigensol. kf The opposite of the numbers are used as elements in the similarity matrix; where m, n, q, k and f are all positive integers, m≠n, n≠q and k≠f.

[0124] To facilitate understanding, the method for determining nearest neighbor relationships is first explained. The method is as follows: Taking the g-th eigensol as the reference point, if there exists only one other eigensol whose first cosine distance to the g-th eigensol is less than or equal to the maximum neighborhood radius of the g-th eigensol, then that other eigensol has a nearest neighbor relationship with the g-th eigensol. Taking the g-th eigensol as the reference point, if there exist two or more other eigensols whose first cosine distance to the g-th eigensol is less than or equal to the maximum neighborhood radius of the g-th eigensol, then among the two or more other eigensols, the eigensol with the smallest first cosine distance to the g-th eigensol has a nearest neighbor relationship with the g-th eigensol.

[0125] Figure 6 A schematic diagram illustrating the principle of the nearest neighbor relationship determination method provided in this embodiment of the disclosure. See also... Figure 6 Taking the intrinsic solution g1 as the reference point, it is assumed that its maximum neighborhood radius is N. g1 At the next frequency, there are two eigensols, eigensol g2 and eigensol g3. The first cosine distance from eigensol g2 to eigensol g1 is greater than N. g1 It is determined that eigensolution g2 and eigensolution g1 are not nearest neighbors. The first cosine distance from eigensolution g3 to eigensolution g1 is less than N. g1 It is determined that the eigensol g3 and the eigensol g1 are close neighbors.

[0126] Taking the intrinsic solution g3 as the reference point, assume that its maximum neighborhood radius is N. g3 At the next frequency, there are three eigensols: eigensol g4, ​​eigensol g5, and eigensol g6. The first cosine distance from eigensol g6 to eigensol g3 is greater than N. g3 It is determined that eigensolution g6 and eigensolution g3 are not nearest neighbors. The first cosine distances from eigensolutions g4 and g5 to eigensolution g3 are both less than N. g3 Furthermore, the first cosine distance between eigensols g4 and g3 is less than the first cosine distance between eigensols g5 and g3. Therefore, eigensols g4 and g3 are determined to be nearest neighbors, while eigensols g5 and g3 are not.

[0127] Based on this, the maximum neighborhood radius corresponding to the m-th eigensol can be set to N. m =b+k(f m -f c ), where f m f is the frequency of the m-th eigensol. c denoted as , where is the center frequency of the target frequency band. b is the reference neighborhood radius, which can be determined based on the finite element model used and its sweep step size. The more complex the finite element model and the larger the sweep step size, the larger the reference neighborhood radius b. k is the slope, which is related to the length of the target frequency band. k(ff) c As a correction term, the larger the frequency span, especially the more high-frequency and / or higher-order modes, the more k(ff) becomes. c The larger the value of k(ff), the higher the frequency of the eigensols, and the larger their corresponding maximum neighborhood radius. It should be noted that, since k(ff) c As a correction term, its value is usually relatively small relative to b. Optionally, b = 0.4, kf c =0.1.

[0128] The finite element model represents the simulation results of the waveguide. In reality, the ultrasonic dynamic characteristics of a waveguide vary depending on its cross-sectional shape, material, and ultrasonic frequency band. The finite element model includes parameters describing the cross-sectional shape, material, and frequency band of the waveguide. By configuring these parameters, a finite element model corresponding to the waveguide under study can be obtained.

[0129] When executing S120, optionally, based on the finite element model, ultrasonic dynamic characteristic data of each of the multiple nodes under different guided wave modes and frequencies can be obtained.

[0130] Assuming that all eigensoles are traversed from low frequency to high frequency (or from high frequency to low frequency, this application does not impose any restrictions on this), if the first cosine distance θ between the m-th eigensoles and the n-th eigensoles... mn If the radius of the largest neighborhood of the m-th eigensol is greater than the radius of the largest neighborhood of the m-th eigensol, then the first cosine distance is used to determine that the m-th and n-th eigensols do not have a nearest neighbor relationship. Based on this, if the q-th and m-th eigensols have a nearest neighbor relationship, and the n-th and q-th eigensols also have a nearest neighbor relationship, then the nearest neighbor relationship can be considered to be propagated from the m-th eigensol to the n-th eigensol. In this case, the first cosine distance θ between the m-th and n-th eigensols is determined. mn The first cosine distance θ between the m-th eigensol and the q-th eigensol. mq and the first cosine distance θ between the q-th eigensol and the n-th eigensol. qn minimum value θmin-mqn ; set the minimum value θ min-mqn The opposite of the similarity is used as an element in the similarity matrix. This ensures that the m-th eigensol and the n-th eigensol can be clustered into one class during subsequent clustering.

[0131] S160. Based on the similarity matrix, cluster all eigensoles to obtain at least one class; each class corresponds to an ultrasonic guided wave mode.

[0132] The essence of this step is to cluster eigensoles with high similarity into a class. The eigensoles in each class belong to the same ultrasonic guided wave mode. Eigensoles in different classes belong to different ultrasonic guided wave modes.

[0133] There are multiple ways to implement this step, and this application does not limit this one. For example, the implementation method of this step includes: based on the similarity matrix, using the AP clustering algorithm to cluster all intrinsic solutions to obtain at least one class.

[0134] The AP clustering algorithm is based on the similarity between data points, treating all data points as potential cluster centers and iteratively selecting the best cluster centers through certain rules. The biggest advantage of the AP clustering algorithm is that it introduces the idea of ​​"voting," eliminating the need to determine the total number of clusters before clustering; the total number of clusters is determined by the properties of the data itself. Since the total number of guided wave modes cannot be determined before clustering, clustering methods that require pre-determining the total number of clusters cannot be applied to ultrasonic guided wave mode clustering, while the AP clustering method can be used for ultrasonic guided wave mode clustering.

[0135] The AP clustering algorithm requires two inputs: a similarity matrix and a reference value p. In this application, the similarity matrix can be obtained through S150. The reference value p can be obtained from the similarity matrix, which is prior art and will not be elaborated upon here.

[0136] In the above technical solution, ultrasonic dynamic characteristic data of each node at different frequencies under different guided wave modes are set to determine the target ultrasonic dynamic characteristic data of all intrinsic solutions within the target frequency band; different intrinsic solutions have different guided wave modes and / or frequencies; based on the target ultrasonic dynamic characteristic data of each intrinsic solution, the first cosine distance between any two intrinsic solutions is determined; based on all first cosine distances and the nearest neighbor propagation algorithm, a similarity matrix is ​​obtained; based on the similarity matrix, all intrinsic solutions are clustered to obtain at least one class; each class corresponds to one ultrasonic guided wave mode. In essence, this involves a deep analysis of the mode shape characteristics of any two intrinsic solutions among multiple intrinsic solutions to determine their correlation. This can solve the problem of low clustering accuracy caused by the overlap between different guided wave modes, data point overlap, and discontinuity of some guided wave modes when clustering multiple intrinsic solutions from the perspective of appearance, thus achieving the goal of improving the clustering accuracy of intrinsic solutions.

[0137] Based on the above technical solutions, the method may optionally further include: plotting all data points corresponding to the intrinsic solutions in a dispersion relation coordinate system; connecting data points belonging to the same class to obtain a dispersion relation curve. Since the above method essentially connects data points based on clustering results, it can ensure the accuracy of the obtained dispersion relation curve.

[0138] Figure 7 This is a dispersion curve of a railway track obtained using the ultrasonic guided wave mode tracking method provided in this embodiment of the disclosure. From... Figure 7 As can be seen from the above, the technical solution provided in this application can accurately handle situations such as the intersection between different waveguide modes, the overlap of data points between different waveguide modes, and the discontinuity of some waveguide modes.

[0139] This method takes into full account the physical background of the ultrasonic guided wave problem when selecting the three feature quantities: particle displacement, angular momentum, and average energy flux density. The selection of these feature quantities enhances the algorithm's ability to extract the characteristics of ultrasonic guided wave modes. Furthermore, since ultrasonic guided wave modes of the same type exhibit gradual changes, the introduction of a nearest neighbor propagation algorithm significantly improves the similarity between eigensoles of the same mode. The combination of these two methods enhances the discriminative power of different modes while simultaneously strengthening the clustering algorithm's ability to aggregate similar guided wave modes. Therefore, this method achieves high clustering accuracy.

[0140] Furthermore, during the clustering process, this method, using the AP intelligent clustering algorithm, does not require pre-determining the total number of guided wave modes in the frequency band of interest, thus automating the entire algorithm process. Therefore, this method has the advantage of high automation.

[0141] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0142] Figure 8 This is a schematic diagram of the structure of an ultrasonic guided wave mode tracking device according to an embodiment of this disclosure, as shown below. Figure 8 As shown, the device includes:

[0143] The node determination module 210 is used to perform finite element mesh generation on the waveguide to form multiple nodes;

[0144] The ultrasonic dynamic characteristic data determination module 220 is used to acquire ultrasonic dynamic characteristic data of each of the plurality of nodes under different guided wave modes and frequencies.

[0145] The target ultrasonic dynamic characteristic data determination module 230 is used to determine the target ultrasonic dynamic characteristic data of all intrinsic solutions within the target frequency band based on the ultrasonic dynamic characteristic data of each node under different guided wave modes and frequencies; the guided wave modes and / or frequencies of different intrinsic solutions are different.

[0146] The cosine distance determination module 240 is used to determine the first cosine distance between any two of the intrinsic solutions based on the target ultrasonic dynamic feature data of each of the intrinsic solutions.

[0147] The nearest neighbor propagation module 250 is used to obtain a similarity matrix based on all the first cosine distances and the nearest neighbor propagation algorithm;

[0148] Clustering module 260 is used to cluster all intrinsic solutions based on the similarity matrix to obtain at least one class; each class corresponds to an ultrasonic guided wave mode.

[0149] Furthermore, the ultrasonic dynamic characteristic data of each node at each frequency in each guided wave mode includes: particle displacement, particle vibration velocity, and average energy flux density.

[0150] The target ultrasonic dynamic feature data determination module is used for:

[0151] Construct a Cartesian coordinate system, within which the waveguide is located;

[0152] The particle vibration velocity of each node at each frequency in each waveguide mode is converted into the angular momentum of each node relative to the origin of the Cartesian coordinate system.

[0153] The particle displacement, angular momentum and average energy flux density of each node at each frequency in each waveguide mode are decomposed to obtain the particle displacement component, angular momentum component and average energy flux density component of each node at each frequency in each waveguide mode on each coordinate axis of the rectangular coordinate system.

[0154] The particle displacement components, angular momentum components, and average energy flux density components of each node at the same frequency in the same guided wave mode are summarized to obtain the target ultrasonic dynamic characteristic data of all eigensols within the target frequency band.

[0155] Furthermore, the cosine distance determination module is used for:

[0156] Based on the target ultrasonic dynamic characteristic data of any two eigensols, calculate the second cosine distance of the target displacement component, the second cosine distance of the target angular momentum component, and the second cosine distance of the target average energy flux density component of the two eigensols under each coordinate axis.

[0157] Based on the target ultrasonic dynamic characteristic data of any two eigensoles, determine the weight values ​​of the target displacement component, the target angular momentum component, and the target average energy flux density component of the two eigensoles under each coordinate axis.

[0158] Based on the second cosine distance between the target displacement components, the target angular momentum components, and the target average energy flux density components of any two intrinsic solutions under each coordinate axis, and the weight values ​​of the target displacement components, the target angular momentum components, and the target average energy flux density components of any two intrinsic solutions under each coordinate axis, the first cosine distance between any two intrinsic solutions is determined.

[0159] Furthermore, the cosine distance determination module is used for:

[0160] Based on the target ultrasonic dynamic characteristic data of any two eigensoles and the definition of cosine distance, calculate the third cosine distance of the target displacement component, the third cosine distance of the target angular momentum component, and the third cosine distance of the target average energy flux density component of any two eigensoles under each coordinate axis.

[0161] If the third cosine distance between the target displacement components of two eigensoles on a coordinate axis is greater than or equal to 0, and less than or equal to... The third cosine distance is taken as the second cosine distance between the target displacement components of the two eigensoles under the coordinate axis; if the third cosine distance between the target displacement components of any two eigensoles under a coordinate axis is greater than... Calculate the difference between π and the third cosine distance, and use the difference as the second cosine distance between the target displacement components of the two eigensols under the coordinate axis;

[0162] If the distance between the third cosine of the target angular momentum components of two eigensoles on a coordinate axis is greater than or equal to 0, and less than or equal to... The third cosine distance is taken as the second cosine distance between the target angular momentum components of the two eigensoles under the coordinate axis; if the third cosine distance between the target angular momentum components of any two eigensoles under a coordinate axis is greater than... Calculate the difference between π and the third cosine distance, and use the difference as the second cosine distance between the target angular momentum components of the two eigensols under the coordinate axis;

[0163] If the third cosine distance between the target average energy flux density components of two eigensoles on a coordinate axis is greater than or equal to 0, and less than or equal to... The third cosine distance is taken as the second cosine distance between the target average energy flux density components of the two eigensoles under the coordinate axis; if the third cosine distance between the target average energy flux density components of any two eigensoles under a coordinate axis is greater than... Calculate the difference between π and the third cosine distance, and use the difference as the second cosine distance between the target average energy flux density components of the two eigensols under the coordinate axis.

[0164] Furthermore, the nearest neighbor propagation module is used for:

[0165] Determine the maximum neighborhood radius corresponding to each of the intrinsic solutions;

[0166] If the first cosine distance θ between the m-th eigensolution and the n-th eigensolution mn If the radius of the neighborhood of the m-th eigensol is greater than the maximum neighborhood radius of the m-th eigensol, but the q-th eigensol and the m-th eigensol are neighbors, and the n-th eigensol and the q-th eigensol are neighbors, then the first cosine distance θ between the m-th eigensol and the n-th eigensol is determined. mn The first cosine distance θ between the m-th eigensol and the q-th eigensol. mq And the first cosine distance θ between the q-th eigensol and the n-th eigensol. qn minimum value θ min-mqn ; the minimum value θ min-mqn The opposite of the value is used as an element in the similarity matrix;

[0167] If the first cosine distance θ between the k-th eigensol and the f-th eigensol is... kf The first cosine distance θ between the k-th and f-th eigensols is less than the maximum neighborhood radius of the m-th eigensol. kf The opposite of the value is used as an element in the similarity matrix;

[0168] Where m, n, q, k and f are all positive integers, m≠n, n≠q and k≠f.

[0169] Furthermore, the clustering module is used for:

[0170] Based on the similarity matrix, the AP clustering algorithm is used to cluster all intrinsic solutions to obtain at least one class.

[0171] The apparatus disclosed in the above embodiments can implement the process flow of the methods disclosed in the above method embodiments and has the same or corresponding beneficial effects. To avoid repetition, it will not be described again here.

[0172] Figure 9 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this disclosure, such as... Figure 9 As shown, the electronic device may include smart terminals such as mobile phones, tablets, and computers. The electronic device includes:

[0173] One or more processors 301, Figure 9 Taking processor 301 as an example;

[0174] Memory 302;

[0175] The electronic device may also include an input device 303 and an output device 304.

[0176] The processor 301, memory 302, input device 303, and output device 304 in the electronic device can be connected via a bus or other means. Figure 9 Taking the example of a connection between China and Israel via a bus.

[0177] The memory 302, as a non-transitory computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the ultrasonic guided wave mode tracking method in the embodiments of this disclosure. The processor 301 executes various functional applications and data processing of the server by running the software programs, instructions, and modules stored in the memory 302, thereby implementing the ultrasonic guided wave mode tracking method of the above method embodiments.

[0178] The memory 302 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the electronic device. Furthermore, the memory 302 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory 302 may optionally include memory remotely located relative to the processor 301, and these remote memories can be connected to the terminal device via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0179] Input device 303 can be used to receive input digital or character information, and to generate key signal inputs related to user settings and function control of the electronic device. Output device 304 may include display devices such as a display screen.

[0180] This disclosure also provides a computer-readable storage medium storing a program or instructions that, when executed by a computer, perform an ultrasonic guided wave mode tracking method, the method comprising:

[0181] The waveguide is divided into finite element meshes to form multiple nodes;

[0182] Acquire ultrasonic dynamic characteristic data of each of the plurality of nodes under different waveguide modes and frequencies;

[0183] Based on the ultrasonic dynamic characteristic data of each node at different frequencies under different guided wave modes, the target ultrasonic dynamic characteristic data of all intrinsic solutions within the target frequency band are determined; the guided wave modes and / or frequencies of different intrinsic solutions are different.

[0184] Based on the target ultrasonic dynamic characteristic data of each of the intrinsic solutions, determine the first cosine distance between any two of the intrinsic solutions;

[0185] Based on all the first cosine distances and the nearest neighbor propagation algorithm, a similarity matrix is ​​obtained;

[0186] Based on the similarity matrix, all intrinsic solutions are clustered to obtain at least one class; each class corresponds to an ultrasonic guided wave mode.

[0187] Optionally, when executed by a computer processor, the computer-executable instructions can also be used to execute the technical solution of the ultrasonic guided wave mode tracking method provided in any embodiment of this disclosure.

[0188] Based on the above description of the implementation methods, those skilled in the art will clearly understand that this disclosure can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of this disclosure.

[0189] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0190] The above are merely specific embodiments of this disclosure, enabling those skilled in the art to understand or implement this disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to these embodiments, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An ultrasonic guided wave mode tracking method, characterized in that, include: The waveguide is divided into finite element meshes to form multiple nodes; Acquire ultrasonic dynamic characteristic data of each of the plurality of nodes under different waveguide modes and frequencies; Based on the ultrasonic dynamic characteristic data of each node at different frequencies under different guided wave modes, the target ultrasonic dynamic characteristic data of all intrinsic solutions within the target frequency band are determined; the guided wave modes and / or frequencies of different intrinsic solutions are different. Based on the target ultrasonic dynamic characteristic data of each of the intrinsic solutions, determine the first cosine distance between any two of the intrinsic solutions; Based on all the first cosine distances and the nearest neighbor propagation algorithm, a similarity matrix is ​​obtained; Based on the similarity matrix, all intrinsic solutions are clustered to obtain at least one class; each class corresponds to an ultrasonic guided wave mode.

2. The method according to claim 1, characterized in that, The ultrasonic dynamic characteristic data of each node at each frequency in each guided wave mode include: particle displacement, particle vibration velocity, and average energy flux density. The determination of target ultrasonic dynamic characteristic data based on ultrasonic dynamic characteristic data of each node at different frequencies in different guided wave modes, including: Construct a Cartesian coordinate system, within which the waveguide is located; The particle vibration velocity of each node at each frequency in each waveguide mode is converted into the angular momentum of each node relative to the origin of the Cartesian coordinate system. The particle displacement, angular momentum and average energy flux density of each node at each frequency in each waveguide mode are decomposed to obtain the particle displacement component, angular momentum component and average energy flux density component of each node at each frequency in each waveguide mode on each coordinate axis of the rectangular coordinate system. The particle displacement components, angular momentum components, and average energy flux density components of each node at the same frequency in the same guided wave mode are summarized to obtain the target ultrasonic dynamic characteristic data of all eigensols within the target frequency band.

3. The method according to claim 2, characterized in that, The determination of the first cosine distance between any two intrinsic solutions based on the target ultrasonic dynamic feature data of each of the intrinsic solutions includes: Based on the target ultrasonic dynamic characteristic data of any two eigensols, calculate the second cosine distance of the target displacement component, the second cosine distance of the target angular momentum component, and the second cosine distance of the target average energy flux density component of the two eigensols under each coordinate axis. Based on the target ultrasonic dynamic characteristic data of any two eigensoles, determine the weight values ​​of the target displacement component, the target angular momentum component, and the target average energy flux density component of the two eigensoles under each coordinate axis. Based on the second cosine distance between the target displacement components, the target angular momentum components, and the target average energy flux density components of any two intrinsic solutions under each coordinate axis, and the weight values ​​of the target displacement components, the target angular momentum components, and the target average energy flux density components of any two intrinsic solutions under each coordinate axis, the first cosine distance between any two intrinsic solutions is determined.

4. The method according to claim 3, characterized in that, The calculation of the second cosine distance between the target displacement component, the second cosine distance between the target angular momentum component, and the second cosine distance between the target average energy flux density component based on the target ultrasonic dynamic characteristic data of any two eigensoles under each coordinate axis includes: Based on the target ultrasonic dynamic characteristic data of any two eigensoles and the definition of cosine distance, calculate the third cosine distance of the target displacement component, the third cosine distance of the target angular momentum component, and the third cosine distance of the target average energy flux density component of any two eigensoles under each coordinate axis. If the third cosine distance between the target displacement components of two eigensoles on a coordinate axis is greater than or equal to 0, and less than or equal to... The third cosine distance is taken as the second cosine distance between the target displacement components of the two eigensoles under the coordinate axis; if the third cosine distance between the target displacement components of any two eigensoles under a coordinate axis is greater than... Calculate the difference between π and the third cosine distance, and use the difference as the second cosine distance between the target displacement components of the two eigensols under the coordinate axis; If the distance between the third cosine of the target angular momentum components of two eigensoles on a coordinate axis is greater than or equal to 0, and less than or equal to... The third cosine distance is taken as the second cosine distance between the target angular momentum components of the two eigensoles under the coordinate axis; if the third cosine distance between the target angular momentum components of any two eigensoles under a coordinate axis is greater than... Calculate the difference between π and the third cosine distance, and use the difference as the second cosine distance between the target angular momentum components of the two eigensols under the coordinate axis; If the third cosine distance between the target average energy flux density components of two eigensoles on a coordinate axis is greater than or equal to 0, and less than or equal to... The third cosine distance is taken as the second cosine distance between the target average energy flux density components of the two eigensoles under the coordinate axis; if the third cosine distance between the target average energy flux density components of any two eigensoles under a coordinate axis is greater than... Calculate the difference between π and the third cosine distance, and use the difference as the second cosine distance between the target average energy flux density components of the two eigensols under the coordinate axis.

5. The method according to claim 1, characterized in that, The similarity matrix obtained based on all the first cosine distances and the nearest neighbor propagation algorithm includes: Determine the maximum neighborhood radius corresponding to each of the intrinsic solutions; If the first cosine distance θ between the m-th eigensolution and the n-th eigensolution mn If the radius of the neighborhood of the m-th eigensol is greater than the maximum neighborhood radius of the m-th eigensol, but the q-th eigensol and the m-th eigensol are neighbors, and the n-th eigensol and the q-th eigensol are neighbors, then the first cosine distance θ between the m-th eigensol and the n-th eigensol is determined. mn The first cosine distance θ between the m-th eigensol and the q-th eigensol. mq and the first cosine distance θ between the q-th eigensol and the n-th eigensol. qn minimum value θ min-mqn ; the minimum value θ min-mqn The opposite of the value is used as an element in the similarity matrix; If the first cosine distance θ between the k-th eigensol and the f-th eigensol is... kf The first cosine distance θ between the k-th and f-th eigensols is less than the maximum neighborhood radius of the m-th eigensol. kf The opposite of the value is used as an element in the similarity matrix; Where m, n, q, k and f are all positive integers, m≠n, n≠q and k≠f.

6. The method according to claim 1, characterized in that, The step of clustering all intrinsic solutions based on the similarity matrix to obtain at least one class also includes: Based on the similarity matrix, the AP clustering algorithm is used to cluster all intrinsic solutions to obtain at least one class.

7. An ultrasonic guided wave mode tracking device, characterized in that, include: The node determination module is used to divide the waveguide into finite element meshes to form multiple nodes; An ultrasonic dynamic characteristic data determination module is used to acquire ultrasonic dynamic characteristic data of each of the plurality of nodes under different guided wave modes and frequencies. The target ultrasonic dynamic characteristic data determination module is used to determine the target ultrasonic dynamic characteristic data of all intrinsic solutions within the target frequency band based on the ultrasonic dynamic characteristic data of each node at different guided wave modes and frequencies; the guided wave modes and / or frequencies of different intrinsic solutions are different. The cosine distance determination module is used to determine the first cosine distance between any two of the intrinsic solutions based on the target ultrasonic dynamic feature data of each of the intrinsic solutions. The nearest neighbor propagation module is used to obtain a similarity matrix based on all the first cosine distances and the nearest neighbor propagation algorithm; The clustering module is used to cluster all intrinsic solutions based on the similarity matrix to obtain at least one class; each class corresponds to an ultrasonic guided wave mode.

8. An electronic device, characterized in that, include: Processor and memory; The processor executes the steps of the method as described in any one of claims 1 to 6 by invoking programs or instructions stored in the memory.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program or instructions that cause a computer to perform the steps of the method as described in any one of claims 1 to 6.

Images