Multi-directional crack quantitative detection method and system based on sparse decomposition

Abstract

The present invention provides a multi-directional crack quantitative detection method based on sparse decomposition, which comprises the following steps: S1. Use a signal generator to emit pulse signals and read ultrasonic guided wave signals through an oscilloscope; S2. Construct an over-complete dictionary set for the mapping relationship between the waveform of ultrasonic guided wave signals and the propagation distance; S3. Separate the signals and establish the mapping relationship between the waveform components of ultrasonic guided wave signals and the propagation distance; S4. Locate and image the cracks through the propagation distance of the waveform components; S5. Locate all cracks and solve all crack boundary points, and connect all crack boundary points to obtain the final crack location and quantitative results. The present invention is based on the sparse decomposition algorithm. By mapping the detection signals to the propagation distances corresponding to each component, the identification of damage components without a baseline signal is realized, the error of common arrival time for distance identification is overcome, and the accuracy of crack location and quantitative analysis is improved.

Description

Technical Field

[0001] The invention relates to ultrasonic guided wave nondestructive testing technology, and in particular to a quantitative detection method for multi-directional crack damage based on sparse decomposition. Background Art

[0002] With the continuous development of modern industry, a large number of engineering structures are continuously put into use, among which a large number of plate-shell structures or plate-shell-like structures have appeared, such as the wing skin of passenger aircraft, the hull of ships, and the body of launch vehicles. Since these structural parts face complex and changeable external working conditions during use, there are many forms of damage during service, among which fatigue cracks are one of the most important forms of damage. Under the combined effects of long-term alternating loads and external environmental media, metal structures induce plastic deformation of the organizational structure of the material itself, generate tiny cracks, and significantly reduce the strength, plasticity, toughness and other mechanical properties of metal materials. As the cracks expand, it will eventually lead to structural failure, triggering structural damage accidents, causing significant economic and personnel losses, and having a great social impact.

[0003] Ultrasonic guided wave non-destructive testing (GWNDT) refers to the detection of discontinuities in the structure to be inspected by ultrasonic guided waves. By sticking piezoelectric materials to the surface of the structure to be inspected and applying a voltage excitation of a certain frequency to the piezoelectric material, due to the existence of the piezoelectric effect, it will be reflected on the upper and lower free surfaces inside the structure to produce stable ultrasonic guided waves. The characteristics of ultrasonic guided waves are small attenuation along the propagation path, and its sound field can be spread throughout the entire structure. When encountering defects during propagation on the surface of the structure, interface reflections will occur or sound speed and energy will be attenuated to reflect the location, size and other information of the defects. It has the advantages of fast and efficient, wide detection range and strong applicability. In response to the problem of multi-directional crack detection involved in this patent, ultrasonic guided waves are easy to operate and can reflect the location, direction and other information of cracks of different shapes. By analyzing the received signals, quantitative analysis of crack damage can be achieved.

[0004] By pasting sensors on the surface of the structure to be inspected, the guided wave signals passing through the damage are picked up. The obtained signals contain waveform components that reflect damage information. However, the complex interaction mechanism between the damage and the guided wave greatly affects the positioning accuracy. In addition, due to the existence of multimodal and dispersion characteristics of ultrasonic guided waves, it is difficult to directly analyze the waveform components corresponding to the damage from them. Usually, the signal containing damage is subtracted from the baseline signal to eliminate the direct wave component and the boundary reflection component in the test signal, so as to identify the damage component. In the actual application scenario, it is difficult to test the baseline signal. At the same time, due to the existence of the dispersion effect, after the waveform propagates a long distance, the existence of waveform distortion will cause a large error in the common method of determining the propagation distance through the propagation time, thus affecting the accuracy of the quantitative result. There is no method in the existing technology that can perform positioning and quantitative analysis on multi-directional cracks. Therefore, it is urgent to study a method that can solve the positioning and quantitative analysis of cracks. Summary of the Invention

[0005] Based on the above deficiencies of the existing technology, the purpose of the present invention is to provide a detection method that can realize the positioning and quantitative analysis of multi-directional crack damage without a baseline signal. This method is based on the sparse decomposition algorithm, which can accurately extract the signal components related to damage in complex signals and reduce the difficulty of signal analysis. It greatly reduces the difficulty of evaluating multi-directional crack damage, so as to conveniently and quickly perform non-destructive testing on plate and shell structures. In this patent, based on sparse decomposition, by constructing a guided wave waveform dictionary set corresponding to different propagation distances, the received signal is decomposed into different propagation distances to realize the separation of damage signal components. Through the extracted distance components, accurately locate the reflection and diffraction signal components related to cracks in the signal, and realize the positioning and quantitative analysis of crack damage. It can not only quickly and accurately determine the direction and boundary points of a single-direction crack, but also quickly and accurately determine the direction of a multi-directional crack such as a broken-line crack and the two outer edge points of the multi-directional crack.

[0006] The present invention provides a multi-directional crack damage quantitative detection method based on sparse decomposition, which includes the following steps:

[0007] S1. Fix the sensor on the surface of the structure to be inspected, use a signal generator to emit a pulse signal, excite the signal of a single sensor, and use the remaining sensors as receiving sensors to receive the ultrasonic guided wave signal and read the ultrasonic guided wave signal through an oscilloscope;

[0008] S2. Construct an over-complete dictionary set of the mapping relationship between the ultrasonic guided wave signal and the propagation distance;

[0009] S3. Decompose the ultrasonic guided wave signal by the sparse decomposition method and establish the mapping relationship between the ultrasonic guided wave signal and the propagation distance;

[0010] S4. Locate and image the crack based on the propagation distance of the ultrasonic guided wave signal, which specifically includes the following sub-steps:

[0011] S41. Divide the ultrasonic guided wave signal into a reflected signal component and a diffracted signal component: Compensate for the amplitude attenuation of the ultrasonic guided wave signal using the propagation distance and perform normalization processing. Divide the ultrasonic guided wave signal into a reflected signal component and a diffracted signal component according to a set threshold. When the amplitude of the ultrasonic guided wave signal is equal to or lower than the threshold, it is a diffracted signal, and when the amplitude of the ultrasonic guided wave signal is greater than the threshold, it is a reflected signal;

[0012] S42. Determine the direction of the crack by analyzing the characteristics of the reflected signal, which specifically includes the following sub-steps:

[0013] S421. Assume that the coordinates of the excitation point T are x1, y1, and the coordinates of T mirror are x2, y2. When the reflected wave reaches the receiving sensor, calculate the propagation distance of the reflected signal using the sparse decomposition method. Taking each receiving point where the receiving sensor of the reflected signal is located as the center of a circle, draw a circle with the propagation distance of the reflected signal as the radius and find the intersection points of multiple circles. This intersection point is the position of the mirror image point T mirror ;

[0014] S422. Calculate the coordinates of the midpoint D of the line connecting the excitation point T and the mirror image excitation point T mirror , and determine the perpendicular bisector of the line connecting the excitation point T and the mirror image excitation point T mirror . Calculate the slope k of the perpendicular bisector to determine the method of the perpendicular bisector. The direction of this perpendicular bisector is the direction of the crack. Among them, the coordinates of the midpoint D of the line connecting the excitation point T and the mirror image excitation point T mirror are calculated by the following formula:

[0015]

[0016]

[0017] The calculation formula for the slope k of the perpendicular bisector is:

[0018]

[0019] S43. Determine the size of the crack by analyzing the characteristics of the diffracted signal: Calculate the propagation distance of the diffracted signal using the sparse decomposition method. Taking the excitation point and each receiving point where the diffracted signal receiving sensor is located as the two foci of an ellipse and using the propagation distance of the diffracted signal as the major axis length of the ellipse to obtain multiple ellipses: The specific method is as follows:

[0020]

[0021] Among them

[0022] (d’) 2 = ||T mirror - T|| 2 = (x2 - r 3x ) 2 + (y2 - r 3y ) 2

[0023] Among them, d1 is the length of the major axis of the ellipse, d′ is the distance between the two foci of the ellipse, r 3x and r 3y are respectively the abscissa and ordinate of the receiving point where the diffraction signal receiving sensor is located;

[0024] The intersection point of the perpendicular bisector obtained in step S42 and the ellipse obtained in step S43 is the boundary point of the crack corresponding to the receiving sensor;

[0025] S5. For the ultrasonic signals received by all receiving sensors for the same crack, solve them respectively according to the above step S4 to obtain the crack direction and the position information of the boundary points corresponding to all receiving sensors, and connect all the boundary points to obtain the final crack direction and the size of the crack.

[0026] Preferably, the dictionary atoms of the dictionary set in step S2 are constructed according to the following formula:

[0027]

[0028] Among them, G(ω, r) is the dictionary atom, r is the propagation distance, ω is the angular frequency, S(ω) is the spectral representation of the excitation signal, k m (ω) is the wave number curve of the corresponding mode, H(ω, r) is the response function of the system, and m is the number of modes; the expression is:

[0029]

[0030] Among them, i is the imaginary number.

[0031] Preferably, the formula for sparse decomposition in step S3 is as follows:

[0032] Z = Φd

[0033] Among them, Φ is the dictionary set constructed in step S2, that is, the sensing matrix, d is the mapping relationship between the decomposed waveform component and the propagation distance, and Z is the signal to be decomposed;

[0034] After that, solve the above equation by the basis pursuit denoising method as follows:

[0035]

[0036] Among them, σ is a coefficient representing the noise level.

[0037] Preferably, in step S1, a signal generator is used to emit a 5-cycle narrowband pulse with a center frequency of 80 kHz.

[0038] Preferably, the threshold value set in step S41 is 40% of the maximum amplitude of the ultrasonic guided wave signal.

[0039] Preferably, the sensor in step S1 is a piezoelectric sensor.

[0040] Preferably, the cracks determined in step S5 include straight cracks and bent cracks.

[0041] Preferably, in step S41, if there are multiple intersection points, the centroid of the multiple intersection points is calculated as the position of the mirror point T. mirror of the position.

[0042] Preferably, on the other hand, the present invention provides a detection system, which includes a signal generator, a signal amplifier, a piezoelectric sensor, a signal excitation device and an oscilloscope. The piezoelectric sensor is bonded to the surface of the structure to be detected. The output end of the signal generator is connected to the input end of the signal amplifier, and the output end of the signal amplifier is connected to the piezoelectric sensor to generate ultrasonic guided waves to propagate in the structure to be detected. The signal excitation device is used to excite a certain piezoelectric sensor once, and the remaining piezoelectric sensors are used as receiving sensors to receive the ultrasonic guided wave signals and display and store them through the oscilloscope.

[0043] Compared with the prior art, the present invention has the following advantages:

[0044] (1) The present invention uses ultrasonic guided waves as the detection method. This detection method conducts experiments by pasting piezoelectric sensors, emits and receives reflected waves. The detection results have high accuracy and reliability, have the potential to be applied in complex environments, can be applied to the directional and quantitative analysis of multi-directional cracks in various scenarios, can accurately extract the signal components related to damage in complex signals in various scenarios, and reduce the difficulty of signal analysis. To a great extent, it reduces the difficulty of evaluating multi-directional crack damage, so as to facilitate and quickly conduct non-destructive testing on plate and shell structures.

[0045] (2) Based on the sparse decomposition algorithm, the present invention realizes the decomposition of a single component of a complex signal by mapping the detection signal to the propagation distances corresponding to each component, and simultaneously realizes the identification of damage components without a baseline signal. Based on sparse decomposition, by constructing a set of guided wave waveform dictionaries corresponding to different propagation distances, the received signal is decomposed into different propagation distances to achieve the separation of damage signal components. Through the extracted distance components, the reflection and diffraction signal components related to cracks in the signal are accurately located to achieve the positioning and quantitative analysis of crack damage. Moreover, this method can not only quickly and accurately determine the direction and boundary points of a single-direction crack, but also quickly and accurately determine the direction of a multi-direction crack such as a broken-line crack and the two outer edge points of the multi-direction crack.

[0046] (3) In the specific process of crack quantitative and directional analysis, the present invention takes into account the error brought by the dispersion effect to the signal arrival time analysis, overcomes the common error of solving the propagation distance by the wave packet arrival time, improves the accuracy of damage positioning and quantification, and thus can ensure the accuracy of crack damage positioning and quantification.

[0047] (4) The method of the present invention makes full use of the interaction mechanism between ultrasonic guided waves and damage, constructs a new information fusion strategy, realizes the positioning of crack reflection / diffraction points, and ensures the accuracy of the final analysis result. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1a is a schematic block diagram of the structure of the detection system of the present invention;

[0049] Figure 1b is a schematic diagram of the settings of each device in the detection system of the present invention;

[0050] Figure 2 is a schematic diagram of the simulation analysis of the present invention;

[0051] Figure 3 is a schematic diagram of the algorithm flow of the present invention;

[0052] Figure 4 is an example diagram of signal decomposition using the sparse decomposition algorithm of the present invention;

[0053] Figure 5 is one of the schematic diagrams of the imaging method of the present invention;

[0054] Figure 6 is another schematic diagram of the imaging method of the present invention;

[0055] Figure 7 is one of the crack imaging result diagrams of the present invention;

[0056] Figure 8 is another crack imaging result diagram of the present invention. Detailed implementation manners

[0057] To better understand the technical solution of the present invention, the following further describes in detail the specific implementation manners of the present invention in conjunction with the accompanying drawings and embodiments. The same reference numerals in the drawings denote elements with the same or similar functions. Although various aspects of the embodiments are shown in the drawings, the drawings do not have to be drawn to scale unless otherwise specified.

[0058] As Figure 1a shown, the detection system of the present invention specifically includes a signal generator 1, a signal amplifier 2, a piezoelectric sensor 3, an oscilloscope, and a signal excitation device 5. The piezoelectric sensor 3 is bonded to the surface of the structure to be detected. The output line of the signal generator is connected to the signal amplifier and then to the piezoelectric sensor 3 to generate ultrasonic guided waves that propagate in the structure to be detected. Single piezoelectric sensors are excited one by one, and the remaining piezoelectric sensors act as receiving sensors to receive signals and display and store them through the oscilloscope. Figure 1b It is a schematic diagram of the experimental sample, where the green markings are the positions of the excitation sensors, and the red dots are the positions of the remaining receiving sensors.

[0059] First, it is necessary to briefly analyze the action mechanism of damage and ultrasonic guided waves. As Figure 2 shown, it can be seen that the ultrasonic waves excited by the excitation end propagate forward and will generate reflection components and diffraction components after encountering the simulated crack damage. It can be seen from Figure 2 that the amplitude of the reflection component is much larger than that of the diffraction component, which also constitutes the basis for distinguishing the diffraction component and the reflection component subsequently. At the same time, under ideal conditions, it can be considered that after the reflection component is reflected by the crack, only the direction is changed, and there is no loss in amplitude. In other words, the position of the reflection signal component can be considered as the signal emitted by the virtual excitation end symmetric to the real excitation end along the crack direction, that is, T mirror in the following text. In addition, the diffraction signal component presents a circular shape during propagation, and its center is the crack tip. At the same time, for the sensor located on the other side of the crack relative to the excitation end, the direct wave component cannot be received, and only the diffraction signal component can be received. Therefore, the position of the crack tip can be judged through the diffraction signal component, and the edge points of the crack can be determined, thereby judging the size of the crack.

[0060] On the other hand, based on the above detection system, the main object of the present invention is to propose a multi-directional crack quantitative detection method based on sparse decomposition, which can especially be applied to the accurate orientation and quantitative measurement of multi-directional cracks in plate and shell structures. Specifically, it includes the following steps:

[0061] S1. Fix the piezoelectric sensor on the surface of the structure to be inspected. Use a signal generator to emit pulse signals, use a signal excitation device to excite a single piezoelectric sensor among them, and use the remaining piezoelectric sensors as receiving sensors to receive the ultrasonic guided wave signals, and read and save the ultrasonic guided wave signals through an oscilloscope.

[0062] S2. Construct an overcomplete dictionary set for the mapping relationship between the ultrasonic guided wave signal waveform and the propagation distance. The dictionary atoms of the dictionary set in step S2 are constructed according to the following formula:

[0063]

[0064] where r is the propagation distance, ω is the angular frequency, S(ω) is the spectral representation of the excitation signal, k m (ω) is the wave number curve of the corresponding mode, and H(ω, r) is the system response function, and the expression is:

[0065]

[0066] where i is the imaginary number.

[0067] S3. Separate the original ultrasonic signal through the sparse decomposition method, decompose the complex multi-component multi-modal signal, obtain the signal positions in the corresponding dictionary set of a single component, and then establish the mapping relationship between the ultrasonic guided wave single signal waveform component and the propagation distance of this component.

[0068] In this step, the formula for sparse decomposition is as follows:

[0069] Z = Φd

[0070] where Φ is the dictionary set constructed in step S2, that is, the sensing matrix, d is the mapping relationship between the decomposed waveform component and the propagation distance, and Z is the signal to be decomposed;

[0071] After that, solve the above equation through the basis pursuit denoising method, as follows:

[0072]

[0073] where σ is the coefficient representing the noise level.

[0074] S4. Define an imaging scheme based on the aforementioned interaction relationship between damage and ultrasonic guided waves, and perform crack location imaging through the propagation distances of different waveform components, specifically including the following sub-steps:

[0075] S41. Split the ultrasonic guided wave signal into a reflected signal component and a diffracted signal component. The specific method is as follows: After compensating for the amplitude attenuation of the ultrasonic guided wave signal according to the propagation distance, perform normalization processing. Set the discrimination threshold between the reflected signal and the diffracted signal to 40% of the maximum amplitude of the ultrasonic guided wave signal. When the amplitude of the signal is equal to or lower than this threshold, it is a diffracted signal; when the amplitude of the signal is greater than this threshold, it is a reflected signal.

[0076] S42. Analyze the characteristics of the reflected signal through the acoustic wave action relationship to determine the direction of the crack. It specifically includes the following sub-steps:

[0077] S421. Assume that the coordinates of the excitation point T are x1, y1, and the coordinates of T mirror are x2, y2. When the reflected wave reaches the receiving sensor, calculate the propagation distance of the reflected signal through the sparse decomposition method. Then, respectively, with the receiving points where multiple receiving sensors are located as the centers, draw multiple circles with the propagation distance of this reflected signal as the radius and find the intersection points of the multiple circles. This intersection point is the position of the mirror image point T mirror .

[0078] S422. Calculate the coordinates of the midpoint D of the line connecting the excitation point T and the mirror image excitation point T mirror , and determine the perpendicular bisector of the line connecting the excitation point T and the mirror image excitation point T mirror . Calculate the slope of this perpendicular bisector to obtain the direction of this perpendicular bisector. The direction of this perpendicular bisector is the direction of this crack. Among them, the calculation formula for the coordinates of the midpoint D of the line connection is:

[0079]

[0080]

[0081] The calculation formula for the slope k of the perpendicular bisector is:

[0082]

[0083] S43. Use the sparse decomposition method to calculate the propagation distance of the diffracted signal. Respectively, take the excitation point and the receiving points of multiple diffracted signals as the two foci of the ellipse and obtain multiple ellipses with the propagation distance of this diffracted signal as the major axis length of the ellipse:

[0084]

[0085] Among them

[0086] (d’) 2 = ||T mirror - T|| 2 = (x2 - r 3x ) 2 + (y2 - r 3y )2

[0087] Among them, d1 is the length of the major axis of the ellipse, d′ is the distance between the two foci of the ellipse, and r 3x and r 3y are the abscissa and ordinate of the receiving point where the diffraction signal receiving sensor is located, respectively;

[0088] The intersection point of the perpendicular bisector obtained in step S42 and the ellipse obtained in step S43 is the boundary point of the crack;

[0089] S5. For the reflection signals and diffraction signals received by all receiving sensors for the same crack, the above-mentioned step S4 is used for solution to obtain the position information of all crack boundary points. All boundary points are connected to obtain the final crack direction and the two outer edge points of the crack. The crack may be a single crack, or a polyline crack or a crack of other shapes composed of cracks in multiple directions. Specific embodiments

[0091] The method and specific application of the present invention will be further described below in conjunction with specific embodiments: As Figure 3 shown, this embodiment provides a method for quantitatively detecting multi-directional cracks in a plate and shell structure based on sparse decomposition, and its specific implementation steps are as follows:

[0092] S1. Bond multiple piezoelectric sensors on the surface of the structure to be inspected. Use a signal generator to emit a 5-cycle narrowband pulse with a center frequency of 80 kHz. Use a signal excitation device to excite a signal for a single piezoelectric sensor, and use the remaining piezoelectric sensors as receiving sensors to receive ultrasonic signals and read and save the ultrasonic guided wave signals through an oscilloscope.

[0093] S2. Construct an over-complete dictionary set for the mapping relationship between ultrasonic signals and propagation distances. The dictionary atoms are constructed according to the following formula:

[0094]

[0095] Among them, r is the propagation distance, ω is the angular frequency, S(ω) is the spectral representation of the excitation signal, k m (ω) is the wave number curve of the corresponding mode, and H(ω, r) is the response function of the system, which can be expressed as

[0096]

[0097] Among them, i is the imaginary number.

[0098] S3. Separate the ultrasonic signals through sparse decomposition and map them to the propagation distance. The formula for sparse decomposition is as follows:

[0099] Z = Φd (3)

[0100] Among them, Φ is the dictionary set constructed in step S2, that is, the sensing matrix. According to the formula shown in step S2, this matrix can be expressed in the following form:

[0101]

[0102] Among them, r is the range of possible propagation distances. This matrix combines the waveforms corresponding to all possible propagation distances to obtain the dictionary set. d is the mapping relationship between the decomposed waveform components and the propagation distance, and Z is the signal to be decomposed. Since the above equation is an underdetermined equation, the above equation needs to be solved by the Basis Pursuit Denoising method. The purpose of this step is to find the optimal signal combination that is closest to the target signal under the assumption that d is sparse. This process can be expressed as:

[0103]

[0104] Among them, σ is the coefficient representing the noise level. The result example is Figure 4 as shown:

[0105] S4. Locate and image the crack through the propagation distance information of the ultrasonic signal. The specific steps are respectively as Figure 5 and Figure 6 shown.

[0106] In this step, first, the ultrasonic signal needs to be classified. First, the amplitude is compensated according to the influence of the propagation distance on the amplitude attenuation of the ultrasonic guided wave signal and then normalized. The discrimination threshold between the reflected signal and the diffracted signal is set to 40% of the maximum amplitude of the ultrasonic guided wave signal, that is, 0.4. When the amplitude of the signal is equal to or lower than this threshold, it is the diffracted signal component. When the amplitude of the signal is greater than this threshold, it is the reflected signal component.

[0107] After that, use the reflected signal component to determine the direction of the crack, and finally determine the starting point and ending point of the crack, that is, the outer boundary points of the crack, through the diffracted signal. The size of the crack can be determined by these points, and the quantitative analysis of the crack is completed.

[0108] As Figure 5 shown, in this embodiment, it is assumed that the ultrasonic guided wave is emitted at point T and then reflected by the crack (as Figure 5As shown by the rectangle in [description], the reflected components are then received by the receiving sensors at points S1, S2, and S3. R1, R2, and R3 correspond to the respective reflection points. Assuming the crack is linear, due to the isotropic material properties of the structure under test, when the waveform interacts with the crack boundary potential, the waveform remains circular. The reflected wave after reflection can be regarded as the waveform generated by the mirror image point of T with the crack direction as the axis of symmetry, while the part interacting with the crack tip generates a new diffracted waveform. Based on this concept, first, a threshold is selected to divide the waveform signal into reflected signal components and diffracted signal components. When the amplitude of the signal is equal to or lower than the threshold, it is a diffracted signal component, and when the amplitude of the signal is greater than the threshold, it is a reflected signal component. By analyzing the characteristics of the reflected signal components, the direction of the crack can be found and the outer boundary points of the crack can be further determined.

[0109] Among them, the steps to determine the crack direction are as follows:

[0110] Assume the excitation point T = (x1, y1), the mirror image excitation point T mirror = (x2, y2), R i = (r i x, r l y), D i = (l i x, l l y)..

[0111] In this method, when the reflected wave reaches the receiving sensor, the propagation distance is calculated by the sparse decomposition method. Taking the spatial position of the receiving sensor that has received all the reflected signals as the center of the circle, and using the propagation distance of the reflected wave corresponding to each received signal as the radius to draw circles respectively, and finding the intersection points of multiple circles. This intersection point is the position of the mirror image point T mirror . Figure 5 There are three receiving sensors in [description], so three circles are obtained. The intersection point of these three circles is the position of the mirror image point T mirror . Figure 5 Only three receiving sensors for receiving reflected signals are shown schematically. Therefore, it is easy to find the intersection point of the three circles. However, in actual situations, the number of receiving sensors for reflected signals is much larger. Therefore, in actual situations, multiple circles may intersect at multiple points instead of simply intersecting at one point. In actual situations, if there are multiple intersection points for multiple circles, then calculate the centroid of the multiple intersection points as the position of the mirror image point T mirror . After that, by finding the perpendicular bisector of the excitation point and the mirror image excitation point, and calculating the slope of this perpendicular bisector, the direction of this perpendicular bisector is the direction of the crack. The calculation formula for the slope of the perpendicular bisector is shown in formula (7), and the calculation formulas for the coordinates of the midpoint of the perpendicular bisector are shown in formulas (8) and (9):

[0112]

[0113] Among them

[0114]

[0115]

[0116]

[0117] Through the calculation of this step, the direction of the crack can be determined. In the next step, the two edge points of the crack are calculated using the diffraction signal to determine the size of the crack. As Figure 6 shown, by taking the excitation point and each receiving point of the receiving sensor of the diffraction signal as the two foci of the ellipse respectively, and taking the propagation distance as the major axis length of the ellipse, multiple ellipses are obtained. Figure 6 The ellipse obtained by taking the diffraction signal receiving sensor S4 and the excitation point as the two foci and taking the diffraction signal propagation distance as the major axis of the ellipse is shown. Taking the excitation section as T and the receiving end as S as an example, the damage is located on the circumference of the ellipse described by the equation. The calculation formula of the ellipse is shown in Equations (10) and (11):

[0118]

[0119] Among them

[0120] (d’) 2 =||T mirror -T|| 2 =(x2 - r 3x ) 2 +(y2 - r 3y ) 2 (11)

[0121] After obtaining the ellipse, there will be intersections between the ellipse and the perpendicular bisector obtained in the previous step. The intersection points of the perpendicular bisector and the ellipse are the crack boundary points. By making ellipses through multiple diffraction signal sensors respectively, the two edge points of the crack, that is, the two tips of the crack, can be obtained. Combining the direction of the crack obtained from the perpendicular bisector and the edge points of the crack obtained in this step, the direction and size of this section of the crack can be obtained.

[0122] S5. Locate and quantitatively calculate all cracks:

[0123] In step S4, the direction and size of the damage (crack) reflected by a propagation distance are determined. In actual work, there are often multiple damages with different propagation distances, and some of the multiple damages of some cracks are connected and the directions of the multiple damages are not the same. Therefore, it is necessary to perform positioning and quantitative analysis and calculation on all damages to obtain the final crack orientation and the final two outer tips. Specifically, for the reflection signals and diffraction signals with different propagation distances received by all receiving sensors, the above step S4 is used for solution to obtain the direction and position information of the boundary points of all crack damages. Then, all the boundary points are connected to obtain the final crack direction and the two outer edge points of the crack.

[0124] In this embodiment, based on the above steps, first, all the signals collected in step S1 are collected. Taking 40% of the maximum amplitude, that is, 0.4, as the threshold, the signals are segmented into reflection signal components and diffraction signal components. With a step size of 0.1 mm and a propagation distance from 5 mm to 200 mm, a dictionary signal set is constructed, and the dictionary contains a total of 1951 signals. Using this dictionary signal set to decompose each signal in the experimental process, solve the propagation distances of different components and segment out the reflection signal components and diffraction components therefrom. Then, according to the method in step S4, first judge the crack orientation based on the reflection signal, and then, based on the known position information of the diffraction signal receiving sensors, solve the outer boundary points of all cracks. Finally, all the obtained boundary points are connected, and the remaining two outer boundary points are the outer edge points of the entire crack, that is, the final crack positioning and quantitative results are obtained. The results for straight cracks and curved cracks are respectively as Figure 7 and Figure 8 shown. Figure 7 is a simple straight crack. Figure 8 is a more complex zigzag curved crack. The rectangular frames shown in the figure are the actual crack positions, and the cross marks are the positions of the imaged crack boundary points obtained according to the method of this patent. It can be seen from the comparison in the figure that the directions and sizes of the two are consistent. Therefore, it can be inferred that the method of the present invention can accurately perform directional and quantitative analysis on cracks with low error, Figure 7 and Figure 8 the two types of cracks in have accurately determined the direction and size of the cracks.

[0125] The present invention uses ultrasonic guided wave signals as the basis for analysis. The entire method is simple and convenient to operate, and can adapt to complex and harsh field conditions. The crack measurement method uses a sparse decomposition algorithm with a fast operation rate, which can decompose complex signals and map them to the distance domain, making up for the shortcomings of inaccurate measurement results caused by the difficulty in obtaining certain data. At the same time, the present invention pioneered a damage detection method without baseline signals, combined with the action mechanism of cracks and ultrasonic guided waves, and proposed a damage imaging algorithm, which can achieve accurate positioning and quantitative evaluation of multi-directional crack damage.

[0126] Finally, it should be noted that the above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some or all of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-directional crack quantitative detection method based on sparse decomposition, characterized in that: It includes the following steps: S1. Fix the sensor on the surface of the structure to be inspected, use a signal generator to emit a pulse signal, excite the signal of a single sensor, use the remaining sensors as receiving sensors to receive the ultrasonic guided wave signal, and read the ultrasonic guided wave signal through an oscilloscope; S2. Construct an over-complete dictionary set of the mapping relationship between the ultrasonic guided wave signal and the propagation distance; S3. Decompose the ultrasonic guided wave signal by the sparse decomposition method, and establish the mapping relationship between the ultrasonic guided wave signal and the propagation distance; S4. Locate and image the crack through the propagation distance of the ultrasonic guided wave signal, which specifically includes the following sub-steps: S41. Divide the ultrasonic guided wave signal into a reflected signal component and a diffracted signal component: compensate the amplitude attenuation of the ultrasonic guided wave signal by the propagation distance and perform normalization processing, divide the ultrasonic guided wave signal into a reflected signal component and a diffracted signal component according to the set threshold, when the amplitude of the ultrasonic guided wave signal is equal to or lower than the threshold, it is a diffracted signal, and when the amplitude of the ultrasonic guided wave signal is greater than the threshold, it is a reflected signal; S42. Determine the direction of the crack by analyzing the characteristics of the reflected signal, which specifically includes the following sub-steps: S421. Assume that the coordinates of the excitation point T are x1, y1, and the coordinates of T mirror are x2, y2. When the reflected wave reaches the receiving sensor, the propagation distance of the reflected signal is calculated by the sparse decomposition method. Taking each receiving point where the receiving sensor of the reflected signal is located as the center of a circle, and using the propagation distance of the reflected signal as the radius to draw a circle, and find the intersection points of multiple circles. This intersection point is the position of the mirror image point T mirror ; S422. Calculate the coordinates of the midpoint D of the line connecting the excitation point T and the mirror excitation point T, and determine the perpendicular bisector of the line connecting the excitation point T and the mirror excitation point T. Calculate the slope k of the perpendicular bisector to determine the method of the perpendicular bisector. The direction of the perpendicular bisector is the direction of the crack. Among them, the coordinates of the midpoint D of the line connecting the excitation point T and the mirror excitation point T are calculated by the following formula: mirror Connect the midpoint D of the line, and determine the excitation point T and the mirror excitation point T mirror Of the perpendicular bisector, calculate the slope k of the perpendicular bisector to determine the method of the perpendicular bisector. The direction of the perpendicular bisector is the direction of the crack. Among them, the excitation point T and the mirror excitation point T mirror The coordinates of the midpoint D of the connection line are calculated by the following formula: The calculation formula of the perpendicular bisector slope k is: S43. Determine the size of the crack by analyzing the characteristics of the diffracted signal: calculate the propagation distance of the diffracted signal by the sparse decomposition method, take the excitation point and the receiving point where each diffracted signal receiving sensor is located as the two foci of the ellipse respectively, and use the propagation distance of the diffracted signal as the major axis length of the ellipse to obtain multiple ellipses. The specific method is as follows: Where (d’) 2 = ||T mirror - T|| 2 = (x2 - r 3x ) 2 + (y2 - r 3y ) 2 ; where d1 is the length of the major axis of the ellipse, d' is the distance between the two foci of the ellipse, r 3x and r 3y are the abscissa and ordinate of the receiving point where the diffraction signal receiving sensor is located, respectively; The intersection point of the perpendicular bisector obtained in step S42 and the ellipse obtained in step S43 is the boundary point of the crack corresponding to the receiving sensor; S5. Solve the ultrasonic wave signals received by all receiving sensors for the same crack respectively according to the above step S4, obtain the crack direction and the position information of the boundary points corresponding to all receiving sensors, and connect all the boundary points to obtain the final crack direction and the size of the crack.

2. The multi-directional crack quantitative detection method based on sparse decomposition according to claim 1, wherein: The dictionary atoms of the dictionary set in step S2 are constructed according to the following formula: Among them, G(ω,r) is the dictionary atom, r is the propagation distance, ω is the angular frequency, S(ω) is the spectral representation of the excitation signal, k m (ω) is the wavenumber curve of the corresponding mode, H m (ω, r) is the response function of the system, m is the number of modes; the expression is: Where, i is an imaginary number.

3. The multi-directional crack quantitative detection method based on sparse decomposition according to claim 1, characterized in that: The formula for sparse decomposition in step S3 is as follows: Z = Φd Where, Φ is the dictionary set constructed in step S2, that is, the sensing matrix, d is the mapping relationship between the decomposed waveform component and the propagation distance, and Z is the signal to be decomposed; After that, solve the above formula for sparse decomposition by the basis pursuit denoising method, as follows: Where, σ is the coefficient representing the noise level.

4. The multi-directional crack quantitative detection method based on sparse decomposition according to claim 1, characterized in that: In step S1, a 5-cycle narrowband pulse with a center frequency of 80 kHz is emitted by the signal generator.

5. The multi-directional crack quantitative detection method based on sparse decomposition according to claim 4, characterized in that: In step S41, the set threshold is 40% of the maximum amplitude of the ultrasonic guided wave signal.

6. The multi-directional crack quantitative detection method based on sparse decomposition according to claim 4, characterized in that: The sensor in step S1 is a piezoelectric sensor.

7. The multi-directional crack quantitative detection method based on sparse decomposition according to claim 4, characterized in that: The cracks determined in step S5 include straight cracks and curved cracks.

8. The multi-directional crack quantitative detection method based on sparse decomposition according to claim 4, characterized in that: In step S41, if there are multiple intersection points, calculate the centroid of the multiple intersection points as the mirror point T mirror 's position.

9. A detection system for the multi-directional crack quantitative detection method based on sparse decomposition according to claim 6, characterized in that: It includes a signal generator, a signal amplifier, a piezoelectric sensor, a signal excitation device and an oscilloscope. The piezoelectric sensor is bonded to the surface of the structure to be inspected. The output end of the signal generator is connected to the input end of the signal amplifier, and the output end of the signal amplifier is connected to the piezoelectric sensor to generate ultrasonic guided waves for transmission in the structure to be inspected. The signal excitation device is used to excite a certain piezoelectric sensor once, and the remaining piezoelectric sensors are used as receiving sensors to receive the ultrasonic guided wave signals and display and store them through the oscilloscope.

Images