Lamb wave excitation and receiving optimization method for hyperbolic glass curtain wall nondestructive testing positioning

Abstract

The invention provides a Lamb wave excitation and receiving optimization method for hyperbolic glass curtain wall nondestructive testing and positioning. The method comprises the following steps: acquiring curtain wall parameter information of a hyperbolic glass curtain wall to be detected and positioned; the curtain wall parameter information at least comprises geometric parameters, thickness parameters, material parameters and environmental parameters; optimizing Lamb wave parameters corresponding to the glass curtain wall according to the curtain wall parameter information; the Lamb wave parameters at least comprise an excitation frequency, a waveform coefficient and an excitation angle; performing signal emission and receiving on the hyperbolic glass curtain wall according to the optimized Lamb wave parameter, and obtaining a damage reflection signal corresponding to the hyperbolic glass curtain wall; performing time-frequency domain decomposition on the damage reflection signal according to wavelet packet transformation to obtain a damage feature vector; and obtaining a neural network positioning model containing hyperbolic geometric constraints, inputting the damage feature vectors into the neural network positioning model, and outputting damage three-dimensional coordinates to complete hyperbolic glass curtain wall nondestructive testing positioning.

Description

Technical Field

[0001] The present application relates to the technical field of glass curtain walls, and in particular to a Lamb wave excitation and reception optimization method for nondestructive detection and positioning of hyperbolic glass curtain walls. Background Art

[0002] With the development of modern architectural design, hyperbolic glass curtain walls, as a unique building facade material, are widely used in large buildings due to their aesthetics and unique shape. The curved shape of the hyperbolic glass curtain wall increases the complexity of its structure and puts higher requirements on damage detection. Traditional glass curtain wall inspection methods, such as visual inspection, vibration detection, thermal imaging and ultrasonic detection, cannot adapt to the special geometric form of the hyperbolic glass curtain wall, resulting in insufficient detection accuracy and positioning accuracy.

[0003] Therefore, a method is urgently needed to solve at least one of the above problems. Summary of the invention

[0004] The present application provides a Lamb wave excitation and reception optimization method for non-destructive detection and positioning of hyperbolic glass curtain walls, aiming to solve the problem that traditional glass curtain wall detection methods, such as visual inspection, vibration detection, thermal imaging and ultrasonic detection, are unable to adapt to the special geometric form of hyperbolic glass curtain walls, resulting in insufficient detection accuracy and positioning accuracy.

[0005] In a first aspect, the present application provides a Lamb wave excitation and reception optimization method for non-destructive testing and positioning of a hyperbolic glass curtain wall, comprising:

[0006] Acquire curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located; the curtain wall parameter information at least includes geometric parameters, thickness parameters, material parameters and environmental parameters;

[0007] Optimizing Lamb wave parameters corresponding to the glass curtain wall according to the curtain wall parameter information; the Lamb wave parameters at least include an excitation frequency, a waveform coefficient, and an excitation angle;

[0008] Transmitting and receiving signals from the hyperbolic glass curtain wall according to the optimized Lamb wave parameters to obtain a damage reflection signal corresponding to the hyperbolic glass curtain wall;

[0009] Decomposing the damage reflection signal in time and frequency domain according to wavelet packet transform to obtain a damage feature vector;

[0010] A neural network positioning model including hyperbolic surface geometric constraints is obtained, the damage feature vector is input into the neural network positioning model, the three-dimensional coordinates of the damage are output, and the non-destructive detection positioning of the hyperbolic glass curtain wall is completed.

[0011] In some embodiments, obtaining curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located includes: obtaining a three-dimensional laser scanning image and a visual image corresponding to the hyperbolic glass curtain wall; transforming feature points corresponding to the three-dimensional laser scanning image and the visual image to the same coordinate system based on a point cloud registration algorithm; and reconstructing the surface of the hyperbolic glass curtain wall according to the registered three-dimensional laser scanning image and the visual image to obtain the curtain wall parameter information.

[0012] In some embodiments, the expression of the excitation frequency includes:

[0013] ;in, represents the excitation frequency, represents the temperature compensation coefficient, is the thickness parameter, is Young's modulus, is the Poisson's ratio, is the material density corresponding to the hyperbolic glass, Used to express the effective stiffness under plane stress state, It is used to represent the influence of Poisson's ratio on the three-dimensional strain of the hyperbolic glass curtain wall.

[0014] In some embodiments, the expression of the excitation angle includes:

[0015] ;in, represents the excitation angle, and They respectively represent the local slopes of the curved surface corresponding to the hyperbolic glass curtain wall, represents the environmental compensation angle corresponding to the hyperbolic glass curtain wall, which is used to correct the wave velocity anisotropy caused by environmental factors. It means that the excitation angle is adjusted according to the local curvature of the hyperbolic surface corresponding to the hyperbolic glass curtain wall to ensure that the lamb wavefront always maintains an optimal incident angle with the surface normal of the hyperbolic glass curtain wall to avoid lamb wave signal scattering.

[0016] Exemplarily, if the ambient temperature corresponding to the hyperbolic glass curtain wall increases, the ambient compensation angle needs to be subtracted from the first compensation angle, and the first compensation angle is within a first compensation angle range of 0.05 to 0.15 rad; if the ambient humidity corresponding to the hyperbolic glass curtain wall increases, the ambient compensation angle needs to be increased by a second compensation angle, and the second compensation angle is within a second compensation angle range of 0.02 to 0.08 rad.

[0017] In some embodiments, the inputting the damage feature vector into the neural network positioning model and outputting the three-dimensional coordinates of the damage includes: inputting the damage feature vector into the neural network positioning model, the neural network positioning model generating geometric damage features through a three-dimensional convolution kernel and generating signal damage features through time-frequency convolution kernel processing; performing feature fusion based on the geometric damage features and the signal damage features, and outputting the three-dimensional coordinates of the damage.

[0018] In some embodiments, before transmitting and receiving signals from the hyperbolic glass curtain wall according to the optimized Lamb wave parameters, it also includes: obtaining the minimum phase velocity, the maximum excitation frequency and the maximum excitation angle corresponding to the Lamb wave; and adjusting the spatial density of the array piezoelectric sensor for receiving the injury reflection signal according to the minimum phase velocity, the maximum excitation frequency and the maximum excitation angle.

[0019] Exemplarily, the expression corresponding to the spatial density includes:

[0020] ;in, is the spatial density, is the minimum phase velocity, is the highest excitation frequency, is the maximum excitation angle.

[0021] In some embodiments, the time-frequency decomposition of the damaged reflection signal according to the wavelet packet transform to obtain the damage feature vector includes: performing time-frequency decomposition of the damaged reflection signal according to the wavelet packet transform to obtain the signal energy attenuation coefficient, the main frequency offset and the wavelet packet node energy entropy; generating the damage feature vector according to the signal energy attenuation coefficient, the main frequency offset and the wavelet packet node energy entropy.

[0022] In some embodiments, after the output of the three-dimensional coordinates of the damage, it also includes: generating a visual detection map according to the three-dimensional coordinates of the damage and a curtain wall model corresponding to the hyperbolic glass curtain wall; the visual detection map includes a hyperbolic surface development diagram of the hyperbolic glass curtain wall, the hyperbolic surface development diagram includes a damage projection corresponding to the three-dimensional coordinates of the damage, a three-dimensional rendering model corresponding to the hyperbolic glass curtain wall, the three-dimensional rendering model includes a defect mark corresponding to the three-dimensional coordinates of the damage, and a stacked display diagram corresponding to the hyperbolic glass curtain wall, and the stacked display diagram includes defect size and depth parameters.

[0023] In a second aspect, the present application provides a Lamb wave excitation and reception optimization method and device for non-destructive testing and positioning of a hyperbolic glass curtain wall, comprising:

[0024] A parameter acquisition unit, used to acquire curtain wall parameter information of the hyperbolic glass curtain wall to be detected and positioned; the curtain wall parameter information at least includes geometric parameters, thickness parameters, material parameters and environmental parameters;

[0025] A parameter optimization unit, configured to optimize Lamb wave parameters corresponding to the glass curtain wall according to the curtain wall parameter information; the Lamb wave parameters at least include an excitation frequency, a waveform coefficient, and an excitation angle;

[0026] A signal acquisition unit, configured to transmit and receive signals from the hyperbolic glass curtain wall according to the optimized Lamb wave parameters, and acquire a damage reflection signal corresponding to the hyperbolic glass curtain wall;

[0027] A signal decomposition unit, used for performing time-frequency domain decomposition on the damage reflection signal according to wavelet packet transform to obtain a damage feature vector;

[0028] The coordinate output unit is used to obtain a neural network positioning model containing hyperbolic surface geometric constraints, input the damage feature vector into the neural network positioning model, output the three-dimensional coordinates of the damage, and complete the non-destructive detection positioning of the hyperbolic glass curtain wall.

[0029] In a third aspect, the present application provides a computer device comprising a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and implement a method provided in any embodiment of the present application when executing the computer program.

[0030] In a fourth aspect, the present application provides a computer-readable storage medium, which stores a computer program. When the computer-readable instructions are executed by the processor, one or more processors execute a method as provided in any embodiment of the present application.

[0031] The embodiment of the present application provides a Lamb wave excitation and reception optimization method for non-destructive detection and positioning of a hyperbolic glass curtain wall. The specific technical process is as follows:

[0032] Get the curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located. Geometric parameters: including the curvature radius and shape parameters of the hyperbolic surface. Thickness parameters: the thickness of the glass curtain wall. Material parameters: the material type of the glass and its physical properties (such as elastic modulus, Poisson's ratio, etc.). Environmental parameters: environmental conditions such as temperature and humidity.

[0033] The Lamb wave parameters corresponding to the glass curtain wall are optimized according to the curtain wall parameter information. Excitation frequency: Select a suitable excitation frequency to ensure the propagation efficiency and resolution of the Lamb wave in the glass. Waveform coefficient: Determine the shape and pulse width of the waveform to adapt to different types of damage. Excitation angle: Adjust the excitation angle so that the Lamb wave can effectively cover the entire hyperbolic area.

[0034] According to the optimized Lamb wave parameters, the hyperbolic glass curtain wall is used to transmit and receive signals, and the damage reflection signal corresponding to the hyperbolic glass curtain wall is obtained. Using the optimized Lamb wave parameters, the hyperbolic glass curtain wall is used to transmit signals through the sensor array. The damage reflection signal captured by the sensor array is received, and the time and intensity of the signal are recorded.

[0035] The damage reflection signal is decomposed in the time-frequency domain according to the wavelet packet transform to obtain the damage feature vector. The damage reflection signal is converted from the time domain to the time-frequency domain using the wavelet packet transform to extract the multi-scale features of the signal. By analyzing the time-frequency domain signal, the feature vector containing the damage location and degree is extracted. A neural network positioning model containing hyperbolic geometric constraints is obtained, the damage feature vector is input into the neural network positioning model, the three-dimensional coordinates of the damage are output, and the non-destructive detection positioning of the hyperbolic glass curtain wall is completed.

[0036] By building a neural network model that includes hyperbolic geometric constraints, the model can handle complex geometric shapes. The extracted damage feature vector is input into the neural network model, and the three-dimensional coordinates of the damage are calculated through the trained model. The results are output to complete the non-destructive detection and positioning of the hyperbolic glass curtain wall.

[0037] The specific implementation of the above steps can be: Data collection and preprocessing: Use a high-precision sensor array to scan the hyperbolic glass curtain wall to collect initial data. Preprocess the collected data to remove noise and interference signals to ensure data quality.

[0038] Parameter optimization: Based on the specific parameters of the curtain wall (geometric parameters, thickness parameters, material parameters and environmental parameters), the excitation frequency, waveform coefficient and excitation angle of the Lamb wave are optimized by combining numerical simulation and experimental verification. Through multiple iterations and feedback adjustments, the optimal combination of Lamb wave parameters is found.

[0039] Signal transmission and reception: Use the optimized Lamb wave parameters to transmit signals to the hyperbolic glass curtain wall through the sensor array. Receive the damage reflection signal captured by the sensor array and record the time and intensity of the signal.

[0040] Signal processing and feature extraction: Use wavelet packet transform to decompose the damage reflection signal in the time and frequency domain, and extract the feature vector containing the damage location and degree. Through the analysis of the time and frequency domain signals, the feature vector is further refined to improve the detection accuracy.

[0041] Construction and application of neural network positioning model: Construct a neural network model with hyperbolic geometric constraints, which can handle complex geometric forms. Use a large amount of labeled data to train the neural network model to ensure that it has good generalization ability. Input the extracted damage feature vector into the trained neural network model to calculate the three-dimensional coordinates of the damage. Output the results to complete the non-destructive detection and positioning of the hyperbolic glass curtain wall.

[0042] The provided method has at least the following beneficial effects:

[0043] Improve detection accuracy: By optimizing Lamb wave parameters, the signal propagation efficiency and resolution in the hyperbolic glass curtain wall are improved, thereby improving detection accuracy.

[0044] Enhanced positioning accuracy: Combining wavelet packet transform and neural network model, it can accurately extract damage features and perform three-dimensional positioning, significantly improving positioning accuracy.

[0045] Adaptability to complex geometries: This method can adapt to the complex geometries of hyperbolic glass curtain walls, solving problems that traditional detection methods cannot effectively address.

[0046] Real-time and automation: Through the sensor array and automated signal processing system, a fast and efficient detection process is achieved, which is suitable for regular maintenance and inspection of large-scale buildings.

[0047] Reduced manual intervention: Automated detection and positioning processes reduce the need for manual intervention, reducing detection costs and time.

[0048] Scalability: This method is not only applicable to hyperbolic glass curtain walls, but can also be extended to the detection of structures with other complex geometric forms, and has wide applicability.

[0049] In summary, the present application provides a Lamb wave excitation and reception optimization method for non-destructive detection and positioning of hyperbolic glass curtain walls, which significantly improves the detection accuracy and positioning accuracy by optimizing Lamb wave parameters, signal processing and neural network positioning model, and provides effective technical support for the maintenance and inspection of hyperbolic glass curtain walls.

[0050] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0052] Figure 1 This is a schematic flow chart of the steps of a Lamb wave excitation and reception optimization method for non-destructive testing and positioning of a hyperbolic glass curtain wall provided in an embodiment of the present application;

[0053] Figure 2 It is a schematic block diagram of the structure of a Lamb wave excitation and reception optimization method device for non-destructive testing and positioning of a hyperbolic glass curtain wall provided in one embodiment of the present application;

[0054] Figure 3 It is a schematic block diagram of the structure of a computer device provided in one embodiment of the present application.

[0055] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0057] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0058] It should be understood that, in order to facilitate the clear description of the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, the words "first", "second", etc. are used to distinguish the same items or similar items with substantially the same functions and effects. Those skilled in the art can understand that the words "first", "second", etc. do not limit the quantity and execution order, and the words "first", "second", etc. do not necessarily limit the difference.

[0059] It should be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0060] It should also be understood that the term “and / or” used in the specification and appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0061] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0062] With the development of modern architectural design, hyperbolic glass curtain walls, as a unique building facade material, are widely used in large buildings due to their aesthetics and unique shape. The curved shape of the hyperbolic glass curtain wall increases the complexity of its structure and puts higher requirements on damage detection. Traditional glass curtain wall inspection methods, such as visual inspection, vibration detection, thermal imaging and ultrasonic detection, cannot adapt to the special geometric form of the hyperbolic glass curtain wall, resulting in insufficient detection accuracy and positioning accuracy.

[0063] Therefore, a method is urgently needed to solve at least one of the above problems.

[0064] To solve the above problems, please refer to Figure 1 ,like Figure 1 As shown, the provided Lamb wave excitation and reception optimization method for non-destructive testing and positioning of hyperbolic glass curtain walls includes steps S101 to S105. The Lamb wave excitation and reception optimization method for non-destructive testing and positioning of hyperbolic glass curtain walls is executed by a computer device, which may be a single server or a server cluster, or may be a handheld terminal, a laptop computer, a wearable device, or a robot, etc.

[0065] like Figure 1 As shown, the details of steps S101-S105 are as follows:

[0066] Step S101. Acquire curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located; the curtain wall parameter information at least includes geometric parameters, thickness parameters, material parameters and environmental parameters.

[0067] Specifically, first of all, it is necessary to collect all relevant parameters about the hyperbolic glass curtain wall. These parameters can be divided into four categories: geometric parameters, thickness parameters, material parameters and environmental parameters. Geometric parameters include but are not limited to the overall size (length and width) of the curtain wall, the radius of curvature, the specific shape of the curved surface and the distribution of curvature, etc., which describe its shape. These parameters can be determined by on-site measurement or by consulting architectural design drawings. The thickness parameter refers to the actual thickness value of the glass layer, which is directly related to the propagation speed and attenuation of the sound wave. It can be accurately measured using an ultrasonic thickness gauge or other non-destructive testing tools. Material parameters involve the specific type of material used to manufacture the curtain wall (such as tempered glass, laminated glass, etc.) and its physical properties (density, elastic modulus, Poisson's ratio, etc.). These data can be obtained from the product specifications provided by the supplier. Environmental parameters cover factors such as temperature, humidity, and wind speed at the installation location, because these conditions affect the behavior of sound waves. Equipment such as thermometers and hygrometers and anemometers can be used to record the current environmental conditions.

[0068] For example, the acquisition of geometric parameters: a high-precision 3D scanner is used to fully scan the hyperbolic glass curtain wall to obtain a detailed geometric model. The scanned data is converted into a digital model through CAD software to extract key geometric parameters such as curvature radius, surface equation, etc. For complex curved surface structures, specialized geometric analysis software can be used for further processing and verification.

[0069] Acquisition of thickness parameters: Use an ultrasonic thickness gauge to measure at multiple points, ensuring that different areas of the entire curtain wall are covered. Record the data at each measurement point, and calculate the average and standard deviation to evaluate the consistency and uniformity of the thickness. For difficult-to-reach locations, a portable ultrasonic thickness gauge or robot-assisted measurement can be used.

[0070] Obtaining material parameters: Check the product specifications provided by the supplier to obtain detailed physical properties of the glass. If the specifications are incomplete, the missing data can be supplemented by laboratory tests (such as tensile tests, compression tests, etc.). For composite materials (such as laminated glass), the interface characteristics between the layers of material also need to be considered.

[0071] Acquisition of environmental parameters: Use thermometers, hygrometers and anemometers to record environmental parameters in different time periods and weather conditions. Upload the data to a computer system for statistical analysis to determine typical working environmental conditions. Consider the impact of seasonal changes and extreme weather and develop corresponding testing strategies.

[0072] Accurately understanding all the above parameters will help to more accurately adjust the relevant settings of the Lamb wave in the subsequent steps, thereby improving the detection accuracy. In addition, detailed parameter information can also be used to establish a more accurate numerical model and further optimize the detection scheme.

[0073] Step S102: Optimizing Lamb wave parameters corresponding to the glass curtain wall according to the curtain wall parameter information; the Lamb wave parameters at least include an excitation frequency, a waveform coefficient, and an excitation angle.

[0074] Specifically, based on the data obtained in the previous step, this step aims to determine the Lamb wave characteristics that are most suitable for use under the current specific conditions. Mainly including: The selection of the excitation frequency needs to take into account factors such as the target defect size and depth to ensure that potential problem areas can be effectively detected. The waveform coefficient determines the shape of the emitted sound wave, affecting the signal penetration ability and resolution. The excitation angle refers to the angle of the sound source relative to the normal direction of the surface of the object being measured, which is particularly important for detection on complex curved surfaces.

[0075] For example, the Rayleigh-Lamb equation is used in combination with finite element analysis software (such as COMSOL Multiphysics) to predict the acoustic field distribution under different parameter combinations. Through numerical simulation, the optimal excitation frequency range is determined so that the sound waves can effectively propagate and reflect back to the sensor. The sound wave propagation characteristics under different waveform coefficients are simulated to select the waveform that can provide the highest resolution and signal-to-noise ratio.

[0076] A simplified hyperbolic glass curtain wall model was built in a laboratory environment for preliminary experimental verification. The reflection signals under different conditions were recorded by changing the excitation frequency, waveform coefficient and excitation angle. The experimental data were analyzed to determine the optimal parameter combination, and repeated experiments were performed multiple times to verify the reliability of the results.

[0077] Due to the characteristics of the hyperbolic structure, it is necessary to introduce an appropriate coordinate transformation algorithm to convert the sound wave propagation model in the plane coordinate system into a model in the curved coordinate system. Mathematical tools (such as tensor analysis) are used for coordinate transformation to ensure that the propagation path and reflection characteristics of the sound wave on the curved surface are accurately described. The effectiveness of the coordinate transformation algorithm is verified through numerical simulation, and necessary adjustments are made.

[0078] The customized Lamb wave configuration in this way can significantly enhance the sensitivity to specific types of damage, while reducing background noise interference and improving the overall signal-to-noise ratio. The optimized Lamb wave parameters can better adapt to the special geometric form of the hyperbolic glass curtain wall and improve the accuracy and reliability of detection.

[0079] Step S103: transmitting and receiving signals to the hyperbolic glass curtain wall according to the optimized Lamb wave parameters, and obtaining a damage reflection signal corresponding to the hyperbolic glass curtain wall.

[0080] Specifically, in this stage, the previously determined optimal Lamb wave parameters will be used to send signals to the hyperbolic glass curtain wall, and the sensor array will be used to capture the returned echo information. In order to cover the entire detection area, a multi-point or multi-angle scanning strategy may be required.

[0081] For example, high-sensitivity piezoelectric transducers can be installed as excitation sources and receivers to ensure the quality of signal transmission and reception. Piezoelectric transducers should be evenly distributed at key locations on the curtain wall to achieve full coverage. For locations that are difficult to reach, robots can be used to assist in installation tasks.

[0082] Set reasonable sampling rate and gain magnification to ensure sufficient dynamic range and signal strength. Use multi-point or multi-angle scanning strategies to ensure that the signal can cover the entire curtain wall area. For each scanning point, record the time domain waveform as a preliminary analysis basis. Through multiple scans and signal superposition, the signal-to-noise ratio and reliability of the signal are improved.

[0083] Use a high-speed data acquisition card to record the reflected signal in real time and store it in the computer system. During the data acquisition process, the stability and consistency of the system should be maintained to avoid external interference. Pre-process the collected data to remove noise and outliers to ensure the quality of the data.

[0084] This method can not only fully reflect the internal state of the inspected object, but also obtain a clearer and more informative reflection signal due to the optimized acoustic parameter settings, which is conducive to the subsequent processing stage. Through multi-point or multi-angle scanning strategies, comprehensive inspection of hyperbolic glass curtain walls can be achieved, improving the coverage and accuracy of inspection.

[0085] Step S104: Decompose the damaged reflection signal in the time-frequency domain according to wavelet packet transform to obtain a damage feature vector.

[0086] Specifically, once the original damage reflection signal is obtained, the next task is to extract useful features from it for identifying and locating defects. Wavelet packet transform technology is recommended for signal processing because it can provide high frequency resolution while maintaining good time resolution.

[0087] Select appropriate wavelet basis functions (such as Daubechies series), and then perform binary tree decomposition according to the predetermined level. Select energy-concentrated segments as candidate features in each level of decomposition results. Combine statistical methods (such as principal component analysis PCA) to further streamline the feature set and remove redundant information. Finally, a feature vector set containing multiple dimensions (such as amplitude, phase, instantaneous frequency, etc.) is formed.

[0088] The original reflected signal is denoised, and filters are used to remove high-frequency noise and low-frequency drift. Through normalization, the signal is compared and analyzed at a uniform scale. The signal is segmented to ensure that each segment has similar length and characteristics.

[0089] Wavelet packet transform is applied to decompose each signal in the time-frequency domain to extract the energy distribution of each frequency band. The statistical features of each frequency band, such as mean, variance, peak value, etc., are calculated. The most representative features are selected through feature selection algorithms (such as mutual information, recursive feature elimination, etc.). The selected features are combined into a high-dimensional feature vector for subsequent damage identification and location.

[0090] The extracted feature vectors are standardized to ensure comparability between different features. A feature vector database is constructed, which contains feature vectors of different types of damage for training and validating the neural network model. The feature vectors are visualized to help understand the distribution of different damage types in the feature space.

[0091] With the powerful localized analysis capability of wavelet packet transform, the part that truly reflects the damage characteristics can be separated from the complex reflection signal, laying a solid foundation for the next step of applying machine learning algorithms to achieve automatic diagnosis. In addition, this process also helps to reduce the computational complexity and speed up the response of the entire system. By constructing a high-dimensional feature vector, the characteristics of the damage can be described more accurately, improving the accuracy of damage identification and location.

[0092] Step S105. Obtain a neural network positioning model containing hyperbolic surface geometric constraints, input the damage feature vector into the neural network positioning model, output the three-dimensional coordinates of the damage, and complete the non-destructive detection positioning of the hyperbolic glass curtain wall.

[0093] Specifically, in this stage, the damage feature vector is input into the trained neural network localization model, which can output the three-dimensional coordinates of the damage according to the geometric constraints and signal characteristics.

[0094] Construction of the neural network positioning model: Design a neural network architecture that includes a three-dimensional convolution kernel and a time-frequency convolution kernel. The three-dimensional convolution kernel is used to process geometric features and generate geometric damage features. The time-frequency convolution kernel is used to process signal features and generate signal damage features. The geometric damage features and signal damage features are fused through the feature fusion layer to generate a comprehensive feature vector. Finally, the three-dimensional coordinates of the damage are output through the fully connected layer and the regression layer.

[0095] Model training: Collect a large amount of hyperbolic glass curtain wall damage sample data, including the three-dimensional coordinates of the damage and the corresponding damage feature vector. Use these data to train the neural network and optimize the model parameters through the back propagation algorithm. Use cross-validation and regularization techniques to prevent overfitting and improve the generalization ability of the model.

[0096] Model application: The extracted damage feature vector is input into the trained neural network positioning model. The model generates geometric damage features through three-dimensional convolution kernels and generates signal damage features through time-frequency convolution kernel processing. The feature fusion layer fuses the geometric damage features and signal damage features to generate a comprehensive feature vector. Finally, the three-dimensional coordinates of the damage are output through the fully connected layer and regression layer.

[0097] Visual inspection map: Generate a visual inspection map based on the 3D coordinates of the damage and the curtain wall model. The visual inspection map includes a hyperbolic expansion map showing the damage projection and a 3D rendering model. The 3D rendering model marks the defect location and provides a stacked display map showing the defect size and depth parameters. The stacked display map can intuitively show the specific situation of the damage.

[0098] The combination of 3D convolution kernel and time-frequency convolution kernel can fully capture the geometric and signal characteristics of the damage. Feature fusion technology improves the accuracy of damage location. The neural network model can automatically learn and identify complex damage patterns, improving the intelligence level of detection. The visual detection map provides intuitive damage distribution and detailed information, which is convenient for maintenance personnel to further process and repair. The 3D rendering model and stacked display help users better understand and analyze the detection results. The readability and practicality of the detection report are improved, which is convenient for maintenance personnel to further process and repair.

[0099] The embodiment of the present application provides a Lamb wave excitation and reception optimization method for non-destructive detection and positioning of a hyperbolic glass curtain wall. The specific technical process is as follows:

[0100] Get the curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located. Geometric parameters: including the curvature radius and shape parameters of the hyperbolic surface. Thickness parameters: the thickness of the glass curtain wall. Material parameters: the material type of the glass and its physical properties (such as elastic modulus, Poisson's ratio, etc.). Environmental parameters: environmental conditions such as temperature and humidity.

[0101] The Lamb wave parameters corresponding to the glass curtain wall are optimized according to the curtain wall parameter information. Excitation frequency: Select a suitable excitation frequency to ensure the propagation efficiency and resolution of the Lamb wave in the glass. Waveform coefficient: Determine the shape and pulse width of the waveform to adapt to different types of damage. Excitation angle: Adjust the excitation angle so that the Lamb wave can effectively cover the entire hyperbolic area.

[0102] According to the optimized Lamb wave parameters, the hyperbolic glass curtain wall is used to transmit and receive signals, and the damage reflection signal corresponding to the hyperbolic glass curtain wall is obtained. Using the optimized Lamb wave parameters, the hyperbolic glass curtain wall is used to transmit signals through the sensor array. The damage reflection signal captured by the sensor array is received, and the time and intensity of the signal are recorded.

[0103] The damage reflection signal is decomposed in the time-frequency domain according to the wavelet packet transform to obtain the damage feature vector. The damage reflection signal is converted from the time domain to the time-frequency domain using the wavelet packet transform to extract the multi-scale features of the signal. By analyzing the time-frequency domain signal, the feature vector containing the damage location and degree is extracted. A neural network positioning model containing hyperbolic geometric constraints is obtained, the damage feature vector is input into the neural network positioning model, the three-dimensional coordinates of the damage are output, and the non-destructive detection positioning of the hyperbolic glass curtain wall is completed.

[0104] By building a neural network model that includes hyperbolic geometric constraints, the model can handle complex geometric shapes. The extracted damage feature vector is input into the neural network model, and the three-dimensional coordinates of the damage are calculated through the trained model. The results are output to complete the non-destructive detection and positioning of the hyperbolic glass curtain wall.

[0105] The specific implementation of the above steps can be: Data collection and preprocessing: Use a high-precision sensor array to scan the hyperbolic glass curtain wall to collect initial data. Preprocess the collected data to remove noise and interference signals to ensure data quality.

[0106] Parameter optimization: Based on the specific parameters of the curtain wall (geometric parameters, thickness parameters, material parameters and environmental parameters), the excitation frequency, waveform coefficient and excitation angle of the Lamb wave are optimized by combining numerical simulation and experimental verification. Through multiple iterations and feedback adjustments, the optimal combination of Lamb wave parameters is found.

[0107] Signal transmission and reception: Use the optimized Lamb wave parameters to transmit signals to the hyperbolic glass curtain wall through the sensor array. Receive the damage reflection signal captured by the sensor array and record the time and intensity of the signal.

[0108] Signal processing and feature extraction: Use wavelet packet transform to decompose the damage reflection signal in the time and frequency domain, and extract the feature vector containing the damage location and degree. Through the analysis of the time and frequency domain signals, the feature vector is further refined to improve the detection accuracy.

[0109] Construction and application of neural network positioning model: Construct a neural network model with hyperbolic geometric constraints, which can handle complex geometric forms. Use a large amount of labeled data to train the neural network model to ensure that it has good generalization ability. Input the extracted damage feature vector into the trained neural network model to calculate the three-dimensional coordinates of the damage. Output the results to complete the non-destructive detection and positioning of the hyperbolic glass curtain wall.

[0110] The provided method has at least the following beneficial effects:

[0111] Improve detection accuracy: By optimizing Lamb wave parameters, the signal propagation efficiency and resolution in the hyperbolic glass curtain wall are improved, thereby improving detection accuracy.

[0112] Enhanced positioning accuracy: Combining wavelet packet transform and neural network model, it can accurately extract damage features and perform three-dimensional positioning, significantly improving positioning accuracy.

[0113] Adaptability to complex geometries: This method can adapt to the complex geometries of hyperbolic glass curtain walls, solving problems that traditional detection methods cannot effectively address.

[0114] Real-time and automation: Through the sensor array and automated signal processing system, a fast and efficient detection process is achieved, which is suitable for regular maintenance and inspection of large-scale buildings.

[0115] Reduced manual intervention: Automated detection and positioning processes reduce the need for manual intervention, reducing detection costs and time.

[0116] Scalability: This method is not only applicable to hyperbolic glass curtain walls, but can also be extended to the detection of structures with other complex geometric forms, and has wide applicability.

[0117] In summary, the present application provides a Lamb wave excitation and reception optimization method for non-destructive detection and positioning of hyperbolic glass curtain walls, which significantly improves the detection accuracy and positioning accuracy by optimizing Lamb wave parameters, signal processing and neural network positioning model, and provides effective technical support for the maintenance and inspection of hyperbolic glass curtain walls.

[0118] In some embodiments, obtaining curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located includes: obtaining a three-dimensional laser scanning image and a visual image corresponding to the hyperbolic glass curtain wall; transforming feature points corresponding to the three-dimensional laser scanning image and the visual image to the same coordinate system based on a point cloud registration algorithm; and reconstructing the surface of the hyperbolic glass curtain wall according to the registered three-dimensional laser scanning image and the visual image to obtain the curtain wall parameter information.

[0119] Use a high-precision 3D laser scanner to fully scan the hyperbolic glass curtain wall to obtain detailed 3D point cloud data. At the same time, use a high-resolution camera to capture visual images from multiple angles to ensure that the entire curtain wall area is covered. Use a point cloud registration algorithm (such as the ICP algorithm) to transform the feature points corresponding to the 3D laser scan image and the visual image into the same coordinate system. By matching feature points (such as corner points, edges, etc.), the transformation matrix is ​​calculated to align the two sets of data in space.

[0120] Surface reconstruction is performed using the registered 3D laser scanning images and visual images. The geometric model of the hyperbolic glass curtain wall is generated using surface reconstruction algorithms (such as Delaunay triangulation, NURBS surface fitting, etc.). Key geometric parameters such as curvature radius and surface equation are extracted.

[0121] By combining 3D laser scanning and visual images, more comprehensive and accurate curtain wall parameter information can be obtained. The point cloud registration algorithm ensures accurate alignment between different data sources, improving data consistency and reliability. Surface reconstruction technology can generate high-precision geometric models, providing reliable basic data for subsequent steps.

[0122] In some embodiments, the expression of the excitation frequency includes:

[0123] ;in, represents the excitation frequency, represents the temperature compensation coefficient, is the thickness parameter, is Young's modulus, is the Poisson's ratio, is the material density corresponding to the hyperbolic glass, Used to express the effective stiffness under plane stress state, It is used to represent the influence of Poisson's ratio on the three-dimensional strain of the hyperbolic glass curtain wall.

[0124] Obtained through experiments or by consulting the material manual , and The ambient temperature is measured in real time by the temperature sensor and the calculation . Measure and record the thickness of the curtain wall .

[0125] Substitute the above parameters into the formula to calculate the optimal excitation frequency. Adjust the sound wave transmitting device to set the calculated excitation frequency.

[0126] By considering the temperature compensation coefficient, the detection effect can be kept stable under different environmental conditions. The physical parameters in the formula make the excitation frequency closely combined with the material properties, improving the accuracy of detection. The optimized excitation frequency can better adapt to the special geometric form of the hyperbolic glass curtain wall and improve the sensitivity and resolution of damage detection.

[0127] In some embodiments, the expression of the excitation angle includes:

[0128] ;in, represents the excitation angle, and They respectively represent the local slopes of the curved surface corresponding to the hyperbolic glass curtain wall, represents the environmental compensation angle corresponding to the hyperbolic glass curtain wall, which is used to correct the wave velocity anisotropy caused by environmental factors. It means that the excitation angle is adjusted according to the local curvature of the hyperbolic surface corresponding to the hyperbolic glass curtain wall to ensure that the lamb wavefront always maintains an optimal incident angle with the surface normal of the hyperbolic glass curtain wall to avoid lamb wave signal scattering.

[0129] Adjust the environmental compensation angle according to the changes in ambient temperature and humidity. If the ambient temperature rises, subtract the first compensation angle (0.05 to 0.15 rad); if the ambient humidity rises, increase the second compensation angle (0.02 to 0.08 rad). Substitute the formula to calculate the final excitation angle. Adjust the angle of the sound wave emission device to ensure that the Lamb wavefront always maintains the best incident angle with the surface normal of the curtain wall.

[0130] By considering the local slope of the surface and environmental factors, the excitation angle can be adjusted more accurately to avoid signal scattering. The introduction of the environmental compensation angle enables the system to maintain a stable detection effect under different environmental conditions. The optimized excitation angle can improve the propagation efficiency and detection accuracy of Lamb waves.

[0131] Exemplarily, if the ambient temperature corresponding to the hyperbolic glass curtain wall increases, the ambient compensation angle needs to be subtracted from the first compensation angle, and the first compensation angle is within a first compensation angle range of 0.05 to 0.15 rad; if the ambient humidity corresponding to the hyperbolic glass curtain wall increases, the ambient compensation angle needs to be increased by a second compensation angle, and the second compensation angle is within a second compensation angle range of 0.02 to 0.08 rad.

[0132] In some embodiments, the inputting the damage feature vector into the neural network positioning model and outputting the three-dimensional coordinates of the damage includes: inputting the damage feature vector into the neural network positioning model, the neural network positioning model generating geometric damage features through a three-dimensional convolution kernel and generating signal damage features through time-frequency convolution kernel processing; performing feature fusion based on the geometric damage features and the signal damage features, and outputting the three-dimensional coordinates of the damage.

[0133] The extracted damage feature vector is input into the neural network positioning model. The input feature vector is processed using a three-dimensional convolution kernel to generate a geometric damage feature. The input feature vector is processed using a time-frequency convolution kernel to generate a signal damage feature. The geometric damage feature and the signal damage feature are fused to output the three-dimensional coordinates of the damage.

[0134] The combination of three-dimensional convolution kernel and time-frequency convolution kernel can fully capture the geometric and signal characteristics of damage. Feature fusion technology improves the accuracy of damage location. The neural network model can automatically learn and identify complex damage patterns, improving the intelligent level of detection.

[0135] In some embodiments, before transmitting and receiving signals from the hyperbolic glass curtain wall according to the optimized Lamb wave parameters, the method further includes: obtaining the minimum phase velocity, the maximum excitation frequency and the maximum excitation angle corresponding to the Lamb wave; adjusting the spatial density of the array piezoelectric sensor for receiving the injury reflection signal according to the minimum phase velocity, the maximum excitation frequency and the maximum excitation angle. Determine the minimum phase velocity, the maximum excitation frequency and the maximum excitation angle of the Lamb wave by experiments or theoretical calculations.

[0136] Exemplarily, the expression corresponding to the spatial density includes:

[0137] ;in, is the spatial density, is the minimum phase velocity, is the highest excitation frequency, is the maximum excitation angle.

[0138] Calculate the spatial density according to the above formula. By reasonably adjusting the spatial density of the piezoelectric sensor, the quality and resolution of signal reception can be improved. Adjust the spatial layout of the array piezoelectric sensor according to the calculation results to ensure sufficient coverage and resolution. Ensure that the reflected signal can be effectively captured at different excitation frequencies and angles. Improve the overall performance and reliability of the detection system.

[0139] In some embodiments, the time-frequency decomposition of the damaged reflection signal according to the wavelet packet transform to obtain the damage feature vector includes: performing time-frequency decomposition of the damaged reflection signal according to the wavelet packet transform to obtain the signal energy attenuation coefficient, the main frequency offset and the wavelet packet node energy entropy; generating the damage feature vector according to the signal energy attenuation coefficient, the main frequency offset and the wavelet packet node energy entropy.

[0140] Perform wavelet packet transform on the damaged reflection signal and decompose it into multiple frequency bands. Calculate the signal energy attenuation coefficient, main frequency offset and wavelet packet node energy entropy of each frequency band. Combine the above features into a high-dimensional damage feature vector.

[0141] Through wavelet packet transform, rich feature information can be extracted from the time-frequency domain. Signal energy attenuation coefficient, main frequency offset and wavelet packet node energy entropy can fully describe the characteristics of damage. The generated damage feature vector helps to improve the accuracy of damage identification and location.

[0142] In some embodiments, after the output of the three-dimensional coordinates of the damage, it also includes: generating a visual detection map according to the three-dimensional coordinates of the damage and a curtain wall model corresponding to the hyperbolic glass curtain wall; the visual detection map includes a hyperbolic surface development diagram of the hyperbolic glass curtain wall, the hyperbolic surface development diagram includes a damage projection corresponding to the three-dimensional coordinates of the damage, a three-dimensional rendering model corresponding to the hyperbolic glass curtain wall, the three-dimensional rendering model includes a defect mark corresponding to the three-dimensional coordinates of the damage, and a stacked display diagram corresponding to the hyperbolic glass curtain wall, and the stacked display diagram includes defect size and depth parameters.

[0143] Generate a visual inspection map based on the 3D coordinates of the damage and the curtain wall model. Includes the damage projection corresponding to the 3D coordinates of the damage. Display the 3D rendering model of the hyperbolic glass curtain wall and mark the defect location. Includes defect size and depth parameters. The overlay display can intuitively show the specific situation of the damage. The visual inspection map provides intuitive damage distribution and detailed information. The 3D rendering model and overlay display help users better understand and analyze the inspection results. Improve the readability and practicality of the inspection report, which is convenient for maintenance personnel to further process and repair.

[0144] See also Figure 2 As shown, Figure 2: It is a structural schematic diagram of the Lamb wave excitation and reception optimization device 200 for non-destructive testing and positioning of hyperbolic glass curtain walls provided in an embodiment of the present application. The Lamb wave excitation and reception optimization device 200 for non-destructive testing and positioning of hyperbolic glass curtain walls is used to execute the steps of the Lamb wave excitation and reception optimization method for non-destructive testing and positioning of hyperbolic glass curtain walls shown in the above embodiments. The Lamb wave excitation and reception optimization device 200 for non-destructive testing and positioning of hyperbolic glass curtain walls can be a single server or a server cluster, or the Lamb wave excitation and reception optimization device 200 for non-destructive testing and positioning of hyperbolic glass curtain walls can be a terminal, which can be a handheld terminal, a laptop computer, a wearable device or a robot, etc.

[0145] like Figure 2 As shown, the Lamb wave excitation and reception optimization method device 200 for non-destructive testing and positioning of a hyperbolic glass curtain wall includes:

[0146] The parameter acquisition unit 201 is used to acquire the curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located; the curtain wall parameter information at least includes geometric parameters, thickness parameters, material parameters and environmental parameters;

[0147] A parameter optimization unit 202, configured to optimize Lamb wave parameters corresponding to the glass curtain wall according to the curtain wall parameter information; the Lamb wave parameters at least include an excitation frequency, a waveform coefficient, and an excitation angle;

[0148] A signal acquisition unit 203 is used to transmit and receive signals from the hyperbolic glass curtain wall according to the optimized Lamb wave parameters, and acquire a damage reflection signal corresponding to the hyperbolic glass curtain wall;

[0149] A signal decomposition unit 204, configured to perform time-frequency domain decomposition on the damage reflection signal according to wavelet packet transform to obtain a damage feature vector;

[0150] The coordinate output unit 205 is used to obtain a neural network positioning model including hyperbolic surface geometric constraints, input the damage feature vector into the neural network positioning model, output the three-dimensional coordinates of the damage, and complete the non-destructive detection positioning of the hyperbolic glass curtain wall.

[0151] In some embodiments, obtaining curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located includes: obtaining a three-dimensional laser scanning image and a visual image corresponding to the hyperbolic glass curtain wall; transforming feature points corresponding to the three-dimensional laser scanning image and the visual image to the same coordinate system based on a point cloud registration algorithm; and reconstructing the surface of the hyperbolic glass curtain wall according to the registered three-dimensional laser scanning image and the visual image to obtain the curtain wall parameter information.

[0152] In some embodiments, the expression of the excitation frequency includes:

[0153] ;in, represents the excitation frequency, represents the temperature compensation coefficient, is the thickness parameter, is Young's modulus, is the Poisson's ratio, is the material density corresponding to the hyperbolic glass, Used to express the effective stiffness under plane stress state, It is used to represent the influence of Poisson's ratio on the three-dimensional strain of the hyperbolic glass curtain wall.

[0154] In some embodiments, the expression of the excitation angle includes:

[0155] ;in, represents the excitation angle, and They respectively represent the local slopes of the curved surface corresponding to the hyperbolic glass curtain wall, represents the environmental compensation angle corresponding to the hyperbolic glass curtain wall, which is used to correct the wave velocity anisotropy caused by environmental factors. It means that the excitation angle is adjusted according to the local curvature of the hyperbolic surface corresponding to the hyperbolic glass curtain wall to ensure that the lamb wavefront always maintains an optimal incident angle with the surface normal of the hyperbolic glass curtain wall to avoid lamb wave signal scattering.

[0156] Exemplarily, if the ambient temperature corresponding to the hyperbolic glass curtain wall increases, the ambient compensation angle needs to be subtracted from the first compensation angle, and the first compensation angle is within a first compensation angle range of 0.05 to 0.15 rad; if the ambient humidity corresponding to the hyperbolic glass curtain wall increases, the ambient compensation angle needs to be increased by a second compensation angle, and the second compensation angle is within a second compensation angle range of 0.02 to 0.08 rad.

[0157] In some embodiments, the inputting the damage feature vector into the neural network positioning model and outputting the three-dimensional coordinates of the damage includes: inputting the damage feature vector into the neural network positioning model, the neural network positioning model generating geometric damage features through a three-dimensional convolution kernel and generating signal damage features through time-frequency convolution kernel processing; performing feature fusion based on the geometric damage features and the signal damage features, and outputting the three-dimensional coordinates of the damage.

[0158] In some embodiments, before transmitting and receiving signals from the hyperbolic glass curtain wall according to the optimized Lamb wave parameters, it also includes: obtaining the minimum phase velocity, the maximum excitation frequency and the maximum excitation angle corresponding to the Lamb wave; and adjusting the spatial density of the array piezoelectric sensor for receiving the injury reflection signal according to the minimum phase velocity, the maximum excitation frequency and the maximum excitation angle.

[0159] Exemplarily, the expression corresponding to the spatial density includes:

[0160] ;in, is the spatial density, is the minimum phase velocity, is the highest excitation frequency, is the maximum excitation angle.

[0161] In some embodiments, the time-frequency decomposition of the damaged reflection signal according to the wavelet packet transform to obtain the damage feature vector includes: performing time-frequency decomposition of the damaged reflection signal according to the wavelet packet transform to obtain the signal energy attenuation coefficient, the main frequency offset and the wavelet packet node energy entropy; generating the damage feature vector according to the signal energy attenuation coefficient, the main frequency offset and the wavelet packet node energy entropy.

[0162] In some embodiments, after the output of the three-dimensional coordinates of the damage, it also includes: generating a visual detection map according to the three-dimensional coordinates of the damage and a curtain wall model corresponding to the hyperbolic glass curtain wall; the visual detection map includes a hyperbolic surface development diagram of the hyperbolic glass curtain wall, the hyperbolic surface development diagram includes a damage projection corresponding to the three-dimensional coordinates of the damage, a three-dimensional rendering model corresponding to the hyperbolic glass curtain wall, the three-dimensional rendering model includes a defect mark corresponding to the three-dimensional coordinates of the damage, and a stacked display diagram corresponding to the hyperbolic glass curtain wall, and the stacked display diagram includes defect size and depth parameters.

[0163] It should be noted that, those skilled in the art can clearly understand that, for the convenience and simplicity of description, the specific working processes of the Lamb wave excitation and reception optimization method device for non-destructive testing and positioning of hyperbolic glass curtain walls and each module described above can refer to the corresponding processes in the embodiments of the Lamb wave excitation and reception optimization method for non-destructive testing and positioning of hyperbolic glass curtain walls described in the above embodiments, and will not be repeated here.

[0164] The above-mentioned Lamb wave excitation and reception optimization method for non-destructive testing and positioning of hyperbolic glass curtain wall can be implemented in the form of a computer program. The computer program can be used in Figure 2 Run on the device shown.

[0165] See also Figure 3 , Figure 3 1 is a schematic block diagram of the structure of a computer device provided in an embodiment of the present application. The computer device includes a processor, a memory and a network interface connected via a device bus, wherein the memory may include a storage medium and an internal memory.

[0166] The storage medium can store an operating device and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any Lamb wave excitation and reception optimization method for non-destructive detection and positioning of hyperbolic glass curtain walls.

[0167] The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.

[0168] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any Lamb wave excitation and reception optimization method for non-destructive detection and positioning of hyperbolic glass curtain walls.

[0169] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the scheme of the present application, and does not constitute a limitation on the terminal to which the scheme of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0170] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0171] In one embodiment, the processor is used to run a computer program stored in the memory to implement the following steps:

[0172] Acquire curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located; the curtain wall parameter information at least includes geometric parameters, thickness parameters, material parameters and environmental parameters;

[0173] Optimizing Lamb wave parameters corresponding to the glass curtain wall according to the curtain wall parameter information; the Lamb wave parameters at least include an excitation frequency, a waveform coefficient, and an excitation angle;

[0174] Transmitting and receiving signals from the hyperbolic glass curtain wall according to the optimized Lamb wave parameters to obtain a damage reflection signal corresponding to the hyperbolic glass curtain wall;

[0175] Decomposing the damage reflection signal in time and frequency domain according to wavelet packet transform to obtain a damage feature vector;

[0176] A neural network positioning model including hyperbolic surface geometric constraints is obtained, the damage feature vector is input into the neural network positioning model, the three-dimensional coordinates of the damage are output, and the non-destructive detection positioning of the hyperbolic glass curtain wall is completed.

[0177] In some embodiments, obtaining curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located includes: obtaining a three-dimensional laser scanning image and a visual image corresponding to the hyperbolic glass curtain wall; transforming feature points corresponding to the three-dimensional laser scanning image and the visual image to the same coordinate system based on a point cloud registration algorithm; and reconstructing the surface of the hyperbolic glass curtain wall according to the registered three-dimensional laser scanning image and the visual image to obtain the curtain wall parameter information.

[0178] In some embodiments, the expression of the excitation frequency includes:

[0179] ;in, represents the excitation frequency, represents the temperature compensation coefficient, is the thickness parameter, is Young's modulus, is the Poisson's ratio, is the material density corresponding to the hyperbolic glass, Used to express the effective stiffness under plane stress state, It is used to represent the influence of Poisson's ratio on the three-dimensional strain of the hyperbolic glass curtain wall.

[0180] In some embodiments, the expression of the excitation angle includes:

[0181] ;in, represents the excitation angle, and They respectively represent the local slopes of the curved surface corresponding to the hyperbolic glass curtain wall, represents the environmental compensation angle corresponding to the hyperbolic glass curtain wall, which is used to correct the wave velocity anisotropy caused by environmental factors. It means that the excitation angle is adjusted according to the local curvature of the hyperbolic surface corresponding to the hyperbolic glass curtain wall to ensure that the lamb wavefront always maintains an optimal incident angle with the surface normal of the hyperbolic glass curtain wall to avoid lamb wave signal scattering.

[0182] Exemplarily, if the ambient temperature corresponding to the hyperbolic glass curtain wall increases, the ambient compensation angle needs to be subtracted from the first compensation angle, and the first compensation angle is within a first compensation angle range of 0.05 to 0.15 rad; if the ambient humidity corresponding to the hyperbolic glass curtain wall increases, the ambient compensation angle needs to be increased by a second compensation angle, and the second compensation angle is within a second compensation angle range of 0.02 to 0.08 rad.

[0183] In some embodiments, the inputting the damage feature vector into the neural network positioning model and outputting the three-dimensional coordinates of the damage includes: inputting the damage feature vector into the neural network positioning model, the neural network positioning model generating geometric damage features through a three-dimensional convolution kernel and generating signal damage features through time-frequency convolution kernel processing; performing feature fusion based on the geometric damage features and the signal damage features, and outputting the three-dimensional coordinates of the damage.

[0184] In some embodiments, before transmitting and receiving signals from the hyperbolic glass curtain wall according to the optimized Lamb wave parameters, it also includes: obtaining the minimum phase velocity, the maximum excitation frequency and the maximum excitation angle corresponding to the Lamb wave; and adjusting the spatial density of the array piezoelectric sensor for receiving the injury reflection signal according to the minimum phase velocity, the maximum excitation frequency and the maximum excitation angle.

[0185] Exemplarily, the expression corresponding to the spatial density includes:

[0186] ;in, is the spatial density, is the minimum phase velocity, is the highest excitation frequency, is the maximum excitation angle.

[0187] In some embodiments, the time-frequency decomposition of the damaged reflection signal according to the wavelet packet transform to obtain the damage feature vector includes: performing time-frequency decomposition of the damaged reflection signal according to the wavelet packet transform to obtain the signal energy attenuation coefficient, the main frequency offset and the wavelet packet node energy entropy; generating the damage feature vector according to the signal energy attenuation coefficient, the main frequency offset and the wavelet packet node energy entropy.

[0188] In some embodiments, after the output of the three-dimensional coordinates of the damage, it also includes: generating a visual detection map according to the three-dimensional coordinates of the damage and a curtain wall model corresponding to the hyperbolic glass curtain wall; the visual detection map includes a hyperbolic surface development diagram of the hyperbolic glass curtain wall, the hyperbolic surface development diagram includes a damage projection corresponding to the three-dimensional coordinates of the damage, a three-dimensional rendering model corresponding to the hyperbolic glass curtain wall, the three-dimensional rendering model includes a defect mark corresponding to the three-dimensional coordinates of the damage, and a stacked display diagram corresponding to the hyperbolic glass curtain wall, and the stacked display diagram includes defect size and depth parameters.

[0189] It should be noted that technical personnel in the relevant field can clearly understand that, for the convenience and simplicity of description, the specific working process of the computer equipment and each module described above can refer to the corresponding process in the embodiment of the Lamb wave excitation and reception optimization method for non-destructive testing and positioning of hyperbolic glass curtain walls described in the above embodiments, and will not be repeated here.

[0190] The present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the processor implements the steps of the Lamb wave excitation and reception optimization method for non-destructive testing and positioning of hyperbolic glass curtain walls as provided in any embodiment of the present application.

[0191] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiment, such as a hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a smart memory card (SmartMedia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc., equipped on the computer device.

[0192] Exemplarily, the medium is used to implement the following steps:

[0193] Acquire curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located; the curtain wall parameter information at least includes geometric parameters, thickness parameters, material parameters and environmental parameters;

[0194] Optimizing Lamb wave parameters corresponding to the glass curtain wall according to the curtain wall parameter information; the Lamb wave parameters at least include an excitation frequency, a waveform coefficient, and an excitation angle;

[0195] Transmitting and receiving signals from the hyperbolic glass curtain wall according to the optimized Lamb wave parameters to obtain a damage reflection signal corresponding to the hyperbolic glass curtain wall;

[0196] Decomposing the damage reflection signal in time and frequency domain according to wavelet packet transform to obtain a damage feature vector;

[0197] A neural network positioning model including hyperbolic surface geometric constraints is obtained, the damage feature vector is input into the neural network positioning model, the three-dimensional coordinates of the damage are output, and the non-destructive detection positioning of the hyperbolic glass curtain wall is completed.

[0198] It should be noted that technical personnel in the relevant field can clearly understand that, for the convenience and conciseness of description, the specific working process of the storage medium and each module described above can refer to the corresponding process in the embodiment of the Lamb wave excitation and reception optimization method for non-destructive testing and positioning of hyperbolic glass curtain walls described in the above embodiments, and will not be repeated here.

[0199] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A Lamb wave excitation and reception optimization method for nondestructive testing and positioning of hyperbolic glass curtain walls, characterized in that: include: Obtaining curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located; The curtain wall parameter information includes at least geometric parameters, thickness parameters, material parameters and environmental parameters; Optimizing Lamb wave parameters corresponding to the glass curtain wall according to the curtain wall parameter information; the Lamb wave parameters at least include an excitation frequency, a waveform coefficient, and an excitation angle; Transmitting and receiving signals from the hyperbolic glass curtain wall according to the optimized Lamb wave parameters to obtain a damage reflection signal corresponding to the hyperbolic glass curtain wall; Decomposing the damage reflection signal in time and frequency domain according to wavelet packet transform to obtain a damage feature vector; A neural network positioning model including hyperbolic surface geometric constraints is obtained, the damage feature vector is input into the neural network positioning model, the three-dimensional coordinates of the damage are output, and the non-destructive detection positioning of the hyperbolic glass curtain wall is completed.

2. The method according to claim 1, characterized in that The step of obtaining curtain wall parameter information of the hyperbolic glass curtain wall to be detected and located includes: Acquire a three-dimensional laser scanning image and a visual image corresponding to the hyperbolic glass curtain wall; Based on a point cloud registration algorithm, the feature points corresponding to the three-dimensional laser scanning image and the visual image are transformed into the same coordinate system; The surface of the hyperbolic glass curtain wall is reconstructed according to the registered three-dimensional laser scanning image and the visual image to obtain parameter information of the curtain wall.

3. The method according to claim 1, characterized in that The expression of the excitation frequency includes: ; in, represents the excitation frequency, represents the temperature compensation coefficient, is the thickness parameter, is Young's modulus, is the Poisson's ratio, is the material density corresponding to the hyperbolic glass, Used to express the effective stiffness under plane stress state, It is used to represent the influence of Poisson's ratio on the three-dimensional strain of the hyperbolic glass curtain wall.

4. The method according to claim 1, characterized in that: The expression of the excitation angle includes: ; in, represents the excitation angle, and They respectively represent the local slopes of the curved surface corresponding to the hyperbolic glass curtain wall, represents the environmental compensation angle corresponding to the hyperbolic glass curtain wall, which is used to correct the wave velocity anisotropy caused by environmental factors. It means that the excitation angle is adjusted according to the local curvature of the hyperbolic surface corresponding to the hyperbolic glass curtain wall to ensure that the lamb wavefront always maintains an optimal incident angle with the surface normal of the hyperbolic glass curtain wall to avoid lamb wave signal scattering.

5. The method according to claim 4, characterized in that If the ambient temperature corresponding to the hyperbolic glass curtain wall increases, the ambient compensation angle needs to be subtracted from the first compensation angle, the first compensation angle is within a first compensation angle range, and the first compensation angle range is 0.05 to 0.15 rad; If the ambient humidity corresponding to the hyperbolic glass curtain wall increases, the ambient compensation angle needs to increase a second compensation angle, and the second compensation angle is within a second compensation angle range, and the second compensation angle range is 0.02 to 0.08 rad.

6. The method according to claim 1, characterized in that The step of inputting the damage feature vector into the neural network positioning model and outputting the three-dimensional coordinates of the damage comprises: The damage feature vector is input into the neural network positioning model, and the neural network positioning model generates geometric damage features through a three-dimensional convolution kernel and generates signal damage features through time-frequency convolution kernel processing; feature fusion is performed based on the geometric damage features and the signal damage features, and the three-dimensional coordinates of the damage are output.

7. The method according to claim 1, characterized in that Before transmitting and receiving signals to the hyperbolic glass curtain wall according to the optimized Lamb wave parameters, the method further includes: Obtaining the minimum phase velocity, the maximum excitation frequency and the maximum excitation angle corresponding to the Lamb wave; The spatial density of the arrayed piezoelectric sensors for receiving the damage reflection signal is adjusted according to the minimum phase velocity, the maximum excitation frequency and the maximum excitation angle.

8. The method according to claim 7, characterized in that The expressions corresponding to the spatial density include: ; in, is the spatial density, is the minimum phase velocity, is the highest excitation frequency, is the maximum excitation angle.

9. The method according to claim 1, characterized in that: The step of performing time-frequency domain decomposition on the damaged reflection signal according to wavelet packet transform to obtain a damage feature vector includes: Decomposing the damaged reflection signal in the time-frequency domain according to wavelet packet transform to obtain the signal energy attenuation coefficient, the main frequency offset and the wavelet packet node energy entropy; The damage feature vector is generated according to the signal energy attenuation coefficient, the main frequency offset and the wavelet packet node energy entropy.

10. The method according to claim 1, characterized in that After outputting the three-dimensional coordinates of the damage, the method further includes: A visual inspection map is generated according to the three-dimensional coordinates of the damage and the curtain wall model corresponding to the hyperbolic glass curtain wall; the visual inspection map includes a hyperbolic surface development diagram of the hyperbolic glass curtain wall, the hyperbolic surface development diagram includes a damage projection corresponding to the three-dimensional coordinates of the damage, a three-dimensional rendering model corresponding to the hyperbolic glass curtain wall, the three-dimensional rendering model includes a defect mark corresponding to the three-dimensional coordinates of the damage, and a stacked display diagram corresponding to the hyperbolic glass curtain wall, the stacked display diagram includes defect size and depth parameters.

Images