Unmanned aerial vehicle full machine symmetry levelness detection method based on 3D scanning and electronic equipment

Abstract

The application discloses a kind of unmanned aerial vehicle symmetry levelness detection methods and electronic equipment based on 3D scanning.The method can include: for the unmanned aerial vehicle to be measured, build the measurement system based on 3D scanning;Based on 3D scanning, the three-dimensional point cloud data of unmanned aerial vehicle surface is collected and the global fusion of local point cloud is carried out, and the original three-dimensional point cloud data set covering the whole machine contour is generated;Based on the original three-dimensional point cloud data set, the registration of measured point cloud and theoretical numerical model is carried out, and the three-dimensional coordinates of feature points are extracted;Based on the three-dimensional coordinates of feature points, the symmetry levelness of the whole aircraft is evaluated.The present application can realize the automation, high-precision extraction of unmanned aerial vehicle horizontal measurement feature points and the accurate evaluation of levelness.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) measurement technology, and more specifically, to a method and electronic device for detecting the overall symmetry level of a UAV based on 3D scanning. Background Technology

[0002] The assembly geometric accuracy of unmanned aerial vehicles (UAVs) directly affects their flight aerodynamic performance and safety. During final assembly and scheduled maintenance, measuring the overall symmetry and levelness of the aircraft is a crucial step in assessing its assembly quality. Traditional levelness measurement methods primarily rely on operators using tools such as total stations, theodolites, laser levels, or plumb bobs, combined with physical markers pre-drilled on the aircraft surface in specific forms (e.g., special rivets, screws, punch marks, drill holes, etc.) for manual alignment, laser ablation, and data reading to calculate the overall symmetry and levelness of the aircraft.

[0003] However, this traditional manual measurement method has shortcomings in practical applications. First, the measurement process is cumbersome and has a low degree of automation. Manually finding reference points and reading by eye are prone to human error, resulting in low accuracy and susceptibility to external environmental interference. Second, as modern medium- and short-range UAVs increasingly adopt composite materials and stealth coatings, aerodynamic and stealth requirements prevent the preservation of a large number of prominent physical measurement reference points on the fuselage surface, making traditional "point finding" difficult. Finally, for medium- and large-scale UAVs, traditional single-point contact or semi-contact measurement methods are insufficient to obtain high-precision global spatial geometric distribution information. If the pre-set measurement reference points are misidentified by human intervention, it will affect the accuracy of the overall quality assessment.

[0004] Therefore, it is necessary to develop a method and electronic equipment for detecting the overall symmetry level of a UAV based on 3D scanning.

[0005] The information disclosed in the background section of this invention is intended only to enhance the understanding of the general background of this invention, and should not be construed as an admission or in any way implying that such information constitutes prior art known to those skilled in the art. Summary of the Invention

[0006] This invention proposes a method and electronic device for detecting the levelness of the entire UAV based on 3D scanning, which can realize the automation, high-precision extraction of level measurement feature points of the UAV, and accurate evaluation of levelness.

[0007] In a first aspect, embodiments of this disclosure provide a method for detecting the overall symmetry levelness of a UAV based on 3D scanning, including: A 3D scanning-based measurement system was built for the drone under test; Based on 3D scanning, three-dimensional point cloud data of the UAV surface is collected and global fusion of local point clouds is performed to generate an original three-dimensional point cloud dataset {Cloud_S} that covers the entire outline of the UAV. Based on the original 3D point cloud dataset, the measurement point cloud and the theoretical digital model are registered, and the 3D coordinates of feature points are extracted. The symmetry level of the entire aircraft is evaluated based on the three-dimensional coordinates of feature points.

[0008] Preferably, the measurement system built for the UAV under test includes: The drone under test is placed in a flat location, and several sets of targets for multi-station stitching are arranged on the ground area below and around the drone, and an optical tracker is set up. Turn on the 3D scanning equipment and graphics workstation to collect point cloud data with absolute coordinate attributes, retrieve the theoretical CAD model of the UAV and the coordinates of all preset feature points. The feature points include two types: one type is the reference feature points used to construct the measurement coordinate system, and the other type is the measurement feature points used to evaluate the symmetry and levelness of the entire UAV.

[0009] Preferably, the registration of the measured point cloud with the theoretical numerical model based on the original 3D point cloud dataset includes: The UAV 3D CAD model is converted into a high-density triangular mesh format using a mesh generation algorithm. Discrete 3D point data is extracted from the surface of the triangular mesh to generate a theoretical model point cloud dataset {Cloud_E}. The original 3D point cloud dataset {Cloud_S} is downsampled using a voxel mesh filter. In the 3D space, tiny voxels of fixed size are divided. The geometric centroid of all original data points in each voxel is calculated and replaced with the centroid to replace all points in the voxel, and the sparse point cloud dataset {Cloud_S} is output. The iterative nearest point algorithm is used to register the sparse point cloud dataset {Cloud_s} with the theoretical model point cloud dataset {Cloud_E} to obtain the optimized point cloud data {Cloud_SS}.

[0010] Preferably, the iterative nearest-point algorithm is used to register the sparse point cloud dataset {Cloud_s} with the theoretical model point cloud dataset {Cloud_E} to obtain optimized point cloud data {Cloud_SS}, including: The sparse point cloud dataset is {Cloud_s}= The theoretical model point cloud dataset is {Cloud_E}= ; The objective function is constructed as follows: , Solving the objective function yields the optimal spatial rotation R and translation parameters T, which are then applied to the original 3D point cloud dataset {Cloud_S} to obtain the optimized point cloud data {Cloud_SS}.

[0011] Preferably, the steps for solving the objective function include: Find the nearest pair of points: For each measurement point in the point cloud P By using a kd-tree accelerated search algorithm, the corresponding point with the closest Euclidean distance is found in the point cloud Q of the CAD model. ; Matrix solution: Solve the objective function using the singular value decomposition method, and calculate the spatial rotation R and translation parameters T under the current iteration step; Coordinate transformation: Apply the obtained R and T to the measured point cloud P to perform a rigid transformation, resulting in the updated point cloud: ; Iterative convergence: Repeat the above steps until the change in the objective function is less than the set convergence threshold, and obtain the optimal spatial rotation R and translation parameters T.

[0012] Preferably, extracting the three-dimensional coordinates of feature points includes: Retrieve the three-dimensional design coordinates of each feature point in the theoretical model, and combine the size of the feature object with a preset tolerance margin to construct a three-dimensional directed bounding box (OBB) that surrounds the feature point in the theoretical coordinate system. Using the predefined OBB spatial boundary equation as a spatial filter, it is applied to optimize the point cloud data {Cloud_SS} for adaptive cutting, selecting a local point cloud subset containing each feature point. Based on a subset of local point clouds, a method combining local mathematical model fitting and analytical solution is used to extract the three-dimensional coordinates of feature points.

[0013] Preferably, based on a subset of local point clouds, a method combining local mathematical model fitting and analytical solution is used to extract the three-dimensional coordinates of feature points, including: For a subset of local point clouds, construct a covariance matrix based on its k-neighbor point set. : , in, Let be the geometric centroid of the neighborhood point set; Based on principal component analysis, the covariance matrix C is decomposed into eigenvalues. The eigenvector corresponding to the smallest eigenvalue is used as the surface normal vector of that point. The point cloud at the edge contour is selected according to the deflection angle of the normal vector space and the degree of curvature change. For the selected set of feature candidate points, a spatial analytical mathematical model matching the local geometry is constructed using the least squares method. The solution objective is to minimize the fitting residuals of all sampling points to the mathematical surface model and obtain the three-dimensional coordinates of the feature points.

[0014] Preferably, the assessment of the overall symmetry level of the aircraft based on the three-dimensional coordinates of feature points includes: Establish a measurement reference coordinate system and perform coordinate transformation on the horizontal measurement points; The symmetry deviation rate is calculated based on the three-dimensional coordinates of feature points to achieve the overall symmetry level assessment of the aircraft.

[0015] Preferably, the symmetry deviation rate is: , Where S is the symmetry deviation rate. L left The orthogonal distance from the three-dimensional coordinates of the corresponding end point of the component on the left side of the aircraft to the reference plane. L right The orthogonal distance from the three-dimensional coordinates of the corresponding end point of the component on the right side of the aircraft to the reference plane, which is the XY plane formed by the longitudinal centerline of the fuselage and the vertical axis.

[0016] Secondly, embodiments of this disclosure also provide an electronic device, the electronic device comprising: Memory, which stores executable instructions; A processor that executes the executable instructions in the memory to implement the 3D scanning-based method for detecting the overall symmetry and levelness of a UAV.

[0017] Its beneficial effects are as follows: This invention features non-contact measurement, high accuracy, high automation, and large amount of information acquired across the entire range. It can be widely applied to high-precision measurement of the external dimensions and levelness of various types of UAVs for medium and short range applications. Furthermore, it eliminates the need for physical markers, meeting the measurement requirements of special fuselage surfaces such as stealth coatings and composite materials. It avoids human reading errors and greatly increases the automation and efficiency of UAV shape and position detection.

[0018] The methods and apparatus of the present invention have other features and advantages that will be apparent from or will be set forth in detail in the accompanying drawings and following detailed description, which together serve to explain the particular principles of the invention. Attached Figure Description

[0019] The above and other objects, features and advantages of the present invention will become more apparent from the more detailed description of exemplary embodiments of the invention in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same parts.

[0020] Figure 1 A flowchart illustrating the steps of a 3D scanning-based method for detecting the overall symmetry level of a UAV according to an embodiment of the present invention is shown.

[0021] Figure 2 A schematic diagram of the overall levelness detection of a drone according to an embodiment of the present invention is shown.

[0022] Figure 3 A schematic diagram of a UAV measurement coordinate system according to an embodiment of the present invention is shown.

[0023] Figure 4 A layout diagram of a three-dimensional scanning measurement system according to an embodiment of the present invention is shown. Detailed Implementation

[0024] Preferred embodiments of the invention will now be described in more detail. While preferred embodiments of the invention are described below, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0025] To facilitate understanding of the solutions and effects of the embodiments of the present invention, two specific application examples are given below. Those skilled in the art should understand that these examples are merely for the purpose of understanding the present invention, and any specific details therein are not intended to limit the present invention in any way.

[0026] Example 1

[0027] Figure 1 A flowchart illustrating the steps of a 3D scanning-based method for detecting the overall symmetry level of a UAV according to an embodiment of the present invention is shown.

[0028] like Figure 1 As shown, the method for detecting the overall symmetry levelness of a UAV based on 3D scanning includes: Step 101: Build a 3D scanning-based measurement system for the UAV under test; Step 102: Collect 3D point cloud data of the UAV surface based on 3D scanning and perform global fusion of local point clouds to generate an original 3D point cloud dataset {Cloud_S} that covers the outline of the entire UAV. Step 103: Based on the original 3D point cloud dataset, register the measured point cloud with the theoretical digital model and extract the 3D coordinates of the feature points; Step 104: Evaluate the overall symmetry level of the aircraft based on the three-dimensional coordinates of the feature points.

[0029] In one example, setting up a measurement system for the drone under test includes: The drone under test is placed in a flat location, and several sets of targets for multi-station stitching are arranged on the ground area below and around the drone, and an optical tracker is set up. Turn on the 3D scanning equipment and graphics workstation to collect point cloud data with absolute coordinate attributes, retrieve the theoretical CAD model of the UAV and the coordinates of all preset feature points. The feature points include two types: one type is the reference feature points used to construct the measurement coordinate system, and the other type is the measurement feature points used to evaluate the symmetry and levelness of the entire UAV.

[0030] In one example, the registration of the measured point cloud with the theoretical digital model based on the original 3D point cloud dataset includes: The UAV 3D CAD model is converted into a high-density triangular mesh format using a mesh generation algorithm. Discrete 3D point data is extracted from the surface of the triangular mesh to generate a theoretical model point cloud dataset {Cloud_E}. The original 3D point cloud dataset {Cloud_S} is downsampled using a voxel mesh filter. In the 3D space, tiny voxels of fixed size are divided. The geometric centroid of all original data points in each voxel is calculated and replaced with the centroid to replace all points in the voxel, and the sparse point cloud dataset {Cloud_S} is output. The iterative nearest point algorithm is used to register the sparse point cloud dataset {Cloud_s} with the theoretical model point cloud dataset {Cloud_E} to obtain the optimized point cloud data {Cloud_SS}.

[0031] In one example, the iterative nearest-point algorithm is used to register the sparse point cloud dataset {Cloud_s} with the theoretical model point cloud dataset {Cloud_E}, resulting in optimized point cloud data {Cloud_SS}, which includes: The sparse point cloud dataset is {Cloud_s}= The theoretical model point cloud dataset is {Cloud_E}= ; The objective function is constructed as follows: , Solving the objective function yields the optimal spatial rotation R and translation parameters T, which are then applied to the original 3D point cloud dataset {Cloud_S} to obtain the optimized point cloud data {Cloud_SS}.

[0032] In one example, the steps to solve the objective function include: Find the nearest pair of points: For each measurement point in the point cloud P By using a kd-tree accelerated search algorithm, the corresponding point with the closest Euclidean distance is found in the point cloud Q of the CAD model. ; Matrix solution: Solve the objective function using the singular value decomposition method, and calculate the spatial rotation R and translation parameters T under the current iteration step; Coordinate transformation: Apply the obtained R and T to the measured point cloud P to perform a rigid transformation, resulting in the updated point cloud: ; Iterative convergence: Repeat the above steps until the change in the objective function is less than the set convergence threshold, and obtain the optimal spatial rotation R and translation parameters T.

[0033] In one example, extracting the 3D coordinates of feature points includes: Retrieve the three-dimensional design coordinates of each feature point in the theoretical model, and combine the size of the feature object with a preset tolerance margin to construct a three-dimensional directed bounding box (OBB) that surrounds the feature point in the theoretical coordinate system. Using the predefined OBB spatial boundary equation as a spatial filter, it is applied to optimize the point cloud data {Cloud_SS} for adaptive cutting, selecting a local point cloud subset containing each feature point. Based on a subset of local point clouds, a method combining local mathematical model fitting and analytical solution is used to extract the three-dimensional coordinates of feature points.

[0034] In one example, based on a subset of local point clouds, a method combining local mathematical model fitting and analytical solution is used to extract the 3D coordinates of feature points, including: For a subset of local point clouds, construct a covariance matrix based on its k-neighbor point set. : , in, Let be the geometric centroid of the neighborhood point set; Based on principal component analysis, the covariance matrix C is decomposed into eigenvalues. The eigenvector corresponding to the smallest eigenvalue is used as the surface normal vector of that point. The point cloud at the edge contour is selected according to the deflection angle of the normal vector space and the degree of curvature change. For the selected set of feature candidate points, a spatial analytical mathematical model matching the local geometry is constructed using the least squares method. The solution objective is to minimize the fitting residuals of all sampling points to the mathematical surface model and obtain the three-dimensional coordinates of the feature points.

[0035] In one example, evaluating the overall symmetry levelness of an aircraft based on the 3D coordinates of feature points includes: Establish a measurement reference coordinate system and perform coordinate transformation on the horizontal measurement points; The symmetry deviation rate is calculated based on the three-dimensional coordinates of feature points to achieve the overall symmetry level assessment of the aircraft.

[0036] In one example, the symmetry deviation rate is: , Where S is the symmetry deviation rate. L left The orthogonal distance from the three-dimensional coordinates of the corresponding end point of the component on the left side of the aircraft to the reference plane. L right The orthogonal distance from the three-dimensional coordinates of the corresponding end point of the component on the right side of the aircraft to the reference plane, which is the XY plane formed by the longitudinal centerline of the fuselage and the vertical axis.

[0037] Figure 2 A schematic diagram of the overall levelness detection of a drone according to an embodiment of the present invention is shown.

[0038] Specifically, such as Figure 2 As shown, this invention provides a method for detecting the overall symmetry levelness of a UAV based on 3D scanning, including the following processing steps: S1: Setup and initialization of the measurement system; The drone under test is parked in a designated flat location, such as a test platform with six degrees of freedom attitude adjustment. According to the pre-determined station switching plan for this model, several sets of targets for multi-station splicing are arranged on the ground area below and around the drone.

[0039] Place the optical tracker in a suitable position next to the drone and adjust its attitude to effectively cover the 3D scanner and the corresponding ground-based target during scanning. Turn on the 3D scanning equipment and graphics workstation, open the scanner's data acquisition software system, and establish a hardware synchronization clock and high-speed data communication link between the devices to ensure that the system can normally acquire point cloud data with absolute coordinate attributes.

[0040] Figure 3 A schematic diagram of a UAV measurement coordinate system according to an embodiment of the present invention is shown.

[0041] Retrieve the theoretical CAD model of the UAV and the coordinates of all preset feature points. These feature points fall into two categories: one category consists of reference feature points used to construct the measurement coordinate system, such as the UAV measurement coordinate system. Figure 3 As shown, another type is the measurement feature points used for evaluating the overall symmetry and levelness of the machine. The selection of reference feature points follows the principles of good rigidity, small deformation, and large distance between points.

[0042] S2: Acquisition of 3D point cloud data of the UAV surface; Figure 4A layout diagram of a three-dimensional scanning measurement system according to an embodiment of the present invention is shown.

[0043] Perform surface point cloud data collection for drones, such as Figure 4 As shown, the tester held a 3D scanner and moved it smoothly across the surface of the drone. Using dynamic tracking technology, without attaching markers, the scanner projected laser lines to collect raw 3D point cloud data of the drone's surface. During the data collection process, it was necessary to ensure that the point cloud completely covered the preset feature point locations, and that the point cloud density met the requirements for subsequent high-precision measurement and evaluation.

[0044] Perform multi-angle and station-switching scans. For medium and large UAVs, when the single-station field of view of the optical tracker cannot cover the entire aircraft, the operator needs to work with a mobile ladder to perform full-coverage scans of various parts such as the wings and fuselage, and move the position of the optical tracker to switch stations for measurement as needed to expand the scanning coverage.

[0045] Global fusion of local point clouds is performed. During the measurement process, the system automatically identifies pre-placed stitching targets on the ground, extracts the spatial relative coordinates of common target points at different stations using algorithms, and calculates the rigid transformation matrix between stations. This matrix is ​​then used to automatically stitch together and unify the fragmented local point clouds collected from each station into a unified coordinate system for the entire machine, ultimately generating a complete high-density 3D point cloud dataset {Cloud_S} covering the entire machine's outline.

[0046] S3: Registration of measured point cloud with theoretical digital model and extraction of feature points; Point cloud extraction from the UAV theoretical model. First, the system uses a mesh generation algorithm to convert the UAV's 3D CAD model into a high-density triangular mesh format, where the mesh step size or patch size needs to be finely set according to the preset measurement accuracy requirements. Then, using sampling algorithms such as uniform sampling or Poisson disk sampling, discrete 3D point data is extracted from the surface of the triangular mesh to generate a theoretical model point cloud dataset, denoted as {Cloud_E}. This point cloud dataset will serve as the mathematical benchmark for subsequent spatial registration and feature point recognition.

[0047] Downsampling of scanned point cloud data. To improve the computational efficiency of registration of massive point cloud data, the system performs downsampling processing on the original 3D point cloud dataset {Cloud_S} using a voxel mesh filter. Fixed-size voxels are divided in 3D space, and the geometric centroid of all original data points within each voxel is calculated. This centroid is then used to replace all points within that voxel. This effectively removes redundant data while accurately preserving the UAV's shape features, resulting in a significantly reduced sparse point cloud dataset, denoted as {Cloud_s}. It is important to emphasize that this sparse point cloud {Cloud_s} is only used for the subsequent spatial rigid transformation matrix operations of the ICP algorithm and does not participate in subsequent feature extraction.

[0048] Registration of the scanned point cloud with the theoretical model point cloud. Since the actual scanned point cloud and the CAD model are in different digital coordinate systems, this step uses the Iterative Closest Point (ICP) algorithm to register {Cloud_s} and {Cloud_E}. The specific steps are as follows: First, let the point cloud set after S32 preprocessing be... .

[0049] The point cloud set obtained by discretizing and sampling the surface of the UAV CAD theoretical model is as follows .

[0050] Secondly, the objective function is constructed. This registration process essentially involves solving for the optimal spatial rotation matrix R and translation vector t, such that after a rigid transformation of the measured point cloud P, the mean square error between it and the target point cloud Q is minimized. Its mathematical model is as follows: .

[0051] The specific solution process is as follows: 1. Find the nearest point pair: For each measurement point in the point cloud P By using a kd-tree accelerated search algorithm, the corresponding point with the closest Euclidean distance is found in the point cloud Q of the CAD model. .

[0052] 2. Matrix solution: Solve the above objective function E(R,t) using the singular value decomposition (SVD) method, and calculate the rotation matrix R and translation vector t under the current iteration step.

[0053] 3. Coordinate Transformation: Apply the obtained R and t to the measured point cloud P to perform a rigid transformation, resulting in the updated point cloud: .

[0054] 4. Iterative convergence: Repeat steps 1 to 3 above until the change in the objective function E(R,t) is less than the convergence threshold set by the system. Right now At this point, the system determines that the point cloud and the CAD model are accurately registered and their coordinate systems are completely aligned. The optimal spatial rotation R and translation parameters T are then determined.

[0055] After the solution converges, the system applies the obtained optimal rotation and translation parameters to the original 3D point cloud dataset {Cloud_S} to perform a global rigid transformation, thereby aligning the original measured point cloud to the theoretical model coordinate system {C-X0Y0Z0}, and the resulting optimized point cloud data is denoted as {Cloud_SS}.

[0056] To reduce the computational complexity of processing massive point cloud data and effectively eliminate interference from surrounding irrelevant data, the system performs ROI segmentation. Specifically, the system first retrieves the 3D design coordinates of each feature point in the theoretical model, and then, combined with the size of the feature object and a preset tolerance margin, constructs a 3D directed bounding box (OBB) enclosing the feature point in the theoretical coordinate system. Since the coordinate system has been completely unified, the system can directly use the pre-defined OBB spatial boundary equation as a spatial filter to adaptively segment the coordinate-transformed point cloud {Cloud_SS}, selecting a local point cloud subset containing each feature point.

[0057] Because the point cloud obtained by 3D scanning is spatially discrete, the actual geometric feature points of the physical structure often cannot be sampled precisely during scanning, resulting in sampling gap errors. To eliminate this error, this step uses a method combining local mathematical model fitting and analytical solution to extract the 3D coordinates of the feature points. Specifically: Feature region identification and candidate point selection: For a subset of local point clouds, construct a covariance matrix based on its k-neighbor point set. : , in, Let C be the geometric centroid of the neighborhood point set. Based on principal component analysis (PCA), eigenvalue decomposition is performed on the covariance matrix C, and the eigenvector corresponding to the smallest eigenvalue is obtained as the surface normal vector of that point. The system selects point clouds at the edge contours based on the deflection angle of the normal vector space and the degree of curvature abrupt change.

[0058] Local spatial analytical model fitting: For the selected set of feature candidate points, a spatial analytical mathematical model matching the local geometry is constructed using the least squares method. Taking an analytical surface as an example, its equation can be expressed as: , The objective is to minimize the fitting residuals from all sampling points to the mathematical surface model, i.e.: .

[0059] After obtaining the local analytical mathematical model, the system performs theoretical analytical calculations based on the continuous mathematical equations, thereby breaking through the grid resolution limitation of discrete point clouds, obtaining high-precision three-dimensional coordinates of feature points, and outputting them to the subsequent evaluation stage.

[0060] S4: Assessment of aircraft-wide symmetry levelness based on feature point coordinates.

[0061] Construct a measurement reference coordinate system. Rigid measurement reference point set. For extraction and verification, the system first extracts multiple specific marker points (such as rivets) distributed in parts of the fuselage structure with high rigidity and small deformation from the registered local point cloud subset, based on the distribution principles specified in the industry standard "HB / Z 103-2023 Aircraft Level Measurement Tolerances". These points are used as leveling reference points. The physical distance between the selected reference points should be as large as possible to reduce the amplification error in attitude calculation. The system automatically obtains the design coordinate values ​​of these reference points under the CAD theoretical model. And the actual measured coordinate values ​​obtained and corrected through 3D scanning. The system then uses a weighted least squares matching algorithm to iteratively optimize and solve for the optimal rotation matrix. Translation vector The objective function is to minimize the weighted sum of squared residuals J between all measurement benchmarks and their theoretical design locations. , in Based on the pre-defined weighting coefficients according to the rigidity and reliability of the reference point location, a unique actual measurement reference coordinate system {CX} is established for the entire machine by solving this matrix. M Y M Z M}

[0062] Perform coordinate transformation of the horizontal measurement points. This is done using a rotation matrix. With translation vector The coordinates of all measured feature points are uniformly transformed from the theoretical model coordinate system {C-X0Y0Z0} to the actual measurement reference coordinate system {CX}. M Y M Z M This provides a unified standard for calculating the geometric tolerances and levelness of subsequent components.

[0063] An evaluation of the overall aircraft levelness is conducted. First, coordinate axes are defined and orthogonally projected onto the feature point space. In this embodiment, the X-axis is defined as the aircraft's longitudinal direction (from nose to tail), the Y-axis as the aircraft's vertical direction (altitude), and the Z-axis as the aircraft's lateral direction (wingspan). According to the definition in the standard "HB / Z 103-2023 Aircraft Level Measurement Tolerances," the system projects the selected feature points onto the reference coordinate axes. Specifically, the system automatically calculates the coordinate differences of the relevant feature points on the altitude axis (Y-axis), longitudinal axis (X-axis), or lateral axis (Z-axis), thereby obtaining the positional deviation (e.g., altitude deviation Δ). Y =| Y 1- Y 2|, Lateral deviation Δ Z =| Z 1- Z 2|) and projection distance (L (i.e., the distance between two points on a reference plane or coordinate axis, such as the longitudinal projection distance) L X =| X 1 -X 2|, Lateral projection distance L Z =| Z 1- Z 2|).

[0064] Overall aircraft symmetry deviation calculation: Using the acquired distance and deviation data, a spatial trigonometric geometric function model is called to solve the symmetry deviation rate. An XY plane, formed by the fuselage longitudinal centerline and the vertical axis, is established as the theoretical symmetry reference plane. The three-dimensional coordinates of the corresponding endpoints on the left side of the aircraft (such as the tip of the left horizontal stabilizer) are extracted, and their orthogonal distances to the reference plane are calculated. L left Similarly, obtain the right-side orthogonal distance. L right The symmetry deviation rate is calculated using a per mille formula: .

[0065] After completing the entire calculation process, the system can evaluate the overall symmetry level of the UAV. The tolerance standard for evaluation can be found in "HB / Z 103-2023 Aircraft Level Measurement Tolerance".

[0066] Example 2

[0067] This disclosure provides an electronic device, comprising: a memory storing executable instructions; and a processor executing the executable instructions in the memory to implement the above-described method for detecting the overall symmetry and levelness of a UAV based on 3D scanning.

[0068] An electronic device according to an embodiment of the present disclosure includes a memory and a processor.

[0069] This memory is used to store non-transitory computer-readable instructions. Specifically, the memory may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may, for example, include random access memory (RAM) and / or cache memory. The non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, etc.

[0070] The processor may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions. In one embodiment of this disclosure, the processor is used to execute computer-readable instructions stored in the memory.

[0071] Those skilled in the art will understand that, in order to solve the technical problem of how to achieve a good user experience, this embodiment may also include well-known structures such as communication buses and interfaces, and these well-known structures should also be included within the protection scope of this disclosure.

[0072] For a detailed description of this embodiment, please refer to the corresponding descriptions in the foregoing embodiments, which will not be repeated here.

[0073] Those skilled in the art should understand that the above description of the embodiments of the present invention is only intended to illustrate the beneficial effects of the embodiments of the present invention, and is not intended to limit the embodiments of the present invention to any of the examples given.

[0074] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.

Claims

1. A method for detecting the overall symmetry levelness of a UAV based on 3D scanning, characterized in that, include: A 3D scanning-based measurement system was built for the drone under test; Based on 3D scanning, three-dimensional point cloud data of the UAV surface is collected and global fusion of local point clouds is performed to generate an original three-dimensional point cloud dataset {Cloud_S} that covers the entire outline of the UAV. Based on the original 3D point cloud dataset, the measurement point cloud and the theoretical digital model are registered, and the 3D coordinates of feature points are extracted. The symmetry level of the entire aircraft is evaluated based on the three-dimensional coordinates of feature points.

2. The method for detecting the overall symmetry levelness of a UAV based on 3D scanning according to claim 1, characterized in that, The measurement system built for the drone under test includes: The drone under test is placed in a flat location, and several sets of targets for multi-station stitching are arranged on the ground area below and around the drone, and an optical tracker is set up. Turn on the 3D scanning equipment and graphics workstation to collect point cloud data with absolute coordinate attributes, retrieve the theoretical CAD model of the UAV and the coordinates of all preset feature points. The feature points include two types: one type is the reference feature points used to construct the measurement coordinate system, and the other type is the measurement feature points used to evaluate the symmetry and levelness of the entire UAV.

3. The method for detecting the overall symmetry and levelness of a UAV based on 3D scanning according to claim 1, characterized in that, Registration of measured point clouds with theoretical numerical models based on the original 3D point cloud dataset includes: The UAV 3D CAD model is converted into a high-density triangular mesh format using a mesh generation algorithm. Discrete 3D point data is extracted from the surface of the triangular mesh to generate a theoretical model point cloud dataset {Cloud_E}. The original 3D point cloud dataset {Cloud_S} is downsampled using a voxel mesh filter. In the 3D space, tiny voxels of fixed size are divided. The geometric centroid of all original data points in each voxel is calculated and replaced with the centroid to replace all points in the voxel, and the sparse point cloud dataset {Cloud_S} is output. The iterative nearest point algorithm is used to register the sparse point cloud dataset {Cloud_s} with the theoretical model point cloud dataset {Cloud_E} to obtain the optimized point cloud data {Cloud_SS}.

4. The method for detecting the overall symmetry levelness of a UAV based on 3D scanning according to claim 3, characterized in that, The iterative nearest-point algorithm is used to register the sparse point cloud dataset {Cloud_s} with the theoretical model point cloud dataset {Cloud_E}, resulting in optimized point cloud data {Cloud_SS}, which includes: The sparse point cloud dataset is {Cloud_s}= The theoretical model point cloud dataset is {Cloud_E}= ; The objective function is constructed as follows: ， Solving the objective function yields the optimal spatial rotation R and translation parameters T, which are then applied to the original 3D point cloud dataset {Cloud_S} to obtain the optimized point cloud data {Cloud_SS}.

5. The method for detecting the overall symmetry and levelness of a UAV based on 3D scanning according to claim 4, characterized in that, The steps to solve the objective function include: Find the nearest pair of points: For each measurement point in the point cloud P By using a kd-tree accelerated search algorithm, the corresponding point with the closest Euclidean distance is found in the point cloud Q of the CAD model. ; Matrix solution: Solve the objective function using the singular value decomposition method, and calculate the spatial rotation R and translation parameters T under the current iteration step; Coordinate transformation: Apply the obtained R and T to the measured point cloud P to perform a rigid transformation, resulting in the updated point cloud: ; Iterative convergence: Repeat the above steps until the change in the objective function is less than the set convergence threshold, and obtain the optimal spatial rotation R and translation parameters T.

6. The method for detecting the overall symmetry levelness of a UAV based on 3D scanning according to claim 3, characterized in that, Extracting the 3D coordinates of feature points includes: Retrieve the three-dimensional design coordinates of each feature point in the theoretical model, and combine the size of the feature object with a preset tolerance margin to construct a three-dimensional directed bounding box (OBB) that surrounds the feature point in the theoretical coordinate system. Using the predefined OBB spatial boundary equation as a spatial filter, it is applied to optimize the point cloud data {Cloud_SS} for adaptive cutting, selecting a local point cloud subset containing each feature point. Based on a subset of local point clouds, a method combining local mathematical model fitting and analytical solution is used to extract the three-dimensional coordinates of feature points.

7. The method for detecting the overall symmetry and levelness of a UAV based on 3D scanning according to claim 6, characterized in that, Based on a subset of local point clouds, a method combining local mathematical model fitting and analytical solution is used to extract the 3D coordinates of feature points, including: For a subset of local point clouds, construct a covariance matrix based on its k-neighbor point set. : ， in, Let be the geometric centroid of the neighborhood point set; Based on principal component analysis, the covariance matrix C is decomposed into eigenvalues. The eigenvector corresponding to the smallest eigenvalue is used as the surface normal vector of that point. The point cloud at the edge contour is selected according to the deflection angle of the normal vector space and the degree of curvature change. For the selected set of feature candidate points, the least squares method is used to construct a spatial analytical mathematical model that matches the local geometry. The solution objective is to minimize the fitting residuals of all sampling points to the mathematical surface model and obtain the three-dimensional coordinates of the feature points.

8. The method for detecting the overall symmetry levelness of a UAV based on 3D scanning according to claim 1, characterized in that, The assessment of the overall symmetry level of an aircraft based on the three-dimensional coordinates of feature points includes: Establish a measurement reference coordinate system and perform coordinate transformation on the horizontal measurement points; The symmetry deviation rate is calculated based on the three-dimensional coordinates of feature points to achieve the overall symmetry level assessment of the aircraft.

9. The method for detecting the overall symmetry levelness of a UAV based on 3D scanning according to claim 8, characterized in that, Symmetry deviation rate: ， Where S is the symmetry deviation rate. L left The orthogonal distance from the three-dimensional coordinates of the corresponding end point of the component on the left side of the aircraft to the reference plane. L right The orthogonal distance from the three-dimensional coordinates of the corresponding end point of the component on the right side of the aircraft to the reference plane, which is the XY plane formed by the longitudinal centerline of the fuselage and the vertical axis.

10. An electronic device, characterized in that, The electronic device includes: Memory, which stores executable instructions; A processor that executes the executable instructions in the memory to implement the method for detecting the overall symmetry level of a UAV based on 3D scanning as described in any one of claims 1-9.

Images