Microcrack positioning method based on improved multiple signal classification

Through the improved multi-signal classification method, the second harmonic signal is extracted in combination with the pulse inversion method, and the covariance matrix and spatial spectrum are calculated, the problem of low microcrack positioning performance in traditional technology is solved, and microcracks and nonlinear damage is achieved The accurate positioning of the detection improves the detection resolution ability.

CN119023799APending Publication Date: 2024-11-26NANJING TECH UNIV
View PDF 0 Cites 4 Cited by

Patent Information

Application Number
CN202411152247.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-21
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The traditional linear Lamb wave technology has low positioning performance for microcracks and round hole damage, and the multi-signal classification algorithm has insufficient spatial spectrum resolution ability under undesirable conditions.

Method used

Using an improved multi-signal classification method, by setting multiple receiving sensor arrays and excitation sensors on a thin metal plate, transmitting regular and inverting Lamb wave signals, extracting second harmonic signals, calculating the covariance matrix of the received signal, and performing characteristics. Value decomposition, construct signal and noise eigenvectors, calculate spatial spectrum, and solve maximum values ​​to locate microcracks.

Benefits of technology

It realizes accurate positioning of microcracks and nonlinear damage, improves the resolution ability of defect detection, is suitable for the detection of conventional linear and nonlinear damage, and has good adaptability and practicality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119023799A_ABST
    Figure CN119023799A_ABST
Patent Text Reader

Abstract

The invention discloses a microcrack positioning method based on improved multiple signal classification, and belongs to the technical field of defect detection. The method comprises the following steps: firstly, arranging an excitation sensor and receiving sensor array on a to-be-detected plate, extracting second harmonics of response signals through superposition of forward excitation and reverse excitation, calculating a covariance matrix of received signals, calculating a noise feature vector of a covariance mean value, and calculating a spatial spectrum of a received signal array; micro-crack direction information is obtained by solving a spatial spectrum extreme value; and finally, a final positioning result is obtained by adopting a double-receiving sensor array and using a cross positioning method. According to the method, the improved multiple signal classification method is combined with the pulse reversal method to extract the second harmonic signal to realize microcrack positioning, the method is not only suitable for detection of conventional linear damage, but also can accurately position nonlinear damage such as microcracks, finite element simulation verification shows that the method has a good positioning effect under different damage types, and the method is suitable for popularization and application. And the practicability and the reliability are relatively high.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a micro-crack positioning method based on improved multiple signal classification, belonging to the technical field of defect detection. Background Art

[0002] Metal sheets are widely used in many engineering fields such as ships, bridges, and aerospace equipment. During long-term operation, due to continuous dynamic loads and external impacts, fatigue cracks may occur in metal plates, leading to damage to the overall structure. In order to prevent catastrophic failures, it is crucial to detect cracks as early as possible in the early stages of their occurrence. Lamb wave-based defects have the characteristics of long-distance propagation and low energy attenuation, and have been widely used in nondestructive testing and structural health monitoring.

[0003] The traditional Lamb wave technology based on linear analysis detects damage in metal sheets by analyzing the changes in signals in the time domain characteristics, such as using wavelet transform to extract the arrival time of scattered signals, and combining it with the elliptical imaging algorithm to perform multi-stage detection to achieve quantitative analysis of damage. Another method uses the extended wave packet at the optimal reconstruction frequency to calculate the damage index, and applies the fine time reversal method and defect probability detection reconstruction algorithm to obtain higher plate damage detection accuracy compared with the classical time reversal method. In recent years, nonlinear ultrasonic detection technology has received more and more attention, such as using nonlinear interaction primary Lamb waves for defect location, and detecting macroscopic surface cracks based on A0 mode Lamb wave linear ultrasonic parameters.

[0004] In the practical application of plate-like structures, two types of defects are usually encountered: microcracks and conventional linear damage. The traditional positioning imaging algorithm based on linear Lamb waves is sensitive to macroscopic linear damage, but the imaging effect of microcrack damage is poor. The microcrack imaging and positioning methods proposed in existing nonlinear research are also not ideal for the identification of microcracks. In view of these limitations of existing linear and nonlinear ultrasonic detection technologies, the method of the present invention realizes the positioning of the two types of damage through the interaction between the single-frequency excited S0 mode Lamb wave and the two different types of damage, microcracks and circular holes, that exist simultaneously in the plate. The traditional multiple signal classification algorithm has a low spatial spectrum resolution under non-ideal conditions. Since the resolution of the spatial spectrum is closely related to the subspace projection vector, the method of the present invention improves the resolution of the defect detection method by weighting the subspace to change the subspace projection vector. Summary of the invention

[0005] Technical problem to be solved by the invention

[0006] Aiming at the problem that the existing methods have low performance in locating linear microcracks and circular hole damage, the present invention proposes a microcrack locating method based on improved multiple signal classification.

[0007] Technical solution

[0008] To achieve the above object, the technical solution provided by the present invention is as follows:

[0009] A micro-crack positioning method based on improved multiple signal classification, comprising the following steps:

[0010] Step 1, set a first sensor array and a second sensor array composed of N receiving sensors at two different positions in the micro-crack damage area of the plate to be detected, set M excitation sensors at the specified position of the plate, satisfying M < N, and symmetrically emit a positive-phase Lamb wave excitation signal and an anti-phase Lamb wave excitation signal on the upper and lower surfaces of the plate at the excitation sensors, and collect the response signals of the first sensor array and the second sensor array;

[0011] Step 2, superimpose the positive-phase excitation response and the anti-phase excitation response of the receiving sensor array, and extract the second harmonic in the response signal;

[0012] Step 3, for the second harmonic signal of the first sensor array, calculate the receiving array direction matrix A(θ), and the method is:

[0013]

[0014] where θ1~θ M is the angle between the direction of the excitation signal emitted by the corresponding excitation sensors 1~M and the normal direction of the first receiving sensor of the receiving sensor array;

[0015] Step 4, construct the transmitted signal matrix S(t) and the noise matrix N(t), where S(t) = [s1(t) s2(t) … s M (t)] T , representing the transmitted signals of M excitation sensors, N(t) = [n1(t) n2(t) … n N (t)], representing the noise signals;

[0016] Calculate the receiving signal matrix X(t) of the sensor array, and the method is:

[0017] X(t) = A(θ)S(t) + N(t).

[0018] Step 5, calculate the covariance matrix R x of the receiving signal matrix X(t) obtained in Step 4, and the method is:

[0019] R x = E[XX H = E[(A(θ)S(t) + N(t))(A(θ)S(t) + N(t)) H

[0020] ​=AE[SS H ]A H +E[NN H ]

[0021] =AR S A H +R N

[0022] Where E[] represents the expected value, H represents the conjugate transpose, and R S =E[SS H ] is the signal correlation matrix, R N =σ 2 I is the noise correlation matrix, σ 2 is the variance of the noise signal, and I is the M×M unit matrix.

[0023] Step 6: The covariance matrix R obtained in step 5 x Perform eigenvalue decomposition and arrange the eigenvalues ​​from large to small to obtain eigenvalues ​​λ1,λ2,...,λ N (λ1≥λ2≥...≥λ N ) and the corresponding eigenvector Q = {q1,q2,...,q N}, where the M larger eigenvalues ​​correspond to the signal space, and the remaining NM smaller eigenvalues ​​correspond to the noise space;

[0024] Construct signal feature vector set Q s ={q1,q2,...,q M}, noise feature vector set Q N = {q M+1 ,q M+2 ,...,q N}, calculate the covariance matrix R x The corresponding noise feature vector E N .

[0025] Step 7: Based on A(θ) obtained in step 3 and E obtained in step 6 N Calculate the spatial spectrum P(θ) of the received sensor array signal as follows:

[0026]

[0027] Step 8, solving the maximum value of the spatial spectrum P(θ) obtained in step 7, the θ direction corresponding to the maximum value is the angle between the microcrack and the array normal direction;

[0028] Step 9, repeating steps 3-8 for the signals received by the second sensor array, and taking the intersection positions of the microcracks with respect to different direction angles of the two arrays as the final positioning results of the microcracks.

[0029] Furthermore, the second harmonic extraction method in step 2 is:

[0030] Assuming that the positive phase excitation is u1(t) and the negative phase excitation is u2(t), the relationship between the two satisfies: u1(t) = -u2(t). The system response obtained by the positive phase excitation and the system response obtained by the negative phase excitation are respectively:

[0031] f + (t)=a1u1(t)+a2u1 2 (t)+a3u1 3 (t)+…

[0032] f-(t)=-a1u1(t)+a2u1 2 (t)-a3u1 3 (t)+…

[0033] where f + (t) represents the system forward response, f - (t) represents the system anti-phase response, u(t) represents the excitation signal, a1, a2, a3 represent the nonlinear coefficients,

[0034] f + (t) and f - After adding (t), we get:

[0035] f + (t)+f - (t) = 2au1 2 (t)+2au1 4 (t)+…

[0036] Then the second harmonic in the response signal containing nonlinear characteristics is extracted.

[0037] Furthermore, the noise feature vector E in step 6 N The calculation method is:

[0038] Get the receiving array covariance matrix R x The inverse of the eigenvalues ​​constructs the inverse diagonal matrix Calculate the covariance matrix R x '=QΛ s -1 Q H , for the covariance matrix R x 'Perform eigenvalue decomposition and take the eigenvectors corresponding to the NM eigenvalues ​​with smaller eigenvalues ​​as the noise eigenvector set E N ={e M+1 ,e M+2 ,...,e N}.

[0039] Beneficial Effects

[0040] The present invention realizes microcrack location by extracting second harmonic signals through an improved multiple signal classification method combined with a pulse inversion method. The method is not only applicable to the detection of conventional linear damage, but also can accurately locate nonlinear damage such as microcracks, and has wide adaptability.

[0041] Finite element simulation verification shows that the method exhibits good positioning effect under different damage types and has strong practicability and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a flowchart of the method of the present invention;

[0043] Figure 2 It is a schematic diagram of a simulation model of an embodiment of the present invention;

[0044] Figure 3 The time domain waveform and spectrum diagram of the excitation signal in the embodiment of the present invention;

[0045] Figure 4 This is a waveform diagram of a time domain signal received by a receiving array according to an embodiment of the present invention;

[0046] Figure 5 Extracting echo signal graph for the receiving array of the present invention;

[0047] Figure 6 This is a result diagram of extracting the second harmonic component by the method of the present invention;

[0048] Figure 7 This is a diagram showing the experimental results of positioning using a dual receiving array according to the method of the present invention;

[0049] Figure 8 A schematic diagram of the results of microcrack positioning using the method of the present invention;

[0050] Fig. 9 This is a schematic diagram of the results of locating damaged circular holes using the method of the present invention. DETAILED DESCRIPTION

[0051] In order to further understand the content of the present invention, the present invention is described in detail in conjunction with the accompanying drawings and specific implementation methods.

[0052] The positioning method of the present invention is as follows: Figure 1 As shown in the figure, the method is mainly based on the multiple information classification algorithm to achieve microcrack location. The multiple signal classification algorithm is an algorithm for processing multiple signals. It can classify multiple signals received at different times, different spaces and different frequencies into different categories, so as to better understand and analyze these signals. This algorithm has a wide range of applications in many fields, such as medicine, biology, engineering, etc.

[0053] The core of the multiple signal classification algorithm is feature extraction. When processing signals, some features need to be extracted from them, which can be used to distinguish different signals. These features can be the frequency, amplitude, time domain characteristics of the signal, etc. By analyzing these features, the signals can be divided into different categories.

[0054] Assume that the linear receiving array consists of N array elements, there are M uncorrelated signal sources in the far field, and the angle between the signal source and the normal direction of the array element is θ, then the array receiving signal matrix X(t) can be expressed as:

[0055] X(t)=A(θ)S(t)+N(t) (1)

[0056] in

[0057]

[0058] Among them, θ1~θ M is the angle between the direction of the excitation signal emitted by the corresponding excitation sensors 1 to M and the normal direction of the first receiving sensor in the receiving sensor array, S(t) = [s1(t)s2(t)…s M (t)] T , represents the transmission signal of M excitation sensors, N(t)=[n1(t)n2(t)…n N (t)], which represents the noise signal.

[0059] Assume the received signal cross-correlation matrix is ​​R x =E[XX H ], assuming that the noise is an uncorrelated zero-mean additive Gaussian white noise, we can get:

[0060] R x =E[(A(θ)S(t)+N(t))(A(θ)S(t)+N(t)) H ]

[0061] =AE[SS H ]A H +E[NN H ]

[0062] =AR S A H +R N (3)

[0063] Where E[] represents the expected value, H represents the conjugate transpose, and R S =E[SS H ] is the signal correlation matrix, R N =σ 2 I is the noise correlation matrix, σ 2 is the variance of the noise signal, and I is the M×M unit matrix.

[0064] Now assume that the covariance matrix is R x with eigenvalues {λ1, λ2, …, λ N}, and the corresponding eigenvectors are {q1, q2, …, q N} in turn. In the case where the multiple signal sources are uncorrelated and the number of signal sources M < N, AR S A H is positive semi - definite, having M larger eigenvalues and N - M smaller eigenvalues with a value of σ 2 . Sort the eigenvalues of the matrix R x in descending order

[0065] λ1 ≥ λ2 ≥ … ≥ λ M ≥ … λ N , where M larger eigenvalues correspond to the signal space, and the remaining N - M smaller eigenvalues correspond to the noise space. The eigenvectors corresponding to the signal eigenvalues and the noise eigenvalues are the signal eigenvector and the noise eigenvector respectively.

[0066] According to the definitions of eigenvalues and eigenvectors, we have R x q i = λ i q i . For M < k < N, we have:

[0067] R x q k = λ k q k = σ 2 q k (4)

[0068] Substitute the representation of the aforementioned covariance matrix R x into the above formula, we get

[0069] R x q k = (ARA S A H + σ 2 I)q k

[0070] = ARA s A H q k + σ 2 q k (5)

[0071] Combining Equation (4) and Equation (5), we get ARA s A H q k = 0. Multiply both sides by to obtain:

[0072]

[0073] Formula (6) shows that the noise feature vector and the array direction selection vector are orthogonal. Using this orthogonal property, the noise feature vector E corresponding to the cross-correlation matrix of the array signal is solved. N Search the arrival azimuth of the signal source, generate a spatial spectrum, and estimate the direction of arrival.

[0074] The spatial spectrum function is defined as

[0075]

[0076] The spatial spectrum image will form a sharp spectrum peak at the incident angle, that is, the θ value corresponding to the maximum value of the spatial spectrum formula is the direction of the signal. The method of the present invention can obtain the angle between the microcrack position and the normal direction of the array by extracting the high-order harmonic signal of the nonlinear Lamb wave and analyzing the nonlinear parameters. In order to realize damage location imaging, it is necessary to use sensor arrays at two different positions to obtain different direction angles of the microcrack relative to the two arrays, and then use the cross-location method to realize the location of the microcrack.

[0077] Simulation Experiment Model

[0078] In this embodiment, the finite element simulation software ABAQUS is used to create a three-dimensional aluminum plate model, and the propagation process of lamb waves in different types of damage in the three-dimensional aluminum plate is solved and analyzed by using display dynamic analysis. The parameters of the aluminum plate model created in the simulation are shown in Table 1.

[0079] Table 1 Dimensions and material properties of the simulation model

[0080]

[0081] The microcracks in the model are set, where the upper and lower surfaces of the microcracks are set to be in hard contact, and the normal is a surface-to-surface interaction without displacement. In the simulation model, the center of the aluminum plate is taken as the origin, the length is taken as the x-axis direction, the width is taken as the y-axis direction, and the thickness is taken as the z-axis direction. In this embodiment, a sensor array 1 with A1 (-50, 50) as the first array element and 5mm as the array element spacing, and a sensor array 2 with B1 (-50, -20) as the first array element and 5mm as the array element spacing are set. Figure 2 shown.

[0082] Figure 3is a lamb wave dispersion curve. In order to reduce the multi-modal and dispersion problems of Lamb waves during propagation, the method of the present invention selects the S0 mode with better dispersion performance and faster propagation speed within the range of double frequency less than the cutoff frequency of other modes as the excitation signal. This embodiment uses a sine wave modulated by a 150kHz Hanning window function with five cycles as the excitation signal. Its waveform in the time domain and the corresponding spectrum are shown in Figure 3 As shown, in order to excite a single S0 mode lamb wave, a symmetrical lamb signal is applied at the coordinate (-50, 0).

[0083] In order to simplify the complexity while ensuring the accuracy and stability of the calculation, it is necessary to appropriately limit the grid size and time step. Assume that the minimum wavelength of the excitation signal is λ min The maximum and minimum sizes of the grid cells are L max and L min At the same time, let the time step be Δt and the speed of Lamb wave in the plate be c g In order to meet the above requirements, the above parameters need to meet:

[0084]

[0085] That is, 10 to 20 grid units are required in a minimum wavelength. This embodiment mainly uses the double frequency of 150kHz as the highest frequency. The maximum grid size calculated by the above formula is about 1.3mm. Based on the following formula, the maximum time step can be calculated to be about 9.6ns. The global grid size used in the simulation of this embodiment is 1mm, the time step is 2ns, and the unit type of the grid is C3D8R.

[0086] Pulse Inversion Method

[0087] The method of the present invention uses the pulse inversion method to extract the second harmonic in the response signal. The pulse inversion method was originally used in medicine to extract the second harmonic to achieve ultrasound contrast. The principle of the method is:

[0088] Since the damage is nonlinear, the excitation and damage of the system can be simplified as

[0089] f(t)=a1u(t)+a2u 2 (t)+a3u 3 (t)+… (10)

[0090] Where f(t) represents the response of the system, u(t) represents the excitation signal, and a1, a2, and a3 represent nonlinear coefficients.

[0091] Assume that the positive phase excitation is u1(t) and the negative phase excitation is u2(t), and the relationship between the two satisfies: u1(t) = -u2(t). Then the system response obtained by the positive phase excitation and the system response obtained by the negative phase excitation are respectively:

[0092] f + (t)=a1u1(t)+a2u1 2 (t)+a3u1 3 (t)+… (11)

[0093] f-(t)=-a1u1(t)+a2u1 2 (t)-a3u1 3 (t)+… (12)

[0094] f + (t) and f - (t) and we get:

[0095] f + (t)+f - (t) = 2au1 2 (t)+2au1 4 (t)+… (13)

[0096] From the above formula, we can see that after the pulse inversion technology is processed, the two system responses obtained by forward and reverse excitation are superimposed, the fundamental frequency and odd harmonic components in the response signal are filtered out, and the amplitude of the even harmonics is doubled. Therefore, the second harmonic in the response signal containing nonlinear characteristics can be successfully extracted by using the pulse inversion technology.

[0097] In order to verify the performance of the method of the present invention, a uniform linear array composed of M antenna elements is used, the axial element spacing is d, the central element of the receiving array is the reference element, the noise in the simulation is additive Gaussian white noise, and the root mean square error is used to measure the estimation error of two irrelevant signals with similar incident angles. The root mean square error of the estimated values ​​of the angles of D incident signals is defined as:

[0098]

[0099] Where P is the number of experiments, represents the estimated value of the incident angle of the Dth signal in the ρth experiment, θ m is the actual angle of the incident signal.

[0100] Damage localization experiment

[0101] The microcracks are set in the aluminum plate as zero-thickness cohesive units with a length l of 5 mm and a height h of 4 mm. First, the finite element simulation results of the microcrack model and the non-destructive model are compared and analyzed to observe the nonlinear effect of the lamb wave. Figure 4 (a) is the time domain signal waveform received by the first array element A1 of array 1 in the microcrack model. Figure 4 (b) in the figure is the time domain signal waveform received by the first array element A1 of array 1 in the lossless plate. It can be seen from the figure that there is an obvious echo signal in the waveform diagram of the microcrack model.

[0102] The results of using microcrack signals and lossless signals to extract echo signals are as follows: Figure 5 As shown in (a) of FIG. 1 , the echo signal segment is transformed by Fourier transform. Figure 5 As shown in (b), it can be found that in addition to the signal at the fundamental frequency, a nonlinear signal caused by microcracks is found at the second frequency.

[0103] The echo signal contains both the fundamental wave and the second harmonic, which is an aliased waveform. If you want to locate the microcracks, you need to filter out the fundamental wave signal that cannot characterize the microcracks. The pulse inversion method is used to extract the signal containing only the second harmonic component. Figure 6 As shown in the figure, after the original signal is processed by the pulse inversion method, not only the fundamental signal is effectively filtered out, but also the second harmonic component is well preserved.

[0104] By applying the method of the present invention to the filtered microcrack echo signal, the angle between the microcrack position and the sensor array normal can be determined. In order to accurately locate the damaged area and form an image, it is necessary to use sensor arrays A and B located at two different positions to receive signals. The two arrays measure the microcracks relative to their respective incident angles, and then the exact position of the microcracks is determined by the cross-positioning method. Figure 7 As shown, the two dotted lines in the figure represent the true incident direction angle of the microcrack at the coordinates (87.5, 7.5) relative to array A and array B. This method uses the information received by the array to calculate and display the location of the damage.

[0105] By using the method of the present invention and the traditional multi-signal classification method to analyze the spatial spectra obtained by arrays at different positions, the microcracks with coordinates of (87.5, 7.5) and (72.5, 2.5) are positioned and imaged. The imaging results are as follows: Figure 8 As shown in (a) to (d), Figure 8 (a) is the result of locating the microcrack at (87.5,7.5) by the method of the present invention; Figure 8 (b) shows the result of locating the microcrack at (87.5,7.5) using the traditional multi-signal classification method; Figure 8(c) in the figure is the result of locating the microcrack at (72.5, 2.5) by the method of the present invention. Figure 8 (d) in the figure is the result of locating the microcrack at (72.5, 2.5) by the traditional multi-signal classification method. The actual damage position in the figure is clearly marked with a '+' sign. It can be seen from the figure that the method of the present invention is an effective method for obtaining the microcrack position.

[0106] Fig. 9 The results of using the method of the present invention to locate the conventional linear damaged circular hole, where the circular holes have coordinates (67.5, -2.5) and (92.5, 2.5). It can be seen from the figure that the method of the present invention can also achieve circular hole positioning well.

[0107] The present invention and its embodiments are described schematically above, and the description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. Therefore, if a person skilled in the art is inspired by it and designs a structural method and an embodiment similar to the technical solution without creativity without departing from the purpose of the invention, they shall all fall within the protection scope of the present invention.

Claims

1. A microcrack location method based on improved multiple signal classification, characterized in that: It includes the following steps: Step S1, set a first sensor array and a second sensor array composed of N receiving sensors at two different positions in the microcrack damage area of the plate to be detected, and set M excitation sensors at the specified positions of the plate, where M < N. At the excitation sensors, a positive-phase Lamb wave excitation signal and an anti-phase Lamb wave excitation signal are symmetrically emitted on the upper and lower surfaces of the plate in sequence, and the response signals of the first sensor array and the second sensor array are collected; Step S2, superimpose the positive-phase excitation response and the anti-phase excitation response of the receiving sensor array, and extract the second harmonic in the response signal; Step S3, for the second harmonic signal of the first sensor array, calculate the receiving array direction matrix A(θ), and the method is as follows: Among them, θ1~θ M is the angle between the direction of the excitation signal emitted by the excitation sensors 1 to M and the normal direction of the first receiving sensor in the receiving sensor array; Step S4, construct the transmission signal matrix S(t) and the noise matrix N(t), where S(t) = [s1(t)s2(t)…s M (t)] T , represents the transmission signal of M excitation sensors, N(t)=[n1(t)n2(t)…n N (t)], represents the noise signal, Calculate the receiving signal matrix X(t) of the sensor array, and the method is as follows: X(t) = A(θ)S(t) + N(t); Step S5, calculating the covariance matrix R of the received signal matrix X(t) obtained in step S4 x , the method is: R x =E[XX H ]=E[(A(θ)S(t)+N(t))(A(θ)S(t)+N(t)) H ] =AE[SS H ]A H +E[NN H ] =ON S A H +R N Where E[] represents the expected value, H represents the conjugate transpose, and R S =E[SS H ] is the signal correlation matrix, R N =σ 2 I is the noise correlation matrix, σ 2 is the variance of the noise signal, I is the M×M unit matrix; Step S6, the covariance matrix R obtained in step S5 is x Perform eigenvalue decomposition and arrange the eigenvalues ​​from large to small to obtain eigenvalues ​​λ1,λ2,...,λ N ,λ1≥λ2≥...≥λ N , and the corresponding eigenvector Q = {q1,q2,...,q N }, where the M larger eigenvalues ​​correspond to the signal space, and the remaining NM smaller eigenvalues ​​correspond to the noise space; Construct signal feature vector set Q s ={q1,q2,...,q M }, noise feature vector set Q N = {q M+1 ,q M+2 ,...,q N }, calculate the covariance matrix R x The corresponding noise feature vector E N ; Step S7, based on A(θ) obtained in step S3 and E obtained in step S6 N Calculate the spatial spectrum P(θ) of the received sensor array signal as follows: Step S8, solve the maximum value of the spatial spectrum P(θ) obtained in Step S7, and the θ direction corresponding to this maximum value is the included angle between the microcrack and the normal direction of the array; Step S9, repeat Steps S3 - S8 for the signals received by the second sensor array, and use the cross position of the different direction angles of the microcrack relative to the two arrays as the final positioning result of the microcrack.

2. A microcrack location method based on improved multiple signal classification as claimed in claim 1, characterized in that: The method for extracting the second harmonic in Step S2 is as follows: Assume that the positive-phase excitation is u1(t), the anti-phase excitation is u2(t), and their relationship satisfies u1(t) = -u2(t). The system responses obtained from the positive-phase excitation and the anti-phase excitation are respectively: f + (t)=a1u1(t)+a2u1 2 (t)+a3u1 3 (t)+… f - (t)=-a1u1(t)+a2u1 2 (t)-a3u1 3 (t)+… where f + (t) represents the system forward response, f - (t) represents the system anti-phase response, u(t) represents the excitation signal, a1, a2, a3 represent the nonlinear coefficients, f + (t) and f - After adding (t), we get: f + (t)+f - (t)=2au1 2 (t)+2au1 4 (t)+… Then extract the second harmonic in the response signal containing non-linear characteristics.

3. A microcrack location method based on improved multiple signal classification as claimed in claim 1, characterized in that: The noise feature vector E in step S6 N The calculation method is: Get the receiving array covariance matrix R x The inverse of the eigenvalues ​​constructs the inverse diagonal matrix Calculate the covariance matrix R x '=QΛ s -1 Q H , for the covariance matrix R x 'Perform eigenvalue decomposition and take the eigenvectors corresponding to the NM eigenvalues ​​with smaller eigenvalues ​​as the noise eigenvector set E N ={e M+1 ,e M+2 ,...,e N }.

Citation Information

Cited By

  • Virtual electrode-based ship surface damaged area positioning method

    CN120522257A

  • Ship surface damaged area positioning method based on virtual electrodes

    CN120522257B

  • Micro-damage tomography method based on nonlinear ultrasonic static component damage index

    CN120870342A

  • Micro-damage tomography method based on nonlinear ultrasonic static component damage index

    CN120870342B