A multi-point stacking three-dimensional imaging method based on impact echo method

Abstract

The present invention discloses a multi-point superposition three-dimensional imaging method based on the impact echo method, which is applied to the field of non-destructive testing technology. The method comprises: using a four-channel sensor impact echo detection system to collect four-channel time domain signals for each measuring point in the detection area; constructing a three-dimensional discrete model of concrete corresponding to the four channels for each measuring point with the center of the detection area as the top surface center, dividing the model into uniform cubic grid units, and calculating the resonant frequency; normalizing the four-channel time domain signals and performing a fast Fourier transform to obtain a normalized frequency domain spectrum; extracting the amplitude of the cubic grid unit based on the resonant frequency, superimposing the amplitude matrices of the four channels at the same measuring point, and obtaining a defect amplitude enhancement matrix for each measuring point; superimposing the enhancement matrices to obtain a global intensity matrix, mapping the matrix to the HSV color space after normalization, and generating a three-dimensional energy density distribution model; and performing multi-dimensional slice analysis on the model. The present invention improves the defect recognition accuracy of the impact echo method.

Description

Technical Field

[0001] The present invention relates to the technical field of nondestructive testing, and in particular to a multi-point superposition three-dimensional imaging method based on an impact echo method. Background Art

[0002] In modern structural engineering, assessing the service life of a structure is a crucial aspect of safety. On-site inspection technologies for concrete structures provide crucial support for quality control of new construction projects, condition assessment of existing buildings, and decision-making regarding the repair of damaged structures. The impact echo method is currently widely used for defect detection, but it has limitations in signal processing. After fast Fourier transforming time-domain data to obtain frequency-domain data, the spectrum often exhibits various artifacts due to interference from complex factors such as internal concrete pores, rebar, and the external environment. This interferes with inspectors' ability to interpret defect frequencies. The current mainstream approach is to construct defect images based on the frequency domain. However, current research on defect visualization has largely focused on two-dimensional imaging or planar tomography. Two-dimensional images can only reflect the wavefield characteristics of a specific cross-section of the structure, making it difficult to analyze the interaction between stress waves and defects in three-dimensional space. This makes it impossible to accurately identify vertically symmetrically distributed damage, and results in inaccurate assessments of internal structural damage. Therefore, it is necessary to expand research on defect visualization to the three-dimensional level.

[0003] Therefore, how to provide a multi-point superposition three-dimensional imaging method based on the impact echo method that can effectively analyze the interaction mechanism between stress waves and defects in three-dimensional space, accurately identify damage symmetrically distributed along the vertical axis, and achieve accurate assessment of internal damage of the structure is an urgent problem that technicians in this field need to solve. Summary of the Invention

[0004] In view of this, the present invention proposes a multi-point superposition three-dimensional imaging method based on the impact echo method.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A multi-point stacking three-dimensional imaging method based on an impact echo method, comprising:

[0007] Step 1: With the four accelerometers as corner points and the impact point as the center, a four-channel sensor impact echo detection system based on an equilateral rectangular array is constructed. The four-channel time domain signal is collected when each measuring point in the detection area is used as the impact point.

[0008] Step 2: Taking the center of the test area as the top surface center, construct four concrete 3D discrete models corresponding to the four channels for each measuring point, divide them into uniform cubic grid units, and calculate the resonant frequency of each cubic grid unit;

[0009] Step 3: Normalize the amplitudes in the four-channel time domain signals and perform fast Fourier transform to obtain four sets of normalized frequency domain spectra corresponding to the four-channel time domain signals. Based on the resonant frequency, extract the amplitude of each cubic grid unit from the normalized frequency domain spectra to obtain the amplitude matrix corresponding to the four channels at each measuring point. Superimpose the amplitude matrices of the four channels at the same measuring point to obtain the defect amplitude enhancement matrix for each measuring point.

[0010] Step 4: Superimpose the defect amplitude enhancement matrices at all measurement points to obtain a global intensity matrix. After normalizing the amplitude in the global intensity matrix again, map it to the HSV color space to generate a three-dimensional energy density distribution model.

[0011] Step 5: Perform multi-dimensional slice analysis on the 3D energy density distribution model to quantitatively analyze the defects.

[0012] Optionally, in step 1, the detection area is a grid detection area arranged in a grid form above the area where defects exist inside the concrete structure; wherein the corner points of each grid are used as measurement points.

[0013] Optionally, in step 2, the center of the detection area is used as the top surface center, and four concrete three-dimensional discrete models corresponding to the four channels are constructed for each measuring point, specifically:

[0014] The top and bottom side lengths of the concrete 3D discrete model are the side lengths of the grid detection area plus the preset length, and the depth is the same as the actual thickness of the concrete structure.

[0015] Optionally, in step 2, the side length of the cube grid unit satisfies the Nyquist sampling criterion, specifically:

[0016]

[0017] in, is the side length of the cubic grid cell; is the propagation velocity of the P wave in the concrete being tested; is the sampling time interval.

[0018] Optionally, in step 2, the resonant frequency of each cubic grid cell is calculated as follows:

[0019]

[0020] in, The center coordinates of the 3D concrete discrete model of the i-th channel collection point at the k-th measurement point are The resonant frequencies of the cubic grid cells, i = 1, 2, 3, 4, correspond to the four-channel accelerometer; is the propagation velocity of the P wave in the concrete being tested; is the Euclidean distance between the center of the cube grid unit and the impact point at the kth measurement point; is the Euclidean distance between the center of the cube grid unit and the collection point of the i-th channel at the k-th measurement point.

[0021] Optionally, in step 3, based on the resonant frequency, the amplitude of each cubic grid unit is extracted from the normalized frequency domain spectrum to obtain the amplitude matrix corresponding to the four channels at each measurement point, specifically:

[0022] Each measurement point has four normalized frequency domain spectra, corresponding to the four-channel time domain signals of the measurement point;

[0023] Each measuring point has four concrete three-dimensional discrete models, which are constructed based on the four channels under the measuring point;

[0024] Each concrete three-dimensional discrete model contains a plurality of cubic grid units, and each cubic grid unit has a resonant frequency;

[0025] For the amplitude of each cubic grid unit in the concrete 3D discrete model at each measuring point, the normalized frequency domain spectrum corresponding to the time domain signal of the channel at the measuring point is determined according to the measuring point and the corresponding channel in the concrete 3D discrete model. The amplitude of the cubic grid unit is extracted from the normalized frequency domain spectrum to obtain the amplitude matrix corresponding to the four channels at each measuring point.

[0026] Optionally, in step 3, the amplitude matrices of the four channels at the same measuring point are superimposed to obtain the defect amplitude enhancement matrix of each measuring point, as follows:

[0027]

[0028] in, is the defect amplitude enhancement matrix of the kth measuring point; For the kth measurement point The amplitude matrix of the channels.

[0029] Optionally, in step 4, the defect amplitude enhancement matrices at all measurement points are superimposed to obtain a global intensity matrix as follows:

[0030]

[0031] in, is the global intensity matrix; is the number of measuring points; is the defect amplitude enhancement matrix of the kth measuring point.

[0032] Optionally, in step 4, the normalized amplitude elements are mapped to the HSV color space, specifically:

[0033] The normalized amplitude elements are mapped to the HSV color space in the order of red-yellow-green-blue, representing the amplitude from high to low, as follows:

[0034]

[0035] in, is the magnitude element in the global intensity matrix.

[0036] Optionally, in step 5, a multi-dimensional slice analysis is performed on the three-dimensional energy density distribution model to quantitatively analyze the defects, specifically:

[0037] Perform orthogonal slicing along the X, Y, and Z axes on the 3D energy density distribution model to extract 2D projections 、 、 , and conduct multi-dimensional analysis of defect conditions through the principle of tomography.

[0038] It can be seen from the above technical solutions that, compared with the prior art, the present invention proposes a multi-point superposition three-dimensional imaging method based on the impact echo method. By synchronously collecting impact echo signals through multiple channels, combined with frequency domain feature extraction and amplitude enhancement mechanism, the expansion of traditional two-dimensional imaging to three-dimensional imaging is achieved, thereby more intuitively and accurately presenting the distribution range and burial depth of internal defects in concrete. At the same time, the generated three-dimensional model can be cut into sections along orthogonal planes in the X, Y, and Z directions to achieve multi-dimensional analysis of the defect status, providing more comprehensive defect information than traditional two-dimensional images, effectively overcoming the situation where traditional two-dimensional imaging only presents the projection of defects in one direction and is insufficient for identifying complex defects, thereby improving the accuracy of defect recognition by the impact echo method. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0040] Figure 1 Schematic diagram of the method of the present invention.

[0041] Figure 2 Schematic diagram of the grid detection area layout and acquisition sequence of the four-channel sensor rectangular array impact echo detection system of the present invention.

[0042] Figure 3 It is a schematic diagram of the concrete three-dimensional discrete model structure of the present invention.

[0043] Figure 4 The present invention takes the kth measurement point, i=1 channel collection point as an example, to calculate the resonant frequency of each cubic grid unit and assign an amplitude schematic diagram.

[0044] Figure 5 This is a schematic diagram of the principle of forming the "spherical effect" of the present invention.

[0045] Figure 6 This is a schematic diagram of the defect amplitude enhancement matrix superposition of the present invention.

[0046] Figure 7 This is a cross-sectional schematic diagram of extracting a three-dimensional imaging model according to the present invention. DETAILED DESCRIPTION

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0048] Example 1:

[0049] Embodiment 1 of the present invention discloses a multi-point superposition three-dimensional imaging method based on the impact echo method, such as Figure 1 Shown, including:

[0050] Step 1: With the four accelerometers as corner points and the impact point as the center, a four-channel sensor impact echo detection system based on an equilateral rectangular array is constructed. The entire system uses the center of the rectangle, i.e., the impact point, as the moving base point and collects four-channel time domain signals when each measuring point in the detection area is used as the impact point.

[0051] like Figure 2 As shown, the inspection area is a grid-like inspection area arranged above the defect area inside the concrete structure, and the inspection area is sufficient to cover the entire defect; wherein the corner points of each grid serve as measurement points. In the present invention, the size of each grid is 3 cm x 3 cm.

[0052] Specifically, based on the orthogonal grid layout principle, a grid detection area of ​​size N×M is laid out above the defect, and the center of the grid detection area is taken as the origin. Establish a rectangular coordinate system and record the coordinates of the impact point each time data is collected for subsequent calculations.

[0053] When collecting data, the four-channel rectangular array sensor acquisition system uses the center of the rectangle (the impact point) as the moving base point. The impact point is placed on the measuring point for impact echo detection, and sampling is performed in the order from measuring point 1 to measuring point n, and measuring line 1 to measuring line n. With the center of the detection area as the coordinate origin, the coordinates of the impact point (the measuring point) and the acquisition point are recorded according to the grid size information for each impact echo detection. Since the distance between the acquisition point and the impact point is fixed, the coordinates of the acquisition point can be calculated from the coordinates of the impact point.

[0054] The coordinates of each channel sensor node are determined by rigid geometric relationships as follows:

[0055]

[0056] in, is the impact point coordinate, which is consistent with the measurement point coordinate; The side length of the rectangular array of four-channel acceleration sensors is 4 cm in this invention. During the test, the impact point is placed on the measuring point, and the impact point coordinates and the four-channel time domain signal are recorded by the synchronous trigger device during each test. , ensuring temporal and spatial consistency.

[0057] Step 2: With the center of the inspection area as the center of the top surface, construct four concrete 3D discrete models corresponding to the four channels for each measuring point and divide them into uniform cubic grid units. Calculate the resonant frequency of each cubic grid unit. All discrete models will be used as empty matrices for subsequent amplitude storage.

[0058] The center of the detection area As the center of the top surface, four concrete three-dimensional discrete models corresponding to the four channels are constructed for each measuring point, specifically:

[0059] The side lengths of the top and bottom surfaces of the concrete three-dimensional discrete model are the side lengths of the grid detection area plus a preset length. In the present invention, the preset length is 5 cm, and the depth is the same as the actual thickness H of the concrete structure.

[0060] The side lengths of the cubic grid cells satisfy the Nyquist sampling criterion, which is:

[0061]

[0062] in, is the side length of the cubic grid cell; is the propagation velocity of the P wave in the concrete being tested; is the sampling time interval. Considering that the propagation speed of P wave in common concrete is about 4000m / s, Usually taken as 2μs, so ,Will The design of 4mm can meet most engineering needs.

[0063] The present invention uses imaging algorithms to discretize concrete into a number of cubic grid cells, with the goal of designing an empty matrix to store the amplitudes extracted from the corresponding frequency domain data for each grid cell. Each cubic grid cell is denoted as , Define an empty matrix (k is the measurement point number, i is the collection point number, are the coordinates of the center of the cube grid, The center coordinates of the 3D concrete discrete model of the i-th channel collection point at the k-th measurement point are Cube grid cells;

[0064] for the reason The local zero matrix of the i-th channel acquisition point under the k-th measurement point is composed of the zero elements at Include elements), which are used to calculate the resonant frequency and store the relevant amplitude in the subsequent steps, and perform matrix superposition.

[0065] Based on the coordinates of each impact point , and the corresponding four-channel acquisition point coordinates , calculate each cubic grid cell in different three-dimensional discrete models Resonant frequency ,as follows:

[0066]

[0067] in, The center coordinates of the 3D concrete discrete model of the i-th channel collection point at the k-th measurement point are The resonant frequencies of the cubic grid cells, i = 1, 2, 3, 4, correspond to the four-channel accelerometer; is the propagation velocity of the P wave in the concrete being tested; is the Euclidean distance between the center of the cube grid unit and the impact point at the kth measurement point; is the Euclidean distance between the center of the cube grid unit and the collection point of the i-th channel at the k-th measurement point.

[0068] like Figure 4 As shown, taking the kth measuring point, i=1 channel acquisition point as an example, based on the impact point coordinates , and the corresponding acquisition point coordinates of channel i=1 , calculate each cube grid unit The resonant frequency of channel i=1 .

[0069] Step 3: Normalize the four-channel time domain signal The amplitude in the four-channel time domain signal is obtained by fast Fourier transform to obtain four sets of normalized frequency domain spectra corresponding to the four-channel time domain signal. (Each measurement point has four sets of normalized frequency domain spectra corresponding to its four-channel time domain signals), based on the resonant frequency , the amplitude of each cubic grid unit is extracted from the normalized frequency domain spectrum to obtain the amplitude matrix corresponding to the four channels at each measuring point. The amplitude matrices of the four channels at the same measuring point are superimposed to obtain the defect amplitude enhancement matrix of each measuring point.

[0070] Based on the resonant frequency, the amplitude of each cubic grid unit is extracted from the normalized frequency domain spectrum, and the amplitude matrix corresponding to the four channels at each measurement point is obtained, specifically:

[0071] Each measurement point has four normalized frequency domain spectra, corresponding to the four-channel time domain signals of the measurement point;

[0072] Each measuring point has four concrete three-dimensional discrete models, which are constructed based on the four channels under the measuring point;

[0073] Each concrete three-dimensional discrete model contains a plurality of cubic grid units, and each cubic grid unit has a resonant frequency;

[0074] For the amplitude of each cubic grid unit in the concrete 3D discrete model at each measuring point, the normalized frequency domain spectrum corresponding to the time domain signal of the channel at the measuring point is determined according to the measuring point and the corresponding channel in the concrete 3D discrete model. The amplitude of the cubic grid unit is extracted from the normalized frequency domain spectrum to obtain the amplitude matrix corresponding to the four channels at each measuring point.

[0075] like Figure 4 As shown, at the kth measurement point, the collection point of channel i=1 For example, the resonant frequency of each cubic grid unit in the three-dimensional discrete model of the i=1 channel acquisition point is calculated. Then, from the normalized frequency domain spectrum Extract the resonant frequency Matched amplitude , and assigned to the corresponding cubic grid cells . Assign these magnitudes to the zero matrix In this way, we can get the amplitude matrix of i=1 at the kth measurement point, and based on this process, we can get the amplitude matrices of the other three i=2, 3, and 4 channel acquisition points respectively. The four amplitude matrices are superimposed to get the defect amplitude enhancement matrix of the kth measurement point. .

[0076] The present invention proposes a dynamic amplitude matching mechanism based on resonant frequency, realizes the automatic association between grid units and corresponding frequency components by traversing frequency domain data, and establishes a multi-dimensional mapping relationship between spatial coordinates, frequency characteristics and signal amplitude.

[0077] By superimposing the amplitude matrices of the four channels at the same measuring point, the defect amplitude enhancement matrix of each measuring point is obtained as follows:

[0078]

[0079] in, is the defect amplitude enhancement matrix of the kth measuring point; is the amplitude matrix of the i-th channel at the k-th measurement point.

[0080] Step 4: Superimpose the defect amplitude enhancement matrices at all measurement points to obtain a global intensity matrix. After normalizing the amplitude in the global intensity matrix again, map it to the HSV color space to generate a three-dimensional energy density distribution model.

[0081] The defect amplitude enhancement matrices at all measurement points are superimposed to obtain the global intensity matrix, as follows:

[0082]

[0083] in, is the global intensity matrix; is the number of measuring points; is the defect amplitude enhancement matrix obtained at the kth measuring point.

[0084] like Figure 5 As shown in the figure, from the calculation formula of the resonant frequency of each grid unit, it can be seen that if only the amplitude is extracted from a single set of frequency domain data to generate the intensity matrix for three-dimensional imaging, when the size of the grid unit is small and When is a certain constant, a large number of cubic grid cells in the discrete model will have the same resonant frequency. The line connecting these grid cells can be approximated as a hemisphere with a radius of R and a center O between the impact point and the collection point. The resonant frequencies of the grid cells passed by the hemisphere trajectory are almost equal. After extracting the amplitude from the spectrum, these grid cells and their adjacent cells will appear the same or similar colors in the energy density distribution diagram, i.e., the three-dimensional imaging of the defect, because the corresponding amplitude differences are small. A sphere will appear in the image, i.e., the "spherical effect". Therefore, when only a single set of frequency domain data is used for imaging, the resonant frequency of a large number of grid cells will be equal to the defect frequency. At this time, a sphere will appear in the image that passes through the grid of the defect area, making it difficult to accurately determine the defect location from the image. The three-dimensional model has a poor visualization effect for internal defects. In order to weaken the interference of the "spherical effect", the characteristics of the defect position can be highlighted by superimposing multiple sets of intensity matrices. The principle is: in the frequency domain data obtained at different measuring points, the defect frequency and amplitude are different, but because the defect position is a strong reflection area, the amplitude of the defect in the frequency domain of each measuring point is higher than that of the defect-free area. Therefore, when a single set of frequency domain data at different measuring points is used for imaging, there will always be a sphere passing through the defect position. When the number of measuring points, that is, the number of sphere centers, is increased to n, it is equivalent to increasing the number of spheres in three-dimensional imaging to n. Changing the measuring point, that is, changing the impact point or acquisition point position, is equivalent to changing the position of the sphere center. The coordinates of each measuring point are the sphere center. The positions are different, so when the intensity matrices are superimposed, these spheres will partially overlap, and the overlapping area is the defect location. The overlap of n spheres is equivalent to the amplitude superposition of the cubic grid cells in the overlapping area, that is, the amplitude at the defect location is enhanced n times. At this time, the reflection intensity at the defect location is significantly different from that at the non-defective location. When the superimposed amplitude is renormalized and color mapped, the defect boundary and burial depth can be clearly presented in the three-dimensional discrete model. Based on the above description, when a four-channel sensor is used to collect signals at a certain measuring point, step 3 is essentially equivalent to changing the position of the sphere center four times at the measuring point. This step can be regarded as a preliminary weakening of the "spherical effect". Therefore, when there are K measuring points, 4K amplitude matrices will actually be generated, that is, 4K "spheres" passing through the defect area will appear in the image. After superposition, the influence of the "spherical effect" can be effectively weakened.

[0085] This method overcomes the limitations of single-data imaging through matrix superposition and fusion, further enhancing the amplitude of defect locations and highlighting the differences between defective and non-defective areas. By coupling acoustic signal intensity with structural geometric parameters in three dimensions, a 3D model containing amplitude gradient information is constructed, which can intuitively display the depth and boundary morphology of defects.

[0086] The normalized amplitude elements are mapped to the HSV color space as follows:

[0087] The normalized amplitude elements are mapped to the HSV color space in the order of red-yellow-green-blue, representing the amplitude from high to low, as follows:

[0088]

[0089] in, is the global intensity matrix Amplitude elements in . A three-dimensional energy density distribution model is generated, with defective areas showing a high saturation red and defect-free areas fading to blue.

[0090] like Figure 6 As shown in the figure, in order to more intuitively demonstrate the matrix superposition effect, the projection of the "spherical effect" on a two-dimensional plane is used, and the impact point and single-channel acquisition point are used to demonstrate it. Figure 6 Center They are all located between the impact point and the collection point. The radius of each circle in the figure is the distance between the center of each cube grid and the impact point and the collection point in step 3 and The blue dotted circle represents the "spherical effect" that appears in the image when imaging with a single set of frequency domain data at each measuring point. For example, when only measuring points are used When a set of data is imaged, the image is The grids that the circular trajectory with the center as the circle passes through all have the defect frequency as the resonant frequency, and Extracting the same amplitude from the frequency domain and performing color mapping will result in a circular ring that closely matches the blue dashed line, making it difficult to discern the actual location of the defect in the image. Increasing the number of measuring points and changing their positions will cause the circular rings to shift. Since all circular rings contain defect information and pass through the defective area, the overlapping portion of the circular rings represents the actual location of the defect. Ring overlap is characterized by the superposition of all amplitude matrices, resulting in an n-fold increase in the grid amplitude of the defective area. This region's amplitude is significantly different from that of the defect-free area. Renormalizing the amplitude and performing color mapping will reveal the image's clear boundaries and depth of the defect.

[0091] Step 5: Perform multi-dimensional slice analysis on the 3D energy density distribution model to quantitatively analyze the defects.

[0092] Perform multi-dimensional slice analysis on the 3D energy density distribution model to quantitatively analyze defects, specifically:

[0093] like Figure 7 As shown, orthogonal slicing is performed on the three-dimensional energy density distribution model along the X, Y, and Z axes to extract the two-dimensional projection 、 、 , and conduct multi-dimensional analysis of defect conditions through the principle of tomography.

[0094] Specifically, by changing the position of the slice along three axes and observing the image from multiple angles, the size and depth of the defect projection in that direction will appear in the 2D slice for each slice of the defect area. By observing all 2D slices that pass through the defect area, the length, width, and depth of the defect in 3D space can be determined.

[0095] Because the defect is internal to the model, the color representing the defect location may be obscured by the color of the defect-free area during color mapping. Furthermore, the frequency domain spectrum obtained after processing each set of time-domain data may show that the amplitude of the defect frequency is smaller than the amplitude of the plate thickness frequency. In this case, the color of the defect location may be difficult to observe in the 3D image. Therefore, by orthogonally slicing the 3D model at different positions on the XYZ axis and observing the projection of the defect location at different positions in the three axes, the imaging situation can be further observed from a 2D perspective, and the defect condition can be further intuitively judged.

[0096] This invention supports multi-dimensional data cross-verification by establishing a dynamic association mechanism between 2D slices and 3D models. It uses spatial projection matrix conversion to achieve rapid reconstruction of any slice. Using tomography principles to construct a defect size and depth calculation model, this method establishes a dual diagnostic model of "3D positioning + 2D quantification."

[0097] The embodiment of the present invention discloses a multi-point superposition three-dimensional imaging method based on the impact echo method. By synchronously collecting impact echo signals through multiple channels, combined with frequency domain feature extraction and amplitude enhancement mechanism, the expansion of traditional two-dimensional imaging to three-dimensional imaging is achieved, thereby more intuitively and accurately presenting the distribution range and burial depth of internal defects in concrete. At the same time, the generated three-dimensional model can be cut into sections along orthogonal planes in the X, Y, and Z directions to achieve multi-dimensional analysis of the defect status, providing more comprehensive defect information than traditional two-dimensional images, effectively overcoming the situation where traditional two-dimensional imaging only presents the projection of defects in one direction and is insufficient for identifying complex defects, thereby improving the recognition accuracy of defects by the impact echo method.

[0098] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0099] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A multi-point superposition three-dimensional imaging method based on the impact echo method, characterized in that: include: Step 1: With the four accelerometers as corner points and the impact point as the center, a four-channel sensor impact echo detection system based on an equilateral rectangular array is constructed. The four-channel time domain signal is collected when each measuring point in the detection area is used as the impact point. Step 2: Taking the center of the detection area as the top surface center, construct four concrete three-dimensional discrete models corresponding to the four channels for each measuring point, divide them into uniform cubic grid units, and calculate the resonant frequency of each cubic grid unit; Step 3: Normalize the amplitudes in the four-channel time domain signals and perform fast Fourier transform to obtain four sets of normalized frequency domain spectra corresponding to the four-channel time domain signals. Based on the resonant frequency, extract the amplitude of each cubic grid unit from the normalized frequency domain spectra to obtain the amplitude matrix corresponding to the four channels at each measuring point. Superimpose the amplitude matrices of the four channels at the same measuring point to obtain the defect amplitude enhancement matrix for each measuring point. Step 4: Superimpose the defect amplitude enhancement matrices at all measurement points to obtain a global intensity matrix, and normalize the amplitudes in the global intensity matrix again, map it to the HSV color space, and generate a three-dimensional energy density distribution model; Step 5: Perform multi-dimensional slice analysis on the three-dimensional energy density distribution model to quantitatively analyze the defects; In step 2, the resonant frequency of each cubic grid cell is calculated as follows: in, is the resonant frequency of the cubic grid unit with the center coordinate (x, y, z) in the 3D concrete discrete model of the i-th channel collection point at the k-th measurement point, i = 1, 2, 3, 4, corresponding to the four-channel acceleration sensor; C p is the propagation velocity of the P wave in the concrete being tested; is the Euclidean distance between the center of the cube grid unit and the impact point at the kth measurement point; is the Euclidean distance between the center of the cube grid unit and the collection point of the i-th channel at the k-th measurement point; In step 3, based on the resonant frequency, the amplitude of each cubic grid unit is extracted from the normalized frequency domain spectrum to obtain the amplitude matrix corresponding to the four channels at each measuring point, specifically: Each measuring point has four normalized frequency domain spectra, which respectively correspond to the four-channel time domain signals of the measuring point; Each measuring point has four concrete three-dimensional discrete models, which are constructed based on the four channels under the measuring point; Each concrete three-dimensional discrete model contains a plurality of cubic grid units, and each cubic grid unit has a resonant frequency; For the amplitude of each cubic grid unit in the three-dimensional discrete concrete model at each measuring point, the normalized frequency domain spectrum corresponding to the time domain signal of the channel at the measuring point is determined based on the measuring point and the corresponding channel in the three-dimensional discrete concrete model. The amplitude of the cubic grid unit is extracted from the normalized frequency domain spectrum to obtain the amplitude matrix corresponding to the four channels at each measuring point.

2. The multi-point superposition three-dimensional imaging method based on the impact echo method according to claim 1, characterized in that: In step 1, the detection area is a grid detection area arranged in a grid form above the area where defects exist inside the concrete structure; wherein the corner points of each grid are used as measurement points.

3. The multi-point superposition three-dimensional imaging method based on the impact echo method according to claim 2, characterized in that: In step 2, the center of the detection area is used as the top surface center, and four concrete three-dimensional discrete models corresponding to four channels are constructed for each measuring point, specifically: The side lengths of the top and bottom surfaces of the concrete three-dimensional discrete model are the side lengths of the grid detection area plus a preset length, and the depth is the same as the actual thickness of the concrete structure.

4. The multi-point superposition three-dimensional imaging method based on the impact echo method according to claim 1, characterized in that: In step 2, the side length of the cube grid unit satisfies the Nyquist sampling criterion, specifically: Where Δx is the side length of the cubic grid unit; C p is the propagation speed of P wave in the concrete being tested; Δt is the sampling time interval.

5. The multi-point superposition three-dimensional imaging method based on the impact echo method according to claim 1, characterized in that: In step 3, the amplitude matrices of the four channels at the same measuring point are superimposed to obtain the defect amplitude enhancement matrix of each measuring point, as follows: Among them, M k is the defect amplitude enhancement matrix of the kth measuring point; M ki is the amplitude matrix of the i-th channel at the k-th measurement point.

6. The multi-point superposition three-dimensional imaging method based on the impact echo method according to claim 1, characterized in that: In step 4, the defect amplitude enhancement matrices at all measurement points are superimposed to obtain the global intensity matrix, as follows: Among them, M G is the global intensity matrix; K is the number of measurement points; M k is the defect amplitude enhancement matrix of the kth measuring point.

7. The multi-point superposition three-dimensional imaging method based on the impact echo method according to claim 1, characterized in that: In step 4, the normalized amplitude elements are mapped to the HSV color space, specifically: The normalized amplitude elements are mapped to the HSV color space in the order of red-yellow-green-blue, representing the amplitude from high to low, as follows: in, is the magnitude element in the global intensity matrix.

8. The multi-point superposition three-dimensional imaging method based on the impact echo method according to claim 1, characterized in that: In step 5, a multi-dimensional slice analysis is performed on the three-dimensional energy density distribution model to quantitatively analyze the defects, specifically: Orthogonal slicing is performed on the three-dimensional energy density distribution model along the X, Y, and Z axes to extract the two-dimensional projection P x (y, z), P y (x, z), P z (x, y), multi-dimensional analysis of defect conditions is performed through the principle of tomography.

Images