A method for optimizing measurement point arrangement in the problem of guided wave dispersion relation recovery

Abstract

The present invention proposes a method for optimizing the arrangement of measuring points in the problem of waveguide dispersion relationship recovery. The method adopts a multi-task complex hierarchical sparse Bayesian model to solve the dispersion relationship of the waveguide, and uses the uncertainty index given by the multi-task complex hierarchical sparse Bayesian model to calculate the differential entropy estimate of the subsequent measurement point arrangement scheme, and obtains the optimal measurement point arrangement by minimizing the differential entropy estimate. The multi-task model of the waveguide dispersion relationship solves the tasks at multiple frequencies simultaneously, and constructs a dictionary matrix so that the weight vectors of the tasks corresponding to the multiple frequencies have consistent sparsity, and filters out the noise points that only appear at a single frequency, so that the dispersion relationship solution result is closer to the true value. In addition, the multi-task complex hierarchical sparse Bayesian model can estimate the hyperparameters in the model based on the data, which reduces the influence of the overfitting phenomenon in the problem of waveguide dispersion relationship recovery, so that the waveguide dispersion relationship can be solved more reliably and the measurement point arrangement scheme can be optimized.

Description

Technical Field

[0001] The invention belongs to the technical field of machine learning and civil engineering, and in particular relates to a method for optimizing measurement point arrangement in the problem of waveguide dispersion relationship recovery. Background Art

[0002] Ultrasonic guided wave nondestructive testing can efficiently detect whether there is damage in the structure, and obtain the damage location and even geometric information, which can provide effective help for the assessment and maintenance of the safety status of the built structure. Compared with the traditional ultrasonic testing method, the nondestructive testing method based on ultrasonic guided wave method has the advantages of larger detection range, smaller detection blind area and higher detection efficiency, which can reduce the manpower and material costs in structural safety assessment and maintenance. Therefore, the application of guided wave nondestructive testing technology to better carry out the safety assessment and maintenance of actual civil engineering structures is a research work that needs to be carried out urgently.

[0003] However, waveguides have dispersion and multimodal characteristics during propagation, which makes the waveguide signals received by the measuring points more complex, increases the difficulty of signal interpretation, and severely limits the difficulty of applying waveguide nondestructive testing technology. In recent years, research on sparse representation of waveguide signals to obtain waveguide dispersion and multimodal parameters (i.e., dispersion relations) has gradually emerged. However, these studies usually adopt a scheme of randomly arranging measuring points, which means that some measuring points with similar arrangement distances will obtain similar signals, resulting in information redundancy in the signal set obtained by the acquisition, which limits the effectiveness of the sparse representation method from the source of information acquisition. For the problem of recovering the dispersion relation of waveguides, how to optimize the arrangement of measuring points is a problem worth studying.

[0004] As an important means of dealing with uncertainty, the Bayesian probability theory method makes full use of data measurement information and prior probability information to infer the posterior probability distribution of unknown parameters, estimates the differential entropy after the measurement points are arranged according to the parameters of the posterior probability distribution, and optimizes the arrangement of the measurement points by minimizing the differential entropy estimate. In the present invention, a multi-task complex hierarchical Bayesian learning algorithm is used to sparsely represent the waveguide signal, and the dispersion relationship of the waveguide is obtained by the position of the non-zero elements of the weight vector. At the same time, the differential entropy after the measurement points are arranged is estimated by the parameters of the posterior distribution, and the optimal measurement point arrangement is obtained by optimizing the differential entropy estimate. Summary of the invention

[0005] The purpose of the present invention is to solve the problems in the prior art and to propose a method for optimizing the arrangement of measuring points in the problem of waveguide dispersion relationship recovery.

[0006] The present invention is implemented by the following technical scheme. The present invention proposes a method for optimizing the arrangement of measuring points in the problem of waveguide dispersion relationship recovery. The method comprises the following steps:

[0007] Step 1: Based on the physical approximate relationship between wave number and frequency, construct a multi-task problem of recovering the dispersion relation by simultaneous multi-frequency operation. Substitute the measured signal of the measuring point into the multi-task problem, use the multi-task complex hierarchical sparse Bayesian model to calculate the multi-task problem, obtain the solution of the dispersion relationship and the value of the corresponding posterior distribution parameter; determine whether the number of measuring points on the current structure reaches the target number, if so, stop, and calculate the number of points according to the mean μ q+n Draw the dispersion curve, otherwise continue with the next steps;

[0008] Step 2: Construct a matrix R consisting of rows of the dictionary matrix to be arranged, with the positions of the measurement points to be arranged as independent variables 2P,q+n ; Use R 2P,q+n and the posterior distribution covariance matrix ∑ obtained in step 1 q+n Construct an estimated value of the differential entropy of the dispersion relation solution after the measurement points are to be arranged;

[0009] Step 3: Optimize the theoretical positions of the measuring points to be arranged so that the estimated value of the differential entropy after the measuring points are arranged is minimized; arrange the measuring points to be arranged on the structure according to the theoretically optimized positions, and return to step 1.

[0010] Furthermore, the step 1 is specifically as follows:

[0011] Step 1.1: The time domain signal set {y i (t)} i＝1,…,D Perform fast Fourier transform and transform it into a frequency domain signal set {Y q} q＝1,…,Q , where Y q is the signal at each measuring point at frequency f q The complex amplitude at (q=1,…,Q);

[0012] Step 1.2: According to the waveguide wave number k q+n and frequency f q+n The physical approximation relationship of is used to obtain a multi-task model for recovering the dispersion relation by multiple frequencies simultaneously.

[0013] Step 1.3: According to the multi-task model constructed in step 1.2, random embedding is performed to transform it into a multi-task complex hierarchical sparse Bayesian learning problem;

[0014] Step 1.4: Use multi-task complex hierarchical sparse Bayesian learning to solve the sparse problem in step 1.3;

[0015] Step 1.5: Determine whether the number of measurement points D on the current structure has reached the target number D end , if it reaches, stop, according to the mean μ q+n Plot the dispersion curve, otherwise continue with the next steps.

[0016] Furthermore, the waveguide wave number k q+n and frequency f q+n The physical approximation relationship is:

[0017]

[0018] Among them, Δf is the rate of change of frequency, which is much smaller than the frequency f q The value of c q is the frequency f q The wave speed at .

[0019] Furthermore, the multi-task model is expressed as follows:

[0020]

[0021] in, is the dictionary matrix under the approximate relationship, and its elements in row i and column j are where ξ q+n,i,j is the frequency f q+n The wave is guided at a speed c q,j Propagation distance d i The amplitude attenuation coefficient that occurs during the process is taken as If it is an axisymmetric tube wave, it is taken as 1.0; {c q,j} j＝1,…,J is the set of possible wave velocity values ​​of guided waves; is the sparse vector to be solved, each element of which corresponds to a possible value of the guided wave velocity. If the element is 0, there is no guided wave mode propagating at this possible velocity value. Otherwise, if the element is not 0, there is a guided wave mode propagating at this velocity value.

[0022] Furthermore, during the conversion process in step 1.3, the uncertain predicted measurement signal It is expressed as:

[0023]

[0024] in, is the uncertain weight vector, is the true weight vector; unknown prediction error e q+n The modeling error p q+n With the measurement error n q+n Composition, of which

[0025] Furthermore, the solution process of step 1.4 is as follows: first, the expression of the sparse problem is transformed as follows:

[0026]

[0027] in,

[0028] The following formula

[0029]

[0030]

[0031]

[0032]

[0033] Perform a loop calculation; first calculate μ q+n and∑ q+n , and then use the calculated μ q+n and∑ q+n calculate and The calculated and To calculate μ q+n and∑ q+n , repeat the iteration until the values ​​of each variable no longer change significantly between the previous and next iterations; finally, w is obtained. q+n The mean μ of the posterior distribution q+n With the covariance matrix ∑ q+n .

[0034] Furthermore, the step 2 is specifically as follows:

[0035] Step 2.1: Assume that the location of the measurement point to be arranged is The new dictionary matrix Relative to the dictionary matrix Φ before the measurement points are to be arranged q+n The new rows of compose a matrix:

[0036] R 2P,q+n =[φ D+1,q+n ,…,φ D+P,q+n ,φ 2D+P+1,q+n ,…,φ 2D+2P,q+n ] T

[0037] in,

[0038] Step 2.2: Set the location of the measuring point to be arranged As the independent variable, the estimated value of the differential entropy of the dispersion relation solution after the measurement points are arranged is used as the objective function T(D n ), expressed as:

[0039]

[0040] Where K is w q+n The number of non-zero elements in .

[0041] Furthermore, the step three is specifically as follows:

[0042] Step 3.1: Use the optimization algorithm to optimize the objective function T(D n ) is optimized to find the objective function T(D n ) The smallest measurement point position D n ;

[0043] Step 3.2: The position D of the measuring point to be arranged is obtained by optimization in step 3.1 n Arrange measurement points on the actual structure to be measured and collect data. Return to step 1 after data collection is completed.

[0044] The present invention proposes an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of a method for optimizing measurement point arrangement in a waveguide dispersion relationship recovery problem are implemented.

[0045] The present invention proposes a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps of the method for optimizing the arrangement of measuring points in the problem of recovering the dispersion relationship of guided waves are implemented.

[0046] Beneficial effects of the present invention:

[0047] 1. The present invention proposes a multi-task model for recovering the waveguide dispersion relationship, so that the weight vectors corresponding to the multiple frequencies obtained by solving the problem have consistent sparsity and filter out the noise points that only appear in a single frequency weight vector.

[0048] 2. The present invention utilizes a multi-task complex hierarchical sparse Bayesian algorithm, which can quantify the uncertainty of the solution results of the multi-task model.

[0049] 3. The present invention provides a method for optimizing the arrangement of measuring points, thereby improving the accuracy of the results of recovering the guided wave dispersion relationship. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0051] Figure 1The present invention provides a flow chart of a method for optimizing the arrangement of measuring points in a problem of recovering the dispersion relation of guided waves.

[0052] Figure 2 Schematic diagram of the connection between the laser ultrasonic scanning system and the test piece used in the embodiment of the present invention.

[0053] Figure 3 Schematic diagram of dispersion relation recovery results after optimizing the arrangement of measurement points in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] The purpose of the present invention is to solve the problem of optimizing the arrangement of measuring points in the process of solving the guided wave dispersion curve, and a method for optimizing the arrangement of measuring points in the process of solving the guided wave dispersion curve based on Bayesian uncertainty quantification is proposed. First, the measured structure is measured using the arranged measuring points. Then, the multi-task complex hierarchical sparse Bayesian algorithm is used to solve the multi-task sparse problem, and the dispersion relationship result and its corresponding covariance matrix are obtained. The differential entropy estimate of the measurement point scheme to be arranged is calculated using the covariance matrix. The optimal scheme to be arranged is obtained by minimizing the differential entropy estimate through the optimization algorithm. Finally, the measuring points are arranged on the measured structure according to the optimal scheme, and after measuring and calculating again, the optimal scheme to be arranged in the next step is obtained, until the number of arranged measuring points reaches the expected number of arranged measuring points.

[0056] Combination Figure 1 The present invention proposes a method for optimizing the arrangement of measuring points in the problem of waveguide dispersion relationship recovery based on Bayesian uncertainty quantification, which specifically includes the following steps:

[0057] Step 1: Based on the physical approximate relationship between wave number and frequency, construct a multi-task problem of recovering the dispersion relation by simultaneous multi-frequency operation. Substitute the measured signal of the measuring point into the problem, use the multi-task complex hierarchical sparse Bayesian model to calculate the multi-task problem, and obtain the solution of the dispersion relationship and the value of the corresponding posterior distribution parameter. Determine whether the number of measuring points on the current structure has reached the target number. If so, stop and calculate according to the mean μ q+n Plot the dispersion curve, otherwise continue with the next steps.

[0058] Step 2: Construct a matrix R consisting of newly added rows of the dictionary matrix with the positions of the measurement points to be arranged as independent variables 2P,q+n Using R2P,q+n and the posterior distribution covariance matrix ∑ obtained in step 1 q+n Construct an estimate of the differential entropy of the dispersion relation solution after the measurement points are arranged.

[0059] Step 3: Optimize the theoretical positions of the measurement points to be arranged so that the estimated value of the differential entropy after the measurement points are arranged is minimized. Arrange the measurement points to be arranged on the structure according to the theoretically optimized positions and return to step 1.

[0060] The step 1 is specifically as follows:

[0061] Step 1.1: The time domain signal set {y i (t)} i＝1,…,D Perform fast Fourier transform and transform it into a frequency domain signal set {Y q} q＝1,…,Q , where Y q is the signal at each measuring point at frequency f q The complex amplitude at (q=1,…,Q).

[0062] Step 1.2: According to the waveguide wave number k q+n and frequency f q+n The physical approximation relationship is:

[0063]

[0064] Among them, Δf is the rate of change of frequency, which is much smaller than the frequency f q The value of c q is the frequency f q The wave speed at .

[0065] Based on the above approximate relationship, a multi-task model for recovering the dispersion relation of multiple frequencies can be constructed, which is expressed as follows:

[0066]

[0067] in, is the dictionary matrix under the approximate relationship, and its elements in row i and column j are where ξ q+n,i,j is the frequency f q+n The wave is guided at a speed c q,j Propagation distance d i The amplitude attenuation coefficient that occurs during the process is taken as If it is an axisymmetric tube wave, it is taken as 1.0. q,j} j＝1,γ,J is the set of possible wave velocity values ​​for the guided wave. is the sparse vector to be solved, each element of which corresponds to a possible value of the guided wave velocity. If the element is 0, there is no guided wave mode propagating at this possible velocity value. Otherwise, if the element is not 0, there is a guided wave mode propagating at this velocity value.

[0068] Step 1.3: According to the multi-task sparse model constructed in step 1.2, perform stochastic embedding to transform it into a multi-task complex hierarchical sparse Bayesian learning problem. It can be expressed as

[0069]

[0070] in, is the uncertain weight vector, is the true weight vector. The unknown prediction error e q+n The modeling error p q+n With the measurement error n q+n Composition, of which

[0071] Step 1.4: Multi-task complex hierarchical sparse Bayesian learning is used to solve the sparse problem in step 1.3. First, the expression of the sparse problem is transformed as follows:

[0072]

[0073] in,

[0074] The following formula

[0075]

[0076]

[0077]

[0078]

[0079] Perform a loop calculation. First calculate μ q+n and∑ q+n , and then use the calculated μ q+n and∑ q+n calculate and The calculated and To calculate μ q+n and∑ q+n , repeat the iteration until the values ​​of each variable no longer change significantly between the previous and next iterations. Finally, we get wq+n The mean μ of the posterior distribution q+n With the covariance matrix ∑ q+n .

[0080] Step 1.5: Determine whether the number of measurement points D on the current structure has reached the target number D end , if it reaches, stop, according to the mean μ q+n Plot the dispersion curve, otherwise continue with the next steps.

[0081] The step 2 is specifically as follows:

[0082] Step 2.1: Assume that the location of the measurement point to be arranged is The new dictionary matrix Relative to the dictionary matrix Φ before the measurement points are to be arranged q+n The new rows of compose a matrix:

[0083] R 2P,q+n =[φ D+1,q+n ,…,φ D+P,q+n ,φ 2D+P+1,q+n ,…,φ 2D+2P,q+n ] T

[0084] in,

[0085] Step 2.2: Set the location of the measuring point to be arranged As the independent variable, the estimated value of the differential entropy of the dispersion relation solution after the measurement points are arranged is used as the objective function T(D n ), which can be expressed as:

[0086]

[0087] Where K is w q+n The number of non-zero elements in .

[0088] The step three is specifically as follows:

[0089] Step 3.1: Use, but not limited to, genetic algorithm, particle swarm optimization algorithm or simulated annealing algorithm to optimize the objective function T(D n ) is optimized to find the objective function T(D n ) The smallest measurement point position D n .

[0090] Step 3.2: The position D of the measuring point to be arranged is obtained by optimization in step 3.1 n Arrange measurement points on the actual structure to be measured and collect data. Return to step 1 after data collection is completed.

[0091] Example

[0092] This embodiment enables the present invention to be applied to the arrangement of measuring points in optimizing the dispersion relationship solution problem of carbon fiber reinforced composite material plates. Figure 1 A flow chart of the method of the present invention is given. Figure 2 A connection diagram of the laser ultrasonic scanning system and the test piece used in this example is given. Figure 3 A schematic diagram of the dispersion relation recovery results after the measurement points are optimized is given.

[0093] The excitation signal is a Hanning window narrowband five-peak sine signal with a center frequency of 180kHz. The room temperature during the measurement is 23 degrees Celsius, and each measurement point is measured 100 times and averaged to improve the signal-to-noise ratio.

[0094] The step 1 is specifically as follows: In this embodiment, 10 measuring points are initially arranged on the composite material plate structure, and the signal y obtained at each measuring point is i (t) Perform fast Fourier transform one by one to obtain the pre-processed signal Y q Substitute the processed signal into the constructed multi-task sparse problem Iteratively calculate μ k ,∑ k and Since the number of existing measuring points is 10, which is far less than the expected number of measuring points of 50, proceed to step 2.

[0095] The step 2 is specifically as follows: As an independent variable, the estimated value T(D n ).

[0096] The step three is specifically: using the optimization algorithm to obtain the differential entropy estimate T(D n ) The arrangement of the measurement points to be arranged corresponding to the minimum value is arranged on the structure to be measured. After measuring and calculating again, the optimal arrangement scheme for the next step is obtained until the number of arranged measurement points reaches the expected number of measurement points 50. The final dispersion relationship result is as follows: Figure 3 shown.

[0097] The present invention proposes an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of a method for optimizing measurement point arrangement in a waveguide dispersion relationship recovery problem are implemented.

[0098] The present invention proposes a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps of the method for optimizing the arrangement of measuring points in the problem of recovering the dispersion relationship of guided waves are implemented.

[0099] The memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus RAM (DRRAM). It should be noted that the memory of the method described in the present invention is intended to include, but is not limited to, these and any other suitable types of memory.

[0100] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a high-density digital video disc (DVD)), or a semiconductor medium (eg, a solid state disc (SSD)).

[0101] In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in a processor or an instruction in the form of software. The steps of the method disclosed in conjunction with the embodiment of the present application can be directly embodied as a hardware processor for execution, or a combination of hardware and software modules in a processor for execution. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in a memory, and the processor reads the information in the memory and completes the steps of the above method in conjunction with its hardware. To avoid repetition, it is not described in detail here.

[0102] It should be noted that the processor in the embodiment of the present application can be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method embodiment can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiment of the present application can be directly embodied as a hardware decoding processor to perform, or the hardware and software modules in the decoding processor can be combined and performed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in a memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0103] The above is a detailed introduction to the method for optimizing the arrangement of measuring points in the problem of waveguide dispersion relationship recovery proposed by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for optimizing the arrangement of measuring points in the problem of waveguide dispersion relation recovery, characterized in that: The method comprises the following steps: Step 1: Based on the physical approximate relationship between wave number and frequency, construct a multi-task problem of recovering the dispersion relation by simultaneous multi-frequency operation. Substitute the measured signal of the measuring point into the multi-task problem, use the multi-task complex hierarchical sparse Bayesian model to calculate the multi-task problem, obtain the solution of the dispersion relationship and the value of the corresponding posterior distribution parameter; determine whether the number of measuring points on the current structure reaches the target number, if so, stop, and calculate the number of points according to the mean μ q+n Draw the dispersion curve, otherwise continue with the next steps; Step 2: Construct a matrix R consisting of rows of the dictionary matrix to be arranged, with the positions of the measurement points to be arranged as independent variables 2P,q+n ; Use R 2P,q+n and the posterior distribution covariance matrix ∑ obtained in step 1 q+n Construct an estimated value of the differential entropy of the dispersion relation solution after the measurement points are to be arranged; Step 3: Optimize the theoretical positions of the measuring points to be arranged so that the estimated value of the differential entropy after the measuring points are arranged is minimized; arrange the measuring points to be arranged on the structure according to the theoretically optimized positions, and return to step 1; The multi-task model is expressed as follows: in, is the dictionary matrix under the approximate relationship, and its elements in row i and column j are where ξ q+n,i,j is the frequency f q+n The wave is guided at a speed c q,j Propagation distance d i The amplitude attenuation coefficient that occurs during the process is taken as If it is an axisymmetric tube wave, it is taken as 1.0; {c q,j } j＝1,…,J is the set of possible wave velocity values ​​of guided waves; is the sparse vector to be solved, each element of which corresponds to a possible value of the guided wave velocity. If the element is 0, there is no guided wave mode propagating at this possible wave velocity value. On the contrary, if the element is not 0, there is a guided wave mode propagating at this wave velocity value. The step 2 is specifically as follows: Step 2.1: Assume that the location of the measurement point to be arranged is The new dictionary matrix Relative to the dictionary matrix Φ before the measurement points are to be arranged q+n The new rows of compose a matrix: R 2P,q+n =[φ D+1,q+n ,…,f D+P,q+n ,f 2D+P+1,q+n ,…,f 2D+2P,q+n ] T in, Step 2.2: Set the location of the measuring point to be arranged As the independent variable, the estimated value of the differential entropy of the dispersion relation solution after the measurement points are arranged is used as the objective function T(D n ), expressed as: Where K is w q+n The number of non-zero elements in .

2. The method according to claim 1, characterized in that The step 1 is specifically as follows: Step 1.1: The time domain signal set {y i (t)} i＝1,…,D Perform fast Fourier transform and transform it into a frequency domain signal set {Y q } q＝1,…,Q , where Y q is the signal at each measuring point at frequency f q The complex amplitude at , q=1,…,Q; Step 1.2: According to the waveguide wave number k q+n and frequency f q+n The physical approximation relationship of is used to obtain a multi-task model for recovering the dispersion relation by multiple frequencies simultaneously. Step 1.3: According to the multi-task model constructed in step 1.2, random embedding is performed to transform it into a multi-task complex hierarchical sparse Bayesian learning problem; Step 1.4: Use multi-task complex hierarchical sparse Bayesian learning to solve the sparse problem in step 1.3; Step 1.5: Determine whether the number of measurement points D on the current structure has reached the target number D end , if it reaches, stop, according to the mean μ q+n Plot the dispersion curve, otherwise continue with the next steps.

3. The method according to claim 2, characterized in that The wave number k q+n and frequency f q+n The physical approximation relationship is: Among them, Δf is the rate of change of frequency, which is much smaller than the frequency f q The value of c q is the frequency f q The wave speed at .

4. The method according to claim 2, characterized in that: In the conversion process of step 1.3, the uncertainty of the predicted measured signal It is expressed as: in, is the uncertain weight vector, is the true weight vector; unknown prediction error e q+n The modeling error p q+n With the measurement error n q+n Composition, of which 5. The method according to claim 2, characterized in that: The solution process of step 1.4 is as follows: First, the expression of the sparse problem is transformed as follows: in, The following formula Perform a loop calculation; first calculate μ q+n and∑ q+n , and then use the calculated μ q+n and∑ q+n calculate and The calculated and To calculate μ q+n and∑ q+n , repeat the iteration until the values ​​of each variable no longer change significantly between the previous and next iterations; finally, w is obtained. q+n The mean μ of the posterior distribution q+n With the covariance matrix ∑ q+n .

6. The method according to claim 1, characterized in that The step three is specifically as follows: Step 3.1: Use the optimization algorithm to optimize the objective function T(D n ) is optimized to find the objective function T(D n ) The smallest measurement point position D n ; Step 3.2: The position D of the measuring point to be arranged is obtained by optimization in step 3.1 n Arrange measurement points on the actual structure to be measured and collect data. Return to step 1 after data collection is completed.

7. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

8. A computer-readable storage medium for storing computer instructions, characterized in that: When the computer instructions are executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images