Laser Ultrasonic Guided Wave Damage Detection Method Based on Interpretable Deep Transfer Learning

Abstract

This invention discloses a laser-ultrasonic guided wave damage detection method based on interpretable deep transfer learning, belonging to the field of non-destructive testing. The method includes using laser detection to inspect the structure under test and calculating its dispersion curve to obtain the analysis frequency band of the detection signal. A basic detection model is then constructed, acquiring the laser detection signal dataset of the structure under test, calculating a multi-source loss function, and constraining the loss function into the basic detection model to obtain a damage depth transfer detection model. This model is iteratively trained until the number of iterations reaches a set threshold or the model reaches convergence, resulting in an interpretable deep learning detection model. This interpretable deep learning detection model is then used to identify laser-ultrasonic guided wave detection samples of the structure under test and outputs the corresponding labels for the detected samples. This invention solves the problems of poor interpretability and insufficient applicability of traditional laser-ultrasonic guided wave detection signal detection models.

Description

Technical Field

[0001] This invention relates to the field of nondestructive testing, and in particular to a laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning. Background Technology

[0002] Non-destructive testing based on ultrasonic guided waves has attracted increasing attention due to its wide monitoring range. Excitation methods for ultrasonic guided waves include electromagnetic ultrasound, piezoelectric ultrasound, and laser ultrasound. Among these, laser ultrasonic guided wave testing technology is considered a highly attractive damage detection technique due to its advantages such as non-contact operation, visualization, long working distance, and high sensitivity.

[0003] Among various laser detection methods, laser spatial light modulation excitation can achieve focusing at any position and has great potential in laser detection. However, the propagation mechanism of the ultrasonic guided wave field of complex structure laser is still unclear, and the characteristics of the detection signal are complex and difficult to distinguish.

[0004] Deep learning directly obtains the mapping relationship between ultrasonic guided wave signals and structural states from the detection data, and constructs a damage detection model by quantifying the high-dimensional feature distance of the detection signals under different states of the tested object. Because it avoids the complex prior knowledge requirements caused by guided wave dispersion and multimodal characteristics, it has attracted much attention in the field of laser ultrasonic guided wave nondestructive testing. However, current deep learning detection methods still mainly focus on model design, and their integration with the nondestructive testing mechanism is not deep enough, making it difficult to guarantee the stability and reliability of the models.

[0005] To address these issues, there is an urgent need for laser ultrasonic guided wave damage detection methods based on interpretable deep transfer learning. Summary of the Invention

[0006] To address the aforementioned issues, this application proposes a laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning. This overcomes the limitation of existing laser ultrasonic detection signal feature mining techniques in mining complex vibration signals, thereby improving the detection accuracy and damage identification precision of laser ultrasonic damage detection. The specific steps are as follows: A laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning includes the following steps: S1: Laser detection is used to detect the structure under test and calculate the dispersion curve of the structure under test to obtain the analysis frequency band of the detection signal. The frequency band with relatively simple modal component type and weak frequency shift phenomenon in the dispersion curve is selected as the frequency band for feature extraction of the detection signal, i.e., the analysis frequency band of the detection signal. The frequency band signal is further analyzed by filtering. S2: Construct a guided wave waveform dictionary using an atomic library, extract guided wave packet information from the analysis frequency band signal based on the guided wave waveform dictionary, and then construct a basic detection model; S3: Obtain the signal dataset of laser detection of the structure under test, calculate the multi-source loss function based on the signal dataset of laser detection, and constrain the loss function into the basic detection model to obtain the damage depth transfer detection model. S4: Iteratively train the damage depth transfer detection model until the number of iterations reaches a set threshold, or the model reaches the convergence condition to obtain an interpretable deep learning detection model. S5: Use an interpretable deep learning detection model to identify laser ultrasonic guided wave detection samples of the structure under test, and output the corresponding labels of the detection samples.

[0007] Preferably, the laser detection method in S1 includes two types: single-point detection and modulation detection. The excitation energy for single-point detection is a Gaussian beam, and the detection signal is used to determine whether the position of the excitation point is damaged. Modulation detection enables damage detection at any specified location by optimizing the excitation shape.

[0008] Preferably, in S1, the dispersion curve of the structure under test is calculated by numerical solution, and the frequency of the target wave velocity change is selected as the analysis frequency band of the detection signal.

[0009] Preferably, the specific content of constructing the guided wave waveform dictionary through the atomic library in S2 is as follows: The atomic library is constructed based on the signal characteristics of the guided wave signal; Let the excitation signal be Waveguide propagation distance is used x It indicates that the spread x The guided wave signal is used This indicates. Then x The guided wave signal at =0 is: ; In the frequency domain, it can be represented as: ; According to the steady-state phase method, The Fourier transform of the signal at point can be expressed as: ; in Indicates frequency as ω The wavenumber function. Taking the inverse Fourier transform of the above equation yields... The time-domain signal at that location.

[0010] ; Given an initial guided wave signal and the dispersion relation of the guided wave in the structure, the guided wave waveform after the initial signal has propagated any distance can be obtained. This waveform can be expressed in terms of propagation distance and wavenumber corresponding to the mode. That is: ; By setting the detection distance and accuracy parameters one by one, a series of guided wave dispersion signals can be obtained as atoms. A set of equally spaced propagation distances is given. X ; ; Then the excitation signal in modal m Method of transmission X The dispersive sub-dictionary is composed of a series of signals obtained from the distance. It can be represented as: ; Given M A set of wavenumber functions for each mode The resulting general dictionary of dispersion It can be represented as: ; Each sub-dictionary consists of n Composed of +1 atomic signals, totaling M Each sub-dictionary, then the dictionary The CCP has ( n +1) M Atom signal.

[0011] Preferably, the specific content of extracting guided wave packet information from the analysis frequency band signal based on the guided wave waveform dictionary, and then constructing the basic detection model, is as follows: The basic detection model is established using a one-dimensional convolutional neural network. Through atomic dictionary extract( n +1) M guided wave packet atomic signals, after determining the time domain signal sampling frequency, obtain the data length and envelope information of the atomic signals; The data length and value of the convolution kernel of the convolutional layer of the basic detection model are determined based on the wave packet data length. The matrix parameters for determining the kernel weights are based on the envelope shape.

[0012] Preferably, the one-dimensional convolutional neural network includes a feedforward layer, a convolutional layer, and a pooling layer; The output features of the convolutional layer are: ; In the formula, It is a convolutional layer. x The input features of the convolutional layer are Ker, where Ker is the kernel weight. The kernel data length of the first two convolutional layers is equal to the atomic dictionary. (in) n +1) The atomic signals of the M guided wave packets are consistent. b For deviation terms, For Ker and x Convolution calculation between them The activation function that generates a nonlinear mapping between the input and output is mainly used to introduce nonlinearity into the data; here, it is the ReLU activation function. The output characteristics of the pooling layer are: ; In the formula, z The output features of the convolutional layer, This is a max pooling layer, whose function is to calculate the maximum value of elements within the pooling window.

[0013] Preferably, the laser detection signal dataset of the structure under test is obtained, and the specific content of the multi-source loss function is calculated based on the laser detection signal dataset as follows: Acquire the signal dataset of laser detection of the structure under test, and extract key frequency band information through filtering; The signal dataset from laser detection is input into the basic detection model to extract high-dimensional features of the signal in the dataset. The signals acquired under the target detection conditions are labeled as different source domains, and the multi-source loss function is calculated. The damage depth migration detection model is obtained by optimizing the basic detection model through loss function feedback, and the model is confirmed by a preset number of iterations.

[0014] Preferably, the multi-source loss function of the damage depth migration detection model is: ; In the formula, E This is the total loss value. Loss for identifying structural health status labels For domain label recognition loss, and These are sample conditional classifiers. Domain Discriminator The parameter vector in; Domain label recognition loss The expression is: ; In the formula, It is the first j Each source domain label loss term, k It is the number of source domains, and the classification loss for a single source domain. Defined as: ; In the formula, n It is the total number of samples. d These are the actual domain tags. The parameters control the weights between the sample state label and the domain label loss during the learning process.

[0015] Preferably, the structural health status label recognition loss is based on a label error weighting factor, expressed as: ; Definition of the first j The first in the source domain n If the first sample is normal, then the second sample is normal. j Loss of predicting state labels for source domain samples for: ; In the formula, the feature vector is generated by the label predictor. Mapping to tags y , The parameter vector representing the mapping. This represents the label error weighting factor. This is the sample state label prediction loss. The structural health state label loss term is calculated using the negative log probability loss, and its calculation formula is as follows: ; Its output is: ; In the formula, and These are calculation parameters. f The activation function is softmax.

[0016] Preferably, the model's output labels include domain labels and sample state labels. The sample state includes normal signals and damaged signals, meaning that the sample state prediction loss includes normal signal prediction loss and damaged signal prediction loss.

[0017] A laser ultrasonic guided wave damage detection device based on interpretable deep transfer learning, used to implement a laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning, comprising: Nanosecond pulsed laser, spatial light modulator, laser receiving probe, laser ultrasound receiving module, and computer; The laser exciter is connected to the laser excitation probe, and the laser ultrasonic receiver module is connected to the computer. Among them, the laser exciter is used to excite the laser signal and excite the object under test through the spatial light modulator; The laser detection probe is used to acquire the vibration signal of the structure under test by laser ultrasonic testing and transmit it to the laser ultrasonic receiving module through optical fiber; The laser ultrasound receiving module is used to receive the vibration signal from the laser ultrasound detection probe and send it to the computer. The computer is used to receive vibration signals from laser ultrasonic testing, perform damage detection on the vibration signals, and output the damage detection results.

[0018] In summary, the laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning of the present invention has the following advantages compared with traditional techniques: This invention first clarifies the key frequency band information of the object under test by calculating the dispersion curve and constructs the guided wave atom library information; secondly, it constructs a feature extraction module for a one-dimensional deep migration detection model using the atom library information; thirdly, it sets the objective function of the multi-source domain migration detection model and aligns different signal feature spaces, solving the problems of poor interpretability and insufficient applicability of traditional laser ultrasonic guided wave detection signal detection models, and improving the recognition accuracy of laser guided wave detection data.

[0019] The technical method of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning in the embodiment. Figure 2 This is a diagram of the damage depth migration detection model architecture in the embodiment; Figure 3 This is a schematic diagram of the laser ultrasonic damage detection device in the embodiment; Figure 4 The image shown is of a U-shaped bend with simulated damage in the embodiment.

[0021] Explanation of icon numbers: 1. Laser exciter; 2. Scanning frame; 3. Laser ultrasonic receiver module; 4. Laser excitation probe; 5. Laser detection probe; 6. Computer. Detailed Implementation

[0022] The technical method of the present invention will be further described below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of this application.

[0023] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the scope of this application and its application or use.

[0024] Techniques, systems, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, they should be considered part of the instruction manual.

[0025] In all the examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0026] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0027] This invention provides a laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning, such as... Figure 1 and Figure 2 As shown, it includes the following steps: S1: Laser detection is used to inspect the structure under test and calculate the value of the copper pipe under test (such as U-shaped bends). Figure 4 The dispersion curve of the structure (as shown) is used to obtain the analysis frequency band of the detection signal.

[0028] Furthermore, the laser detection method in S1 is a single-point detection method: the excitation energy for single-point detection is a Gaussian beam, and the detection signal is used to determine whether the position of the excitation point is damaged.

[0029] Furthermore, in S1, the dispersion curve of the copper pipe structure under test is calculated using a numerical solution method, and the frequency of the target wave velocity change is selected as the analysis frequency band of the detection signal. Here, the three main guided wave modes in the pipe are mainly considered: the first-order torsional mode T0, the bending mode B0, and the longitudinal mode L0.

[0030] S2: Construct a guided wave waveform dictionary using an atomic library, extract guided wave packet information from the analysis frequency band signal based on the guided wave waveform dictionary, and then construct a basic detection model.

[0031] Furthermore, the specific content of constructing the guided wave waveform dictionary through the atomic library in S2 is as follows: The atomic library is constructed based on the signal characteristics of the guided wave signal; Let the excitation signal be Waveguide propagation distance is used x This indicates that the guided wave signal after propagation x is used... This indicates. Then x The guided wave signal at =0 is: ; In the frequency domain, it can be represented as: ; According to the steady-state phase method, The Fourier transform of the signal at point can be expressed as: ; in Indicates frequency as ωThe wavenumber function. Taking the inverse Fourier transform of the above equation yields... The time-domain signal at that location, ; Given an initial guided wave signal and the dispersion relation of the guided wave in the structure, the guided wave waveform after the initial signal has propagated any distance can be obtained. This waveform can be expressed in terms of propagation distance and wavenumber corresponding to the mode. That is: ; Here, the detection distance is set to 50mm and the positional accuracy to 2mm, allowing the acquisition of a series of guided wave dispersion signals as atoms. Given a set X of equally spaced propagation distances; ; Then the excitation signal in modal m A dispersive sub-dictionary composed of a series of signals obtained by propagating over a distance of 50mm. It can be represented as: ; Given a set of wavenumber functions for a certain mode The resulting total dictionary of dispersion of the three guided wave modes—torsional mode T0, bending mode B0, and longitudinal mode L0—is then obtained. It can be represented as: .

[0032] Each sub-dictionary consists of 26 atomic signals, and there are 3 sub-dictionaries in total. There are a total of 78 atomic signals.

[0033] Furthermore, the specific content of extracting guided wave packet information from the frequency band signal based on the guided wave waveform dictionary, and then constructing the basic detection model, is as follows: The basic detection model is established using a one-dimensional convolutional neural network. Through atomic dictionary extract( n +1) M guided wave packet atomic signals, after determining the time domain signal sampling frequency, obtain the data length and envelope information of the atomic signals; The data length and value of the convolution kernel of the convolutional layer of the basic detection model are determined based on the wave packet data length. The matrix parameters for determining the kernel weights are based on the envelope shape.

[0034] Furthermore, the one-dimensional convolutional neural network includes a feedforward layer, a convolutional layer, and a pooling layer, and the output features of the convolutional layer are: .

[0035] In the formula, It is a convolutional layer. x The input features of the convolutional layer are Ker, where Ker is the kernel weight. The kernel data length of the first two convolutional layers is equal to the atomic dictionary. (in) n +1) The atomic signals of the M guided wave packets are consistent. b For deviation terms, For Ker and x Convolution calculation between them The activation function that generates a nonlinear mapping between the input and output is mainly used to introduce nonlinearity into the data; here, it is the ReLU activation function.

[0036] The output characteristics of the pooling layer are: ; In the formula, z represents the output feature of the convolutional layer. This is a max pooling layer, whose function is to calculate the maximum value of elements within the pooling window.

[0037] The one-dimensional convolutional network consists of six layers. The kernel size of the first layer is the same as the length of the atomic signal, and it has 78 channels. The convolutional layers and pooling layers of the last five layers are configured similarly. The specific network structure of the copper pipe detection network in this invention is shown in Table 1.

[0038] Table 1 Network Structure

[0039] This approach allows for the rapid extraction of modal features from the network in the first convolutional layer, ensuring that the features of each modal unit are quickly and fully extracted. This significantly accelerates the network's convergence speed and improves its feature extraction capabilities.

[0040] S3: Obtain the signal dataset of laser detection of the structure under test, calculate the multi-source loss function based on the signal dataset of laser detection, and constrain the loss function into the basic detection model to obtain the damage depth transfer detection model.

[0041] The specific content of the multi-source loss function calculated based on the laser detection signal dataset of the structure under test is as follows: the laser detection signal dataset of the structure under test is obtained, and the laser is set into multiple source domains according to the incident angle of the laser. Key frequency band information is extracted by filtering.

[0042] This application uses a U-shaped tube for experiments, and the laser incident angle error of the U-shaped tube in the experiment is 90°±15.0°.

[0043] The experiment included two labels: normal and damaged. The damaged pipes were of the same size and located at random positions on the circumference.

[0044] The acquired dataset is as follows: 1) Training set: A total of 9 U-shaped pipes were used as experimental research objects, of which 1 was normal and 8 had cracks near the bend of the pipe. Detection signals were collected from the pipes under laser excitation at incident angles of 90°, 75°, and 105°, with 20 samples collected at each location. The collected data were labeled as different source domains according to different locations. Therefore, the experiment contained three source domains.

[0045] 2) Test set: A total of 6 U-shaped pipes were used for testing, of which 2 were normal and 4 had cracks and damage near the bends of the pipes.

[0046] Each pipe collects detection signals at two excitation locations: one with an incident angle of 82° and the other with an incident angle of 98°. Twenty samples are collected at each location.

[0047] To achieve consistency between positive and negative samples, the number of normal samples is increased to match that of damaged samples by adding 5% Gaussian noise. The number of samples collected for each laser incident angle in this invention is shown in Table 2.

[0048] Table 2. Number of samples collected for each laser incident angle in the case study experiment.

[0049] The signal dataset from laser detection is input into the basic detection model to extract high-dimensional features of the signals in the dataset.

[0050] The signals acquired under the target detection conditions are labeled as different source domains, and the multi-source loss function is calculated.

[0051] The damage depth migration detection model is obtained by optimizing the basic detection model through loss function feedback, and the model is confirmed by a preset number of iterations.

[0052] During the training phase of the detection model, the goal is to minimize the source domain label prediction loss, which ensures good prediction performance of the label predictor on the overall samples of the source domain. The loss of the proposed method is divided into domain label recognition loss and structural health status label recognition loss.

[0053] Furthermore, the multi-source loss function of the damage depth migration detection model is: .

[0054] In the formula, E This is the total loss value. Loss for identifying structural health status labels For domain label recognition loss, and These are sample conditional classifiers. Domain Discriminator The parameter vector in.

[0055] As can be seen from the formula, minimizing the target object is to obtain the saddle point parameters of the function. , , The calculation method is as follows: .

[0056] .

[0057] Domain label recognition loss The expression is: .

[0058] In the formula, It is the first j Each source domain label loss term, k It is the number of source domains, and the classification loss for a single source domain. It can be defined as: .

[0059] In the formula, n It is the total number of samples. d These are the actual domain tags. The parameters control the weights between the sample state label and the domain label loss during the learning process. Since the parameters in the equation cannot simultaneously achieve the desired results... minimize and through maximize Therefore, in Adding a minus sign at the beginning makes the optimization direction consistent.

[0060] Furthermore, the structural health status label recognition loss is based on a label error weighting factor, expressed as: .

[0061] To adjust the prediction loss contribution of normal signal and damaged signal tags, this application proposes a tag error weighting factor. If the first n samples in the j-th source domain are defined as normal, then the state label prediction loss for the j-th source domain sample is... for: .

[0062] In the formula, the feature vector is generated by the label predictor. Mapping to tags y , The parameter vector representing the mapping. This represents the label error weighting factor. This is the sample state label prediction loss. The structural health state label loss term is calculated using the negative log probability loss, and its calculation formula is as follows: .

[0063] Its output is: .

[0064] In the formula, and These are calculation parameters. f The activation function is softmax.

[0065] Furthermore, the sample state prediction loss includes normal signal prediction loss and damaged signal prediction loss.

[0066] The weights for different source domains are determined using adaptive weight coefficients in the formula. During model training, the training objective is to minimize the total loss function value through stochastic gradient descent. After training, the model's performance is tested using test samples from the target domain.

[0067] S4: Iteratively train the damage depth transfer detection model until the number of iterations reaches a set threshold, or the model reaches the convergence condition to obtain an interpretable deep learning detection model.

[0068] S5: An interpretable deep learning detection model is used to identify laser ultrasonic guided wave detection samples of the tested structure and output the corresponding labels for the detected samples. The simulated damage inside the pipe is crack damage, and the labels are divided into two categories: normal samples and damaged samples. The effectiveness of the method is not explained, but the proposed method is compared with current classical methods. The identification results (test signal detection accuracy) of different methods are shown in Table 3.

[0069] Table 3. Signal detection accuracy (%) of different testing methods

[0070] A laser ultrasonic guided wave damage detection device based on interpretable deep transfer learning is provided to implement a laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning, such as... Figure 3 As shown, it includes: a nanosecond pulsed laser, a spatial light modulator, a laser receiving probe, a laser ultrasound receiving module 3, and a computer 6.

[0071] The laser exciter 1 is connected to the laser excitation probe 4, and the laser ultrasonic receiver module 3 is connected to the computer 6.

[0072] Among them, the laser exciter 1 is used to excite the laser signal and excite the object under test through the spatial light modulator.

[0073] The laser detection probe 5 is used to acquire the vibration signal of the structure under test by laser ultrasonic testing and transmit it to the laser ultrasonic receiving module 3 through optical fiber.

[0074] The laser ultrasound receiving module 3 is used to receive the vibration signal of laser ultrasound detection sent by the laser detection probe 5 and send it to the computer 6.

[0075] Computer 6 is used to receive vibration signals from laser ultrasonic testing, perform damage detection on the vibration signals, and output the damage detection results.

[0076] An electronic device includes a memory and a processor, wherein the memory stores a computer program that can run on the processor, characterized in that the processor executes the computer program to implement the steps of a laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning.

[0077] The electronic device of the present invention can execute the non-contact laser ultrasonic damage detection method based on deep transfer learning of the present invention, and can execute any combination of implementation steps of the method embodiments, and has the corresponding functions and beneficial effects of the method.

[0078] A computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of a laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning.

[0079] The computer storage medium of the present invention stores instructions or programs that can execute the non-contact laser ultrasonic damage detection method based on deep transfer learning of the present invention. When the instructions or programs are run, any combination of implementation steps of the method embodiment can be executed, and the method has the corresponding functions and beneficial effects.

[0080] The technical solution of this invention can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0081] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical methods of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical methods of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical methods to deviate from the spirit and scope of the technical methods of the present invention.

Claims

1. A laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning, characterized in that, Includes the following steps: S1: Laser detection is used to detect the structure under test and the dispersion curve of the structure under test is calculated to obtain the analysis frequency band of the detection signal; S2: Construct a guided wave waveform dictionary using an atomic library, extract guided wave packet information from the analysis frequency band signal based on the guided wave waveform dictionary, and then construct a basic detection model; S3: Obtain the signal dataset of laser detection of the structure under test, calculate the multi-source loss function based on the signal dataset of laser detection, and constrain the loss function into the basic detection model to obtain the damage depth transfer detection model. S4: Iteratively train the damage depth transfer detection model until the number of iterations reaches a set threshold, or the model reaches the convergence condition to obtain an interpretable deep learning detection model. S5: Use the interpretable deep learning detection model to identify laser ultrasonic guided wave detection samples of the structure under test, and output the corresponding labels of the detection samples; The specific content of constructing the guided wave waveform dictionary using the atomic library in S2 is as follows: The atomic library is constructed based on the signal characteristics of the guided wave signal; The preset excitation signal is The guided wave propagation distance is x ,spread x The subsequent guided wave signal is , t It is a time-domain guided wave signal; but x The guided wave signal at =0 is: ； x The guided wave signal at =0 can be represented in the frequency domain as: ； According to the steady-state phase method, for any point p, The Fourier transform of the signal at point is expressed as: ； in Indicates frequency as ω wavenumber function, ω Angular frequency, i The imaginary unit, This is the location where the guided wave signal propagates; Taking the inverse Fourier transform of the above equation yields... Time-domain signal at: ； Given an initial guided wave signal and the dispersion relation of the guided wave in the structure, the guided wave waveform after the initial signal has propagated any distance is obtained. This waveform is expressed in terms of propagation distance and wavenumber corresponding to the mode. That is: ； in, For scalar potential, By setting the detection distance and accuracy parameters one by one, a series of guided wave dispersion signals are obtained as atoms, given a set of propagation distances with equal spacing. X : ； in, p For any discrete point location, the excitation signal is modally... m Method of transmission X The dispersive sub-dictionary is composed of a series of signals obtained from the distance. Represented as: ； Given M A set of wavenumber functions for each mode The resulting general dictionary of dispersion Represented as: ； Each sub-dictionary consists of n Composed of +1 atomic signals, totaling M Each sub-dictionary, then the dictionary The CCP has ( n +1) M One atomic signal; The specific content of extracting guided wave packet information from the frequency band signal based on the guided wave waveform dictionary, and then constructing the basic detection model, is as follows: The basic detection model is established using a one-dimensional convolutional neural network. Through atomic dictionary extract( n +1) M For each guided wave packet atomic signal, after determining the time-domain signal sampling frequency, the data length and envelope information of the atomic signal are obtained; The data length and value of the convolution kernel of the convolutional layer of the basic detection model are determined based on the wave packet data length. The matrix parameters for determining the kernel weights are based on the envelope shape.

2. The laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning according to claim 1, characterized in that, The laser detection methods in S1 include two types: single-point detection and modulation detection. The excitation energy for single-point detection is a Gaussian beam, and the detection signal is used to determine whether the position of the excitation point is damaged. Modulation detection enables damage detection at any specified location by optimizing the excitation shape.

3. The laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning according to claim 1, characterized in that, The one-dimensional convolutional neural network includes a feedforward layer, a convolutional layer, and a pooling layer; The output features of the convolutional layer are: ； In the formula, It is a convolutional layer. x The input features are the features of the convolutional layer, and Ker is the weight of the convolutional kernel. b For deviation terms, For Ker and x Convolution calculation between them An activation function that generates a non-linear mapping between inputs and outputs; The output characteristics of the pooling layer are: ； In the formula, z The output features of the convolutional layer, This is a max pooling layer, whose function is to calculate the maximum value of elements within the pooling window.

4. The laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning according to claim 1, characterized in that, The laser detection signal dataset of the structure under test is obtained, and the specific content of the multi-source loss function is calculated based on the laser detection signal dataset: Acquire the signal dataset of laser detection of the structure under test, and extract key frequency band information through filtering; The signal dataset from laser detection is input into the basic detection model to extract high-dimensional features of the signal in the dataset. The signals acquired under the target detection conditions are labeled as different source domains, and the multi-source loss function is calculated. The damage depth migration detection model is obtained by optimizing the basic detection model through loss function feedback, and the model is confirmed by a preset number of iterations.

5. The laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning according to claim 4, characterized in that, The multi-source loss function of the damage depth migration detection model is: ； In the formula, E This is the total loss value. Loss for identifying structural health status labels For domain label recognition loss, and These are sample conditional classifiers. Domain Discriminator The parameter vector in; Domain label recognition loss The expression is: ； In the formula, It is the first j There are source domain label loss terms, where k is the number of source domains and the classification loss per source domain is... Defined as: ； In the formula, N It is the total number of samples. d These are the actual domain tags. The parameters control the weights between the sample state label and the domain label loss during the learning process. It is the signal of the i-th sample in the j-th source domain. It is the first j Domain label loss values ​​in the source domain.

6. The laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning according to claim 5, characterized in that, The structural health status label recognition loss is based on a label error weighting factor, expressed as follows: ； Definition of the first j If the first n samples in the n source domains are normal, then the nth sample... j Loss of predicting state labels for source domain samples for: ； In the formula, The output feature vector is generated by the sample conditional classifier. Mapped to label y, The parameter vector representing the mapping. This represents the label error weighting factor. This is the sample state label prediction loss. The structural health state label loss term is calculated using the negative log probability loss, and its calculation formula is as follows: ； Its output is: ； In the formula, and These are calculation parameters. f The activation function is softmax.

7. A laser ultrasonic guided wave damage detection device based on interpretable deep transfer learning, used to implement the laser ultrasonic guided wave damage detection method based on interpretable deep transfer learning as described in any one of claims 1-6, characterized in that, include: Nanosecond pulsed laser, spatial light modulator, laser receiving probe, laser ultrasound receiving module, and computer; The laser exciter is connected to the laser excitation probe, and the laser ultrasonic receiver module is connected to the computer. Among them, the laser exciter is used to excite the laser signal and excite the structure under test through the spatial light modulator; The laser detection probe is used to acquire the vibration signal of the structure under test by laser ultrasonic testing and transmit it to the laser ultrasonic receiving module through optical fiber; The laser ultrasound receiving module is used to receive the vibration signal from the laser ultrasound detection probe and send it to the computer. The computer is used to receive vibration signals from laser ultrasonic testing, perform damage detection on the vibration signals, and output the damage detection results.

Images