A method and system for quantitative non-destructive monitoring of safety in aircraft concrete structures

Abstract

This invention relates to the field of civil engineering structural safety monitoring and aircraft image detection technology, and discloses a method and system for quantitative non-destructive monitoring of concrete structure safety based on aircraft. The method includes: constructing a multi-objective path optimization model to generate a flight path point sequence; the aircraft acquiring images and calculating image reliability scores; using a deep learning model to identify cracks and spalling defects; calculating the structural risk index (SFI) based on a partial least squares regression model; and generating a monitoring report and uploading it to the cloud. Compared to existing technologies that rely on manual inspection or static image recognition, especially in irregular curved surfaces, complex occlusion areas, or conditions with drastic lighting changes, this invention addresses the technical problem of failing to achieve quantitative assessment of structural safety based on in-flight images. By integrating three-dimensional path optimization, image reliability scoring mechanisms, and AI defect recognition, this invention achieves intelligent identification of structural surface coverage and crack spalling, improving the automation level of concrete structure monitoring.

Description

Technical Field

[0001] This invention relates to the field of civil engineering structural safety monitoring and aircraft image detection technology, and in particular to a method and system for quantitative non-destructive monitoring of aircraft concrete structure safety. Background Technology

[0002] Currently, safety inspections of concrete structures primarily rely on manual inspections, localized image acquisition, and post-processing analysis. However, these methods have significant shortcomings. For instance, manual inspections are inefficient, easily influenced by subjective experience, and struggle to systematically record the evolution of structural defects. Traditional static image recognition methods depend on single-angle or low-coverage images, often failing to comprehensively capture early signs of degradation such as minute cracks, spalling, and exposed reinforcement on the concrete surface. In engineering practice, especially for structures with complex geometric surfaces or obstructed environments, such as bridge piers, tunnel linings, and tall towers, aircraft are limited by issues like imperfect trajectory planning, unstable flight attitude, and blurred images. This often leads to inconsistent image data quality, unclear crack boundaries, and incomplete defect feature extraction, severely impacting subsequent image data processing and the accuracy of structural risk assessment. Existing technologies still have significant shortcomings in image acquisition stability, adaptive image data processing, structural defect parameter quantification, and automatic reporting linkage, failing to fully meet the practical needs for efficient, accurate, quantitative, and closed-loop non-destructive monitoring of concrete structures in complex environments. Therefore, there is an urgent need for an aircraft monitoring method that integrates three-dimensional structural modeling, adaptive trajectory optimization, in-flight image quality assessment, AI defect identification, and structural risk quantification model. This method should be able to achieve high-quality image data processing and intelligent quantitative expression of structural damage even under conditions of attitude disturbances, lighting changes, and complex structural morphology, so as to comprehensively improve the intelligence level, assessment accuracy, and remote linkage capability of concrete structure monitoring systems. Summary of the Invention

[0003] To address the aforementioned technical shortcomings, the purpose of this invention is to propose a non-destructive monitoring method for the safety quantification of concrete structures in aircraft. This method aims to solve the technical problem that existing technologies rely on manual inspection or static image recognition, especially in the case of irregular curved surfaces, complex occlusion areas, or conditions with drastic changes in lighting, where it is impossible to achieve a quantitative assessment of structural safety based on in-flight images.

[0004] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a method for quantitative non-destructive monitoring of the safety of concrete structures of aircraft.

[0005] The method for quantitative non-destructive monitoring of the safety of aircraft concrete structures includes:

[0006] Step S10: Acquire 3D laser point cloud data or BIM model data of the target concrete structure, extract surface morphology data of the target concrete structure, and discretize the surface morphology data into a set of facets S; set a candidate track point sequence based on the angle between the normal vector of each facet in the set S and the preset viewing angle of the aircraft camera, the minimum visible distance, and occlusion conditions. Construct a multi-objective path optimization model, in which the objective function includes the first objective function. Second objective function and the third objective function The multi-objective path optimization model is solved by combining a non-dominated sorting genetic algorithm, and a set of optimized waypoint sequences is output. ;

[0007] Step S20: Based on the optimized waypoint sequence The image at time t was obtained by taking pictures using an aircraft. According to the captured images Perform an image optimization task to obtain an optimized set of captured images;

[0008] Step S30: Extract crack and peeling defect regions from the optimized image set using model inference, and output defect parameters; wherein, the defect parameters include the bounding box of the defect region k. Confidence level of defect region boundaries Crack length Average crack width Number of cracks Number of peeling pieces and boundary roughness ;

[0009] Step S40: Calculate the structural risk index SFI based on the defect parameters using partial least squares regression.

[0010] Step S50: If the structural risk index SFI exceeds the preset high-risk threshold, a non-destructive monitoring risk report is generated based on the structural risk index SFI, defect parameters, and optimized image set, and uploaded to the cloud database using the aircraft's WIFI module and preset API interface.

[0011] Preferably, in step S20, based on the captured image The steps for performing image optimization tasks to obtain an optimized set of captured images specifically include: obtaining the angular velocity of the inertial measurement unit of the spacecraft at time t. and angular acceleration and total amplitude of attitude disturbance Regarding the captured images Image sharpness metrics, including Laplacian variance, are derived from pixel extraction. Average brightness and exposure offset And combined with angular velocity and angular acceleration and total amplitude of attitude disturbance Comprehensive calculation of image reliability score When image reliability score Less than the preset image reliability score threshold At that time, mark the image For unreliable images, record time t as the location to be reshot; perform a reshot task based on the location to be reshot, combining the captured images. Generate an optimized set of captured images.

[0012] Preferably, in step S10, the first objective function The first objective function is the image coverage function, used to maximize the effective coverage of the target patch set by the range of the waypoints; the second objective function is... The first objective function is the path length minimization function, used to minimize the cumulative distance traveled by the aircraft from start to finish; the second objective function is... This is an imaging viewpoint quality function used to optimize the angle between the camera optical axis and the normal of the surface to be close to perpendicular, thereby improving the geometric resolution of the image and the ability to identify defects.

[0013] Preferably, step S30, which involves extracting crack and peeling defect regions from the optimized image set using model inference and outputting defect parameters, specifically includes: using a lightweight U-Net model to perform pixel-level segmentation on the images in the image set and outputting a crack mask. With peeling mask The YOLOv8 object detection model is called in parallel to detect the boundaries of defect regions in the image set, and the bounding box of defect region k is output. Confidence level of defect region boundaries ; for crack masking Extract the centerline and calculate the crack length based on pixels. The average crack width was calculated using the sliding window vertical width measurement method. ; Calculate the number of cracks Number of peeling pieces and crack boundary roughness .

[0014] Preferably, in step S30, the boundary roughness The "torsionalism" and "nonlinearity" used to effectively characterize crack edges are image geometric feature parameters for judging the degree of structural damage evolution. Their calculation steps specifically include: using Canny edge detection to mask the crack. Edge extraction is performed to obtain crack edge pixels, and the image is divided into multiple... Count the number of grid cell boxes of different sizes that can cover the pixels at the edge of the crack. The size of the mesh cell Box is changed and the iteration is repeated. During the iteration of the mesh cell Box size, the boundary roughness is obtained by fitting logarithmic coordinates. Boundary roughness ,in, The value ranges from 1 to 2. This indicates that the crack is straight; if This indicates that the crack has many twists and turns, and the risk is higher.

[0015] Preferably, in step S40, the step of calculating the structural risk index SFI based on the defect parameters using partial least squares regression specifically includes: normalizing the defect parameters to form a feature vector. Input feature vector The input is fed into a trained partial least squares regression model to obtain the structural risk index SFI. ,in, These are the normalized variables for each parameter. and SFI represents the regression coefficients obtained through training with historical crack and spalling labeled data. SFI represents the degree of structural risk; a higher SFI value indicates more severe structural degradation.

[0016] Preferably, in step S20, the image reliability score is... The calculation formula is: ,in, For a moment Capture images Image reliability score; For angular velocity and angular acceleration The calculated total amplitude of attitude disturbance during image capture; The Laplacian variance in the image sharpness index; image The average brightness; This is a preset ideal brightness exposure reference point; Exposure offset; To take the exponential function; and Let be the weighting coefficient, satisfying It is determined by experience or training data; These are the sensitivity attenuation factors that control the total amplitude of attitude disturbance and the average brightness, respectively.

[0017] This invention also provides a non-destructive monitoring system for the safety quantification of concrete structures in aircraft, comprising:

[0018] The 3D modeling and path optimization module is used to acquire 3D laser point cloud data or BIM model data of the target concrete structure, extract surface morphology data of the target concrete structure, and discretize the surface morphology data into a set of facets S. Based on the angle between the normal vector of each facet in the set S and the preset viewing angle of the aircraft camera, the minimum visible distance, and occlusion conditions, a sequence of candidate waypoints is set. Construct a multi-objective path optimization model, in which the objective function includes the first objective function. Second objective function and the third objective function The multi-objective path optimization model is solved by combining a non-dominated sorting genetic algorithm, and a set of optimized waypoint sequences is output. ;

[0019] The image acquisition and optimization module is used to optimize the waypoint sequence. The image at time t was obtained by taking pictures using an aircraft. According to the captured images Perform an image optimization task to obtain an optimized set of captured images;

[0020] The defect identification and parameter extraction module is used to extract crack and peeling defect regions from an optimized set of captured images using model inference, and outputs defect parameters; among which, the defect parameters include the bounding box of the defect region k. Confidence level of defect region boundaries Crack length Average crack width Number of cracks Number of peeling pieces and boundary roughness ;

[0021] The structural risk calculation module is used to calculate the structural risk index (SFI) based on defect parameters using partial least squares regression.

[0022] The risk warning and upload module is used to generate a non-destructive monitoring risk report based on the structural risk index SFI, defect parameters, and optimized image set when the SFI exceeds the preset high-risk threshold. The report is then uploaded to the cloud database using the aircraft's WIFI module and preset API interface.

[0023] The present invention also provides a device for quantitative non-destructive monitoring of the safety of concrete structures for aircraft, comprising: a memory, a processor, and a program for quantitative non-destructive monitoring of the safety of concrete structures for aircraft stored in the memory and executable on the processor. When the program for quantitative non-destructive monitoring of the safety of concrete structures for aircraft is executed by the processor, a method for quantitative non-destructive monitoring of the safety of concrete structures for aircraft is implemented.

[0024] The present invention also provides a computer program product, including a program for quantitative non-destructive monitoring of the safety of concrete structures based on aircraft, wherein the program for quantitative non-destructive monitoring of the safety of concrete structures based on aircraft is executed by a processor to implement the method for quantitative non-destructive monitoring of the safety of concrete structures based on aircraft.

[0025] The beneficial effects of this invention are as follows: by constructing a trajectory optimization and image data processing mechanism that integrates structural geometry, flight attitude and image quality indicators, high coverage and high stability of structural surface image acquisition are achieved, solving the problems of multiple blind spots and blurry and unstable images in the prior art.

[0026] By introducing a deep learning-based image data processing model and a structural risk index regression method, it is possible to automatically identify and extract parameters of concrete cracks and spalling defects, and to achieve a quantitative assessment of the structural safety status, thereby improving the intelligence and assessment accuracy of the monitoring system. Attached Figure Description

[0027] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0028] Figure 1 This is a flowchart illustrating the first embodiment of a method for quantitative non-destructive monitoring of the safety of concrete structures for aircraft, according to the present invention.

[0029] Figure 2 This is a schematic diagram of the equipment for a method of quantitative non-destructive monitoring of the safety of concrete structures for aircraft, based on the present invention. Detailed Implementation

[0030] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0031] Example 1: As Figure 1 The diagram shown is a flowchart of the first embodiment of the method for quantitative non-destructive monitoring of the safety of concrete structures for aircraft according to the present invention. The first embodiment of the method for quantitative non-destructive monitoring of the safety of concrete structures for aircraft according to the present invention is presented.

[0032] In the first embodiment, the method for quantitative non-destructive monitoring of aircraft concrete structure safety includes:

[0033] Step S10: Acquire 3D laser point cloud data or BIM model data of the target concrete structure, extract surface morphology data of the target concrete structure, and discretize the surface morphology data into a set of facets S; set a candidate track point sequence based on the angle between the normal vector of each facet in the set S and the preset viewing angle of the aircraft camera, the minimum visible distance, and occlusion conditions. Construct a multi-objective path optimization model, in which the objective function includes the first objective function. Second objective function and the third objective function The multi-objective path optimization model is solved by combining a non-dominated sorting genetic algorithm, and a set of optimized waypoint sequences is output. ;

[0034] It should be noted that in step S10, the first objective function The first objective function is the image coverage function, used to maximize the effective coverage of the target patch set by the range of the waypoints; the second objective function is... The first objective function is the path length minimization function, used to minimize the cumulative distance traveled by the aircraft from start to finish; the second objective function is... This is an imaging viewpoint quality function used to optimize the angle between the camera optical axis and the normal of the surface to be close to perpendicular, thereby improving the geometric resolution of the image and the ability to identify defects.

[0035] It should be noted that in setting the candidate waypoint sequence, the orientation of each facet extracted from the 3D model in space is analyzed first, i.e., its surface normal direction is obtained. Based on the normal direction of each facet, multiple combination points of the aircraft's shooting position and attitude direction are preset within a certain range in front of it. Then, each combination point is judged to meet the following three conditions in turn: First, it is judged whether the angle between the shooting direction of the aircraft camera and the normal vector of the facet is close to perpendicular, for example, whether it is between 75 degrees and 105 degrees, to ensure that the camera is basically facing the target surface when the image is acquired, reducing distortion and enhancing the defect recognition effect; Second, it is judged whether the distance between the shooting point and the facet is within the set effective imaging range, for example, whether it is between 1 and 3 meters, so as to ensure that the image will not be out of focus due to excessive distance, nor will the field of view be insufficient due to insufficient distance; Finally, it is also necessary to judge whether the line of sight from the shooting point to the target facet is blocked by other structures. Occlusion detection can be performed based on spatial visibility analysis of the positional relationship of other facets in 3D space. If occlusion exists, the candidate point is excluded. Only shooting points that meet all three conditions will be included in the candidate track point sequence for subsequent path optimization module to filter and sort.

[0036] It should be understood that, compared with traditional uniform linear flight or regular grid imaging methods, the above-mentioned optimized path generation method can not only achieve high coverage imaging of concrete structure surfaces in complex shapes (such as arches, curved surfaces, and T-sections), but also significantly reduce the decrease in image resolution and defect identification errors caused by angular deviations, thereby providing highly reliable data support for subsequent image quality assessment and defect extraction.

[0037] Furthermore, based on the principle of an electrolytic cell, the flow and changes of current in different regions are monitored. During corrosion, the conduction of current is affected by the distribution of electrolyte concentration. Utilizing this diffusion phenomenon (after a certain service life of reinforced concrete structures, the interaction of ions in the concrete protective layer of the steel reinforcement over many years can be approximated as satisfying the diffusion phenomenon), changes in current and voltage caused by corrosion can be detected in real time, thereby enabling quantitative analysis of the degree of corrosion. By combining image processing methods with electrochemical signals, the accuracy of detection results for defects caused by corrosion can be further improved. Correspondingly, the aircraft in this invention employs a spring-type adaptive flight control corrosion quantification non-destructive monitoring equipment. This equipment uses a spring-type IoT sensor, equipped with nine equally sized 5cm high trapezoidal spring contacts. The ultimate elastic force of each trapezoidal spring... With the lift of drones The optimal relationship is set using the following formula: ,in It uses a constant spring constant. This design allows for early warning of structural defects caused by rust pits, which can be anticipated more than a year in advance.

[0038] Step S20: Based on the optimized waypoint sequence The image at time t was obtained by taking pictures using an aircraft. According to the captured images Perform an image optimization task to obtain an optimized set of captured images;

[0039] It should be noted that in step S20, based on the captured image... The steps for performing image optimization tasks to obtain an optimized set of captured images specifically include: obtaining the angular velocity of the inertial measurement unit of the spacecraft at time t. and angular acceleration and total amplitude of attitude disturbance Regarding the captured images Image sharpness metrics, including Laplacian variance, are derived from pixel extraction. Average brightness and exposure offset And combined with angular velocity and angular acceleration and total amplitude of attitude disturbance Comprehensive calculation of image reliability score When image reliability score Less than the preset image reliability score threshold At that time, mark the image For unreliable images, record time t as the location to be reshot; perform a reshot task based on the location to be reshot, combining the captured images. Generate an optimized set of captured images. Image reliability score. The calculation formula is: ,in, For a moment Capture images Image reliability score; For angular velocity and angular acceleration The calculated total amplitude of attitude disturbance during image capture; The Laplacian variance in the image sharpness index; image The average brightness; This is a preset ideal brightness exposure reference point; Exposure offset; To take the exponential function; and Let be the weighting coefficient, satisfying It is determined by experience or training data; These are the sensitivity attenuation factors that control the total amplitude of attitude disturbance and the average brightness, respectively.

[0040] Understandably, by introducing a dual-factor evaluation mechanism of attitude perturbation and image content, this invention can proactively identify unreliable images caused by flight jitter, focus failure, or abnormal lighting immediately after image acquisition, greatly reducing the incidence of subsequent recognition errors. The image reliability score not only serves as an image selection criterion but also constitutes an important credibility foundation for subsequent defect identification tasks, helping to ensure the stability of defect detection input data.

[0041] Step S30: Extract crack and peeling defect regions from the optimized image set using model inference, and output defect parameters; wherein, the defect parameters include the bounding box of the defect region k. Confidence level of defect region boundaries Crack length Average crack width Number of cracks Number of peeling pieces and boundary roughness ;

[0042] It should be noted that step S30, which involves using model inference to extract crack and peeling defect regions from the optimized image set and outputting defect parameters, specifically includes: using a lightweight U-Net model to perform pixel-level segmentation on the images in the image set and outputting a crack mask. With peeling mask The YOLOv8 object detection model is called in parallel to detect the boundaries of defect regions in the image set, and the bounding box of defect region k is output. Confidence level of defect region boundaries ; for crack masking Extract the centerline and calculate the crack length based on pixels. The average crack width was calculated using the sliding window vertical width measurement method. ; Calculate the number of cracks Number of peeling pieces and crack boundary roughness Boundary roughness The "torsionalism" and "nonlinearity" used to effectively characterize crack edges are image geometric feature parameters for judging the degree of structural damage evolution. Their calculation steps specifically include: using Canny edge detection to mask the crack. Edge extraction is performed to obtain crack edge pixels, and the image is divided into multiple... Count the number of grid cell boxes of different sizes that can cover the pixels at the edge of the crack. The size of the mesh cell Box is changed and the iteration is repeated. During the iteration of the mesh cell Box size, the boundary roughness is obtained by fitting logarithmic coordinates. Boundary roughness ,in, The value ranges from 1 to 2. This indicates that the crack is straight; if This indicates that the crack has many twists and turns, and the risk is higher.

[0043] It should be noted that during the defect region extraction process on the optimized image set, a lightweight U-Net model is first invoked to perform pixel-level segmentation of the images. This U-Net model takes the original RGB image as input and employs an encoder-decoder structure to extract semantic and boundary features layer by layer from the image content. The encoder consists of multiple convolutional, batch normalization, and max-pooling modules to progressively extract spatial semantic features of cracks and spalling areas. The decoder uses deconvolution and skip connections to progressively restore the high-dimensional feature map to the original image resolution, achieving pixel-level segmentation of cracks and spalling areas. The output layer consists of two independent channels, representing the probability maps of the crack mask and the spalling mask, respectively. Higher pixel values ​​in the mask indicate higher confidence that the corresponding location belongs to that defect category. Simultaneously, to quickly obtain the bounding boxes and confidence information of structural defects, a pre-trained YOLOv8 object detection model is invoked in parallel. This model employs a multi-scale feature fusion structure, enabling the regression of defect candidate boxes and classification confidence evaluation of the entire image in a single forward inference process. The detection process of the YOLOv8 model includes: taking the original image as input, extracting deep features through the Backbone; feeding the feature maps into the Neck network for multi-scale fusion to adapt to defects of different sizes; the Head network performing regression and classification operations on each candidate detection box, outputting: the coordinates of the bounding box of the defect region; the defect category identifier; and the defect boundary confidence score, which indicates the model's confidence that the detection result is a valid defect region. Finally, the segmentation mask result of the U-Net model is aligned with the bounding box detection result of the YOLOv8 model for positional alignment and category fusion.

[0044] Understandably, by deploying pixel-level segmentation and target detection models in parallel, it ensures that the boundaries of cracks and spalling areas can be accurately tracked (U-Net); on the other hand, it ensures that defect areas can be quickly and accurately located in complex backgrounds (YOLOv8). The two complement each other, improving the overall robustness and computational efficiency of structural defect identification.

[0045] Step S40: Calculate the structural risk index SFI based on the defect parameters using partial least squares regression.

[0046] It should be noted that in this step, a multi-dimensional feature vector of the structural state is constructed based on the multiple defect parameters extracted in step S30. These parameters include: the bounding box size of the defect area, boundary confidence, total crack length, average width, number of cracks, number of spalled pieces, and crack boundary roughness. To eliminate the influence of different parameters in terms of dimensions and scale, all defect parameters are normalized and arranged in a preset order to form a unified input vector. Subsequently, partial least squares regression is used to model and analyze the above feature vector. This method extracts the principal directional features that best explain the output variables from the input features, establishes a mapping relationship between features and structural risk, and thus obtains the risk index SFI under the current structural state. SFI, as a comprehensive evaluation index, reflects the overall safety level of the target concrete structure under its current state. The numerical range of this risk index is typically divided into three levels: low risk, medium risk, and high risk, to support the decision-making logic for whether to trigger on-site verification or special reinforcement.

[0047] Understandably, partial least squares regression has the advantage of handling high-dimensional and highly correlated defect parameters. It can effectively avoid the problem of unstable prediction in traditional linear models under the background of variable collinearity, making the model more sensitive to abnormal changes in image parameters. By establishing a quantitative relationship between image defect parameters and structural safety status, it realizes the automated quantitative assessment of structural health status.

[0048] Step S50: If the structural risk index SFI exceeds the preset high-risk threshold, a non-destructive monitoring risk report is generated based on the structural risk index SFI, defect parameters, and optimized image set, and uploaded to the cloud database using the aircraft's WIFI module and preset API interface.

[0049] It should be noted that by introducing an aircraft communication interface and a standardized API structure, this invention enables remote synchronization of image data processing results, supports cross-platform risk visualization and subsequent big data management, and further enhances the engineering practicality and data closed-loop management capabilities of the image-driven non-destructive monitoring system.

[0050] Example 2: Furthermore, the present invention provides a non-destructive testing system for the safety quantification of concrete structures for aircraft, employing a non-destructive testing method for the safety quantification of concrete structures for aircraft as described in the above embodiments, which can solve a technical problem related to non-destructive testing for the safety quantification of concrete structures for aircraft. Compared with the prior art, the beneficial effects of the non-destructive testing system for the safety quantification of concrete structures for aircraft provided by the present invention are the same as those of the non-destructive testing method for the safety quantification of concrete structures for aircraft provided in the above embodiments, and other technical features of the non-destructive testing system for the safety quantification of concrete structures for aircraft provided are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0051] Example 3: This invention provides a non-destructive monitoring device for the safety quantification of concrete structures in aircraft. Please refer to... Figure 2A non-destructive testing device for the safety quantification of concrete structures based on aircraft includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the non-destructive testing method for the safety quantification of concrete structures based on aircraft as described in Embodiment 1 above. The non-destructive testing device for the safety quantification of concrete structures based on aircraft in this embodiment of the invention may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), vehicle terminals (e.g., vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. This non-destructive testing device for the safety quantification of concrete structures based on aircraft is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the invention. A non-destructive testing device for the safety quantification of aircraft concrete structures may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the non-destructive testing device for the safety quantification of aircraft concrete structures. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows a non-destructive testing device for the safety quantification of aircraft concrete structures to exchange data with other devices wirelessly or via wired means. Although the figure shows a non-destructive testing device for the safety quantification of aircraft concrete structures with various systems, it should be understood that it is not required to implement or possess all of the systems shown. More or fewer systems may be implemented alternatively.

[0052] Example 4: This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described method for quantitative non-destructive monitoring of the safety of aircraft concrete structures. The computer program product provided by this invention can solve a technical problem related to quantitative non-destructive monitoring of the safety of aircraft concrete structures. Compared with the prior art, the beneficial effects of the computer program product provided by this invention are the same as those of the method for quantitative non-destructive monitoring of the safety of aircraft concrete structures provided in the above embodiments, and will not be repeated here.

[0053] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this invention.

[0054] It should be understood that the various parts disclosed in this invention can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0055] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for quantitative non-destructive monitoring of the safety of concrete structures in aircraft, characterized in that, The methods include: Step S10: Acquire 3D laser point cloud data or BIM model data of the target concrete structure, extract surface morphology data of the target concrete structure, and discretize the surface morphology data into a set of facets S; set a candidate track point sequence based on the angle between the normal vector of each facet in the set S and the preset viewing angle of the aircraft camera, the minimum visible distance, and occlusion conditions. Construct a multi-objective path optimization model, in which the objective function includes the first objective function. Second objective function and the third objective function The multi-objective path optimization model is solved by combining a non-dominated sorting genetic algorithm, and a set of optimized waypoint sequences is output. ; Step S20: Based on the optimized waypoint sequence The image at time t was obtained by taking pictures using an aircraft. According to the captured images Perform an image optimization task to obtain an optimized set of captured images; Among them, based on the captured images The steps for performing image optimization tasks to obtain an optimized set of captured images specifically include: obtaining the angular velocity of the inertial measurement unit of the spacecraft at time t. and angular acceleration and total amplitude of attitude disturbance Regarding the captured images Image sharpness metrics, including Laplacian variance, are derived from pixel extraction. Average brightness and exposure offset And combined with angular velocity and angular acceleration and total amplitude of attitude disturbance Comprehensive calculation of image reliability score When image reliability score Less than the preset image reliability score threshold At that time, mark the image For unreliable images, record time t as the location to be reshot; perform a reshot task based on the location to be reshot, combining the captured images. Generate an optimized set of captured images; image reliability scoring. The calculation formula is: ,in, For a moment Capture images Image reliability score; For angular velocity and angular acceleration The calculated total amplitude of attitude disturbance during image capture; The Laplacian variance in the image sharpness index; This represents the normalized lower bound of the Laplacian variance of image sharpness; This represents the normalized upper boundary of the Laplacian variance of image sharpness; image The average brightness; This is a preset ideal brightness exposure reference point; Exposure offset; To take the exponential function; and Let be the weighting coefficient, satisfying It is determined by experience or training data; These are the sensitivity attenuation factors that control the total amplitude of attitude disturbance and the average brightness, respectively. Step S30: Extract crack and peeling defect regions from the optimized image set using model inference, and output defect parameters; wherein, the defect parameters include the bounding box of the defect region k. Confidence level of defect region boundaries Crack length Average crack width Number of cracks Number of peeling pieces and boundary roughness ; Step S40: Calculate the structural risk index SFI based on the defect parameters using partial least squares regression. Step S50: If the structural risk index SFI exceeds the preset high-risk threshold, a non-destructive monitoring risk report is generated based on the structural risk index SFI, defect parameters, and optimized image set, and uploaded to the cloud database using the aircraft's WIFI module and preset API interface.

2. The method for quantitative non-destructive monitoring of aircraft concrete structure safety as described in claim 1, characterized in that, In step S10, the first objective function The first objective function is the image coverage function, used to maximize the effective coverage of the target patch set by the range of the waypoints; the second objective function is... The first objective function is the path length minimization function, used to minimize the cumulative distance traveled by the aircraft from start to finish; the second objective function is... This is an imaging viewpoint quality function used to optimize the angle between the camera optical axis and the normal of the surface to be close to perpendicular, thereby improving the geometric resolution of the image and the ability to identify defects.

3. The method for quantitative non-destructive monitoring of aircraft concrete structure safety as described in claim 1, characterized in that, In step S30, the step of extracting crack and peeling defect regions from the optimized image set using model inference and outputting defect parameters specifically includes: using a lightweight U-Net model to perform pixel-level segmentation on the images in the image set and outputting a crack mask. With peeling mask The YOLOv8 object detection model is called in parallel to detect the boundaries of defect regions in the image set, and the bounding box of defect region k is output. Confidence level of defect region boundaries ; for crack masking Extract the centerline and calculate the crack length based on pixels. The average crack width was calculated using the sliding window vertical width measurement method. ; Calculate the number of cracks Number of peeling pieces and crack boundary roughness .

4. The method for quantitative non-destructive monitoring of aircraft concrete structure safety as described in claim 3, characterized in that, In step S30, boundary roughness The "torsionalism" and "nonlinearity" used to effectively characterize crack edges are image geometric feature parameters for judging the degree of structural damage evolution. Their calculation steps specifically include: using Canny edge detection to mask the crack. Edge extraction is performed to obtain crack edge pixels, and the image is divided into multiple... Count the number of grid cell boxes of different sizes that can cover the pixels at the edge of the crack. The size of the mesh cell Box is changed and the iteration is repeated. During the iteration of the mesh cell Box size, the boundary roughness is obtained by fitting logarithmic coordinates. Boundary roughness ,in, The value ranges from 1 to 2. This indicates that the crack is straight; if This indicates that the crack has many twists and turns, and the risk is higher.

5. The method for quantitative non-destructive monitoring of aircraft concrete structure safety as described in claim 1, characterized in that, Step S40, which involves calculating the structural risk index SFI based on the defect parameters using partial least squares regression, specifically includes: normalizing the defect parameters to form a feature vector. Input feature vector The input is fed into a trained partial least squares regression model to obtain the structural risk index SFI. ,in, The bounding box of the defect region k Confidence level of defect region boundaries Crack length Average crack width Number of cracks Number of peeling pieces and boundary roughness Variables after normalization and SFI represents the regression coefficients obtained through training with historical crack and spalling labeled data. SFI represents the degree of structural risk; a higher SFI value indicates more severe structural degradation.

6. A non-destructive testing system for the safety quantification of concrete structures for aircraft, applied to the non-destructive testing method for the safety quantification of concrete structures for aircraft as described in any one of claims 1 to 5, characterized in that, The aircraft-based concrete structure safety quantitative non-destructive monitoring system includes: The 3D modeling and path optimization module is used to acquire 3D laser point cloud data or BIM model data of the target concrete structure, extract surface morphology data of the target concrete structure, and discretize the surface morphology data into a set of facets S. Based on the angle between the normal vector of each facet in the set S and the preset viewing angle of the aircraft camera, the minimum visible distance, and occlusion conditions, a sequence of candidate waypoints is set. Construct a multi-objective path optimization model, in which the objective function includes the first objective function. Second objective function and the third objective function The multi-objective path optimization model is solved by combining a non-dominated sorting genetic algorithm, and a set of optimized waypoint sequences is output. ; The image acquisition and optimization module is used to optimize the waypoint sequence. The image at time t was obtained by taking pictures using an aircraft. According to the captured images Perform an image optimization task to obtain an optimized set of captured images; Among them, based on the captured images The steps for performing image optimization tasks to obtain an optimized set of captured images specifically include: obtaining the angular velocity of the inertial measurement unit of the spacecraft at time t. and angular acceleration and total amplitude of attitude disturbance Regarding the captured images Image sharpness metrics, including Laplacian variance, are derived from pixel extraction. Average brightness and exposure offset And combined with angular velocity and angular acceleration and total amplitude of attitude disturbance Comprehensive calculation of image reliability score When image reliability score Less than the preset image reliability score threshold At that time, mark the image For unreliable images, record time t as the location to be reshot; perform a reshot task based on the location to be reshot, combining the captured images. Generate an optimized set of captured images; Image reliability score The calculation formula is: ,in, For a moment Capture images Image reliability score; For angular velocity and angular acceleration The calculated total amplitude of attitude disturbance during image capture; The Laplacian variance in the image sharpness index; This represents the normalized lower bound of the Laplacian variance of image sharpness; This represents the normalized upper boundary of the Laplacian variance of image sharpness; image The average brightness; This is a preset ideal brightness exposure reference point; Exposure offset; To take the exponential function; and Let be the weighting coefficient, satisfying It is determined by experience or training data; These are the sensitivity attenuation factors that control the total amplitude of attitude disturbance and the average brightness, respectively. The defect identification and parameter extraction module is used to extract crack and peeling defect regions from an optimized set of captured images using model inference, and outputs defect parameters; among which, the defect parameters include the bounding box of the defect region k. Confidence level of defect region boundaries Crack length Average crack width Number of cracks Number of peeling pieces and boundary roughness ; The structural risk calculation module is used to calculate the structural risk index (SFI) based on defect parameters using partial least squares regression. The risk warning and upload module is used to generate a non-destructive monitoring risk report based on the structural risk index SFI, defect parameters, and optimized image set when the SFI exceeds the preset high-risk threshold. The report is then uploaded to the cloud database using the aircraft's WIFI module and preset API interface.

7. A non-destructive monitoring device for the safety quantification of concrete structures in aircraft, characterized in that, The aircraft-based concrete structure safety quantitative non-destructive monitoring device includes: a memory, a processor, and an aircraft-based concrete structure safety quantitative non-destructive monitoring program stored in the memory and executable on the processor. When the aircraft-based concrete structure safety quantitative non-destructive monitoring program is executed by the processor, it implements the aircraft-based concrete structure safety quantitative non-destructive monitoring method according to any one of claims 1 to 5.

8. A computer program product, characterized in that, The computer program product includes a non-destructive monitoring program for the safety of concrete structures based on aircraft. When the non-destructive monitoring program for the safety of concrete structures based on aircraft is executed by a processor, it implements a non-destructive monitoring method for the safety of concrete structures based on aircraft as described in any one of claims 1 to 5.

Images