Method for locating damage of non-reference guided wave based on path time domain self-matching

Abstract

The present invention relates to the technical field of ultrasonic guided waves, and in particular provides a reference-free guided wave damage location method based on path time-domain self-matching. The method includes building an ultrasonic guided wave structural damage detection system; according to the system, performing grouping of co-directional and different-length sensing paths and constructing a signal time-domain matching matrix; through path grouping and the signal time-domain matching matrix, identifying healthy paths and reconstructing effective reference signals based on the scattered signal energy; using healthy path identification and effective reference signal reconstruction to perform imaging of the damage location based on the damage path probability distribution. This method better adapts to changes in the external detection environment, improves the calculation efficiency and the accuracy of damage location identification.

Description

Technical Field

[0001] The present invention relates to the technical field of ultrasonic guided waves, and particularly relates to a reference-free guided wave damage location method based on path time domain self-matching. Background Art

[0002] In the field of structural health monitoring, ultrasonic guided waves are a research hotspot in the non-destructive testing field due to their advantages such as long propagation distance, low attenuation, large monitoring area, and high sensitivity, and are widely used in fields such as aerospace, energy pipelines, and composite materials. Currently, most of the damage location technologies based on ultrasonic guided waves judge the damage location according to the differences (such as amplitude attenuation and time difference change) between the guided wave signals collected under the damaged state of the structure (usually referred to as damage signals) and the guided wave signals collected under the healthy state of the structure (usually referred to as reference signals). However, in practical applications, the interference of non-damage factors such as changes in the detection environment temperature, changes in the surface stress of the structure, and structure aging will also cause changes in the amplitude and time difference of the guided waves. Therefore, if accurate location of structural damage is to be achieved, the acquisition of the reference signal before damage occurs is particularly important. However, in practical applications, due to the unpredictability of the occurrence time of structural damage, it is difficult to accurately obtain the reference signal before the occurrence of structural damage. Summary of the Invention

[0003] In view of this, the present invention provides a reference-free guided wave damage location method based on path time domain self-matching to adapt to changes in the external detection environment and improve the calculation efficiency and damage location recognition accuracy.

[0004] In a first aspect, the present invention provides a reference-free guided wave damage location method based on path time domain self-matching, and the method includes:

[0005] Step 1: Build an ultrasonic guided wave structure damage detection system;

[0006] Step 2: According to the system in Step 1, perform grouping of co-directional and different-length sensing paths and construct a signal time domain matching matrix;

[0007] Step 3: Through the path grouping and the signal time domain matching matrix in Step 2, identify healthy paths and reconstruct effective reference signals based on the scattered signal energy;

[0008] Step 4: Use the healthy path identification and effective reference signal reconstruction in Step 3 to perform imaging of the damage location based on the damage path probability distribution.

[0009] Optionally, the Step 1 includes:

[0010] An ultrasonic guided wave structural damage detection system is built using a piezoelectric sensor array, an ultrasonic guided wave detector, a host computer, and a structure to be measured. Among them, the piezoelectric sensor array is composed of multiple piezoelectric sensors, which are evenly pasted on the surface of the structure to be measured in a circular or square shape at equal intervals, and are used to transmit and receive guided wave signals in the structure; one end of the ultrasonic guided wave detector is connected to the piezoelectric sensor, and the other end is connected to the host computer, and is used to transmit the excited and collected guided wave signals; the host computer is used for setting the excitation signal parameters and displaying and storing the collected signals; the structure to be measured is a flat plate or a circular tube made of aluminum alloy or carbon fiber.

[0011] Optionally, step 2 includes:

[0012] a. Grouping of co-directional and different-length sensing paths. According to the different propagation directions of the paths, all sensing paths are divided into N groups, namely path group 1 to path group N; among them, different path groups represent different propagation directions, namely direction 1 to direction N; the sensing paths within the same group have the same direction, and the distances are the same or different; the propagation directions between different path groups are different, and the number of paths included is the same or different; use to represent the number of sensing paths in the k-th group, then is the number of all sensing paths in the sensing array;

[0013] b. Acquisition and grouping of healthy signals. Under the healthy state of the structure to be measured, use the ultrasonic guided wave detector to collect a set of guided wave signals, denoted as the healthy signal HS, and combine the path grouping results to divide HS into N sets of guided wave signals, namely HS = { {HS1}, {HS2}, …, {HS N}}, the N signal sets respectively correspond to N path groups; if the number of sampling points of the guided wave is represented by and represents the number of sensing paths in the k-th group, then the dimension of the k-th signal set {HS k} is ;

[0014] c. Construction of the signal time-domain matching matrix. For each set of guided wave signals {HS k}, respectively based on the dynamic time warping (DTW) algorithm, calculate the time-domain matching matrix W k and the matching error matrix E k inside it; among them, the signal set {HS k} contains guided wave signals of n k paths, which are respectively denoted as , , …, ; ~ are all column vectors with a dimension of m×1, k = 1, 2, …, N.

[0015] Optionally, take the signal set {HSk} The first column signal within = [x1, x2,..., x m T and the second column signal = [y1, y2,..., y m T , and use them as the reference signal and the target signal respectively. The process of calculating the time-domain matching matrix by the DTW algorithm is specifically introduced as follows: Process:

[0016] ① First, construct an m×m-dimensional matrix D to represent the Euclidean distance between the reference signal and the target signal . Any element in the matrix , 1 ≤ i ≤ m, 1 ≤ j ≤ m; then, calculate the cumulative distance matrix C according to Equation (1) m×m , and its expression is:

[0017] (1)

[0018] where the parameter β is a warping coefficient introduced to avoid signal overcompensation, β ≥ 1, and is used for non-linear constraint of the signal;

[0019] ② After obtaining the cumulative distance matrix C m×m , start from the last element c mm of the matrix and backtrack along the minimum path to c 11 . The selection principle for each element on the minimum path is: if c ij is an element on the minimum path, then the next element on the path is the smallest one among {c (i-1)j , c i(j-1) , c (i-1)(j-1)}, and so on, to obtain the minimum path P from element c mm to c 11 in the cumulative distance matrix C;

[0020] ③ The non-linear mapping matrix between the reference signal and the target signal , m ≤ q ≤ 2m + 1, is the set of row and column numbers of each element on the minimum path P; any variable w o = [i o , j o in the matrix is the data in the o-th column of , indicating that the i -th element o in matches the j -th element o in , that is:​​ The \(i\)th o element is assigned as the \(j\)th o element in, based on the reference signal to obtain a target estimated signal matching the target signal ;

[0021] ④ Calculate the error between the target estimated signal and the target signal ;

[0022] ⑤ Repeat processes ① - ④ to obtain the signal set \(\{HS\) k \}, which is the time - domain matching matrix and the matching error matrix of path group \(k\), and their expressions are respectively:

[0023] , ;

[0024] where represents the non - linear mapping matrix from the reference signal to the target signal ; represents the difference signal between the target estimated signal obtained according to the reference signal and the non - linear mapping matrix and the target signal has a dimension of \(m\times1\), where \(m\) is the number of guided - wave sampling points;

[0025] Repeat processes ① - ⑤ to obtain the healthy signal \(HS\), which is the time - domain matching matrix and the corresponding matching error matrix .

[0026] Optionally, step 3 includes:

[0027] e. Acquisition and grouping of damage signals. Under the damaged state of the structure to be measured, use an ultrasonic guided - wave detector to collect the guided - wave signals in the structure, denoted as the original damage signal \(DS\); combined with the path - grouping results, divide the original damage signal \(DS\) into \(N\) original signal sets, i.e., \(DS = \{\{DS1\}, \{DS2\}, \ldots, \{DS\) N \}\}, and the \(N\) signal sets respectively correspond to \(N\) path groups; the number of signals in the signal set \(\{DS\) k \} and the signal set \(\{HS\) k \} is the same as the data dimension, \(k = 1, 2, \ldots, N\);

[0028] f. Benchmark estimation signal calculation: For the original damage signal DS, combine the time-domain matching matrix W and the matching error matrix E to calculate the benchmark estimation signal ; Take the k-th damage signal set {DS k}, and specifically introduce the calculation process of the benchmark estimation signal :

[0029] The signal set {DS k} contains m×1 dimensional guided wave signals, which are respectively denoted as , , …, , that is ; The signal set {DS k} combines the time-domain matching matrix {W k} and the matching error matrix {E k} to obtain the benchmark estimation signal , and its expression is:[[]]

[0030] ;

[0031] Among them, represents taking as the reference signal to obtain the benchmark estimation signal corresponding to , with a dimension of m×1; f, u = 1, 2, …, n k , and the specific calculation process is as follows:

[0032] For any variable w k in the element in the f-th row and u-th column of {W o} = [i o , j o , it means that the i o -th element in path f matches the j o -th element in path u; Assign the i -th element in the guided wave signal o on path f to the j -th element in the guided wave signal o on path u, and subtract the matching error from this result, that is, taking the guided wave signal on path f as the reference signal to obtain the benchmark estimation signal matching the signal ; m ≤ q ≤ 2m + 1; 1 ≤ o ≤ q;

[0033] g. Scattered signal energy calculation: Based on the original damage signal set {DS k} and the reference estimation signal , calculate the scattering energy matrix ; The scattering energy matrix is a two-dimensional matrix with dimensions of n k ×n k , where n k represents the number of sensing paths in the k-th path group, and its expression is:

[0034] ;

[0035] where represents the difference signal energy between the reference estimation signal obtained based on the reference signal and the signal , and its expression is:

[0036] (2)

[0037] Repeat the above process to calculate the scattering energy matrices of all path groups ;

[0038] h. Identification of healthy paths and effective damage paths based on the scattering energy matrix. According to the scattering energy matrix , identify the healthy paths and effective damage paths in each path group under the damaged state of the structure to be measured, 1 ≤ k ≤ N;

[0039] i. Reconstruction of the effective reference signal. Denote the reference estimation signal of the effective damage path set DP obtained based on the healthy path set HP as the effective reference signal , where represents the effective reference signal in the k-th path group; if there are u healthy paths and v effective damage paths in the k-th path group, then for the healthy path and the effective damage path , consider the guided wave signal of the healthy path as a healthy signal not affected by damage; according to the reference estimation signal of the effectively matched effective damage path is the effective reference signal of the effective damage path ; Denote the effective reference signals of all the effectively damaged paths obtained from the healthy paths as the effective reference signal reconstruction of the effectively damaged paths in the k-th path group, k ∈ [1, N]; u + v = n k ;

[0040] Among them, 。

[0041] Optionally, take the k-th path group, and specifically introduce the recognition process of healthy paths and effective damage paths:

[0042] Ⅰ. Parameter initialization: path set AP k =[1, 2,..., n k ; the number of healthy paths f = n k ; effective damage path set DP k =[ ];

[0043] Ⅱ. Calculate the energy of each path , i, j ∈ AP k \DP k , and its expression is:

[0044] (3)

[0045] Ⅲ. If , then execute step Ⅳ, otherwise execute step Ⅴ; where max{·} and average{·} represent the maximum value and the average value respectively;

[0046] Ⅳ. Let , then the j-th path is an effective damage path; let f = f - 1, DP k =[DP k , j], and continue to execute step Ⅱ to calculate the energy of the remaining f paths;

[0047] Ⅴ. Healthy path set HP k =AP k \DP k , and the paths included in HP k are the healthy paths within the k-th path group;

[0048] Repeat steps Ⅰ - Ⅴ to obtain the healthy path sets HP = {HP1, HP2,..., HP N} and the effective damage path sets DP = {DP1, DP2,..., DP N}.

[0049] Optionally, step 4 includes:

[0050] k. Extraction of the effective damage signal set. According to the original damage signal and the effective damage path set DP = {DP1, DP2,..., DP N}, extract the guided wave signals of the effective damage paths, denoted as the effective damage signal set , where The guided wave signal representing the effective damage path in the k-th path group, where k ∈ [1, N]; if there are v damage paths in the k-th path group , then the damage signal in the path corresponding damage signal is denoted as the effective damage signal of the k-th path group ;

[0051] l. Calculation of the effective scattering signal, based on the effective reference signal and the effective damage signal , calculate the effective scattering signal ; where represents the scattering signal of the effective damage path in the k-th path group, and its expression is:

[0052] (4)

[0053] where represents the i-th healthy path in the k-th path group, represents the j-th effective damage path in the k-th path group, and u and v respectively represent the number of healthy paths and effective damage paths in the k-th path group, 1 ≤ i ≤ u; 1 ≤ j ≤ v;

[0054] m. Time of flight TOF matrix of search points on the effective damage path Construction: According to the damage identification accuracy requirement, the monitoring area of the structure to be measured is evenly discretized into r × g search points, and the position matrix PS of the search points, its expression is:

[0055] ;

[0056] where represents the position coordinates of the -th search point, 1 ≤ i ≤ r, 1 ≤ j ≤ g;

[0057] Calculate the TOF matrix of search points on the effective damage path ; where represents the TOF matrix of search points on the effective damage path in the k-th path group , represents the number of effective damage paths in the k-th path group; The expression of

[0058] (5)

[0059] where j = 1, 2, …, v, represents the coordinates of the excitation sensor on the effective damage path ; The effective damage path The coordinates of the receiving sensor; Represents the search point position PS and The Euclidean distance of Represents the search point position PS and The Euclidean distance of It represents the propagation speed of the guided wave in the structure to be tested;

[0060] n. Flight time of damage on the effective damage path Identify, identify the damage TOF on the effective damage path, and obtain the TOF matrix on the effective damage path ,in represents the damage TOF on the effective damage path in the kth path group, and its expression is:

[0061] ;

[0062] in Represents the effective scattered signal The envelope peak moment; 1 ≤ i ≤u, 1≤ j≤v; u is the number of healthy paths in the kth path group, and v is the number of effective damaged paths in the kth path group;

[0063] o. Calculation of damage probability on effective damage paths and total damage probability in the monitoring area , where N is the number of path groups; is the damage probability on the effective damage path in the kth path group, and its expression is:

[0064] ;

[0065] Where u and v are the number of healthy paths and effective damaged paths in the kth path group, respectively; According to the effective scattered signal The effective damage path obtained The damage probability on is expressed as:

[0066] (6)

[0067] in, is the time threshold, which is set to 1.5 times of the waveguide excitation period;

[0068] p, damage location imaging, the total damage probability PD is a two-dimensional matrix with a dimension of r×g, where r and g are the number of rows and columns of discrete points in the monitoring area, respectively, and its expression is:

[0069] ;

[0070] Among them, represents the probability of damage occurring at the position of the th search point. The area with the maximum damage probability concentration is the damage position.

[0071] In the technical solution provided by the present invention, the method includes building an ultrasonic guided wave structure damage detection system; grouping the same-direction and different-length sensing paths according to the system, and constructing a signal time-domain matching matrix; through path grouping and the signal time-domain matching matrix, identifying healthy paths and reconstructing effective reference signals based on the scattered signal energy; using healthy path identification and effective reference signal reconstruction, imaging the damage position based on the damage path probability distribution. This method better adapts to the changes in the external detection environment, improves the calculation efficiency and the accuracy of damage position identification. Description of the Drawings

[0072] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0073] Figure 1 is a flowchart of the no-reference guided wave damage location method based on path time-domain self-matching provided by the embodiment of the present invention;

[0074] Figure 2 is a flowchart of another no-reference guided wave damage location method based on path time-domain self-matching provided by the embodiment of the present invention. Detailed Embodiments

[0075] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0076] It should be clear that the described embodiments are only some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0077] The terms used in the embodiments of the present invention are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The singular forms "a", "said", and "the" used in the embodiments of the present invention are also intended to include the plural forms unless the context clearly dictates otherwise.

[0078] It should be understood that the term "and / or" used herein is merely a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after.

[0079] Depending on the context, the word "if" as used herein can be interpreted as "when" or "while" or "in response to determining" or "in response to detecting". Similarly, depending on the context, the phrase "if determined" or "if detected (stated condition or event)" can be interpreted as "when determined" or "in response to determining" or "when detected (stated condition or event)" or "in response to detecting (stated condition or event)".

[0080] The present invention provides a method for non-reference guided wave damage location based on path time-domain self-matching, as Figure 1 and Figure 2 shown, the method includes:

[0081] Step 1: Build an ultrasonic guided wave structure damage detection system.

[0082] In the embodiments of the present invention, Step 1 includes:

[0083] Build an ultrasonic guided wave structure damage detection system using a piezoelectric sensor array, an ultrasonic guided wave detector, a host computer, and a structure to be measured. Among them, the piezoelectric sensor array is composed of multiple piezoelectric sensors, which are evenly pasted on the surface of the structure to be measured in a circular or square shape at equal intervals, and are used to transmit and receive guided wave signals in the structure. The straight-line connection path between two sensors is called a sensing path; one end of the ultrasonic guided wave detector is connected to the piezoelectric sensor, and the other end is connected to the host computer, and is used to transmit the excited and collected guided wave signals, and has functions such as filtering and amplification; the host computer is used for setting the excitation signal parameters (such as excitation frequency, amplitude, filter upper and lower limits, etc.) and displaying and storing the collected signals; the structure to be measured is a flat plate or a circular tube made of aluminum alloy or carbon fiber.

[0084] Step 2: According to the system in Step 1, perform grouping of co-directional and different-length sensing paths, and construct a signal time-domain matching matrix.

[0085] In the embodiments of the present invention, Step 2 includes:

[0086] a. Grouping of co-directional and different-length sensing paths: According to the different propagation directions of the paths, all sensing paths are divided into N groups, namely path group 1 to path group N. Among them, different path groups represent different propagation directions, namely direction 1 to direction N. The sensing paths within the same group have the same direction, and the distances can be the same or different. The propagation directions between different path groups are different, and the number of included paths can be the same or different. Let represent the number of sensing paths in the k-th group, then is the number of all sensing paths in the sensing array;

[0087] b. Acquisition and grouping of healthy signals: Under the healthy state of the structure to be measured, use an ultrasonic guided wave detector to collect a set of guided wave signals, denoted as healthy signals (Healthy Signal, HS), and combine the path grouping results to divide HS into N sets of guided wave signals, namely HS = { {HS1}, {HS2}, …, {HS N}}, and the N signal sets respectively correspond to the N path groups. If the number of sampling points of the guided wave is represented by , represents the number of sensing paths in the k-th group, then the dimension of the k-th signal set {HS k} is ;

[0088] c. Construction of the signal time-domain matching matrix: For each set of guided wave signals {HS k}, respectively based on the Dynamic Time Warping (DTW) algorithm, calculate the time-domain matching matrix W k and the matching error E k inside it. Among them, the signal set {HS k} contains guided wave signals of n k paths, which are respectively denoted as , , …, ; ~ are all column vectors with a dimension of m×1, and k = 1, 2, …, N.

[0089] In the embodiment of the present invention, take the first column signal k = [x1, x2, …, x m and the second column signal T = [y1, y2, …, y m in the signal set {HS T}, and use them as the reference signal and the target signal respectively to specifically introduce the process of calculating the time-domain matching matrix :

[0090] ​​① First, construct an m×m - dimensional matrix D to represent the reference signal and the target signal of the Euclidean distance. Any element in the matrix , where 1 ≤ i ≤ m, 1 ≤ j ≤ m; then, calculate the cumulative distance matrix C according to Equation (1) m×m , and its expression is

[0091] (1)

[0092] where the parameter β is a bending coefficient introduced to avoid signal over - compensation, β ≥ 1, and is used for non - linear constraint of the signal;

[0093] ② After obtaining the cumulative distance matrix C m×m , start from the last element c mm of the matrix and backtrack along the minimum path to c 11 . The selection principle for each element on the minimum path is: if c ij is an element on the minimum path, then the next element on the path is the smallest one among {c (i-1)j , c i(j-1) , c (i-1)(j-1)}, and so on, to obtain the minimum path P from the element c mm to c 11 in the cumulative distance matrix C;

[0094] ③ The non - linear mapping matrix between the reference signal and the target signal , where m ≤ q ≤ 2m + 1, is the set of row and column numbers of each element on the minimum path P; any variable w o = [i o , j o in the matrix is the data in the o - th column of , indicating that the i -th element o in matches the j -th element o in , that is: assign the i -th element o in to the j -th element in o to obtain the target estimated signal matching the target signal ; ;

[0095] ④ Calculate the target estimated signal and the target signal the error between ;

[0096] ⑤ Repeat steps ① - ④ to obtain the signal set {HS k}, which is the time - domain matching matrix of path group k and the matching error matrix , and their expressions are respectively:

[0097] , ;

[0098] wherein, represents the non - linear mapping matrix from the reference signal to the target signal ; represents the difference signal between the target estimated signal obtained according to the reference signal and the non - linear mapping matrix and the target signal , has a dimension of m×1, where m is the number of guided - wave sampling points;

[0099] Repeat steps ① - ⑤ to obtain the healthy signal HS, which is the time - domain matching matrix of the path group and the corresponding matching error matrix .

[0100] Step 3: Based on the path grouping and the signal time - domain matching matrix in Step 2, identify healthy paths and reconstruct effective reference signals based on the scattered signal energy.

[0101] In the embodiment of the present invention, Step 3 includes:

[0102] e. Acquisition and grouping of damage signals. In the damaged state of the structure to be measured, use an ultrasonic guided - wave detector to collect the guided - wave signals in the structure, denoted as the original damage signal (Damage Signal, DS); combined with the path grouping results, divide the original damage signal DS into N original signal sets, i.e., DS = { {DS1}, {DS2}, …, {DS N}}, and the N signal sets respectively correspond to N path groupings; the number of signals and the data dimensions in the signal set {DS k} and the signal set {HS k} are the same, k = 1, 2, …, N;

[0103] f. Calculation of the reference estimated signal. For the original damage signal DS, combined with the time - domain matching matrix W and the matching error matrix E, calculate the reference estimated signal ; Take the k - th damage signal set {DS k}, and specifically introduce the reference estimation signal The calculation process is as follows:

[0104] The signal set {DS k} contains m×1-dimensional guided wave signals, which are respectively denoted as , , …, , that is ; The signal set {DS k} combines the time-domain matching matrix {W k} and the matching error matrix {E k} to obtain the reference estimation signal , and its expression is:

[0105] ;

[0106] Among them, represents using as the reference signal to obtain the reference estimation signal corresponding to , with a dimension of m×1; f, u = 1, 2, …, n k , and the specific calculation process is as follows:

[0107] For any variable w k =[i in the element o of the f-th row and u-th column of {W o , j o , it represents the i o -th element in path f matching the j o -th element in path u; Assign the i -th element in the guided wave signal o on path f to the j -th element in the guided wave signal o on path u, and subtract the matching error from this result, that is, using the guided wave signal on path f as the reference signal to obtain the reference estimation signal matching the signal ; m ≤ q ≤ 2m + 1; 1 ≤ o ≤ q;

[0108] g. Scattering signal energy calculation, based on the original damage signal set {DS k} and the reference estimation signal , calculate the scattering energy matrix ; The scattering energy matrix is a matrix with a dimension of n k ×n kTwo-dimensional matrix, n k Represents the number of sensing paths in the k-th path group, and its expression is:

[0109] ;

[0110] Wherein, Represents the reference estimation signal obtained based on the reference signal And the difference signal energy between the signal And the signal Its expression is:

[0111] (2)

[0112] Repeat the above process to calculate the scattering energy matrix of all path groups ;

[0113] h. Identification of healthy paths and effective damage paths based on the scattering energy matrix. According to the scattering energy matrix , identify the healthy paths (i.e., the paths where the received guided wave signals are hardly affected by damage due to being far from the damage) and effective damage paths (i.e., the paths where the received guided wave signals are affected by damage due to being close to the damage) in each path group under the damage state of the structure to be measured, 1 ≤ k ≤ N;

[0114] In the embodiment of the present invention, take the k-th path group, and specifically introduce the identification process of healthy paths and effective damage paths:

[0115] Ⅰ. Parameter initialization: Path set AP k =[1, 2,..., n k ; The number of healthy paths f = n k ; Effective damage path set DP k =[ ];

[0116] Ⅱ. Calculate the energy of each path , i, j ∈ AP k \DP k , and its expression is:

[0117] (3)

[0118] Ⅲ. If , then execute step Ⅳ, otherwise execute step Ⅴ; where max{·} and average{·} represent the maximum value and the average value respectively;

[0119] Ⅳ. Let , then the j-th path is an effective damage path; Let f = f - 1, DP k =[DP k, j], and continue to execute Step II to calculate the energy of the remaining f paths;

[0120] Ⅴ. Healthy path (health path) set HP k = AP k \ DP k , HP k The paths included in it are the healthy paths within the k-th path group;

[0121] Repeat Steps I - V to obtain the healthy path sets HP = {HP1, HP2,..., HP N} and the effective damage path sets DP = {DP1, DP2,..., DP N}.

[0122] i. Reconstruction of the effective reference signal. Denote the reference estimation signal of the effective damage path set DP obtained based on the healthy path set HP as the effective reference signal , where represents the effective reference signal in the k-th path group; if there are u healthy paths in the k-th path group , v effective damage paths , then for the healthy path and the effective damage path , it is considered that the guided wave signal of the healthy path is a healthy signal not affected by damage; according to the matched effective damage path the reference estimation signal is the effective reference signal of the effective damage path ; the effective reference signals of the effective damage paths obtained from all healthy paths are denoted as the reconstruction of the effective reference signals of the effective damage paths in the k-th path group, k ∈ [1, N]; u + v = n k ;

[0123] Among them, .

[0124] Step 4. Utilize the healthy path recognition and effective reference signal reconstruction in Step 3 to perform imaging of the damage location based on the damage path probability distribution.

[0125] In the embodiment of the present invention, Step 4 includes:

[0126] k. Extraction of the effective damage signal set. According to the original damage signal and the effective damage path set DP = {DP1, DP2,..., DP N, extract the guided wave signals of the effective damage paths, denoted as the effective damage signal set , where represents the guided wave signal of the effective damage path in the k-th path group, k ∈ [1, N]; if there are v damage paths in the k-th path group , then the damage signal in the path corresponding damage signal is denoted as the effective damage signal of the k-th path group ;

[0127] l. Calculation of effective scattering signals, based on the effective reference signal and the effective damage signal , calculate the effective scattering signal (Scatter Signal) ; where represents the scattering signal of the effective damage path in the k-th path group, and its expression is:

[0128] (4)

[0129] where represents the i-th healthy path in the k-th path group, represents the j-th effective damage path in the k-th path group, u and v respectively represent the number of healthy paths and effective damage paths in the k-th path group, 1 ≤ i ≤ u; 1 ≤ j ≤ v;

[0130] m. Time-Of-Flight (TOF) matrix of search points on the effective damage path Construction, according to the damage identification accuracy requirement, evenly discretize the monitoring area of the structure to be measured into r × g search points, and the position matrix PS of the search points (positionof search point) has the following expression:

[0131] ;

[0132] where represents the position coordinates of the -th search point, 1 ≤ i ≤ r, 1 ≤ j ≤ g;

[0133] Calculate the TOF matrix of the search points on the effective damage path ; where represents the TOF matrix of the search points on the effective damage path in the k-th path group , represents the number of effective damage paths in the k-th path group; The expression of

[0134] (5)

[0135] where \(j = 1, 2, \ldots, v\) represents the coordinates of the excitation sensor on the effective damage path ; and represents the coordinates of the receiving sensor on the effective damage path ; represents the Euclidean distance between the search point position \(PS\) and ; represents the Euclidean distance between the search point position \(PS\) and ; represents the propagation speed of the guided wave in the structure to be measured

[0136] n, the flight time of the damage on the effective damage path Identify the TOF of the damage on the effective damage path to obtain the TOF matrix on the effective damage path , where represents the TOF of the damage on the effective damage path in the \(k\)th path group, and its expression is:

[0137] ;

[0138] where represents the envelope peak time of the effective scattering signal ; \(1\leq i\leq u\), \(1\leq j\leq v\); \(u\) is the number of healthy paths in the \(k\)th path group, and \(v\) is the number of effective damage paths in the \(k\)th path group

[0139] o, calculation of the damage probability on the effective damage path, the total damage probability in the monitoring area (probability of damage) , where \(N\) is the number of path groups is the damage probability on the effective damage path in the \(k\)th path group, and its expression is:

[0140] ;

[0141] where \(u\) and \(v\) are the numbers of healthy paths and effective damage paths in the \(k\)th path group, respectively is the damage probability on the effective damage path obtained according to the effective scattering signal and its expression is:

[0142] (6)

[0143] where is the time threshold, set to 1.5 times the guided wave excitation period

[0144] p, damage location imaging, the total damage probability PD is a two-dimensional matrix with a dimension of r×g, where r and g are the number of rows and columns of discrete points in the monitoring area, respectively, and its expression is:

[0145] ;

[0146] in, Indicates Search point locations The probability of damage occurring at a location is the maximum concentration area of damage probability, which is the damage location.

[0147] Compared with the prior art, the present invention has the following beneficial effects:

[0148] (1) No reference dependence: Existing damage location methods rely on the collection of fixed reference signals. However, in practical applications, when the external detection environment changes, it is difficult to pre-collect the waveguide signal of the healthy state of the structure under the current detection environment. The reference signal used in the damage location identification method proposed in the present invention is a reference estimation signal reconstructed based on the collected damage signal and the pre-constructed time domain matching matrix, which can better adapt to changes in the external detection environment.

[0149] (2) Lightweight data, improve computing efficiency and damage location identification accuracy: In the present invention, by processing the scattered energy matrix, the healthy path and the effective damaged path in each path group can be identified. The damage location is subsequently identified based only on the waveguide signal of the effective damaged path, and the waveguide signal of the healthy path is no longer processed and calculated, which effectively reduces the amount of calculation; and the healthy path signal does not contain damage information, but contains interference information such as the environment. By filtering out the signal of the healthy path, the interference of factors such as the environment on the damage location identification result can be effectively reduced.

[0150] In the technical solution provided by the present invention, the method includes building an ultrasonic guided wave structure damage detection system; according to the system, grouping isotropic and heterogeneous sensing paths and constructing a signal time domain matching matrix; through path grouping and signal time domain matching matrix, healthy path identification and effective reference signal reconstruction are performed based on scattered signal energy; using healthy path identification and effective reference signal reconstruction, imaging of the damage position is performed based on the probability distribution of the damage path. This method better adapts to changes in the external detection environment and improves calculation efficiency and damage position identification accuracy.

[0151] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0152] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for damage location of benchmark-free guided wave based on path time-domain self-matching, characterized in that The method includes the following steps: Step 1: Build an ultrasonic guided wave structure damage detection system; Step 2: According to the system in Step 1, perform grouped sensing paths with the same direction but different lengths, and construct a signal time-domain matching matrix; Step 3: Through the path grouping and the signal time-domain matching matrix in Step 2, identify healthy paths and reconstruct effective reference signals based on the scattered signal energy; Step 4: Utilize the healthy path identification and effective reference signal reconstruction in Step 3 to perform imaging of the damage location based on the damage path probability distribution; The said Step 3 includes: e. Damage signal acquisition and grouping. Under the damaged state of the structure to be measured, use an ultrasonic guided wave detector to collect the guided wave signals in the structure, denoted as the original damage signal DS; combined with the path grouping results, divide the original damage signal DS into N original signal sets, i.e., DS = { {DS1}, {DS2}, …, {DS N}}, where the N signal sets correspond to N path groupings respectively; the number of signals and the data dimensions in the signal set {DS k} and the signal set {HS k} are the same, k = 1, 2, …, N; Among them, the health signal is denoted as HS. Combining the path grouping results, HS is divided into N guided wave signal sets, that is, HS = { {HS1}, {HS2}, …, {HS N}}, and the N signal sets respectively correspond to N path groups. {HS k} represents the k-th signal set; f. Benchmark estimation signal calculation: For the original damage signal DS, combine the time-domain matching matrix W and the matching error matrix E to calculate the benchmark estimation signal Take the k-th damage signal set {DS k}, and the calculation process of the benchmark estimation signal is as follows: The signal set {DS k} contains n k m×1 dimensional guided wave signals, denoted as That is The signal set {DS k} combined with the time-domain matching matrix {W k} and the matching error matrix {E k} yields the reference estimation signal Its expression is: Among them, denotes taking as the reference signal to obtain the reference estimation signal corresponding to , with a dimension of m×1; f, u = 1, 2, …, n k , and the calculation process is as follows: For the element in the f-th row and u-th column of {W k}, for any variable w in o = [i o , j o , it means that the i-th o element in path f matches the j-th o element in path u ; assign the i-th element in the guided wave signal on path f o to be the j-th element in the guided wave signal on path u o , and subtract the matching error on this result, that is, using the guided wave signal on path f as the reference signal to obtain the reference estimation signal that matches the signal m ≤ q ≤ 2m + 1; 1 ≤ o ≤ q;​ g. Scattering signal energy calculation, based on the original damage signal set {DS k}, and the reference estimated signal to calculate the scattering energy matrix A k ; The scattering energy matrix A k is a two-dimensional matrix with dimensions of n k ×n k , where n k represents the number of sensing paths in the k-th path group, and its expression is: Among them, represents the reference estimation signal obtained based on the reference signal and the difference signal energy between the signal is expressed as: Repeat the above process to calculate the scattering energy matrix A = { {A1}, {A2}, …, {A N}}; h. Identification of healthy paths and effective damage paths based on the scattering energy matrix. According to the scattering energy matrix A k , identify the healthy paths and effective damage paths within each path group in the damaged state of the structure to be measured, where 1 ≤ k ≤ N; i. Reconstruction of effective reference signal: The reference estimation signal of the effective damaged path set DP obtained based on the healthy path set HP is recorded as the effective reference signal in represents the valid reference signal in the kth path group; if there are u healthy paths in the kth path group {p h1 ,p h2 ,…,p hu }, v effective damage paths {p d1 ,p d2 ,…,p dv }, then for the healthy path p hi and the effective damage path p dj , it is believed that the health path p hi The guided wave signal is a healthy signal that is not affected by damage; The matched effective damage path p dj The baseline estimation signal is the effective damage path p dj The effective reference signal; all healthy paths {p h1 ,p h2 ,…,p hu The effective damage path obtained is {p d1 ,p d2 ,…,p dv }effective reference signal Denoted as the effective reference signal reconstruction of the effective damaged path in the kth path group, k∈[1,N]; u+v=n k ; Among them, 2. The method according to claim 1, characterized in that, The said Step 1 includes: Build an ultrasonic guided wave structure damage detection system by using a piezoelectric sensor array, an ultrasonic guided wave detector, a host computer, and a structure to be measured. Among them, the piezoelectric sensor array is composed of multiple piezoelectric sensors, which are evenly pasted on the surface of the structure to be measured in a circular or square shape at equal intervals, and are used to transmit and receive guided wave signals in the structure; one end of the ultrasonic guided wave detector is connected to the piezoelectric sensor, and the other end is connected to the host computer, and is used to transmit the excited and collected guided wave signals; the host computer is used for setting the excitation signal parameters and displaying and storing the collected signals; the structure to be measured is a flat plate or a circular tube made of aluminum alloy or carbon fiber.

3. The method according to claim 1, wherein The said Step 2 includes: a. Grouping of sensing paths with the same direction but different lengths. According to the different propagation directions of the paths, all sensing paths are divided into N groups, namely path group 1 to path group N. Among them, different path groups represent different propagation directions, namely direction 1 to direction N. The sensing paths within the same group have the same direction, and the distances can be the same or different. The propagation directions between different path groups are different, and the number of paths included can be the same or different. Let n k represent the number of sensing paths in the k-th group, then is the number of all sensing paths in the sensing array; b. Health signal acquisition and grouping. Under the healthy state of the structure to be measured, use an ultrasonic guided wave detector to collect a set of guided wave signals, denoted as the health signal HS. Combining with the path grouping results, divide HS into N guided wave signal sets, that is, HS = { {HS1}, {HS2}, …, {HS N}}, and the N signal sets respectively correspond to N path groups; if the number of sampling points of the guided wave is represented by m, and n k represents the number of sensing paths in the k-th group, then the dimension of the k-th signal set {HS k} is m×n k ; c. Construction of the signal time-domain matching matrix. For each guided wave signal set {HS k}, respectively based on the dynamic time warping (DTW) algorithm, calculate the time-domain matching matrix W k and the matching error matrix E k between every two signals within it; where the signal set {HS k} contains guided wave signals of n k paths, which are respectively denoted as Both are column vectors of dimension m×1, and k = 1, 2, …, N.

4. The method according to claim 3, wherein Take the first column signal in the signal set {HS k}, and the second column signal Respectively use them as the reference signal and the target signal, and the process of calculating the time-domain matching matrix by the DTW algorithm is as follows: ① First, construct an m×m-dimensional matrix D to represent the Euclidean distance between the reference signal and the target signal . Any element d in the matrix ij = ||x i - y j ||², 1 ≤ i ≤ m, 1 ≤ j ≤ m; then, calculate the cumulative distance matrix C m×m according to Equation (1), and its expression is: Among them, the parameter β is a bending coefficient introduced to avoid signal over-compensation, β≥1, and is used for non-linear constraint of the signal; ② Obtain the cumulative distance matrix C m×m After that, starting from the last element c of the matrix mm Backtrack along the minimum path in reverse to c 11 , and the selection principle for each element on the minimum path is: if c ij is an element on the minimum path, then the next element on the path is the smallest one among {c (i-1)j , c i(j-1) , c (i-1)(j-1)}, and so on, to obtain the minimum path P from element c mm to c 11 in the cumulative distance matrix C; ③Reference signal and the target signal The non - linear mapping matrix between m ≤ q ≤ 2m + 1, which is the row - column number set of each element on the minimum path P; any variable w in the matrix o =[i o , j o is The data in the o - th column of, indicating The i - th o element in matches The j - th o element in, that is: assign the i - th element in to be o the j - th element in, so as to obtain a target estimated signal matching the target signal o based on the reference signal ​​ ④ Calculate the target estimated signal and the target signal to obtain the error ⑤ Repeat the process ① - ④ to obtain the signal set {HS k}, which is the time-domain matching matrix W of path group k k and the matching error matrix E k . Their expressions are respectively: Among them, represents the reference signal to the target signal nonlinear mapping matrix; represents the target estimated signal obtained according to the reference signal and the nonlinear mapping matrix the difference signal between and the target signal has a dimension of m×1, where m is the number of guided wave sampling points; Repeat the process ①-⑤ to obtain the healthy signal HS, i.e., the time-domain matching matrix W of the path group = { {W1}, {W2}, …, {W N}} and the corresponding matching error matrix E = { {E1}, {E2}, …, {E N}}.

5. The method according to claim 1, characterized in that, Take the k-th path group, and the identification process of healthy paths and effective damage paths is as follows: Ⅰ. Parameter initialization: path set AP k = [1, 2, …, n k ; number of healthy paths f = n k ; Effective damage path set DP k = []; Ⅱ. Calculate the energy of each path i, j ∈ AP k \DP k , and its expression is: Ⅲ. If then execute Step Ⅳ, otherwise execute Step Ⅴ; where max{·} and average{·} represent the maximum value and the average value respectively; Ⅳ. Let Then the j-th path is a valid damage path; let f = f - 1, DP k = [DP k , j], and continue to execute Step II to calculate the energies of the remaining f paths. Ⅴ. Health Path Set HP k = AP k \DP k , HP k The paths included in HP are the health paths within the k-th path group; Repeat steps I - V to obtain the healthy path set HP = {HP1, HP2,..., HP N} and the effective damage path set DP = {DP1, DP2,..., DP N} for all path groups.

6. The method according to claim 1, characterized in that, The said Step 4 includes: k, extraction of the effective damage signal set. According to the original damage signal DS = { { DS1}, { DS2}, …, { DS N}} and the effective damage path set DP = { DP1, DP2, …, DP N}, extract the guided wave signals of the effective damage paths, denoted as the effective damage signal set where represents the guided wave signal of the effective damage path in the k-th path group, k ∈ [1, N]; if there are v damage paths { p d1 , p d2 , …, p dv} in the k-th path group, then the damage signals in the path { p d1 , p d2 , …, p dv} are denoted as the effective damage signals of the k-th path group ​ l. Calculation of effective scattering signal, based on effective reference signal and effective damage signal to calculate the effective scattering signal wherein represents the scattering signal of the effective damage path in the k-th path group, and its expression is: Among them, p hi represents the i-th healthy path in the k-th path group, and p dj represents the j-th effective damage path in the k-th path group. u and v respectively represent the numbers of healthy paths and effective damage paths in the k-th path group, where 1 ≤ i ≤ u; 1 ≤ j ≤ v; m, the time-of-flight (TOF) matrix TS of search points on the effective damage path d Construct. According to the damage identification accuracy requirement, the monitoring area of the structure to be measured is evenly discretized into r×g search points, and the position matrix PS of the search points, whose expression is: Among them, (x i , y j ) represents the position coordinates of the ((i - 1)×g + j)-th search point, where 1 ≤ i ≤ r and 1 ≤ j ≤ g; Calculate the TOF matrix of the search points on the effective damage path Among them, represents the TOF matrix of the search points on the effective damage path in the k-th path group v represents the number of effective damage paths in the k-th path group; The expression of is: where j = 1, 2, …, v, represents the coordinates of the excitation sensor on the effective damage path p dj represents the coordinates of the receiving sensor on the effective damage path p dj represents the Euclidean distance between the search point position PS and represents the Euclidean distance between the search point position PS and ; V wave represents the propagation speed of the guided wave in the structure to be measured;​​​ n, the time of flight TD of the damage on the effective damage path d Identify the TOF of the damage on the effective damage path, and obtain the TOF matrix on the effective damage path where represents the TOF of the damage on the effective damage path in the k-th path group, and its expression is: wherein represents the envelope peak time of the valid scattering signal ; 1 ≤ i ≤ u, 1 ≤ j ≤ v; u is the number of healthy paths in the k-th path group, and v is the number of valid damaged paths in the k-th path group o. Calculation of the damage probability on the effective damage path. The total damage probability within the monitoring area is where N is the number of path groups; is the damage probability on the effective damage path in the k-th path group, and its expression is: where \(u\) and \(v\) are the numbers of healthy paths and effective damaged paths in the \(k\)th path group, respectively; is the effective damaged path \(p\) obtained according to the effective scattering signal and the damage probability on it is expressed as: dj ​ where t th is the time threshold, which is set to 1.5 times the guided wave excitation period; p. Imaging of the damage location. The total damage probability PD is a two-dimensional matrix with dimensions of r×g, where r and g are the number of rows and columns of discrete points in the monitoring area respectively, and its expression is: Among them, pd ij represents the probability of damage occurring at the position of the (i - 1)×g + j-th search point (x i , y j ). The area with the highest concentration of the maximum damage probability is the damage position.

Images