A crack evaluation method based on linear frequency modulation lamb waves

Abstract

The application provides a crack evaluation method based on a linear frequency modulation Lamb wave, and is characterized by comprising the following steps: step S1, two piezoelectric transducers are pasted on a plate-shaped structural material to be detected, and are respectively used for transmitting an excitation signal x(t) and collecting a Lamb wave receiving signal y(t); step S2, Hilbert transform is performed on the excitation signal x(t) and the receiving signal y(t) to obtain analytical forms z x (t) and z y (t); step S3, mutual Wigner distribution calculation is performed to obtain a time-frequency distribution diagram; step S4, rearrangement and smoothing processing are performed on the time-frequency distribution diagram; step S5, a ridge line is extracted by using a Crazy-Climber algorithm, and then the ridge line connection algorithm is used to obtain a ridge line corresponding to each Lamb wave mode; step S6, according to a mode separation algorithm, each Lamb wave mode is separated from useless components to obtain a symmetric mode S0 and an asymmetric mode A0; step S7, two damage factors DI E and DI ρ are calculated; and step S8, according to linear relationships between the damage factors and crack length and crack width respectively, crack length and crack width are obtained.

Description

Technical Field

[0001] This invention relates to a crack evaluation method based on linear frequency modulated Lamb waves. Background Technology

[0002] A Lamb wave is a wave packet formed by the superposition of transverse and longitudinal waves propagating in thin plates or tubular materials. Due to its advantages such as low attenuation and sensitivity to structural changes, Lamb waves are widely used in materials health monitoring and non-destructive testing. However, the dispersion and multi-mode characteristics of Lamb waves make the propagation of the excitation signal in plate-like materials quite complex.

[0003] Current research primarily utilizes narrowband excitation to ensure mode purity, thereby simplifying waveform analysis. However, this method often results in limited signal information and requires multiple measurements of specific Lamb wave modes generated by narrowband excitation. In narrowband excitation, the signal used is limited to a fluctuating signal near a specific frequency. The excited guided wave signal is concentrated on the specific mode of interest to the engineer. Designing the narrowband excitation signal requires consideration of both frequency selection and window function modulation. When using a window function for windowing, it's crucial to ensure the response is primarily concentrated near the center frequency while also maximizing the acquisition of modally dispersed observation signals in the time domain. When modulating a sinusoidal signal, the number of periods must be carefully considered. Too few periods can lead to spectral leakage. While a large number of periods can maintain spectral characteristics, it can cause aliasing between different modes and even between sequentially generated modes in the received signal in the time domain, hindering signal analysis and processing.

[0004] Zhou Kai et al. controlled specific symmetric or antisymmetric modes by installing ultrasonic transducers at the two boundary surfaces of a flat material and controlling the phase. However, in some cases, it is desirable to obtain a complete response over a wider frequency band. Modal responses obtained over a wider bandwidth provide more complete observational information about the structure, thus helping us extract more valuable information from the received signal.

[0005] Wang Bo et al. utilized the characteristic that different modes of Lamb waves have varying sensitivities to different types of defects (cracks, indentations, pores, etc.). By exciting different Lamb wave signals for various defect detection scenarios, they then used linear frequency modulated (LFM) signals to perform structural flaw detection under different conditions. However, due to the large bandwidth and signal length of LFM signals, the dispersion phenomenon of Lamb wave signals is quite severe, resulting in relatively complex signals received by the sensor. This makes subsequent signal feature extraction and information conversion relatively complicated. Summary of the Invention

[0006] To address the aforementioned problems, this invention provides a crack evaluation method based on linear frequency modulated Lamb waves. The technical solution adopted in this invention is as follows:

[0007] This invention provides a crack evaluation method based on linear frequency modulated Lamb waves for evaluating the length and width of cracks in plate-shaped structural materials. The method comprises the following steps: Step S1, attaching two piezoelectric transducers to the plate-shaped structural material to be tested. One piezoelectric transducer transmits an excitation signal x(t), which is a linear frequency modulated signal, and the other piezoelectric transducer collects the received signal y(t) formed after the excitation signal x(t) passes through the plate-shaped structural material; Step S2, performing a Hilbert transform on the excitation signal x(t) and the received signal y(t) to obtain the analytical form z of the excitation signal x(t). x The analytical forms of z(t) and the received signal y(t) y (t); Step S3, the analytical form z of the excitation signal x(t) x The analytical forms of z(t) and the received signal y(t) y (t), perform mutual Wigner distribution calculation to obtain the time-frequency distribution map; Step S4, perform rearrangement and smoothing processing on the time-frequency distribution map to obtain the rearranged smoothed time-frequency map; Step S5, use the Crazy-Climber algorithm to extract ridges from the rearranged smoothed time-frequency map, and then use the ridge connection algorithm to obtain each Lamb wave mode of the rearranged smoothed time-frequency map; Step S6, use the mode separation algorithm to separate each Lamb wave mode from the useless components to obtain the pure Lamb wave mode, namely the symmetric mode S0 and the asymmetric mode A0; Step S7, calculate the mode energy coefficient DI based on the symmetric mode S0 and the asymmetric mode A0. E And mode energy ratio coefficient DI ρ Step S8, based on the mode energy coefficient DI E Mode energy ratio coefficient DI ρ By analyzing the linear relationships between crack length and crack width, the crack length and crack width can be obtained.

[0008] The crack evaluation method based on linear frequency modulated Lamb waves provided by this invention also has the following technical feature, wherein the specific expression of the excitation signal x(t) is: In the formula, A is the amplitude of the signal, B is the bandwidth of the linear frequency modulated signal, f0 is the starting frequency, t is the time variable, T is the duration of the linear frequency modulated signal, and ω(t) is the descending cosine window. The specific expression is as follows:

[0009] The crack evaluation method based on linear frequency modulated Lamb waves provided by this invention also has the following technical feature: the analytical form z of the excitation signal x(t) x The analytical forms of z(t) and the received signal y(t) y (t), the specific expression is: In the formula, z x (t) is the analytic signal of the excitation signal, z y Let (t) be the analytic signal of the received signal, and H[x(t)] be the Hilbert transform of the excitation signal x(t), specifically expressed as: In the formula, τ is the transform factor, and H[y(t)] is the Hilbert transform of the received signal y(t), specifically expressed as:

[0010] The crack evaluation method based on linear frequency modulated Lamb waves provided by this invention also has the following technical feature: in step S3, the calculation formula for the mutual Wigner distribution is as follows: In the formula, w is a frequency domain variable. The analytic signal after time shifting of the y component. It is the conjugate function of the x-component after time shift.

[0011] The crack evaluation method based on linear frequency modulated Lamb waves provided by this invention also has the following technical features: In step S4, the coordinates of each point in the time-frequency distribution map are sequentially represented as (n, k). After rearrangement and smoothing, the corresponding new coordinates of the rearranged and smoothed time-frequency map are obtained as follows: The calculation formula is:

[0012]

[0013]

[0014] In the formula, 'a' represents the infinitesimal element on the horizontal axis of the time-frequency distribution plot, 'b' represents the infinitesimal element on the vertical axis of the time-frequency distribution plot, and the unit element of the time-frequency distribution plot takes values ​​in the horizontal axis range of... The range of values ​​for the vertical axis is XWVD(n+a,k+b) is the time-frequency distribution amplitude of the point (n+a,k+b) on the time-frequency distribution map.

[0015] This invention provides a crack evaluation method based on linear frequency modulated Lamb waves, which also has the following technical features: Step S5 includes a Crazy-Climber algorithm and a ridge connection algorithm. The Crazy-Climber algorithm includes the following steps: First, randomly distribute a large number of climbers on the time-frequency matrix; second, according to the same Markov chain motion rules, make each climber climb towards the local energy maximum point with a high probability; third, reduce the temperature according to the idea of ​​simulated annealing to make the climbers lose their ability to move, and record the number of times each pixel is traversed to determine the ridge point. The ridge connection algorithm determines the connection method of the ridges according to the estimated ridge direction. Then, starting from the bottom of the time-frequency graph, after determining the maximum single search step size, the ridge points are connected to form ridges.

[0016] The crack evaluation method based on linear frequency modulated Lamb waves provided by this invention also has the following technical features: the mode separation algorithm extracts pure Lamb wave modes (XWVD) from multiple ridges extracted by the ridge connection algorithm. k (w,t), the calculation formula is: In the formula, ridge k Let p be the ridge line corresponding to a certain Lamb wave mode, p be the time-domain coordinate corresponding to the local minimum point on the left side of the ridge line, and q be the time-domain coordinate corresponding to the local minimum point on the right side of the ridge line.

[0017] The crack evaluation method based on linear frequency modulated Lamb waves provided by this invention also has the following technical features, wherein the mode energy coefficient DI E The specific expression is: In the formula, Let σ represent the time-frequency region corresponding to the asymmetric mode A0, σ be the integral element of the time-frequency region, XWVD(w,t) be the amplitude of each time-frequency point in the time-frequency distribution diagram, and C be the energy normalization coefficient.

[0018] The crack evaluation method based on linear frequency modulated Lamb waves provided by this invention also has the following technical feature, wherein the mode energy ratio coefficient DI ρ The specific expression is: In the formula, The integration interval represents the time-frequency region corresponding to the symmetric mode S0.

[0019] Invention Function and Effect

[0020] According to the crack evaluation method based on linear frequency modulated Lamb waves of the present invention, firstly, the present invention uses a linear frequency modulated signal as the excitation signal, and after signal analysis and mutual Wigner distribution calculation, obtains various Lamb wave modes, which can collect more crack information and thus provide a greater guarantee for structural safety.

[0021] Secondly, this invention utilizes the similarity between the received signal and the excitation signal, suppresses interference caused by negative frequency components by analyzing the signal, performs cross-Wigner distribution calculations using the analyzed signals of the excitation and received signals to obtain the time-frequency diagram, and finally calculates the mode energy coefficient DI. E And mode energy ratio coefficient DI ρ By analyzing cracks, the problem of time-domain shift caused by the mutual Wigner distribution can be solved.

[0022] Finally, this invention obtains the ridges corresponding to each Lamb wave mode using the Crazy-Climber algorithm and the ridge connection algorithm, and then uses a mode separation algorithm to separate the pure Lamb wave modes, obtaining pure direct S0 and A0 modes. This ensures the integrity and correctness of the separated signals. Attached Figure Description

[0023] Figure 1 This is a flowchart of a crack evaluation method based on linear frequency modulated Lamb wave in an embodiment of the present invention;

[0024] Figure 2 This is a schematic diagram of crack evaluation of a plate-like structure in an embodiment of the present invention;

[0025] Figure 3 This is a schematic diagram of the aluminum plate used in the experiment of this invention embodiment;

[0026] Figure 4 This is a schematic diagram of the excitation signal and the received signal in an embodiment of the present invention;

[0027] Figure 5 This is the original time-frequency diagram before mode separation in the case of crack 7 (crack width 1.5mm, depth 1.0mm, length 10.0mm) in this embodiment of the invention;

[0028] Figure 6 This is the time-frequency diagram obtained by mode separation in the case of crack 7 (crack width 1.5mm, depth 1.0mm, length 10.0mm) in the embodiment of the present invention;

[0029] Figure 7 This is a correlation diagram of crack size and damage factor in an embodiment of the present invention. Detailed Implementation

[0030] To make the technical means, creative features, objectives and effects of this invention easy to understand, the following describes in detail a crack evaluation method based on linear frequency modulated Lamb waves.

[0031] <Example>

[0032] Figure 1 This is a flowchart of a crack evaluation method based on linear frequency modulated Lamb waves, as described in an embodiment of the present invention. Figure 2 This is a schematic diagram of crack evaluation in a plate-like structure according to an embodiment of the present invention. Figure 3 This is a schematic diagram of the aluminum plate used in the experiment of this invention.

[0033] like Figure 1 , Figure 2 and Figure 3 As shown in Table 1, in step S1, this embodiment uses a 6061 aluminum plate with dimensions of 1000mm*1000mm*2mm. Eight different scribed lines are fabricated on the aluminum plate to simulate cracks of varying lengths and widths, thereby studying crack evaluation methods. The linear frequency modulated signal used in the experiment has a bandwidth of 200kHz, a starting frequency of 200kHz, a signal duration of 100μs, and a voltage of 10V. The waveforms of the received Lamb wave signals are averaged after 40 samplings. The specifications of the fabricated cracks are shown in Table 1.

[0034] Table 1 Crack Size Table

[0035] Crack parameter \ size Crack 1 Crack 2 Crack 3 Crack 4 Length 5.0 mm 10.0 mm 15.0 mm 20.0 mm Width 1.0 mm 1.0 mm 1.0 mm 1.0 mm Depth 1.0 mm 1.0 mm 1.0 mm 1.0 mm Crack parameter \ size Crack 5 Crack 6 Crack 7 Crack 8 Length 10.0 mm 10.0 mm 10.0 mm 10.0 mm Width 0.5 mm 1.0 mm 1.5 mm 2.0 mm Depth 1.0 mm 1.0 mm 1.0 mm 1.0 mm

[0036] Figure 4 This is a schematic diagram of the excitation signal and the received signal in an embodiment of the present invention.

[0037] like Figure 4 As shown, the experiment uses a one-transmitter-one-receiver detection method. Two piezoelectric transducers (PZTs) are attached to an aluminum plate. A computer generates an excitation signal, which is then sent to a waveform generator. The waveform generator amplifies the generated electrical excitation signal, and a piezoelectric sensor (PZT-5, φ8mm*1mm) at one end of the crack transmits the acoustic signal into the aluminum plate via the inverse piezoelectric effect. When the Lamb wave passes through the crack, its waveform changes accordingly. The piezoelectric sensor at the other end of the crack converts the received acoustic signal into an electrical signal via the direct piezoelectric effect, which is then acquired and received by an oscilloscope. Finally, the computer performs subsequent signal processing. Figure 4 (b) is the excitation signal waveform. Figure 4 (c) shows the received signal waveform.

[0038] Step S2: Perform a Hilbert transform on the excitation signal x(t) and the received signal y(t) to obtain the analytical form z of the excitation signal x(t). x The analytical forms of z(t) and the received signal y(t) y (t).

[0039] z x (t)=x(t)+j×H[x(t)]

[0040] z y (t)=y(t)+j×H[y(t)]

[0041] In the formula, z x (t) is the analytic signal of the excitation signal, z y H[x(t)] is the analytic signal of the received signal, H[x(t)] is the Hilbert transform of the excitation signal x(t), and H[y(t)] is the Hilbert transform of the received signal y(t).

[0042] Step S3, for z x (t) and z y (t) Perform the Cross-Wigner-Ville distribution (XWVD) calculation, the formula is as follows:

[0043]

[0044] In the formula, w is a frequency domain variable. The analytic signal after time shifting of the y component. It is the conjugate function of the x-component after time shift.

[0045] Step S4 involves rearranging and smoothing the time-frequency distribution plot to obtain a rearranged and smoothed time-frequency plot. Rearranging is used to smooth the time-frequency plot calculated from the pegged Wigner distribution. This involves determining the energy center of each unit cell in each time-frequency region and then reassigning the energy values. The aim is to reduce errors in the time-frequency plot caused by various factors such as instrument noise, PZT quality differences, wire welding gaps, and temperature drift.

[0046] Step S5 involves extracting ridges from the rearranged smoothed time-frequency graph using the Crazy-Climber algorithm, and then using a ridge connection algorithm to obtain the ridges corresponding to each Lamb wave mode in the rearranged smoothed time-frequency graph. This step includes the Crazy-Climber algorithm and the ridge connection algorithm.

[0047] For the Crazy-Climber algorithm: First, a large number of climbers are randomly distributed on the time-frequency matrix; second, following the same Markov chain motion rules, each climber is made to climb towards the local energy maximum point with a high probability; third, the temperature is reduced according to the idea of ​​simulated annealing to make the climbers lose their ability to move, and the number of times the pixel is traversed is recorded to determine the ridge point.

[0048] The ridge connection algorithm determines the connection method of the ridges according to the predicted ridge direction. Then, it starts searching from the bottom of the time-frequency graph, and after determining the maximum single search step size, it connects the ridge points to form ridges.

[0049] Step S6: Separate each Lamb wave mode from the useless components according to the mode separation algorithm to obtain the pure Lamb wave mode and two pure direct modes, namely the symmetric mode S0 and the asymmetric mode A0.

[0050] The formula for calculating the pure mode extraction is as follows:

[0051]

[0052] In the formula, ridge k Let p be the ridge line corresponding to a certain Lamb wave mode, p be the time-domain coordinate corresponding to the local minimum point on the left side of the ridge line, and q be the time-domain coordinate corresponding to the local minimum point on the right side of the ridge line.

[0053] Figure 5 This is the original time-frequency diagram before mode separation in the case of crack 7 (crack width 1.5mm, depth 1.0mm, length 10.0mm) in an embodiment of the present invention.

[0054] like Figure 5 As shown, Figure 5 (a) is the original three-dimensional time-frequency plot before mode separation. Figure 5 (b) is a two-dimensional time-frequency plot before mode separation. Figure 5 (c) is the time-domain plot before mode separation. From Figure 5 Parts (b) and (c) show that the S0 and A0 modes, each Lamb wave mode component excited by the linear frequency modulated signal, exhibits a sloping band shape. Since the S0 and A0 modes are relatively complex in the time domain diagram, they are not easy to separate. However, two bright spots can be seen in the two-dimensional time-frequency diagram, which makes them easy to separate. Therefore, the S0 and A0 modes, useless components, and the parts where the two direct modes overlap can be further separated in the two-dimensional time-frequency diagram.

[0055] Figure 6 This is the time-frequency diagram obtained by mode separation in the case of crack 7 (crack width 1.5mm, depth 1.0mm, length 10.0mm) in the embodiment of the present invention.

[0056] like Figure 6 As shown, the original time-frequency diagram ( Figure 5 (b) Partial pattern separation is performed to obtain Figure 6 (a) Part and Figure 6 (b) Part Figure 6 (a) shows the pure S0 mode extracted using the mode separation algorithm. Figure 6 (b) Pure A0 mode obtained by using the pattern separation algorithm.

[0057] Step S7: Calculate two damage factors (mode energy coefficients DI) based on the symmetric mode S0 and the asymmetric mode A0. E And mode energy ratio coefficient DI ρ To assess cracks.

[0058] After extracting the symmetric mode S0 and the asymmetric mode A0, DI is calculated respectively. E and DI ρ Based on the relationship between the damage factor and the crack parameters, Table 2 was obtained.

[0059] Table 2. Results of Damage Factors Corresponding to Crack Parameters

[0060] Crack length / mm 5.0 10.0 15.0 20.0 DI ρ ]] 2.9 7.2 9.55 15.52 Crack width / mm 0.5 1.0 1.5 2.0 DI E ]] 0.8869 0.6798 0.5803 0.3145

[0061] Step S8, based on the mode energy coefficient DI E Mode energy ratio coefficient DI ρ By analyzing the linear relationships between crack length and crack width, the crack length and crack width can be obtained.

[0062] Figure 7 This is a graph showing the relationship between crack size and damage factor in an embodiment of the present invention.

[0063] like Figure 7 As shown in part (a), the crack length is related to DI. ρ The data exhibits a linear positive correlation, with a slope of 0.804 for the fitted curve, and a Pearson correlation coefficient of 0.9864 between the original and fitted data. Figure 7 As shown in section (b), the crack width and DI E A linear negative correlation was observed, with the slope of the fitted curve being -0.363. The Pearson correlation coefficient between the original data and the fitted data was 0.9863. Finally, based on the model energy coefficient DI... E And mode energy ratio coefficient DI ρ The length and width of the crack in the aluminum plate were calculated.

[0064] Functions and effects of the embodiments

[0065] According to the crack evaluation method based on linear frequency modulated Lamb waves provided in this embodiment, firstly, this embodiment uses a linear frequency modulated signal as the excitation signal. After signal analysis and mutual Wigner distribution calculation, various Lamb wave modes are obtained, which can collect more crack information and thus provide structural safety assurance to a greater extent.

[0066] Secondly, this embodiment utilizes the similarity between the received signal and the excitation signal, suppresses interference caused by negative frequency components by analyzing the signal, performs cross-Wigner distribution calculations using the analyzed signals of the excitation and received signals to obtain the time-frequency diagram, and finally calculates the mode energy coefficient DI. E And mode energy ratio coefficient DI ρ By analyzing cracks, the problem of time-domain shift caused by the mutual Wigner distribution can be solved.

[0067] Finally, this embodiment obtains each Lamb wave mode using the Crazy-Climber algorithm and the ridge connection algorithm, and then uses a mode separation algorithm to separate the useless components of each Lamb wave mode into pure Lamb wave modes, obtaining pure direct S0 and A0 modes. This ensures the integrity and correctness of the separated signals.

[0068] The above embodiments are only used to illustrate specific implementations of the present invention, and the present invention is not limited to the scope of the description of the above embodiments.

Claims

1. A crack evaluation method based on linear frequency modulated Lamb waves, used to evaluate the length and width of cracks in plate-shaped structural materials, characterized in that, Includes the following steps: Step S1: Two piezoelectric transducers are attached to the plate-shaped structural material to be tested, one of which is used to transmit an excitation signal. The excitation signal is a linear frequency modulated signal, and another piezoelectric transducer is used to acquire the excitation signal. The received signal is formed after passing through the plate-shaped structural material. ; Step S2, for the excitation signal and the received signal The excitation signal is obtained by performing a Hilbert transform. parse form and the received signal parse form ; Step S3, for the excitation signal parse form and the received signal parse form We then performed cross-Wigner distribution calculations to obtain the time-frequency distribution map. Step S4: The time-frequency distribution map is rearranged and smoothed to obtain a rearranged and smoothed time-frequency map; Step S5: Use the Crazy-Climber algorithm to extract ridge points from the rearranged smooth time-frequency map, and then use the ridge line connection algorithm to obtain the ridge lines corresponding to each Lamb wave mode of the rearranged smooth time-frequency map. Step S6: Separate each Lamb wave mode from the useless components using the mode separation algorithm to obtain the pure Lamb wave mode, i.e., the symmetric mode. Asymmetric patterns ; Step S7, according to the symmetry pattern and the aforementioned asymmetric mode Calculate the mode energy coefficient and mode energy ratio coefficient ; Step S8, based on the mode energy coefficient The mode energy ratio coefficient The crack length and crack width are obtained by relating them to the linear relationship between the crack length and crack width, respectively.

2. The crack evaluation method based on linear frequency modulated Lamb wave according to claim 1, characterized in that: in, The excitation signal The specific expression is: ， In the formula, A is the amplitude of the signal, and B is the bandwidth of the linear frequency modulated signal. Let be the starting frequency, t be the time variable, and T be the duration of the linear frequency modulated signal. To lower the cosine window, the specific expression is: 。 3. The crack evaluation method based on linear frequency modulated Lamb wave according to claim 1, characterized in that: in, The excitation signal parse form and the received signal parse form The specific expression is: ， In the formula, The analytic signal of the excitation signal. The parsed signal of the received signal. The excitation signal The Hilbert transform is expressed as follows: ， In the formula, For transformation factor, For the received signal The Hilbert transform is expressed as follows: 。 4. The crack evaluation method based on linear frequency modulated Lamb wave according to claim 1, characterized in that: in, In step S3, the formula for calculating the mutual Wigner distribution is: ， In the formula, For frequency domain variables, The analytic signal after time shifting of the y component. It is the conjugate function of the x-component after time shift.

5. The crack evaluation method based on linear frequency modulated Lamb wave according to claim 4, characterized in that: in, In step S4, the coordinates of each point in the time-frequency distribution map are sequentially represented as (n, k). After rearrangement and smoothing, the corresponding new coordinates of the rearranged and smoothed time-frequency map are obtained as (n, k). The calculation formula is: ， In the formula, a is the infinitesimal element of the horizontal axis of the time-frequency distribution plot, b is the infinitesimal element of the vertical axis of the time-frequency distribution plot, and the unit element of the time-frequency distribution plot takes values ​​in the range of [ ]. ], the range of values ​​on the vertical axis is [ ], For the points on the time-frequency distribution map ( The time-frequency distribution amplitude of ).

6. The crack evaluation method based on linear frequency modulated Lamb wave according to claim 1, Its features are: in, Step S5 includes the Crazy-Climber algorithm and the ridge connection algorithm. The Crazy-Climber algorithm Includes the following steps: The first step is to randomly distribute a large number of Climbers on the time-frequency matrix; The second step is to follow the same Markov chain motion rules so that each climber is likely to climb toward the local energy maximum point; The third step involves lowering the temperature using simulated annealing to render the climber immobile, and then recording the number of times each pixel is traversed to determine the ridge points. The ridge connection algorithm determines the connection method of the ridges according to the estimated ridge direction, and then starts searching from the bottom of the time-frequency graph. After determining the maximum single search step size, the ridge points are connected to form ridges.

7. The crack evaluation method based on linear frequency modulated Lamb wave according to claim 6, characterized in that: in, The mode separation algorithm separates each Lamb wave mode from multiple ridges generated by the ridge connection algorithm, resulting in a pure Lamb wave mode. The calculation formula is: ， In the formula, Let p be the ridge line corresponding to a certain Lamb wave mode, p be the time-domain coordinate corresponding to the local minimum point on the left side of the ridge line, and q be the time-domain coordinate corresponding to the local minimum point on the right side of the ridge line.

8. The crack evaluation method based on linear frequency modulated Lamb wave according to claim 1, characterized in that: in, The mode energy coefficient The specific expression is: ， In the formula, The integration interval represents the time-frequency region corresponding to the asymmetric mode A0. For the integral infinitesimal element in the time-frequency region, Let C be the amplitude at each time-frequency point in the time-frequency distribution diagram, and C be the energy normalization coefficient.

9. The crack evaluation method based on linear frequency modulated Lamb wave according to claim 1, characterized in that: in, The mode energy ratio coefficient The specific expression is: ， In the formula, The integration interval represents the symmetry pattern. The corresponding time-frequency region.

Images