Structural damage identification method based on digital twinning and feature migration

The damage sample data of the bridge structure is generated through digital twin technology and the feature migration component analysis is carried out, which solves the problems of insufficient data and high computing resources in damage detection of traditional bridge structures, improves the accuracy and efficiency of detection, and is suitable for target domain data with limited data scene.

CN119027741AActive Publication Date: 2024-11-26CHANGSHA UNIVERSITY OF SCIENCE AND TECHNOLOGY

Patent Information

Application Number
CN202411234132.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-04
Publication Date
2024-11-26
Estimated Expiration
2044-09-04

AI Technical Summary

Technical Problem

Traditional bridge structure damage detection methods have problems such as low detection accuracy, low efficiency, insufficient data and high computing resource requirements, resulting in reduced model accuracy and increased cost.

Method used

Using a structural damage recognition method based on digital twins and feature migration, the source and target domain data sets are constructed, and the acceleration response signal is generated using the digital twin model, feature extraction and migration component analysis are carried out, and the damage recognition model is constructed to reduce the annotated data. Rely on and improve the adaptability and accuracy of the model.

Benefits of technology

It effectively reduces the cost of data acquisition of damage samples, improves the accuracy and robustness of the model, is suitable for scenarios with limited target domain data, and improves the efficiency and accuracy of damage detection of bridge structures.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119027741A_ABST
    Figure CN119027741A_ABST
Patent Text Reader

Abstract

The invention discloses a structural damage identification method based on digital twinning and feature migration, and the method comprises the steps: carrying out the feature extraction of a first time-frequency image of a source domain data set and a second time-frequency image of a target domain data set, and obtaining a first feature vector and a second feature vector; respectively combining all the first feature vectors and all the second feature vectors to obtain a first feature matrix and a second feature matrix; splicing the first feature matrix and the second feature matrix to obtain a combined feature matrix; carrying out migration component analysis on the joint feature matrix to obtain a source domain feature vector and a target domain feature vector; training a damage identification model by using the source domain feature vector and the damage type of the source domain data set; and performing training and performance evaluation on the trained damage identification model by using the target domain feature vector and the damage type of the target domain data set. The method solves the problem that a finite element model is used for obtaining sample data, and the requirement for computing resources is high.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of structural damage identification, and in particular relates to a structural damage identification method, device, storage medium and product based on digital twin and feature migration. Background Art

[0002] Existing structural health monitoring technologies mainly rely on the collection and analysis of sensor data to identify structural damage through vibration signals, strain data, etc. However, traditional methods usually require a large amount of labeled data for training, with high data collection costs and unstable data quality, and are easily disturbed by environmental noise. With the rapid development of digital twin technology, a new method for structural health monitoring is provided by building virtual digital models to simulate and monitor the state of physical structures. At the same time, transfer learning, as a machine learning technology, can use the knowledge of trained models to quickly adapt and optimize new tasks, thereby reducing dependence on a large amount of labeled data and improving the generalization ability and accuracy of the model.

[0003] As an important part of transportation infrastructure, the health of bridges is directly related to public safety and economic development. As the service life of bridges increases, problems such as material aging and fatigue damage gradually emerge. Efficient and accurate damage detection of bridge structures has become a technical problem that needs to be solved urgently. At present, traditional bridge structure damage detection methods mainly rely on manual inspections and simple monitoring equipment. These methods have the following shortcomings:

[0004] (1) Low detection accuracy: Manual inspections are easily limited by the experience and subjective judgment of inspectors, resulting in missed detections and misjudgments;

[0005] (2) Low efficiency: Manual inspection takes a long time and it is difficult to achieve real-time monitoring and rapid response of bridge structures;

[0006] (3) Insufficient data: Traditional methods have difficulty in obtaining sufficient damage sample data, which affects the training effect and accuracy of the detection model;

[0007] (4) High computing resource requirements: The establishment of finite element models and simulation calculations require a lot of professional knowledge and computing resources, which increases the complexity and cost of the detection system. Summary of the invention

[0008] The purpose of the present invention is to provide a structural damage identification method based on digital twins and feature migration to solve at least one of the problems in traditional methods, namely, high cost of damage sample data collection, reduced accuracy due to insufficient damage sample data, and high demand for computing resources when obtaining sample data using finite element models.

[0009] The present invention solves the above technical problems through the following technical solutions: a structural damage identification method based on digital twins and feature migration, comprising:

[0010] Constructing a source domain dataset and a target domain dataset; wherein each sample in the source domain dataset includes a first time-frequency image of the structure and its damage type, and each sample in the target domain dataset includes a second time-frequency image of the structure and its damage type, and the first time-frequency image and the second time-frequency image are both converted from the acceleration response signal of the structure; the source domain dataset is constructed using a digital twin model of the structure;

[0011] Performing feature extraction on each of the first time-frequency images to obtain a first feature vector; combining all the first feature vectors to obtain a first feature matrix; performing feature extraction on each of the second time-frequency images to obtain a second feature vector; combining all the second feature vectors to obtain a second feature matrix;

[0012] Concatenating the first feature matrix and the second feature matrix to obtain a joint feature matrix;

[0013] Performing migration component analysis on the joint feature matrix to obtain a source domain feature vector and a target domain feature vector in a common feature space;

[0014] Constructing a damage recognition model, and performing supervised training on the damage recognition model using the source domain feature vector and the damage type of the source domain data set;

[0015] Using the target domain feature vector to perform unsupervised training on the supervised trained damage recognition model;

[0016] The target domain feature vector and the damage type of the target domain data set are used to evaluate the performance of the damage recognition model after unsupervised training to obtain a target recognition model.

[0017] Furthermore, the digital twin model of the structure is used to construct the source domain dataset, including:

[0018] Step A1: construct a parametric finite element model of the structure;

[0019] Step A2: performing dynamic analysis on the parameterized finite element model and exporting an Inp file;

[0020] Step A3: using a Python script to open the Inp file, and parsing the unit set and its material properties in the Inp file;

[0021] Step A4: modify the material properties of the unit set according to different damage types of the structure, update the material properties of different damage locations, and obtain Inp files of different damage types;

[0022] Step A5: Use the subprocess module to perform ABAQUS simulation calculations on the Inp files of different damage types to obtain ODB result files of different damage types;

[0023] Step A6: Use Python script to open the ODB result files of different damage types, extract the acceleration response signals of the structural nodes from the ODB result files of different damage types, and obtain the acceleration response signals under different damage types;

[0024] Step A7: converting the acceleration response signals under different damage types into images to obtain first time-frequency images under different damage types;

[0025] Step A8: construct a source domain dataset according to the first time-frequency images under different damage types.

[0026] Furthermore, the acceleration response signal of the structure is converted into a first time-frequency image or a second time-frequency image by using a spectral Markov transfer field, including:

[0027] The frequency spectrum signal of the acceleration response signal is calculated using the following formula:

[0028]

[0029] Wherein, s(k) represents the kth component in the spectrum signal, x(i) represents the value of the i-th sampling point in the acceleration response signal, n represents the number of sampling points of the acceleration response signal, abs() represents the modulus value function, and j represents the imaginary unit;

[0030] Performing normalization processing on each component in the spectrum signal to obtain a normalized component of the spectrum signal;

[0031] The value range of the amplitude set is evenly divided to obtain Q sub-intervals; wherein the amplitude set includes the amplitudes of all normalized components of the spectrum signal;

[0032] Each amplitude in the amplitude set is mapped to a subinterval; wherein the specific mapping formula is:

[0033]

[0034] Among them, A k represents the kth amplitude in the amplitude set, q j represents the jth subinterval, argmin j It means finding j that makes the function reach the minimum value;

[0035] Generate a spectral Markov transition field matrix according to the probability of each amplitude being transferred from the current subinterval to other subintervals;

[0036] The spectral Markov transfer field matrix is ​​converted into a two-dimensional image to obtain a first time-frequency image or a second time-frequency image.

[0037] Furthermore, a pre-trained convolutional neural network model is used to perform feature extraction on the first time-frequency image and the second time-frequency image; wherein the pre-trained convolutional neural network model does not include a fully connected layer.

[0038] Furthermore, performing migration component analysis on the joint feature matrix includes:

[0039] A joint Gaussian kernel matrix is ​​constructed according to the joint feature matrix. The specific formula is:

[0040]

[0041] Among them, k ij represents the element in the i-th row and j-th column of the joint Gaussian kernel matrix; x i 、x j represents the two eigenvectors in the joint feature matrix; σ represents the bandwidth of the Gaussian kernel function;

[0042] The joint centralized kernel matrix is ​​constructed according to the joint Gaussian kernel matrix. The specific formula is:

[0043]

[0044] Among them, K c represents the joint centralized kernel matrix; K represents the joint Gaussian kernel matrix; H represents the centralized matrix; I and 1 both represent the all-1 matrix; n s Represents the number of samples in the source domain dataset, n t Indicates the number of samples in the target domain dataset;

[0045] A maximum mean difference matrix is ​​constructed, which is expressed as:

[0046]

[0047] Where M represents the maximum mean difference matrix; 1 a×b Represents a matrix of size a×b with all 1s;

[0048] The objective function is constructed according to the joint centralized kernel matrix and the maximum mean difference matrix; wherein the expression of the objective function is:

[0049]

[0050] Where A represents the projection matrix, superscript T represents the transpose, λ represents the regularization parameter, and tr() represents the trace of the matrix; It means solving the optimal projection matrix so that the distribution difference between the source domain and the target domain in the common feature space is minimized;

[0051] Solving the objective function to obtain a projection matrix;

[0052] constructing a first centralized kernel matrix according to the first feature matrix and the joint feature matrix, and obtaining a source domain feature vector according to the first centralized kernel matrix and the projection matrix;

[0053] A second centralized kernel matrix is ​​constructed according to the second feature matrix and the joint feature matrix, and a target domain feature vector is obtained according to the second centralized kernel matrix and the projection matrix.

[0054] Further, generalized eigenvalue decomposition is used to solve the objective function, including:

[0055] Convert the objective function into a generalized eigenvalue problem:

[0056] (K c MK c +λK c )A=K c AΛ;

[0057] Where Λ represents the eigenvalue diagonal matrix, which contains the first k smallest eigenvalues;

[0058] The eigenvector corresponding to the minimum eigenvalue is selected as the column vector of the projection matrix to obtain the projection matrix.

[0059] Furthermore, the identification method further includes:

[0060] Acquire a real-time acceleration response signal of the structure, and convert the real-time acceleration response signal into a time-frequency image;

[0061] The target recognition model is used to perform online damage recognition on the time-frequency image.

[0062] Based on the same concept, the present invention also provides an electronic device, including a memory, a processor, and a computer program / instruction stored in the memory, wherein the processor executes the computer program / instruction to implement the structural damage identification method as described above.

[0063] Based on the same concept, the present invention also provides a computer-readable storage medium on which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the structural damage identification method as described above is implemented.

[0064] Based on the same concept, the present invention also provides a computer program product, including a computer program / instruction, which implements the structural damage identification method as described above when executed by a processor.

[0065] Beneficial Effects

[0066] Compared with the prior art, the advantages of the present invention are:

[0067] The present invention uses digital twin technology to generate a large number of acceleration response signals under different damage types, and then constructs a source domain data set, reducing the dependence on a large amount of actual annotated data, ensuring the accuracy of the model, and solving the problems of high cost of damage sample data collection, reduced model accuracy due to insufficient damage sample data, and high computing resource requirements for obtaining sample data using finite element models.

[0068] Through feature migration component analysis, a large amount of source domain data and limited target domain data are mapped to a common feature space, so that the distribution difference between the source domain features and the target domain features in the common feature space is minimized, which effectively improves the adaptability and performance of the recognition model in new tasks, improves the accuracy and robustness of the model, and is particularly suitable for scenarios with limited target domain data. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the technical solution of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only one embodiment of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0070] Figure 1 is a flow chart of a method for identifying structural damage in an embodiment of the present invention;

[0071] Figure 2 is a digital twin model of the bridge structure generated by the Inp file in an embodiment of the present invention;

[0072] Figure 3 is a partial acceleration response signal of the bridge structure extracted from the ODB result file in an embodiment of the present invention;

[0073] Figure 4 is an acceleration response signal obtained from a field test model of a bridge structure in an embodiment of the present invention;

[0074] Figure 5 1 is a comparison of modal parameters of the field test model of the bridge structure and the digital twin model in different orders in the embodiment of the present invention; wherein, from left to right, they are first order, second order and third order;

[0075] Figure 6 is a frequency domain signal of the field test model of the bridge structure in an embodiment of the present invention;

[0076] Figure 7 are frequency domain signals of different digital twin models in the embodiments of the present invention;

[0077] Figure 8 is a schematic diagram of a process of converting an acceleration response signal into a two-dimensional time-frequency image in an embodiment of the present invention;

[0078] Fig. 9 is a two-dimensional time-frequency image obtained by using a spectral Markov transition field in an embodiment of the present invention;

[0079] Fig.10 is a visualization diagram of damage recognition results of the target recognition model for the first time-frequency image in an embodiment of the present invention;

[0080] Fig.11 is a visualization diagram of damage identification results of the second time-frequency image without using the TCA method in an embodiment of the present invention;

[0081] Fig.12 is a visualization diagram of the damage identification result of the second time-frequency image by the method of the present invention in an embodiment of the present invention;

[0082] Fig.13 is the recognition accuracy of 8 damage categories on the target domain dataset without using TCA and with using TCA in the embodiment of the present invention;

[0083] Fig.14 It is the damage identification confusion matrix of the method of the present invention on the target domain data set in the embodiment of the present invention. DETAILED DESCRIPTION

[0084] The following is a clear and complete description of the technical solutions in the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0085] The technical solution of the present application is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0086] Example 1

[0087] like Figure 1 As shown, a structural damage identification method based on digital twin and feature migration provided by an embodiment of the present invention includes the following steps:

[0088] Step 1: Construct source domain dataset and target domain dataset.

[0089] Each sample in the source domain dataset includes the first time-frequency image of the structure and its damage type, and each sample in the target domain dataset includes the second time-frequency image of the structure and its damage type. Both the first time-frequency image and the second time-frequency image are converted from the acceleration response signal of the structure.

[0090] In order to solve the problems of high cost of collecting damage sample data, reduced model accuracy due to insufficient damage sample data, and high demand for computing resources for obtaining sample data using finite element models, the present invention uses digital twin technology to generate a large number of structural damage samples. In a specific embodiment of the present invention, a digital twin model of a structure is used to construct a source domain data set, including:

[0091] Step A1: Construct a parametric finite element model of the structure.

[0092] In this embodiment, the specific steps of constructing the parametric finite element model of the structure are:

[0093] Step A1.1: According to the geometric dimensions of the structure, a finite element model of the structure is constructed using the ABAQUS finite element simulation platform;

[0094] Step A1.2: Mesh the finite element model and apply loads and boundary conditions to simulate the working state of the structure in the actual environment; in this embodiment, the mesh size is 0.1m to ensure accurate details;

[0095] Step A1.3: Define the components in different regions of the meshed finite element model as unit sets, set appropriate material properties for each unit set, and generate a parameterized finite element model.

[0096] In this embodiment, the material properties include elastic modulus, density and Poisson's ratio. Taking the bridge structure as an example, the elastic modulus is set to 2.1×10 5 N / mm 2 , the density is set to 7.9×10 -6 kg / mm 3 , Poisson’s ratio is set to 0.3.

[0097] Step A2: Define the node set for field variable output, perform dynamic implicit analysis on the parameterized finite element model, and export the Inp file.

[0098] In this embodiment, the field variable output is a vertical acceleration response signal. The node of the field variable output is consistent with the position of the acceleration sensor of the actual structure. The Inp file contains material parameters, dimensional parameters, boundary conditions and loads, etc. Based on the Inp file obtained using finite element technology, a large number of numerical models similar to the structural model (including experimental structural models and real structures) are automatically modeled using digital twin technology, and then automated simulation calculations are performed to provide reasonable and sufficient training samples.

[0099] Step A3: Use Python script (or Python software) to open the Inp file and parse the unit set and its material properties in the Inp file.

[0100] The Python code to open the Inp file and read its content is: with open('model.inp','r')asfile:content=file.read()

[0101] Step A4: modify the material properties of the unit set according to the damage type of the structure, update the material properties of the damage location, and obtain the Inp file of the damage type (ie, the updated Inp file).

[0102] Taking bridge structure as an example, Figure 2 The figure shows the transformation from the finite element model of the bridge structure to the digital twin model generated using the Inp file. According to the damage type of the structure, the damage location (i.e., the unit set) and the corresponding material properties that need to be updated are searched, the material properties of the damage location are set, the material update of the damage location is realized, and the synchronization of the digital twin model and the physical structure is ensured. The specific Python code is: content = content.replace(old_material, new_material)

[0103] Step A5: Use the subprocess module to perform ABAQUS simulation calculations on the updated Inp file to obtain an ODB result file.

[0104] The subprocess module in the Python programming language is used to run commands in the command line environment of the operating system to control the execution of external programs. In this embodiment, the subprocess module is used to execute the command of the ABAQUS simulation software on the updated Inp file to perform finite element simulation calculations on the updated Inp file to obtain an ODB result file, which includes the acceleration response signal under the damage type. The specific Python code is: subprocess.run(['abaqus', 'job=new_model', 'interactive'])

[0105] Step A6: After the simulation calculation is completed, use the Python script to open the ODB result file and extract the acceleration response signal of the structural node from the ODB result file.

[0106] After the simulation calculation is completed, the generated ODB result file is automatically read using the Python API of ABAQUS, the node set output by the predefined field variables is selected, the acceleration response signal is extracted, and the extracted acceleration response signal is written into the CSV file to complete the automatic data extraction. For example, the truss node set output by the predefined field variables is selected, and the acceleration response signal of the truss node set is extracted. Figure 3 The following figure shows some acceleration response signals extracted from the ODB result file. The specific Python code is:

[0107] odb=visualization.openOdb('model.odb')

[0108] extract_acceleration_data('model.odb','Truss-NodeS')

[0109] Step A7: For different damage types, repeat steps A3 to A6 to obtain acceleration response signals under different damage types.

[0110] For different damage types, the material properties are automatically updated to generate Inp files of different damage types. The specific Python code is: new_inp_file = update_inp_file ('model.inp', material_changes)

[0111] The Python script is used to automatically submit the ABAQUS simulation calculation of the Inp files of different damage types by calling the subprocess module; the ABAQUS simulation calculation is run in the background, multiple simulation tasks are processed in parallel, and the output log is captured to monitor the calculation progress and results, thereby improving the efficiency of the automated modeling process.

[0112] The acceleration response signals under different damage types are integrated into a large data set to facilitate the construction of the source domain data set. The specific code is:

[0113] combined_data.to_csv('combined_data.csv',index=False)

[0114] The present invention uses digital twin technology to obtain acceleration response signals under different damage types. It only needs to build a finite element model once and then generate a corresponding Inp file. Python is used to programmatically modify the Inp file, read the file content, write various damage type data corresponding to the actual structure, and then run the script through the Abaqus-Command command window to perform the calculation. This method does not need to open the Abaqus-CAE software, and the calculation can be automatically completed by calling the solver, thereby improving the calculation efficiency.

[0115] Step A8: converting the acceleration response signals under different damage types into images to obtain first time-frequency images under different damage types;

[0116] Step A9: construct a source domain dataset according to the first time-frequency images under different damage types.

[0117] In a specific embodiment of the present invention, the target domain data set is constructed by a structural field test model. Taking a bridge structure as an example, a bridge structure field test model is designed, and acceleration sensors are arranged at key positions of the bridge structure field test model to collect acceleration response signals of the bridge structure under different damage types or damage conditions. Figure 4 The acceleration response signal collected by the acceleration sensor in the field test model of the bridge structure is shown. During the field test, the damage condition is simulated by changing the cable force of the bridge cable, as shown in Table 1.

[0118] Table 1 Different damage conditions of cable-stayed bridge cables

[0119]

[0120] This test uses the percentage of material elastic modulus damage to model the cable damage of the bridge structure in order to effectively identify the damage; vehicles with different weights and speeds are used as external loads. In Table 1, S and N represent the initials of the cables on both sides, respectively. The damage conditions of this embodiment include damage location, vehicle speed, vehicle weight, and degree of damage. The moving vehicle is used as the external load of the bridge. The vehicle is driven by a motor, and the speed is controlled by controlling the speed of the motor. Damage condition 1: healthy, different combinations of vehicle speeds and vehicle weights; Damage condition 2: the first cable S1 on one side is damaged, different combinations of vehicle speeds and vehicle weights, the damage degree is 10%, 20% or 50%; Damage condition 3: the third cable S3 on one side is damaged, different combinations of vehicle speeds and vehicle weights, the damage degree is 10%, 20% or 50%; Damage condition 4: the fifth cable S5 on one side is damaged, different combinations of vehicle speeds and vehicle weights, the damage degree is 10%, 20% or 50%; Damage condition 5: the first and fourth cables S1 / S4 on one side are damaged, different combinations of vehicle speeds and vehicle weights The combination of vehicle weight and the degree of damage is 10%, 20% or 50%; Damage condition 6: The second and fifth cables S2 / S5 on one side are damaged, and the combination of different vehicle speeds and different vehicle weights, the degree of damage is 10%, 20% or 50%; Damage condition 7: The fourth cable S4 on one side is damaged and the second cable N2 on the other side is damaged, and the combination of different vehicle speeds and different vehicle weights, the degree of damage is 10%, 20% or 50%; Damage condition 8: The sixth cable S6 on one side is damaged and the fifth cable N5 on the other side is damaged, and the combination of different vehicle speeds and different vehicle weights, the degree of damage is 10%, 20% or 50%. Each damage condition forms multiple sub-conditions through different combinations of vehicle speed, vehicle mass and degree of damage.

[0121] By selecting widely distributed damage, the robustness and accuracy of the present invention in different scenarios can be more comprehensively evaluated. The data of the operating status and all damage conditions are from the same excitation scenario to ensure the consistency and comparability of the data. The acceleration response signals under different damage conditions are obtained through the field test model of the bridge structure, and the acceleration response signals under different damage conditions are converted into time-frequency images to obtain the second time-frequency images under different damage conditions. The target domain data set is constructed according to the second time-frequency images under different damage conditions.

[0122] Figure 5 The figure shows the comparison of the modal parameters of the field test model of the bridge structure and the modal parameters of the digital twin model of the bridge structure. Figure 5 It can be seen that the modal parameters of the field test model of the bridge structure are very similar to the modal parameters of the digital twin model. Figure 6 and Figure 7The error values ​​of the first three characteristic frequencies of the field test model of the bridge structure and the digital twin model of the bridge structure are listed (the error is expressed in percentage), among which the maximum difference is only 4.10% and does not exceed 5.00%. Therefore, the modal parameters ( Figure 4 ) and characteristic frequencies (Table 2) indicate that the digital twin model can effectively reflect the dynamic characteristics of the test bridge structure.

[0123] Table 2 Characteristic frequencies of the field test model and digital twin model of the bridge structure

[0124]

[0125] The acceleration response signals obtained by the digital twin model and the field test model are preprocessed, and then the preprocessed acceleration response signals are converted into time-frequency images. The acceleration response signals obtained by the digital twin model correspond to the first time-frequency image, and the acceleration response signals obtained by the field test model correspond to the second time-frequency image. In this embodiment, the preprocessing includes normalization processing and filtering processing. The normalization processing process is: calculating the maximum and minimum values ​​of the acceleration response signal, normalizing the amplitude of each acceleration response signal, and mapping it to the [-1,1] interval to ensure consistency. Filtering processing is used to eliminate noise and outliers.

[0126] There are many ways to convert a one-dimensional time series signal (i.e., acceleration response signal) into a two-dimensional time-frequency image, such as Markov Transfer Field (MTF), Gram's Angle Field (GADF), Recursive Graph (RT), Short-Time Fourier Transform (STFT), etc. Conventional methods of converting time series to two-dimensional time-frequency images often face the limitation of batch processing when processing time series of different lengths. Since the length of the acceleration response signal is not consistent under different external loads, and the length of the acceleration response signal is also different during each measurement, the signal length becomes an important influencing factor in feature classification.

[0127] In order to solve the above problems, the present invention uses spectral Markov transfer field SMTF to convert the acceleration response signal into a time-frequency image, which specifically includes:

[0128] Step 1.1: Calculate the frequency spectrum of the acceleration response signal. The calculation formula is:

[0129]

[0130] Where s(k) represents the kth component in the spectrum signal, x(i) represents the value of the i-th sampling point in the acceleration response signal, n represents the number or length of sampling points of the acceleration response signal, abs() represents the modulus value function, and j represents the imaginary unit. The kth component s(k) in the spectrum signal is a complex number, and the kth component s(k) represents the amplitude and phase information at frequency k. represents the Fourier kernel function.

[0131] Step 1.2: Perform normalization processing on each component in the spectrum signal to obtain the normalized component of the spectrum signal.

[0132] In this embodiment, a minimum-maximum normalization method is used to normalize each component in the spectrum signal. The specific formula is:

[0133]

[0134] Where u(k) represents the kth normalized component of the spectrum signal, s max Represents the maximum component in the spectrum signal, s min Represents the minimum component in the spectral signal.

[0135] Step 1.3: Evenly divide the value range of the amplitude set to obtain Q sub-intervals; wherein the amplitude set includes the amplitudes of all normalized components of the spectrum signal.

[0136] Since each amplitude in the amplitude set is normalized, the value range of the amplitude set is [0,1]. The value range of the amplitude set is evenly divided, that is, the logarithmic interval [0,1] is evenly divided to obtain Q sub-intervals, each of which can be expressed as q j , j∈[1,Q].

[0137] Step 1.4: Map each amplitude in the amplitude set to a subinterval; the specific mapping formula is:

[0138]

[0139] Among them, A k represents the kth amplitude in the amplitude set; q j represents the jth subinterval; argmin j It means finding j that makes the function reach the minimum value, that is, finding The minimum j. Formula (3) can be used to determine which sub-interval each amplitude in the amplitude set is currently in.

[0140] Step 1.5: Generate a spectral Markov transition field matrix based on the probability of each amplitude transferring from the current subinterval to other subintervals.

[0141] In this embodiment, the spectral Markov transition field matrix of size n×n can be expressed as:

[0142]

[0143]

[0144] Among them, p(A a ∈q i ∣A b ∈q j ) indicates that the amplitude is at qj at frequency k = b, and after multiple state changes, it is at q at k = a i The conditional probability of .

[0145] Step 1.6: Convert the spectral Markov transfer field matrix into a two-dimensional time-frequency image.

[0146] The element value of each position in the spectral Markov transfer field matrix represents the gray value of the corresponding pixel point of the image, thereby generating a two-dimensional time-frequency image. Before converting the spectral Markov transfer field matrix into a two-dimensional time-frequency image, the spectral Markov transfer field matrix is ​​first standardized so that the gray value of each pixel point of the image is between 0 and 255, which is convenient for subsequent image processing and pattern recognition. The time-frequency image is saved in PNG image format for subsequent analysis and processing.

[0147] Figure 8 The process of converting the acceleration response signal into a two-dimensional time-frequency image is shown. Fig. 9 The two-dimensional time-frequency image obtained by converting different acceleration response signals using spectral Markov transfer fields is shown. Converting one-dimensional time series signals into two-dimensional time-frequency images can better capture the time and frequency characteristics of the signal through time-frequency analysis. These features may not be directly or fully represented in the original one-dimensional time series signal, and the time-frequency diagram can reveal the local characteristics and trends of the signal.

[0148] Step 2: Extract features from each first time-frequency image to obtain a first feature vector; combine all first feature vectors to obtain a first feature matrix.

[0149] Step 3: Extract features from each second time-frequency image to obtain a second feature vector; combine all second feature vectors to obtain a second feature matrix.

[0150] Directly performing transfer component analysis (TCA) on the source and target datasets may not fully capture the differences between the source and target domains, or may result in unstable results. This is because the original data usually contains noise or redundant information, which may interfere with the effectiveness of TCA. Before performing TCA, feature extraction is performed first, especially using deep learning models (such as convolutional neural network models CNN), which can extract more stable and discriminative features, so that TCA can better align the data distribution of the source and target domains. The dimension of the original data is usually very high, and direct TCA may result in excessive computation. Feature extraction can reduce the dimension of the data, thereby reducing the computational complexity of TCA and improving computational efficiency. Directly performing TCA on high-dimensional raw data may cause the damage identification model to overfit to the source domain dataset and fail to generalize well to the target domain; by extracting features from time-frequency images, the data can be simplified and the risk of overfitting can be reduced. Therefore, performing feature extraction before performing TCA can make TCA more effective and reliable. Especially when dealing with high-dimensional and complex raw data, feature extraction can greatly improve the overall performance and robustness of transfer learning.

[0151] In a specific embodiment of the present invention, a pre-trained convolutional neural network model is used to extract features from the first time-frequency image and the second time-frequency image, that is, a pre-trained convolutional neural network model is used to extract features from the first time-frequency image to obtain a first feature vector, and a pre-trained convolutional neural network model is used to extract features from the second time-frequency image to obtain a second feature vector. The API provided by PyTorch is called to load the convolutional neural network model. In this embodiment, the convolutional neural network model uses the ResNet-50 model, and the features of the first time-frequency image or the second time-frequency image are extracted through the ResNet-50 model. During the feature extraction process, the parameters of all layers in the ResNet-50 model are frozen to avoid parameter modification during feature extraction. The ResNet-50 model has been pre-trained on the ImageNet dataset and has a strong image feature extraction capability.

[0152] The original ResNet-50 model consists of multiple convolutional layers, pooling layers, fully connected layers, etc., and has the ability to extract multi-level image features. In order to use the ResNet-50 model as a feature extractor, the last fully connected layer is deleted, that is, the ResNet-50 model of the present invention does not include a fully connected layer. The ResNet-50 model of the present invention includes multiple convolutional layers and a global average pooling layer. The 2048-dimensional feature vector output by the global average pooling layer is the feature vector required by the present invention.

[0153] In this embodiment, the image processing library PIL is used to load the time-frequency image from which features need to be extracted, ensuring that the loaded image format is correct and not damaged, and the loaded time-frequency image is adjusted to the input size (224×224) required by the ResNet-50 model. In addition, the time-frequency image needs to be normalized using the mean and standard deviation of ImageNet.

[0154] The convolutional neural network model can learn richer feature expressions from time-frequency images; when extracting image features, the convolutional neural network model often automatically performs dimensionality reduction to retain the most important features and reduce noise interference.

[0155] Combine all the first eigenvectors to get the first eigenmatrix Combine all the second eigenvectors to get the second characteristic matrix Among them, n s Represents the number of samples in the source domain dataset, n t represents the number of samples in the target domain data set, and d represents the dimension of the first feature vector or the second feature vector. In this embodiment, d is 2048.

[0156] Step 4: Concatenate the first feature matrix and the second feature matrix to obtain a joint feature matrix.

[0157] The joint feature matrix X can be expressed as

[0158] Step 5: Perform migration component analysis on the joint feature matrix to obtain the source domain feature vector and target domain feature vector in the common feature space.

[0159] The goal of transfer component analysis is to find a projection matrix A that minimizes the distribution difference between the source domain and the target domain in the common feature space after projection while retaining the discriminative information of the data.

[0160] In a specific embodiment of the present invention, performing migration component analysis on the joint feature matrix includes:

[0161] Step 5.1: Construct a joint Gaussian kernel matrix based on the joint feature matrix X. The specific formula is:

[0162]

[0163] Among them, k ij represents the element in the i-th row and j-th column of the joint Gaussian kernel matrix; x i 、x j represents the two eigenvectors in the joint feature matrix; σ represents the bandwidth of the Gaussian kernel function. The joint Gaussian kernel matrix can be expressed as K ij =kij (x i ,x j ), the joint Gaussian kernel matrix K is obtained by calculating the kernel function value between any two eigenvectors in the joint feature matrix.

[0164] In order to capture complex feature relationships, a kernel TCA variant is used to map the original feature space to a high-dimensional feature space. The similarity between samples is calculated by the kernel function. Since the image feature vector is high-dimensional, TCA faces computational challenges when processing this data. Therefore, the present invention uses a Gaussian kernel function to help process complex nonlinear relationships.

[0165] Step 5.2: Construct a joint centralized kernel matrix based on the joint Gaussian kernel matrix. The specific formula is:

[0166] K c =HKH (7)

[0167]

[0168] Among them, K c represents the joint centralized kernel matrix, that is, the distribution difference matrix between the source domain and the target domain; H represents the centralized matrix; I and 1 both represent all-1 matrices.

[0169] Step 5.3: In order to measure the distribution difference between the source domain and the target domain, the maximum mean difference is used as the metric to construct the maximum mean difference matrix, which is expressed as:

[0170]

[0171] Where M represents the maximum mean difference matrix; 1 a×b Represents an all-one matrix of size a×b.

[0172] Step 5.4: Construct the objective function based on the joint centered kernel matrix and the maximum mean difference matrix.

[0173] In this embodiment, the expression of the objective function is:

[0174]

[0175] Where A represents the projection matrix; the superscript T represents the transpose; λ represents the regularization parameter; tr() represents the trace of the matrix; It means solving the optimal projection matrix so that the distribution difference between the source domain and the target domain in the common feature space is minimized. The first term on the right side of the objective function equal sign represents the distribution difference between the source domain and the target domain, and the second term is the regularization term, which is used to avoid overfitting.

[0176] Step 5.5: Solve the objective function and obtain the projection matrix.

[0177] In this embodiment, generalized eigenvalue decomposition is used to solve the objective function, including:

[0178] Convert the objective function into a generalized eigenvalue problem:

[0179] (K c MK c +λK c )A=K c AΛ (11)

[0180] Where Λ represents the eigenvalue diagonal matrix, which contains the first k smallest eigenvalues. The eigenvector corresponding to the smallest eigenvalue is selected as the column vector of the projection matrix to obtain the projection matrix.

[0181] Step 5.6: Construct a first centralized kernel matrix according to the first feature matrix and the joint feature matrix, and obtain the source domain feature vector according to the first centralized kernel matrix and the projection matrix.

[0182] The source domain dataset and the target domain dataset are projected into the common feature space through the projection matrix. In order to ensure that the distribution of the source domain and the target domain in the common feature space is closer, the centralized kernel matrix of the source domain (i.e., the first centralized kernel matrix) and the centralized kernel matrix of the target domain (i.e., the second centralized kernel matrix) are projected. The construction process of the first centralized kernel matrix is:

[0183] First, construct the source domain Gaussian kernel matrix based on the first feature matrix and the joint feature matrix. The specific formula is:

[0184]

[0185] Among them, Ks represents the source domain Gaussian kernel matrix, (K s ) ij Represents the features of the i-th row and j-th column in the source domain Gaussian kernel matrix; represents the i-th first eigenvector in the first characteristic matrix, x j represents the jth eigenvector (first eigenvector or second eigenvector) in the joint feature matrix. The source domain Gaussian kernel matrix K s Represents the kernel matrix between the first feature matrix and the joint feature matrix, the source domain Gaussian kernel matrix K s Each row corresponds to a first eigenvector of the first characteristic matrix, and each column corresponds to an eigenvector of the first characteristic matrix or the second characteristic matrix. In this way, the source domain Gaussian kernel matrix K s The similarity between the first feature matrix (i.e., source domain) and the joint feature matrix (i.e., source domain + target domain) is captured.

[0186] Referring to formula (7) and formula (8), the source domain Gaussian kernel matrix Ks Perform centralization processing to obtain the first centralized kernel matrix Project the first centralized kernel matrix to obtain the source domain eigenvector. The projection formula is:

[0187]

[0188] Among them, Z s Represents the source domain feature vector.

[0189] Step 5.7: Construct a second centralized kernel matrix based on the second feature matrix and the joint feature matrix, and obtain the target domain feature vector based on the second centralized kernel matrix and the projection matrix.

[0190] The construction process of the second centralized core matrix is:

[0191] First, construct the target domain Gaussian kernel matrix based on the second feature matrix and the joint feature matrix. The specific formula is:

[0192]

[0193] Among them, Kt represents the Gaussian kernel matrix of the target domain, (K t ) ij Represents the features of the i-th row and j-th column in the Gaussian kernel matrix of the target domain; represents the i-th second eigenvector in the second characteristic matrix, x j represents the jth eigenvector (first eigenvector or second eigenvector) in the joint feature matrix. The target domain Gaussian kernel matrix K t Represents the kernel matrix between the second feature matrix and the joint feature matrix, the target domain Gaussian kernel matrix K t Each row corresponds to a second eigenvector of the second feature matrix, and each column corresponds to an eigenvector of the first feature matrix or the second feature matrix. In this way, the target domain Gaussian kernel matrix K t The similarity between the second feature matrix (i.e., target domain) and the joint feature matrix (i.e., source domain + target domain) is captured.

[0194] Referring to formula (7) and formula (8), the Gaussian kernel matrix K of the target domain is t Perform centralization to obtain the second centralized kernel matrix Project the second centralized kernel matrix to obtain the target domain eigenvector. The projection formula is:

[0195]

[0196] Among them, Z t Represents the target domain feature vector.

[0197] Step 6: Construct a damage identification model and use the source domain feature vector and the damage type of the source domain dataset to perform supervised training on the damage identification model.

[0198] In a specific embodiment of the present invention, the damage identification model uses a support vector machine classifier. The support vector machine classifier is supervisedly trained using the source domain feature vector and the damage type of the source domain data set. The specific expression is:

[0199] Classifier = SVM().fit(Z s ,Y s ) (16)

[0200] Among them, Classifier represents the variable name of the support vector machine classifier, SVM() represents the constructor of the support vector machine classifier, fit represents calling the fit method of the support vector machine classifier to train the support vector machine classifier, and Y s Represents the damage type of the source domain dataset, that is, the true label corresponding to the source domain feature vector. Represents the source domain feature vector Z s Each row of represents a sample. Formula (16) represents the creation of a support vector machine classifier using the source domain feature vector Z s and the true label Y s The trained damage identification model is saved in the variable Classifier.

[0201] Step 7: Use the target domain feature vector to perform unsupervised training on the damage recognition model after supervised training.

[0202] Step 8: Use the target domain feature vector and the damage type of the target domain dataset to evaluate the performance of the damage recognition model after unsupervised training to obtain the target recognition model.

[0203] The damage recognition model after unsupervised training is used to identify damages on the target domain feature vector to obtain the predicted label, and the predicted label is compared with the corresponding damage type in the target domain data set to determine the recognition accuracy, and the performance is evaluated based on the recognition accuracy. The target recognition model after supervised and unsupervised training has a high damage recognition accuracy.

[0204] Step 9: Online identification of structural damage.

[0205] The real-time acceleration response signal of the structure is obtained and converted into a real-time time-frequency image. The pre-trained convolutional neural network model is used to extract features from the real-time time-frequency image to obtain a real-time feature vector. The target recognition model is used to perform online damage identification on the real-time feature vector, which can identify the damage type of the real-time acceleration response signal from the real structure, effectively improving the accuracy and automation of real structural health monitoring.

[0206] In order to verify the effectiveness of the method of the present invention, the target recognition model (i.e., the trained support vector machine classifier) ​​is used to perform damage recognition on the first time-frequency image of the source domain data set. T-distributed random neighbor embedding is a visualization tool that can be used to obtain a clear intuition of the data feature space. In order to realize the visualization of the damage recognition results of the target recognition model (i.e., the trained support vector machine classifier) ​​on the first time-frequency image, t-distributed random neighbor embedding is used to realize the display of damage recognition results, such as Fig.10 shown. Fig.10 In the figure, solid circles of different colors mark different damage types. It can be seen that the target recognition model of the present invention can well identify the damage category of the first time-frequency image.

[0207] In order to illustrate the advantages of the present invention, the damage recognition results of the target domain dataset by the method without TCA are compared with the damage recognition results of the target domain dataset by the method of the present invention. Fig.11 and Fig.12 As shown. Fig.11 and Fig.12 It can be seen that when TCA is not used, the damage identification results are messy, the distribution overlaps seriously, and there is no clustering, which means that the features extracted by CNN do not contribute much to category separation, resulting in poor recognition accuracy; while the method of the present invention obviously clusters and separates the damage identification results, indicating that the method of the present invention has good damage identification ability.

[0208] Fig.13 The figure shows the recognition accuracy of 8 damage categories on the target domain dataset without TCA and with TCA (i.e., the method of the present invention). Fig.13 It can be seen that the recognition accuracy of the method of the present invention in each damage category is better than that of the method without using TCA. Fig.14 The damage recognition confusion matrix of the method of the present invention on the target domain data set is shown. The number of samples on the diagonal of the recognition confusion matrix is ​​the number of correctly predicted samples. The target recognition model of the present invention exhibits a high damage recognition accuracy.

[0209] Example 2

[0210] An embodiment of the present invention also provides an electronic device, which includes: a memory, a processor, and a computer program / instructions stored in the memory, and the processor executes the computer program / instructions to implement the structural damage identification method in the embodiment of the present application.

[0211] Although not shown, the electronic device includes a processor, which can perform various appropriate operations and processes according to the program and / or data stored in the read-only memory (ROM) or the program and / or data loaded from the storage portion into the random access memory (RAM). The processor can be a multi-core processor, or it can include multiple processors. In some embodiments, the processor can include a general main processor and one or more special coprocessors, such as a central processing unit, a graphics processing unit (GPU), a neural network processor (NPU), a digital signal processor (DSP), etc. In RAM, various programs and data required for device operation are also stored. The processor, ROM and RAM are connected to each other via a bus. The input / output (I / O) interface is also connected to the bus.

[0212] The processor and the memory are used together to execute the program / instructions stored in the memory. When the program / instructions are executed by the computer, the methods, steps or functions described in the above embodiments can be implemented.

[0213] Although not shown, an embodiment of the present invention further provides a computer-readable storage medium on which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the structural damage identification method in the embodiment of the present application is implemented.

[0214] Storage media in embodiments of the present invention include permanent and non-permanent, removable and non-removable items that can be implemented by any method or technology to store information. Examples of storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission medium that can be used to store information that can be accessed by a computing device.

[0215] Readable storage media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0216] Although not shown, an embodiment of the present invention further provides a computer program product, including: a computer program / instruction, which implements the structural damage identification method in the embodiment of the present application when executed by a processor.

[0217] What is disclosed above is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or modifications within the technical scope disclosed in the present invention, which should be covered within the protection scope of the present invention.

Claims

1. A structural damage identification method based on digital twin and feature migration, characterized in that: The identification method comprises: Constructing a source domain dataset and a target domain dataset; wherein each sample in the source domain dataset includes a first time-frequency image of the structure and its damage type, and each sample in the target domain dataset includes a second time-frequency image of the structure and its damage type, and the first time-frequency image and the second time-frequency image are both converted from the acceleration response signal of the structure; the source domain dataset is constructed using a digital twin model of the structure; Performing feature extraction on each of the first time-frequency images to obtain a first feature vector; combining all the first feature vectors to obtain a first feature matrix; performing feature extraction on each of the second time-frequency images to obtain a second feature vector; combining all the second feature vectors to obtain a second feature matrix; Concatenating the first feature matrix and the second feature matrix to obtain a joint feature matrix; Performing migration component analysis on the joint feature matrix to obtain a source domain feature vector and a target domain feature vector in a common feature space; Constructing a damage recognition model, and performing supervised training on the damage recognition model using the source domain feature vector and the damage type of the source domain data set; Using the target domain feature vector to perform unsupervised training on the supervised trained damage recognition model; The target domain feature vector and the damage type of the target domain data set are used to evaluate the performance of the damage recognition model after unsupervised training to obtain a target recognition model.

2. The structural damage identification method according to claim 1, characterized in that: Use the digital twin model of the structure to build the source domain dataset, including: Step A1: construct a parametric finite element model of the structure; Step A2: performing dynamic analysis on the parameterized finite element model and exporting an Inp file; Step A3: using a Python script to open the Inp file, and parsing the unit set and its material properties in the Inp file; Step A4: modify the material properties of the unit set according to different damage types of the structure, update the material properties of different damage locations, and obtain Inp files of different damage types; Step A5: Use the subprocess module to perform ABAQUS simulation calculations on the Inp files of different damage types to obtain ODB result files of different damage types; Step A6: Use Python script to open the ODB result files of different damage types, extract the acceleration response signals of the structural nodes from the ODB result files of different damage types, and obtain the acceleration response signals under different damage types; Step A7: converting the acceleration response signals under different damage types into images to obtain first time-frequency images under different damage types; Step A8: construct a source domain dataset according to the first time-frequency images under different damage types.

3. The structural damage identification method according to claim 1, characterized in that: The acceleration response signal of the structure is converted into a first time-frequency image or a second time-frequency image by using a spectral Markov transfer field, including: The frequency spectrum signal of the acceleration response signal is calculated using the following formula: Wherein, s(k) represents the kth component in the spectrum signal, x(i) represents the value of the i-th sampling point in the acceleration response signal, n represents the number of sampling points of the acceleration response signal, abs() represents the modulus value function, and j represents the imaginary unit; Performing normalization processing on each component in the spectrum signal to obtain a normalized component of the spectrum signal; The value range of the amplitude set is evenly divided to obtain Q sub-intervals; wherein the amplitude set includes the amplitudes of all normalized components of the spectrum signal; Each amplitude in the amplitude set is mapped to a subinterval; wherein the specific mapping formula is: Among them, A k represents the kth amplitude in the amplitude set, q j represents the jth subinterval, argmin j It means finding j that makes the function reach the minimum value; Generate a spectral Markov transition field matrix according to the probability of each amplitude being transferred from the current subinterval to other subintervals; The spectral Markov transfer field matrix is ​​converted into a two-dimensional image to obtain a first time-frequency image or a second time-frequency image.

4. The structural damage identification method according to claim 1, characterized in that: A pre-trained convolutional neural network model is used to perform feature extraction on the first time-frequency image and the second time-frequency image; wherein the pre-trained convolutional neural network model does not include a fully connected layer.

5. The structural damage identification method according to claim 1, characterized in that: Performing migration component analysis on the joint feature matrix includes: A joint Gaussian kernel matrix is ​​constructed according to the joint feature matrix. The specific formula is: Among them, k ij represents the element in the i-th row and j-th column of the joint Gaussian kernel matrix; x i 、x j represents the two eigenvectors in the joint feature matrix; σ represents the bandwidth of the Gaussian kernel function; The joint centralized kernel matrix is ​​constructed according to the joint Gaussian kernel matrix. The specific formula is: K c =HKH, Among them, K c represents the joint centralized kernel matrix; K represents the joint Gaussian kernel matrix; H represents the centralized matrix; I and 1 both represent the all-1 matrix; n s Represents the number of samples in the source domain dataset, n t Indicates the number of samples in the target domain dataset; A maximum mean difference matrix is ​​constructed, which is expressed as: Where M represents the maximum mean difference matrix; 1 a×b Represents a matrix of size a×b with all 1s; The objective function is constructed according to the joint centralized kernel matrix and the maximum mean difference matrix; wherein the expression of the objective function is: Where A represents the projection matrix, superscript T represents the transpose, λ represents the regularization parameter, and tr() represents the trace of the matrix; It means solving the optimal projection matrix so that the distribution difference between the source domain and the target domain in the common feature space is minimized; Solving the objective function to obtain a projection matrix; constructing a first centralized kernel matrix according to the first feature matrix and the joint feature matrix, and obtaining a source domain feature vector according to the first centralized kernel matrix and the projection matrix; A second centralized kernel matrix is ​​constructed according to the second feature matrix and the joint feature matrix, and a target domain feature vector is obtained according to the second centralized kernel matrix and the projection matrix.

6. The structural damage identification method according to claim 5, characterized in that: The objective function is solved by generalized eigenvalue decomposition, including: Convert the objective function into a generalized eigenvalue problem: (K c MK c +λK c )A=K c AL; Where Λ represents the eigenvalue diagonal matrix, which contains the first k smallest eigenvalues; The eigenvector corresponding to the minimum eigenvalue is selected as the column vector of the projection matrix to obtain the projection matrix.

7. The structural damage identification method according to any one of claims 1 to 6, characterized in that: The identification method further comprises: Acquiring a real-time acceleration response signal of the structure, and converting the real-time acceleration response signal into a real-time time-frequency image; Extracting features from the real-time time-frequency image to obtain a real-time feature vector; The target recognition model is used to perform online damage recognition on the real-time feature vector.

8. An electronic device comprising a memory, a processor, and a computer program / instruction stored in the memory, characterized in that: The processor executes the computer program / instructions to implement the structural damage identification method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the structural damage identification method according to any one of claims 1 to 7 is implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the structural damage identification method according to any one of claims 1 to 7 is implemented.

Citation Information

Patent Citations

  • Residential car elevator fault diagnosis method and device, electronic equipment and storage medium

    CN114676741A

  • Rolling bearing fault diagnosis method of digital twin drive deep transfer learning model

    CN116952583A

  • Waste household appliance recovery cost evaluation system based on online migration component analysis

    CN118096213A

  • Cable-stayed bridge damage identification method and device, storage medium and product

    CN118585919A

  • Bridge damage identification method considering uncertainty

    US11709979B1

Cited By

  • Cross-domain structure damage identification method based on digital twinning and time-space feature fusion

    CN119646751A