Composite material wallboard damage identification and evaluation method based on Lamb wave and auto-encoder

Abstract

The invention discloses a composite material wallboard damage identification and evaluation method based on Lamb waves and auto-encoders, and the method comprises the steps: arranging a Lamb wave signal collection system on a first wallboard in a healthy state, and training a depth auto-encoder model at each receiver; for a to-be-detected second wall plate, acquiring a Lamb wave response signal by adopting a Lamb wave signal acquisition system; determining the Lamb wave flight time at each receiver on the second wall plate through the differential signal; a geometric model equation set is constructed and solved in combination with a Lamb wave anisotropic propagation velocity distribution model, and the damage position in the second wall plate is positioned; and reconstructing the Lamb wave response signals acquired by the plurality of receivers on the second wallboard by using the depth auto-encoder model, and carrying out standardization processing to obtain a final comprehensive damage assessment score. The method has the advantages of being high in positioning precision, high in evaluation sensitivity, good in damage grade distinguishing capacity and the like, and is suitable for nondestructive testing and evaluation of the composite material structure wall plate.

Description

Technical Field

[0001] The present invention relates to the field of panel damage identification and assessment, and in particular to a method for identifying and assessing composite panel damage based on Lamb waves and autoencoders. Background Art

[0002] With the rapid development of high-end equipment manufacturing fields such as aerospace, rail transit, ocean engineering, and new energy, lightweight and high-strength advanced composite materials have been widely used in various key load-bearing structures. However, during actual service, composite materials are extremely vulnerable to various factors such as impact, fatigue, delamination, and foreign object penetration, resulting in invisible or minor internal damages, such as delamination, debonding, matrix cracks, and fiber fractures. These damages often have concealment and development characteristics. If not detected and identified in a timely and accurate manner, it is very likely to cause structural degradation or even catastrophic failure, seriously threatening personnel safety and system reliability. Therefore, the damage detection, identification, and assessment technology of composite panels is a research hotspot in the field of structural health monitoring. Among them, the non-destructive testing technology based on guided waves (such as Lamb waves) has received extensive attention in recent years due to its high sensitivity, long-distance propagation ability, and adaptability to various damage types, and has gradually been applied to the damage identification of composite structures.

[0003] Although a series of achievements have been made in the current Lamb wave detection technology, methods such as probability imaging algorithms (CN114910562A, CN113376250A), data-driven modeling (CN114925716A), and signal time-frequency analysis (CN113390967A) all show good performance in damage location and identification. However, from the overall development trend, most of the existing research focuses on the detection and judgment of damage locations, and the quantitative assessment of damage degree is significantly insufficient. In other words, the current damage identification methods emphasize more on "whether there is" and "where", while the judgment ability of "how serious" is still in the initial exploration stage. Especially when facing minor damages, multi-level damages, or continuous damage evolution situations during service, it is difficult to obtain reliable and quantifiable assessment results only relying on traditional means based on characteristic quantities such as time difference, waveform change, or wave velocity distribution. Therefore, it is necessary to develop a Lamb wave detection technology that can not only accurately locate the damage of composite panels but also comprehensively evaluate the damage degree of composite panels. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for identifying and assessing composite panel damage based on Lamb waves and autoencoders, which is used for non-destructive testing and assessment of composite panels, and has the characteristics of high positioning accuracy, strong assessment sensitivity, and good discrimination ability for damage levels.

[0005] To achieve the above tasks, the present solution adopts the following technical solutions:

[0006] A method for damage identification and evaluation of composite panels based on Lamb waves and autoencoders, comprising:

[0007] Arrange a Lamb wave signal acquisition system on the first panel in a healthy state; under a preset excitation signal, use receivers at different positions on the first panel to obtain Lamb wave response signals; based on the Lamb wave response signals obtained by each receiver, train a deep autoencoder model corresponding to the receiver;

[0008] For the second panel to be detected, arrange a Lamb wave signal acquisition system in the same manner as the first panel, and use the receivers to obtain Lamb wave response signals; wherein, the second panel is of the same type as the first panel;

[0009] Calculate the differential signal using the Lamb response signals obtained from the first panel and the second panel, and determine the Lamb wave flight time at each receiver on the second panel through the differential signal;

[0010] Based on the Lamb wave flight times at some receivers, construct a geometric model equation set in combination with the Lamb wave anisotropic propagation velocity distribution model; construct an objective function to solve the geometric model equation set, so as to locate the damage position in the second panel;

[0011] Select the Lamb wave response signals obtained by multiple receivers on the second panel as the original signals, and input them into the corresponding deep autoencoder models respectively to obtain the reconstructed signals corresponding to each receiver; use the original signals and the reconstructed signals to determine the average reconstruction error as the damage evaluation value at each receiver and perform normalization processing; based on the normalized damage evaluation values at the selected multiple receivers, determine the comprehensive damage evaluation score.

[0012] Further, the Lamb wave signal acquisition system includes an exciter and multiple receivers arranged around the exciter; construct a damage location coordinate system with the position of the exciter as the origin, define the positive direction of the 𝑥-axis along the fiber main axis direction of the first panel, and the positive direction of the y-axis is in the same plane as the positive direction of the x-axis;

[0013] The preset excitation signal adopts a single Lamb wave mode, and the center frequency is set in the non-dispersive section of the dispersion curve; the form of the excitation signal is a sine pulse.

[0014] Further, the deep autoencoder model includes an encoder and a decoder; the encoder is used to map the Lamb wave response signal to a low-dimensional feature space to form a low-dimensional feature vector at the hidden layer;

[0015] The decoder takes the low-dimensional feature vector at the hidden layer as input and outputs the reconstructed signal of the Lamb wave response signal; the training objective of the deep autoencoder model is to minimize the reconstruction error between the Lamb wave response signal and the reconstructed signal.

[0016] Further, the differential signal is calculated using the Lamb response signals obtained from the first wall panel and the second wall panel, and the Lamb wave flight time at each receiver on the second wall panel is determined through the differential signal, including:

[0017] The Lamb wave response signal collected by the th receiver on the second wall panel is , and the average response signal at the th receiver is calculated; the Lamb wave response signal collected by the th receiver on the first wall panel is , and the average response signal at the th receiver is calculated;

[0018] The differential signal at the th receiver is calculated;

[0019] The complex Morlet wavelet transform is applied to the differential signal . The result of the wavelet transform presents a time-scale spectrum, and the moment corresponding to the peak of the envelope spectrum is extracted as the arrival time of the Lamb wave response signal at the th receiver on the second wall panel;

[0020] The Lamb wave flight time at the th receiver on the second wall panel is determined; represents the starting time of the excitation signal.

[0021] Further, when locating the damage position in the second wall panel, first, the Lamb wave flight times of all the receivers on the second wall panel are sorted from short to long, and the first Q Lamb wave flight times are selected;

[0022] Based on the selected Q Lamb wave flight times, combined with the Lamb wave anisotropic propagation velocity distribution model, a geometric model equation set based on Lamb waves is constructed;

[0023] The least squares method is used to construct an objective function to solve the geometric model equation set; by minimizing the sum of the squared residuals between the calculated value of the Lamb wave flight time from the damage position to the i-th receiver and the actual Lamb wave flight time , the position coordinates of the optimal estimate of the damage position are obtained.

[0024] Furthermore, the geometric model equations are expressed as follows:

[0025] ;

[0026] Among them, the sixth equation of the equations is the Lamb wave anisotropic propagation velocity distribution model;

[0027] In the above formula, represents the Lamb wave flight time at the i-th receiver among the selected first Q Lamb wave flight times. There is a damage position D on the second wall panel, and its position coordinates are ; The position coordinates of the i-th receiver are ; , respectively represent the propagation paths of the Lamb wave from the actuator to the damage position D and from the damage position D to the i-th receiver , of the lengths, , respectively represent the included angles between the propagation paths , and the positive x-axis direction of the damage location coordinate system, , are respectively the propagation velocities of the Lamb wave in the included angle , directions; The dimensionless parameter characterizes the ratio of the Lamb wave velocity in the direction forming an angle with the fiber main axis direction of the second wall panel to the Lamb wave velocity in the fiber main axis direction , is obtained through actual measurement.

[0028] Furthermore, the least squares method is used to construct an objective function to solve the geometric model equations, so as to obtain the position coordinates of the optimal estimate of the damage position; its objective function is expressed as follows:

[0029] ;

[0030] Among them, represents the position coordinates of the optimal estimate, , respectively represent the Euclidean distances between the damage position D and the origin O of the damage location coordinate system, and between the damage position D and the i-th receiver ; is the calculated value of the Lamb wave flight time.

[0031] Furthermore, the expression of the dimensionless parameter is:

[0032] ;

[0033] Among them, is the equivalent Young's modulus in the x-axis direction of the damage location coordinate system on the second wall panel, is the equivalent Poisson's ratio along the y-axis on the plane with the x-axis as the normal; and are respectively the equivalent Young's modulus and equivalent Poisson's ratio in the direction forming an angle with the x-axis.

[0034] Furthermore, the average reconstruction error is determined using the original signal and the reconstructed signal as the damage evaluation value at each receiver and is standardized; based on the standardized damage evaluation values at the selected multiple receivers, a comprehensive damage evaluation score is determined, including:

[0035] Using the smallest average reconstruction error among the damage evaluation values calculated based on the Lamb wave response signals collected by the selected multiple receivers as the normalization benchmark, the damage evaluation values are standardized;

[0036] The multiple standardized damage evaluation values are arithmetically averaged to obtain the average damage degree evaluation value in the current damage state of the second wall panel; introducing the standardized damage evaluation value corresponding to the reconstruction error of the Lamb wave response signal obtained on the first wall panel in the healthy state as the benchmark, a second standardization process is performed to obtain the comprehensive damage evaluation score.

[0037] A terminal device includes a processor, a memory, and a computer program stored in the memory; when the processor executes the computer program, the method for identifying and evaluating damage to a composite wall panel based on Lamb waves and an autoencoder is implemented.

[0038] A computer-readable storage medium stores a computer program; when the computer program is executed by a processor, the method for identifying and evaluating damage to a composite wall panel based on Lamb waves and an autoencoder is implemented.

[0039] Compared with the prior art, the present invention has the following technical features:

[0040] The present invention effectively solves the problem commonly existing in the prior art that only the damage location can be identified, but it is difficult to finely and quantitatively evaluate the damage degree. Through the means of differential enhancement and continuous wavelet transform, the present solution effectively extracts the flight time of Lamb waves, improving the signal processing accuracy. At the same time, by introducing an autoencoder model trained based on the healthy state, the quantitative evaluation of different damage degrees is realized by using the reconstruction error, solving the problem that it is difficult for traditional methods to measure "how serious it is". The method of the present solution can achieve the precise positioning and degree evaluation of composite material panel damage, and has strong application prospects and popularization value. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a schematic flow chart of the method of the present invention;

[0042] Figure 2 is a schematic diagram of the damage location and evaluation system of the method of the present invention;

[0043] Figure 3 is a schematic diagram for solving the Lamb wave velocity in each direction of the present invention;

[0044] Figure 4 is a schematic diagram of the deep autoencoder model used in the embodiment of the present invention;

[0045] Figure 5 is the dispersion curve of the orthogonally laminated composite material panel in the 0° direction in the embodiment of the present invention;

[0046] Figure 6 is a schematic diagram of the experimental test equipment and experimental arrangement in the embodiment of the present invention;

[0047] Figure 7 is the normalized Lamb wave response signal collected by Receiver 2 when the panel is in a non-damaged state in the embodiment of the present invention;

[0048] Figure 8 is the ratio of the Lamb wave velocity in each direction derived in the embodiment of the present invention to the Lamb wave velocity in the main axis fiber direction distribution model;

[0049] Figure 9 is the result graph after differential processing and continuous wavelet transform of the damaged Lamb wave response signals collected by each receiver in the damage case of the present invention;

[0050] Figure 10 is the visual comparison of the estimated damage location and the actual damage location in the embodiment of the present invention;

[0051] Figure 11 is the comparison of the comprehensive evaluation values of the damage degree under each damage level in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] The present invention provides a method for identifying and evaluating damage to composite panels based on Lamb waves and autoencoders. By combining technical means such as the propagation characteristics of Lamb waves, wave velocity anisotropy modeling, wavelet transform to extract the flight time, and a deep autoencoder model to calculate the reconstruction error between damaged and healthy signals, the method realizes the identification of the damage location and the evaluation of the damage degree of composite panels.

[0053] See Figure 1 , a method for identifying and evaluating damage to composite panels based on Lamb waves and autoencoders provided by this solution includes:

[0054] Step 1, arrange a Lamb wave signal acquisition system on the first panel in a healthy state; under a preset excitation signal, use receivers at different positions on the first panel to obtain Lamb wave response signals; train a deep autoencoder model corresponding to the receiver based on the Lamb wave response signals obtained by each receiver.

[0055] (1-1) The preset excitation signal.

[0056] To achieve damage location and degree evaluation of composite panels based on Lamb waves, it is necessary to first determine the parameters of the excitation signal used to stimulate the propagation of Lamb waves, mainly including: the Lamb wave mode excited, the center frequency and the signal form.

[0057] To reduce the interference caused by multimodal superposition, this solution selects a single Lamb wave mode with good propagation characteristics, weak dispersion effect, and relatively stable wave velocity (such as the symmetric mode or the antisymmetric mode ) as the main propagation mode of the excitation signal; the selection of the Lamb wave mode can be determined according to factors such as the thickness of the composite plate and the dispersion curve.

[0058] The center frequency should be set in the non-dispersive section of the dispersion curve where the wave velocity is relatively stable to improve the signal propagation stability and facilitate the subsequent extraction of the flight time and time-domain response characteristics.

[0059] The signal form of the excitation signal adopts a 5-cycle Hanning amplitude-modulated sine pulse, and its expression is as follows:

[0060] (1);

[0061] where, A = 1 represents the amplitude of the excitation signal, is the angular frequency of the excitation signal, is the time parameter.

[0062] (1-2) Lamb wave signal acquisition system.

[0063] In this solution, the panel for damage identification is a composite panel. First, a healthy panel needs to be selected as the first panel to collect relevant prior data. Since the propagation speed of Lamb waves is different along different fiber directions of the panel, in order to facilitate the positioning of the location of damage in the panel, this solution constructs a Lamb wave signal acquisition system on the panel and establishes a damage location coordinate system, specifically as follows:

[0064] This solution constructs a Lamb wave signal acquisition system; see Figure 2 , this system includes an exciter and multiple receivers arranged around the exciter, jointly constituting a sensing network for Lamb wave signal excitation and reception; both the exciter and the receivers use the same transducer element, such as piezoelectric ceramics, etc.; the damage location coordinate system is a local Cartesian coordinate system established with the location of the exciter as the origin O(0,0); it is defined that the positive direction of the x-axis is along the fiber main axis direction of the first panel, and the positive direction of the y-axis is in the same plane as the positive direction of the x-axis and perpendicular to the positive direction of the x-axis.

[0065] The area enclosed by all the receivers arranged on the first panel is the damage detection area of the Lamb wave; parameters such as the location, size, and shape of the damage detection area of the Lamb wave are determined by the user according to actual needs by adjusting the number and location of the receivers; the subsequent work of this solution is to predict the position coordinates of the damage location in the damage location coordinate system and evaluate its damage degree within the damage detection area of the second panel to be detected; Figure 2 Schematic diagram of the rectangular damage detection area formed when arranging 4 receivers.

[0066] (1-3) Deep autoencoder model.

[0067] This solution uses an unsupervised learning method based on the deep autoencoder model (Deep Autoencoder, DAE) to achieve quantitative evaluation of the damage degree without preset labels; by establishing a deep autoencoder model that only learns the "characteristics of the healthy state of the panel", the deep autoencoder model is trained to only "understand" the Lamb wave response characteristics in the healthy state. Once the input is a Lamb wave response signal containing damage, its reconstruction result will produce a large reconstruction error. Based on this principle, this solution uses the reconstruction error of the trained deep autoencoder model for Lamb wave response signals under different damage degrees as the damage degree evaluation index to achieve quantitative evaluation of the severity of damage to the composite panel.

[0068] The deep autoencoder model includes an encoder and a decoder, where:

[0069] The encoder is responsible for mapping the Lamb wave response signal X into a low-dimensional feature space to form a low-dimensional feature vector at the hidden layer; in this process, the encoder extracts the main features of the signal and compresses redundant information, thereby obtaining a concise but information-rich representation.

[0070] The decoder takes the low-dimensional feature vector at the hidden layer as input and attempts to reconstruct a reconstructed signal X that is as close as possible to the original high-dimensional Lamb wave response signal R ; the role of the decoder is to evaluate the expressive ability of the encoded features.

[0071] As shown in the Figure 4 appendix, in the training phase of this scheme, the input of the autoencoder model is the Lamb wave response signal X obtained by each receiver on the first wall panel = ; the output is the reconstructed signal X R = ; in addition,

[0072] In addition, the forward propagation process of the autoencoder model can be described as follows:

[0073] Encoding process from the input layer to the hidden layer:

[0074] (2);

[0075] Decoding process from the hidden layer to the output layer:

[0076] (3);

[0077] Among them, represents the encoding process, represents the decoding process, represents the activation function, 、 are the weight matrices of the encoding and decoding processes respectively, 、 are the bias terms, is the feature representation of the hidden layer.

[0078] The training objective of the deep autoencoder model is to minimize the reconstruction error Loss between the Lamb wave response signal and the reconstructed signal:

[0079] (4);

[0080] Among them, N is the length of the Lamb wave response signal X, , are the time-domain amplitudes of the Lamb wave response signal X and the reconstructed signal at the nth sampling point respectively; the number of training times of the deep autoencoder model is 1000 times, the learning rate is 0.02, and the optimizer is Adam.

[0081] In this solution, the deep autoencoder model is trained using unsupervised learning. The input Lamb wave response signals do not require pre-set labels and are only learned based on the Lamb wave response signals in the healthy state. Through the training of a large number of Lamb wave response signal samples, the model can master the basic feature distribution of the healthy state Lamb wave signals and reconstruct them stably.

[0082] To improve the recognition accuracy and adaptability, in this solution, for each receiver a separate deep autoencoder model with the same model framework is constructed and trained with the Lamb wave response signal X collected by this receiver to learn the feature distribution of the Lamb wave response signals obtained by this receiver at the position of the first panel in the healthy state, ensuring the reliability and regional sensitivity of the evaluation results.

[0083] Step 2: For the second panel to be detected, arrange the Lamb wave signal acquisition system in the same way as the first panel, and use the receiver to obtain the Lamb wave response signal; where the second panel is the same type of panel as the first panel.

[0084] In actual application, the second panel to be detected is the same type of panel as the first panel, that is, a composite material panel with the same material, layup, and size; first, arrange the Lamb wave signal acquisition system on the second panel, and the number, position of the exciters and receivers, and the preset excitation signals in this system are the same as those set on the first panel.

[0085] After arranging the Lamb wave signal acquisition system, the exciter generates Lamb waves based on the preset excitation signals, and the corresponding Lamb wave response signals are obtained at each receiver.

[0086] Step 3: Calculate the differential signal using the Lamb response signals obtained from the first panel and the second panel, and determine the Lamb wave flight time at each receiver on the second panel through the differential signal.

[0087] This solution provides a signal processing method for extracting the Lamb wave flight time by combining differential operation and wavelet transform technology Through differential operation on the Lamb wave response signals collected from the first panel and the second panel, the characteristics of the scattered signals caused by damage are enhanced, and then using the time-frequency focusing ability of the complex Morlet wavelet, time-frequency transformation is performed on the differential signal, so as to accurately extract the arrival time of the scattered wave and calculate the Lamb wave flight time.

[0088] (3-1) Denote the The Lamb wave response signal collected by a receiver is , and the number of sampling times is set to , the th receiver's Lamb wave response signal obtained at the th sampling is denoted as , where ; calculate the average response signal at the receiver:

[0089] (5);

[0090] (3-2) The Lamb wave response signal collected by the th receiver on the first wall panel is , then the average response signal at the receiver is:

[0091] (6);

[0092] (3-3) To enhance the signal difference caused by damage, calculate the differential signal at the th receiver:

[0093] (7);

[0094] (3-4) Apply the complex Morlet wavelet transform to the differential signal , and its wavelet function expression is:

[0095] (8);

[0096] Among them, is the imaginary unit, is the time variable, is the natural constant, is the bandwidth, is the center frequency.

[0097] The wavelet transform result presents a time-scale spectrum, and the moment corresponding to the peak of the envelope spectrum is extracted as the arrival time of the Lamb wave response signal at the th receiver on the second wall panel, denoted as .

[0098] (3-5) Determine the Lamb wave flight time at the th receiver on the second wall panel:

[0099] (9);

[0100] Among them, Indicates the starting time of the excitation signal.

[0101] Step 4: Based on the Lamb wave flight time at some receivers, construct a geometric model equation set in combination with the Lamb wave anisotropic propagation velocity distribution model; construct an objective function to solve the geometric model equation set, so as to locate the damage position in the second wall panel.

[0102] (4-1) Damage location model.

[0103] See Figure 2 , taking the second wall panel as an example for illustration; assume that there is a damage position D on the second wall panel, and its position coordinates in the damage location coordinate system are ; the black dotted line in the figure represents the propagation direction of the Lamb wave. The Lamb wave is emitted by the exciter, scattered after passing through the damage position D, and then propagates to the receiver ; the receiver 's position coordinates are .

[0104] Let the propagation time of the Lamb wave from the exciter to the damage position D be , and the propagation time from the damage position D to the i-th receiver be , then the flight time experienced by the i-th receiver from the excitation of the Lamb wave by the exciter until it receives the Lamb wave satisfies:

[0105] (10);

[0106] Among them, , respectively represent the propagation paths of the Lamb wave from the exciter to the damage position D and from the damage position D to the i-th receiver , 's lengths, , respectively represent the included angles between the propagation paths , and the positive x-axis direction of the damage location coordinate system, , are respectively the propagation velocities of the Lamb wave in the included angle , directions.

[0107] At the same time, in the Lamb wave signal acquisition system, the lengths , of the propagation paths , The geometric relationship between them can be expressed as:

[0108] (11);

[0109] The included angle 、 can be expressed as:

[0110] (12);

[0111] Through the above damage location model, namely equations (10) to (12), using multiple receivers with known position coordinates and Lamb wave velocities in different directions , an inverse equation system for the damage location can be constructed to inversely deduce the position coordinates of the damage location.

[0112] (4 - 2) Lamb wave propagation velocity distribution model in all directions.

[0113] Considering that the Lamb wave propagation velocity in the composite material wall panel is anisotropic, it is rather cumbersome to obtain the Lamb wave velocities in all directions through actual measurement; therefore, to reflect the variation of the Lamb wave propagation velocity in different directions in the second wall panel, see Figure 3 , a dimensionless parameter is introduced to characterize the Lamb wave velocity in the direction forming an angle with the fiber main axis direction of the second wall panel, which is the ratio of the Lamb wave velocity in the direction forming an angle with the fiber main axis direction of the second wall panel (i.e., 0°, consistent with the positive x-axis direction of the damage location coordinate system established in this scheme) to the Lamb wave velocity in the fiber main axis direction of the second wall panel, as shown in formula (13); thus, based on , it is only necessary to measure the Lamb wave velocity along the fiber main axis direction of the second wall panel, and then the Lamb wave velocities

[0114] in other

[0115] directions forming an angle with the fiber main axis direction of the second wall panel can be deduced. Here, is the Lamb wave propagation velocity along the fiber main axis direction of the second wall panel, is the equivalent Young's modulus in the x-axis direction of the damage location coordinate system on the second wall panel, is the equivalent Poisson's ratio along the y-axis on the plane with the normal being the x-axis; and are the equivalent Young's modulus and equivalent Poisson's ratio in the direction forming an angle with the x-axis respectively:

[0116] (14);

[0117] Among them, and are the equivalent Young's moduli in the x and y axis directions of the wall panel respectively, is the equivalent shear modulus along the y axis on the plane with the x axis as the normal.

[0118] (4 - 3) Lamb wave velocity in the fiber principal axis direction is obtained through actual measurement, and the specific steps are as follows:

[0119] A receiver is arranged at a position along the fiber principal axis direction (0°) of the second wall panel and at a specified distance L from the exciter (coordinate origin O). Then, the exciter is used to emit a preset excitation signal, and the arranged receiver is used to receive the Lamb wave response signal. Subsequently, the flight time t of the Lamb wave is extracted. Finally, it is calculated according to the following formula :

[0120] (15);

[0121] Through the Lamb wave anisotropic propagation velocity distribution models of formulas (13) to (15), the wave velocity of the Lamb wave in a specified mode in the composite wall panel along the direction at an angle of with the fiber principal axis direction can be determined, which is used for the calculation of damage location.

[0122] (4 - 4) Locate the damage position in the second wall panel.

[0123] In order to accurately identify the damage position in the second wall panel, this scheme fully considers the flight time error caused by factors such as boundary reflection, material nonlinearity, and noise during the propagation of Lamb waves. By selecting the data at the receiver with a shorter flight time, a geometric equation set is constructed in combination with the Lamb wave anisotropic propagation velocity distribution model, and the least - square optimization method is introduced to solve it. Finally, the optimal spatial position estimation value of the damage position is obtained.

[0124] (4 - 4 - 1) During the actual detection process, as the propagation path of the Lamb wave extends, the Lamb response signal collected by the receiver farther from the damage position is more likely to be affected by factors such as boundary reflection, material nonlinearity, and noise interference, thereby reducing the accuracy of extracting the flight time of the Lamb wave.

[0125] Based on this, sort the Lamb wave flight times of all the receivers on the second wall panel obtained in step 3 from short to long, and select the receivers corresponding to the first Q Lamb wave flight times for positioning modeling; among them, the value of Q can be flexibly set by the user according to the detection accuracy requirements and the actual situation of the structure.

[0126] (4-4-2) Based on the selected Q Lamb wave flight times , combined with the Lamb wave anisotropic propagation velocity distribution model, construct a geometric model equation set based on Lamb waves; this equation set describes the functional relationship between the propagation path from the damage location to each receiver and the Lamb wave flight time and Lamb wave velocity, as shown in Equation (16):

[0127] (16);

[0128] In the above formula, the Q Lamb wave flight times screened for the second wall panel and the corresponding receiver position coordinates , dimensionless parameters at each angle direction and the Lamb wave velocity in the direction of the fiber main axis are known quantities.

[0129] (4-4-3) To obtain the optimal estimation result of the damage location, this solution uses the least squares method to solve the above geometric model equation set (16); by minimizing the sum of the squared residuals between the calculated value of the Lamb wave flight time from the damage location to the i-th receiver and the actual Lamb wave flight time, the position coordinates of the optimal estimation of the damage location are obtained; its objective function is expressed as follows:

[0130] (17);

[0131] Among them, represents the Euclidean distance between point A and point B in the damage location coordinate system, that is, ; , represent the Lamb wave velocities in the directions making angles of , with the fiber main axis direction of the second wall panel.

[0132] By solving the above objective function, the position coordinates of the optimal estimation of the damage location in the second wall panel can be obtained, realizing the damage location of the composite material wall panel.

[0133] Step 5: Select the Lamb wave response signals acquired by multiple receivers on the second wall panel as the original signals, and input them into the corresponding deep autoencoder models respectively to obtain the reconstructed signals corresponding to each receiver; use the original signals and the reconstructed signals to determine the average reconstruction error as the damage evaluation value at each receiver and perform standardization processing; based on the standardized damage evaluation values at the selected multiple receivers, determine the comprehensive damage evaluation score.

[0134] Based on the positioning result of the damage location on the second wall panel, further evaluate its damage degree. Considering the complex factors such as energy attenuation, boundary reflection, and material nonlinearity during the propagation of Lamb waves, it may lead to the weakening of effective damage features in the signals collected by receivers far from the damage location, thus affecting the accuracy of damage degree evaluation; therefore, this solution selects M receivers closest to the damage location (usually M = Q), based on the trained deep autoencoder models corresponding to these receivers reconstruct the collected Lamb wave response signals, calculate the average reconstruction error, and through normalization and comprehensive processing, quantify the deviation degree of the health state between the second wall panel and the first wall panel to achieve accurate evaluation of the damage degree.

[0135] (5 - 1) Take the Lamb wave response signals collected by the selected M receivers on the second wall panel as the original signals and input them into the trained deep autoencoder models corresponding to each of them to obtain their reconstructed signals ; take the average reconstruction error between the original signal and the reconstructed signal as the damage evaluation value at the i-th receiver , and its calculation formula is as follows:

[0136] (18);

[0137] where N represents the length of the Lamb wave response signal, represents the number of sampling times of the i-th receiver; represents the average reconstruction error between the original signal and the reconstructed signal at the i-th receiver, that is, the damage evaluation value; is the time-domain amplitude of the Lamb wave response signal at the n-th sampling point in the k-th sampling of the i-th receiver ; is the deep autoencoder model according to the input signal , the time-domain amplitude of the reconstructed n-th sampling point.

[0138] (5 - 2) To eliminate the differences in absolute error values between different receivers and improve the comparability of evaluation results, use M damage evaluation values The minimum average reconstruction error As a normalization reference, for the damage assessment value Perform normalization processing:

[0139] (19);

[0140] Among them, Indicates the damage assessment value after normalization of the average reconstruction error of the th receiver.

[0141] (5 - 3) Arithmetically average the M normalized damage assessment values To obtain the average damage degree assessment value of the second wall panel under the current damage state; to quantify the deviation degree of this average damage degree assessment value from the healthy state, this solution introduces the damage assessment value after normalization corresponding to the reconstruction error of the Lamb wave response signal obtained on the first wall panel in the healthy state as a reference, and performs a second normalization processing to determine the comprehensive damage assessment score :

[0142] (20);

[0143] Among them, Indicates that the Lamb wave response signal collected by the th receiver on the first wall panel is used as the original signal, and is input into the corresponding trained deep autoencoder model According to the same method as in step (5 - 1), the average reconstruction error

[0144] between the reconstructed signal and the original signal is obtained, and then the damage assessment value after normalization is calculated according to the same method as in step (5 - 2). It should be noted that when calculating this step, the M receivers selected on the first wall panel are arranged in the same way as those on the second wall panel. The obtained comprehensive damage assessment score

[0145] Numerically reflects the deviation degree of the current second wall panel from the healthy state of the first wall panel; the larger its value, the more serious the damage degree. ) of a carbon fiber composite wall panel, the damage location and assessment experiment based on Lamb waves and deep autoencoder models was carried out using the method of the present invention; the size of this wall panel is 500×500×1.5mm 3 , and its equivalent mechanical parameters are shown in Table 1.

[0146] Table 1: Equivalent mechanical parameters of the orthogonally laminated carbon fiber composite wall panel;

[0147]

[0148] See the appendix Figure 5 , the dispersion curve of the orthogonally laminated composite panel in the 0° (fiber main axis) direction is plotted; it can be seen from the figure that when the frequency-thickness product is 0 - 0.7 MHz·mm, the S0 mode phase velocity value is relatively stable; while when it is 0.2 - 0.7 MHz·mm, the A0 mode phase velocity value is relatively stable; when the frequency-thickness product exceeds 0.7 MHz·mm, multiple modes appear, resulting in complex signal processing. Considering the thickness of the panel is 1.5 mm, in this embodiment, the center frequency kHz of the S0 mode Lamb wave is selected as the parameter of the excitation signal.

[0149] See the appendix Figure 6 , in this embodiment, the Lamb wave signal acquisition system uses piezoelectric ceramic chips (receivers) as the exciter (receiver 0) and receivers (receivers 1 - receivers 8) of the Lamb wave; among them, the distance between two adjacent piezoelectric ceramics is L = 150 mm. The excitation signal is generated by the exciter, amplified by the power amplifier and then loaded onto the surface of the panel; after the Lamb wave propagates in the panel, the Lamb wave response signal is collected by each receiver and transmitted to the computer through the data collector for storage and analysis; the sampling frequency is set to 50 MHz, and the sampling duration is 200 ; in the experiment, a magnet is selected as the additional mass to simulate the damage of the panel, and different levels of damage are simulated according to the weight of the magnet.

[0150] See the appendix Figure 7 , which shows the normalized Lamb wave response signal received by receiver receiver 2 in the undamaged state; since the distance between the excitation and the receiver is L = 150 mm, according to formula (15), the Lamb wave propagation speed along the fiber main axis direction is calculated to be approximately = 6000 m / s; further, the Lamb wave propagation velocity distribution model derived based on formula (13) is as shown in the appendix Figure 8 .

[0151] The Lamb wave response signal acquisition and the construction and training of the deep autoencoder model are respectively carried out on the panel in the healthy state (i.e., the first panel mentioned above). In this embodiment, the sampling times of each receiver are set to times.

[0152] To verify the positioning and evaluation capabilities of the proposed method, a set of damage cases are set in the damage detection area shown in Figure 6 ; the specific setting parameters of the damage cases are shown in Table 2. Among them, the sampling times of each receiver are set to times.

[0153] Table 2: Specific setting parameters of damage cases;

[0154]

[0155] Calculate the Lamb wave flight time at each receiver in the above damage case (i.e., equivalent to the second wall panel in the previous text) according to Step 3; refer to Figure 9 , which gives the Lamb wave response signals collected by each receiver under the above damage case The results after differential processing and continuous wavelet transform; according to Figure 5 In the dispersion curve information, the S0 mode has the fastest propagation speed at the frequency-thickness product ( ), so it is determined that the first wave packet in the wavelet transform result corresponds to the S0 mode. In Figure 9 , the arrival times of the Lamb wave response signals received by each receiver are clearly marked; considering that the waveform and time sequence of the excitation signal are known, combined with the experimental settings, it can be known that the excitation time of the Lamb wave is about ; accordingly, the flight time results of each receiver are calculated as shown in Table 3.

[0156] Table 3: Calculation of Lamb wave flight time at each receiver;

[0157]

[0158] During the damage location process, based on the Lamb wave flight time obtained in Table 3, in this embodiment, Receiver 1 and Receiver 2 are selected as the main basis for damage location; the least squares method is used to solve the objective function in combination with the geometric model equations, and the position coordinates of the damage location are estimated to be mm; this result is close to the true damage location D mm, with a small error. Refer to Appendix Figure 10 , and the estimated damage location and the true damage location are visually compared to verify the accuracy and effectiveness of the proposed method.

[0159] During the damage degree assessment process, for each level of damage, the Lamb wave response signals on the damaged state wall panel (i.e., the second wall panel in the previous text) collected by Receiver 1 and Receiver 2 are respectively extracted and , and input into the trained deep autoencoder models DAE1 and DAE2, and the final comprehensive damage assessment scores are as shown in Appendix Figure 11 .

[0160] The experimental results show that for each level of damage with relatively small differences in the damage degree simulated by magnets, the proposed method can effectively distinguish different damage degrees and exhibits good monotonicity, that is, the comprehensive damage assessment score gradually increases with the increase of the damage level, verifying that the method has the advantages of high positioning accuracy, strong assessment sensitivity, and good ability to distinguish damage levels in the quantitative assessment of damage degree.

[0161] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for damage identification and assessment of composite panels based on Lamb waves and autoencoders, characterized in that, Including: Arranging a Lamb wave signal acquisition system on the first wall panel in a healthy state; Under a preset excitation signal, using receivers at different positions on the first wall panel to obtain Lamb wave response signals; training a deep autoencoder model corresponding to each receiver based on the Lamb wave response signals obtained by each receiver; For the second wall panel to be detected, arranging a Lamb wave signal acquisition system in the same manner as the first wall panel, and using the receivers to obtain Lamb wave response signals; wherein, the second wall panel is of the same type as the first wall panel; Calculating a differential signal using the Lamb response signals obtained from the first wall panel and the second wall panel, and determining the Lamb wave flight time at each receiver on the second wall panel through the differential signal; Based on the Lamb wave flight times at some receivers, combining with the Lamb wave anisotropic propagation velocity distribution model to construct a geometric model equation set; constructing an objective function to solve the geometric model equation set, so as to locate the damage position in the second wall panel; Selecting the Lamb wave response signals obtained by multiple receivers on the second wall panel as the original signals, respectively inputting them into the corresponding deep autoencoder models to obtain the reconstructed signals corresponding to each receiver; using the original signals and the reconstructed signals to determine the average reconstruction error as the damage evaluation value at each receiver and performing normalization processing; based on the normalized damage evaluation values at the selected multiple receivers, determining the comprehensive damage evaluation score.

2. The method for damage identification and evaluation of composite panels based on Lamb waves and autoencoders according to claim 1, characterized in that The Lamb wave signal acquisition system includes an exciter and multiple receivers arranged around the exciter; constructing a damage location coordinate system with the position where the exciter is located as the origin, defining the positive direction of the 𝑥-axis along the fiber main axis direction of the first wall panel, and the positive direction of the y-axis is in the same plane as the positive direction of the 𝑥-axis; The preset excitation signal adopts a single Lamb wave mode, and the center frequency is set in the non-dispersive section of the dispersion curve; the form of the excitation signal adopts a sine pulse.

3. The method for damage identification and evaluation of composite panels based on Lamb waves and autoencoders according to claim 1, wherein The deep autoencoder model includes an encoder and a decoder; the encoder is used to map the Lamb wave response signal to a low-dimensional feature space to form a low-dimensional feature vector at the hidden layer; The decoder takes the low-dimensional feature vector at the hidden layer as the input, and the output is the reconstructed signal of the Lamb wave response signal; the training objective of the deep autoencoder model is to minimize the reconstruction error between the Lamb wave response signal and the reconstructed signal.

4. The method for damage identification and evaluation of composite panels based on Lamb waves and autoencoders according to claim 1, wherein Calculating a differential signal using the Lamb response signals obtained from the first wall panel and the second wall panel, and determining the Lamb wave flight time at each receiver on the second wall panel through the differential signal, including: The Lamb wave response signal collected by the th receiver on the second wall panel is , and calculate the average response signal at the th receiver ; The Lamb wave response signal collected by the th receiver on the first wall panel is , and calculate the average response signal at the th receiver ; Calculate the differential signal at the th receiver; Apply complex Morlet wavelet transform to the differential signal. The result of the wavelet transform presents a time-scale spectrogram. Extract the moment corresponding to the peak of the envelope spectrum as the arrival time of the Lamb wave response signal at the th receiver on the second wall panel. ;​ Determine the Lamb wave flight time at the th receiver on the second wall panel ; represents the start time of the excitation signal.

5. The method for identifying and evaluating damage of composite panels based on Lamb waves and autoencoders according to claim 1, wherein When locating the damage position in the second wall panel, first sort the Lamb wave flight times of all receivers on the second wall panel from short to long, and select the first Q Lamb wave flight times; Based on the selected Q Lamb wave flight times, combining with the Lamb wave anisotropic propagation velocity distribution model, constructing a geometric model equation set based on Lamb waves; The least squares method is used to construct an objective function to solve the geometric model equations; by minimizing the sum of the squared residuals between the calculated Lamb wave flight time from the damage location to the i-th receiver and the actual Lamb wave flight time to obtain the position coordinates of the optimal estimate of the damage location.

6. The method for damage identification and evaluation of composite panels based on Lamb waves and autoencoders according to claim 5, wherein The geometric model equation set is expressed as follows: ； Among them, the sixth equation of the equation set is the Lamb wave anisotropic propagation velocity distribution model; In the above formula, represents the Lamb wave flight time at the i-th receiver among the selected first Q Lamb wave flight times. There is a damage position D on the second wall panel, and its position coordinates are ; the position coordinates of the i-th receiver are ; , respectively represent the propagation paths of the Lamb wave from the actuator to the damage position D and from the damage position D to the i-th receiver , the lengths of, , respectively represent the propagation paths , the angles between and the positive x-axis direction of the damage location coordinate system, , are respectively the Lamb wave propagation speeds in the angles , directions; the dimensionless parameter characterizes the Lamb wave speed in the direction at an angle to the fiber main axis direction of the second wall panel and the Lamb wave speed in the fiber main axis direction, is obtained through actual measurement.

7. The method for damage identification and evaluation of composite panels based on Lamb waves and autoencoders according to claim 6, wherein The least squares method is used to construct an objective function to solve the geometric model equations, thereby obtaining the position coordinates of the optimal estimate of the damage location; the expression of the objective function is as follows: ； Among them, represents the position coordinates of the optimal estimate, , respectively represent the Euclidean distances between the damage position D and the origin O of the damage location coordinate system, and between the damage position D and the i-th receiver ; is the calculated value of the Lamb wave flight time.

8. The method for damage identification and evaluation of composite panels based on Lamb waves and autoencoders according to claim 6, characterized in that Dimensionless parameter has the following expression: ； Among them, is the equivalent Young's modulus in the x-axis direction of the damage location coordinate system on the second wall panel, is the equivalent Poisson's ratio along the y-axis on the plane with the x-axis as the normal; and are respectively the equivalent Young's modulus and equivalent Poisson's ratio in the direction making an angle of with the x-axis.

9. The method for damage identification and evaluation of composite panels based on Lamb waves and autoencoders according to claim 1, characterized in that The average reconstruction error is determined using the original signal and the reconstructed signal as the damage evaluation value at each receiver and is standardized. Based on the standardized damage evaluation values at multiple selected receivers, a comprehensive damage evaluation score is determined, including: Based on the smallest average reconstruction error among the damage evaluation values calculated from the Lamb wave response signals collected at multiple selected receivers as the normalization reference, the damage evaluation values are standardized. The multiple standardized damage evaluation values are arithmetically averaged to obtain the average damage degree evaluation value in the current damage state of the second wall panel; the standardized damage evaluation value corresponding to the reconstruction error of the Lamb wave response signal obtained on the first wall panel in the healthy state is introduced as the reference for the second standardization process to obtain the comprehensive damage evaluation score.

10. A terminal device, comprising a processor, a memory, and a computer program stored in the memory; characterized in that, When the processor executes the computer program, it implements the method for identifying and evaluating the damage of the composite wall panel based on Lamb waves and autoencoders according to any one of claims 1-9.

Images