Virtual acceptance method and device for steel pipe arch rib

Abstract

The invention provides a virtual acceptance method and device for a steel pipe arch rib, and relates to the technical field of bridge engineering construction. The method comprises the following steps: performing multidirectional scanning on a prefabricated steel pipe arch rib through a three-dimensional laser scanner pre-installed in a site to obtain multidirectional arch rib point cloud data; and constructing a BIM real scene model and a point cloud real scene model in sequence, and meanwhile, importing drawing data to construct a BIM design model and a point cloud design model. In the aspect of model construction, the real state of the arch rib can be highly restored, and complete mapping of the design and the actual model in the geometrical shape and information level can be achieved. During acceptance analysis, a BIM real scene and a design model are qualitatively analyzed to obtain a first shape difference result, a point cloud real scene and the design model are quantitatively analyzed to obtain a second shape difference result, qualitative and quantitative combination is achieved, the arch rib is deeply evaluated, the acceptance result is scientific and reliable, the defects that a traditional method is large in labor consumption, large in error and the like are overcome, and the acceptance efficiency and accuracy are improved; the engineering quality is ensured.

Description

Technical Field

[0001] The invention mainly relates to the technical field of bridge engineering construction, and in particular to a virtual acceptance method and device for steel pipe arch ribs. Background Art

[0002] The steel tube concrete arch bridge is a type of steel-concrete composite structure. It fills the steel tube with concrete and uses the radial constraint of the steel tube to limit the expansion of the compressed concrete, so that the concrete is in a three-dimensional compression state, thereby significantly improving the compressive strength of the concrete. The steel tube concrete arch bridge often uses a circular steel tube section because the circular section can better play the role of the steel tube as a hoop. When the span of the arch bridge is large, a truss section is generally used to enhance the bearing capacity of the arch bridge.

[0003] The production and acceptance of steel tubes for steel tube concrete arch bridges is a complex and rigorous process that requires strict compliance with relevant standards and specifications to ensure the quality and safety of the steel tubes.

[0004] The axis of the steel pipe is generally designed to be a catenary or parabola. The current processing technology is to use multiple sections of straight steel pipes to fit into a curved steel pipe. When making steel pipes, the cutting must be accurate and the diameter error of the finished pipe must be within the specified range. The splicing of the arch ribs should be carried out on a sample table, and the quality should be strictly guaranteed during welding.

[0005] In the traditional steel pipe acceptance process, a large number of feature points are often marked on the surface of the steel pipe, and then a total station is used for measurement. The size and line shape of the steel pipe are accepted by observing the coordinates of the feature points. This method has the following shortcomings: high labor consumption. The traditional acceptance method requires manual marking of a large number of feature points and manual operation of the total station for measurement, which is time-consuming and laborious; limited data. The feature point data measured by the total station can only reflect a small amount of feature point information on the surface of the steel pipe, and cannot fully reflect the overall size and shape of the steel pipe; large measurement errors. Due to the reliance on manual operation during the measurement process, human errors are prone to occur, especially in the process of marking feature points, setting up the total station and reading data, the error accumulation is more serious; inefficient and inaccurate. Due to the limited number of feature points and the long distance between each measurement point, the measurement accuracy and precision are low, and the overall geometric shape of the steel pipe cannot be effectively reflected.

[0006] With the technological development of 3D laser scanners, drone oblique photography and radar scanning, the improvement of algorithms and the increase in computing power, point cloud models can be generated efficiently, and the measurement accuracy is also increasing. Among them, 3D laser scanner scanning measurement can increase the accuracy to mm level, which can meet the accuracy requirements of steel pipe acceptance.

[0007] In view of this, it is urgent to develop a virtual acceptance method and device based on steel tube arch ribs. Summary of the invention

[0008] The technical problem to be solved by the present invention is to provide a virtual acceptance method and device for a steel tube arch rib in view of the deficiencies in the prior art.

[0009] The technical solution of the present invention to solve the above technical problem is as follows: A virtual acceptance method for steel pipe arch ribs comprises the following steps:

[0010] S1. Perform multi-directional scanning on the prefabricated steel tube arch ribs by using a 3D laser scanner pre-installed in the site to obtain multi-directional arch rib point cloud data;

[0011] S2. Construct a BIM real-scene model and a point cloud real-scene model according to the multi-directional arch rib point cloud data, and construct a BIM design model and a point cloud design model according to the prefabricated steel pipe arch rib drawing data, and merge the BIM real-scene model and the BIM design model into the same file, and merge the point cloud real-scene model and the point cloud design model into the same file;

[0012] S3. Performing a qualitative analysis on the BIM real-scene model and the BIM design model in the same file to obtain a first appearance difference result, and performing a quantitative analysis on the point cloud real-scene model and the point cloud design model in the same file to obtain a second appearance difference result, thereby completing the virtual acceptance of the steel tube arch rib.

[0013] Another technical solution of the present invention to solve the above technical problem is as follows: a virtual acceptance device for steel pipe arch ribs, comprising:

[0014] A data import module, used for importing multi-directional arch rib point cloud data, wherein the multi-directional arch rib point cloud data is obtained by scanning the prefabricated steel pipe arch rib in multiple directions by a three-dimensional laser scanner pre-installed in the site;

[0015] A model construction module, used to construct a BIM real-scene model and a point cloud real-scene model according to the multi-directional arch rib point cloud data, and to construct a BIM design model and a point cloud design model according to the prefabricated steel pipe arch rib drawing data, and to merge the BIM real-scene model and the BIM design model into the same file, and to merge the point cloud real-scene model and the point cloud design model into the same file;

[0016] The virtual acceptance module is used to perform a qualitative analysis on the BIM real-scene model and the BIM design model in the same file to obtain a first appearance difference result, and to perform a quantitative analysis on the point cloud real-scene model and the point cloud design model in the same file to obtain a second appearance difference result, thereby completing the virtual acceptance of the steel tube arch rib.

[0017] Another technical solution of the present invention to solve the above-mentioned technical problem is as follows: a virtual acceptance device for steel tube arch ribs, characterized in that it includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the virtual acceptance method for steel tube arch ribs as described above is implemented.

[0018] The beneficial effects of the present invention are as follows: in terms of model construction, a BIM real-scene model and a point cloud real-scene model are constructed based on multi-directional arch rib point cloud data, which can highly restore the real state of the prefabricated steel tube arch rib. At the same time, the BIM design model and the point cloud design model are constructed by importing the prefabricated steel tube arch rib drawing data, so that the actual constructed model and the design model have a basis for direct comparison. The model constructed in this way is not only accurate in geometric shape, but also can realize a complete mapping from design to practice at the information level, which is helpful to comprehensively evaluate the situation of the prefabricated steel tube arch rib at various stages; in terms of acceptance analysis, qualitative and quantitative analysis are combined to realize a comprehensive and in-depth evaluation of the prefabricated steel tube arch rib, making the acceptance results more scientific and reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic flow chart of a virtual acceptance method for a steel tube arch rib provided by an embodiment of the present invention;

[0020] Figure 2 A schematic connection diagram of the functional modules of the virtual acceptance device for steel tube arch ribs provided by an embodiment of the present invention;

[0021] Figure 3 A schematic diagram of a BIM real-scene model provided by an embodiment of the present invention;

[0022] Figure 4 A schematic diagram of a BIM design model provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0024] The traditional total station measurement method has high labor consumption, limited data, large measurement errors, and is inefficient and inaccurate. This method is based on three-dimensional laser scanning and BIM technology, with a high degree of automation, which reduces the errors caused by manual operation. Multi-directional point cloud data acquisition and precise model construction and analysis can complete the acceptance work more quickly and accurately, greatly improving the acceptance efficiency, while also improving the accuracy of the acceptance results, helping to timely discover and solve problems in the production process of prefabricated steel pipe arch ribs, and ensuring the quality of the project. The present invention is described in detail below through a number of embodiments.

[0025] The test in the embodiment is based on a bridge reconstruction project in a certain province. The bridge adopts a bottom-supported steel tube concrete tied arch bridge with a main span of 191m, a calculated span of 180m, a bridge deck width of 51.1m, a calculated rise of 40m, and a rise-span ratio of 1 / 4.5. The arch axis adopts a catenary, and the arch axis coefficient m=1.3. The arch rib is a steel tube concrete dumbbell-shaped structure. A single-span bridge has two arch ribs, and the arch ribs are 18.9m apart from each other in the transverse direction of the bridge. The single arch rib adopts a dumbbell-shaped section, the dumbbell height is 3.2m, the upper and lower chord diameters are 1300mm, and the wall thickness is 24-28mm.

[0026] Example 1: Figure 1 As shown, an embodiment of the present invention provides a virtual acceptance method for a steel tube arch rib, comprising the following steps:

[0027] S1. Perform multi-directional scanning on the prefabricated steel tube arch ribs by using a 3D laser scanner pre-installed in the site to obtain multi-directional arch rib point cloud data;

[0028] S2. Construct a BIM real-scene model and a point cloud real-scene model according to the multi-directional arch rib point cloud data, and construct a BIM design model and a point cloud design model according to the prefabricated steel pipe arch rib drawing data, and merge the BIM real-scene model and the BIM design model into the same file, and merge the point cloud real-scene model and the point cloud design model into the same file;

[0029] S3. Performing a qualitative analysis on the BIM real-scene model and the BIM design model in the same file to obtain a first appearance difference result, and performing a quantitative analysis on the point cloud real-scene model and the point cloud design model in the same file to obtain a second appearance difference result, thereby completing the virtual acceptance of the steel tube arch rib.

[0030] Specifically, S1, the prefabricated steel tube arch ribs are scanned by a three-dimensional laser scanner pre-installed in multiple positions on the site to obtain multi-directional arch rib point cloud data, including:

[0031] The 3D laser scanning technology is used to obtain the point cloud data of the prefabricated steel tube arch ribs in the prefabrication factory site, and a multi-scale point cloud data collection solution is adopted: a 3D laser scanner is used to scan the prefabricated steel tube arch ribs of the steel tube arch bridge placed in the prefabrication factory site, and data is collected from multiple angles such as the top of the component, the cross-section at both ends, the inner arc and the outer arc, to obtain a large amount of point cloud data containing 3D coordinate points, reflection intensity, color and other additional information.

[0032] The advantages of Example 1 are: in terms of model construction, the BIM real-scene model and the point cloud real-scene model are constructed based on the multi-directional arch rib point cloud data, which can highly restore the real state of the prefabricated steel tube arch rib. At the same time, the prefabricated steel tube arch rib drawing data is imported to construct the BIM design model and the point cloud design model, so that the actual constructed model and the design model have a basis for direct comparison. The model constructed in this way is not only accurate in geometric shape, but also can achieve a complete mapping from design to practice at the information level, which is helpful to comprehensively evaluate the situation of the prefabricated steel tube arch rib at various stages; in terms of acceptance analysis, the combination of qualitative and quantitative analysis realizes a comprehensive and in-depth evaluation of the prefabricated steel tube arch rib, making the acceptance results more scientific and reliable.

[0033] Example 2: Figure 3 , Figure 4 As shown, in S2, a BIM real scene model and a point cloud real scene model are constructed according to the multi-directional arch rib point cloud data, and a BIM design model and a point cloud design model are constructed according to the prefabricated steel pipe arch rib drawing data, and the BIM real scene model and the BIM design model are merged into the same file, and the point cloud real scene model and the point cloud design model are merged into the same file, including:

[0034] S2.1. Importing the multi-directional arch rib point cloud data into a point cloud analysis tool for data splicing to obtain point cloud models of multiple prefabricated arch rib segments, and virtually assembling the point cloud models of the multiple prefabricated arch rib segments to obtain a point cloud real-scene model of the prefabricated steel tube arch rib, and converting the point cloud real-scene model into a BIM real-scene model through a BIM design tool;

[0035] S2.2, importing the prefabricated steel tube arch rib drawing data, inputting the prefabricated steel tube arch rib drawing data into the BIM design tool to generate a BIM design model, and converting the BIM design model into a point cloud design model through the point cloud analysis tool;

[0036] S2.3. Import the BIM real-scene model and the BIM design model into a BIM design tool as the same BIM file to merge them into the same file; and import the point cloud real-scene model and the point cloud design model into the same file.

[0037] The advantage of Example 2 is that the multi-directional arch rib point cloud data is imported into the point cloud analysis tool for data splicing and virtual assembly to obtain a point cloud real-life model of the prefabricated steel tube arch rib, which is then converted into a BIM real-life model. This process can accurately restore the actual shape of the arch rib and ensure a high degree of consistency between the real-life model and the actual structure.

[0038] By inputting the prefabricated steel tube arch rib drawing data into the BIM design tool to generate a BIM design model, and then converting it into a point cloud design model, a rapid conversion from design drawings to models is achieved. The design model and the real-scene model are constructed based on the same data source (point cloud data or drawing data), which facilitates subsequent comparative analysis, improves the efficiency and accuracy of model construction, and lays the foundation for the smooth progress of virtual acceptance.

[0039] Embodiment 3: In S3, the BIM real scene model and the BIM design model in the same file are qualitatively analyzed to obtain a first appearance difference result, including:

[0040] S3.1. Overlap the BIM real-scene model and the BIM design model in the same file to obtain a common spatial coordinate system;

[0041] S3.2, based on the spatial coordinate system, respectively select the arch rib bottom plane in the BIM real scene model and the BIM design model, the arch rib bottom plane includes two upper and lower circles and the middle connecting part of the two circles, take the center of the upper circle in the arch rib bottom plane as the centroid and set the lengths of the three sides as fixed values ​​to construct the characteristic triangle corresponding to the BIM real scene model and the characteristic triangle corresponding to the BIM design model;

[0042] S3.3. Taking the characteristic triangle corresponding to the BIM real scene model and the three corners of the characteristic triangle corresponding to the BIM design model as three characteristic point pairs, and based on the principle of determining the position of a plane in space by three points that are not in the same straight line, determine the positions of the end planes of the BIM real scene model and the BIM design model in space through the three characteristic point pairs and make them coincide with each other;

[0043] S3.4. Calculate the difference between the overlapped BIM real-scene model and the BIM design model using a Boolean operation method, and obtain a first appearance difference result according to the difference.

[0044] Specifically, the real-life BIM model and the design BIM model are matched in the spatial coordinate system of this BIM file, so that the coordinates of the feature points of the real-life BIM model and the design BIM model in the spatial coordinate system of this BIM file completely overlap. The matching scheme of the real-life BIM model and the design BIM model is as follows: a triangle with a fixed length of three sides is proposed as a feature triangle, and the same feature triangle is made on the plane at one end of the real-life BIM model and the design BIM model (that is, the bottom plane of the arch rib, which is dumbbell-shaped and consists of two upper and lower circles and a connecting part in the middle), with the center of the upper circle as the centroid of the feature triangle, and the vertices of each corner of the two feature triangles are used as a pair of feature points (which can be understood as specific positions marked on the surface of the model). Based on the principle that three points that are not in the same straight line determine the position of a plane in space, three pairs of feature points can determine the position of the end planes of the real-life model and the design BIM model in space. If the positions of the three pairs of feature points in space are completely overlapped, the positions of the end planes of the real-life model and the designed BIM model in space will completely overlap, and thus the positions of the real-life model and the designed BIM model in space will completely overlap.

[0045] When the real-life model and the design BIM model are completely overlapped, Boolean operations are used to calculate the "difference" between the two to determine the difference in appearance between the design model and the actual model, that is, to conduct a qualitative analysis of the differences in shape, size, etc. between the two.

[0046] The advantage of Example 3 is that by importing the BIM real-scene model and the BIM design model into the same BIM file, characteristic triangles are constructed using the plane features at the bottom of the arch ribs, and the two models are accurately matched in the spatial coordinate system based on the overlap of characteristic point pairs, thereby ensuring the consistency of the models in spatial position and providing a prerequisite for subsequent accurate analysis of appearance differences.

[0047] By using Boolean operations to calculate the difference set of the overlapped models, the differences between the design model and the actual model in terms of shape, size, etc. can be clearly identified, so that the first appearance difference result can be obtained intuitively, which provides an effective qualitative basis for evaluating the quality of prefabricated steel tube arch ribs, helps to quickly determine whether the arch ribs meet the design requirements, and can effectively perform qualitative difference analysis.

[0048] Embodiment 4: In S3, the point cloud real scene model and the point cloud design model in the same file are quantitatively analyzed to obtain a second appearance difference result, including:

[0049] S3.5, selecting a plurality of point cloud pairs corresponding to the same position from the point cloud real scene model and the point cloud design model in the same file, and performing preliminary registration processing on the selected plurality of point cloud pairs by using the RANSAC random sampling consensus algorithm;

[0050] S3.6, perform precise registration processing on the point cloud real scene model and the point cloud design model after preliminary registration through ICP iterative closest point algorithm;

[0051] S3.7. Perform posture difference detection on the point cloud real scene model and the point cloud design model after precise registration through the KNN neighbor algorithm to obtain a second appearance difference result.

[0052] It should be understood that the registration of the point cloud reality model and the point cloud design model is divided into two steps: coarse registration and precise registration.

[0053] Since the point cloud real scene model and the point cloud design model have a low correlation in space, first select the feature points with the same position on the real point cloud model and the design point cloud model (in this step, it is not necessary to select the points with the same position accurately, as long as they are roughly corresponding), and use the Random Sample Consensus (RANSAC) algorithm to roughly align them so that they roughly overlap; the second step is precise alignment: use the existing algorithm, the Iterative Closest Point (ICP) algorithm to accurately align the real point cloud model and the design point cloud model. The point cloud model is essentially a set of spatial data points obtained by a 3D scanning device, and the role of the ICP algorithm is to optimize the alignment between the two point cloud models through continuous iteration, so that the design model and the actual model completely overlap in space. In this way, the position and posture of the two point cloud models in space can be ensured to be very accurate.

[0054] The advantage of Example 4 is that the alignment of the point cloud real-scene model and the point cloud design model is divided into two steps: coarse alignment and precise alignment. The RANSAC algorithm is first used for preliminary alignment, and then the ICP algorithm is used for precise alignment. This strategy fully considers the low correlation of the model in the initial state, gradually optimizes the alignment of the model, and ensures that the position and posture of the two point cloud models in space are very precise, which lays the foundation for accurate posture difference detection and quantitative analysis.

[0055] By using the KNN neighbor algorithm to detect the posture difference of the accurately aligned models, we can comprehensively obtain the second appearance difference results between the models, including information such as the spatial position deviation between each point cloud pair, and realize the quantitative analysis of the appearance differences of prefabricated steel tube arch ribs, providing detailed data support for more accurate evaluation of the production quality of the arch ribs.

[0056] Embodiment 5: In S3.5, multiple point cloud pairs corresponding to the same position are selected from the point cloud real scene model and the point cloud design model, and a preliminary registration process is performed on the selected multiple point cloud pairs using the RANSAC random sampling consensus algorithm, including:

[0057] S3.5.1. Taking each point cloud in the point cloud design model as a source point cloud, collecting all the source point clouds to obtain a source point cloud set P; taking each point cloud in the point cloud real scene model as a target point cloud, collecting all the target point clouds to obtain a target point cloud set Q; selecting multiple source point clouds and target point clouds corresponding to the same position to obtain multiple groups of point cloud pairs;

[0058] S3.5.2, calculating normal vectors of point clouds in the plurality of point clouds to obtain eigenvalues ​​of each point cloud pair;

[0059] S3.5.3. Randomly select a set of point cloud pairs from multiple sets of point cloud pairs, denoted as Among them, p i is the source point cloud, p i ∈P,q i is the target point cloud, q i ∈Q, K is the number of point cloud pairs;

[0060] S3.5.4. Construct a rigid transformation matrix T according to the eigenvalues ​​corresponding to the selected point cloud pair, wherein the rigid transformation matrix T includes a rotation matrix R and a translation vector t, such that:

[0061] T(P i )=R·p i +t≈q i ;

[0062] S3.5.5, based on the rigid transformation matrix T applied to all source point clouds in the source point cloud set P for transformation processing, a transformed source point cloud set P' is obtained, wherein the source point cloud set P' is composed of the transformed source point clouds, and the distance difference between each transformed source point cloud in the source point cloud set P' and the target point cloud belonging to the same point cloud pair in the target point cloud set Q is calculated, each distance difference is compared with a preset difference, and point cloud pairs corresponding to distance differences less than the preset difference are obtained, and the number of point cloud pairs obtained by comparison is counted, and the rigid transformation matrix T is updated according to the number of point cloud pairs obtained by comparison;

[0063] S3.5.6, repeat S3.5.3 to S3.5.5 according to a preset number of iterations, until the preset number of iterations is reached, thereby obtaining an updated rigid transformation matrix T;

[0064] S3.5.7. Apply the updated rigid transformation matrix T to all source point clouds in the source point cloud set P for transformation processing, and complete the preliminary registration processing of the point cloud real scene model and the point cloud design model.

[0065] For example, in S3.5.1, some point cloud pairs are selected at key positions of the arch rib (such as the arch crown, arch foot, etc.). These point cloud pairs appear to be in similar positions in the two models, but there may be certain deviations.

[0066] For example, in S3.5.2, the normal vectors of the point clouds in the selected multiple point clouds are calculated, and the eigenvalues ​​of each point cloud pair are calculated using a suitable algorithm (such as estimation based on the local geometric shape of the point cloud). Assume that the normal vectors of one of the point cloud pairs (p1, q1) are and The eigenvalue f1=0.99 of this set of point cloud pairs is obtained by an eigenvalue calculation method (such as the cosine value of the normal vector angle, etc.).

[0067] For example, in S3.5.3, a point cloud pair is randomly selected from multiple point cloud pairs. Assume that (p3, q3) is selected, and its eigenvalue is f3 = 0.98. The rigid transformation matrix T is constructed according to the eigenvalues ​​corresponding to the selected point cloud pair, where the rotation matrix R and the translation vector t are calculated by the relevant formula,

[0068]

[0069] So that q3=R·p3+t holds within a certain error range (the error range here can be set according to actual needs, such as the eigenvalue error is less than 0.01, etc.).

[0070] For example, in S3.5.4, the rigid transformation matrix T is applied to all source point clouds in the source point cloud set P to perform transformation processing, and the transformed source point cloud set P' is obtained. For each transformed source point cloud P' in P' i , calculate the target point cloud q that belongs to the same point cloud pair with the target point cloud set Q i The distance difference Δd i =‖P′ i -q i ‖. For example, for P′4 and q4, Δd is calculated i =0.03m.

[0071] For example, in S3.5.5, each distance difference is compared with a preset difference (assuming the preset difference is 0.05m) to obtain the number of point cloud pairs corresponding to the distance difference less than the preset difference. Assuming that after comparison, there are 30 groups of point cloud pairs with distance differences less than 0.05m, the rigid transformation matrix is ​​updated based on the information of these 30 groups of point cloud pairs (the update method can adopt strategies such as weighted average, and assign different weights to adjust R and t according to the matching degree of the point cloud pairs).

[0072] For example, in S3.5.6, according to the preset number of iterations (assuming 10 times), the above steps of randomly selecting point cloud pairs, constructing rigid transformation matrices, transforming source point clouds, calculating distance differences, and updating matrices are repeated. In each iteration, the rigid transformation matrix is ​​continuously adjusted so that the matching degree between the point cloud real scene model and the point cloud design model is gradually improved.

[0073] For example, in S3.5.7, when the preset number of iterations is 10, the updated rigid transformation matrix T is obtained. The updated rigid transformation matrix T is applied to all source point clouds in the source point cloud set P for final transformation processing. At this time, the point cloud real scene model and the point cloud design model roughly overlap, completing the preliminary registration processing of the point cloud real scene model and the point cloud design model, providing a good foundation for subsequent precise registration.

[0074] The advantages of Example 5 are: by reasonably selecting the source point cloud and the target point cloud set, and based on a series of steps such as randomly selecting point cloud pairs, calculating eigenvalues, constructing a rigid transformation matrix, applying transformations, calculating distance differences, updating the matrix and repeating iterations, it is possible to effectively find the optimal transformation matrix for preliminary alignment when the correlation between the point cloud models is low, so that the point cloud real-scene model and the point cloud design model are roughly overlapped, providing good initial conditions for subsequent precise alignment, and improving the accuracy and stability of the entire alignment process.

[0075] Embodiment 6: In S3.6, the point cloud real scene model and the point cloud design model after preliminary registration are accurately registered by using the ICP iterative closest point algorithm, including:

[0076] S3.6.1. Select point cloud pairs containing eigenvalue c from all point cloud pairs of the point cloud real scene model and the point cloud design model after preliminary registration, and construct a feature data set based on the eigenvalues ​​corresponding to each point cloud in the selected point cloud pairs. The feature data set includes x p ,c∈x p and x q ,c∈x q , where x p ,c and x q ,c is the matching feature point pair;

[0077] S3.6.2. The error between the source point cloud feature value and the target point cloud feature value in the feature data set is represented by an error function, wherein the error function is:

[0078]

[0079] Among them, R is the rotation matrix of the source point cloud during transformation, t is the translation vector of the source point cloud during transformation, and x p,h and x q,h are the hth eigenvalues ​​in the matching feature point pairs, a is the number of matching feature point pairs,

[0080] The error function is solved by the matrix singular value decomposition method to obtain the rotation matrix R and the translation vector t to minimize the error function; the error function is solved by the matrix singular value decomposition method to obtain the rotation matrix R and the translation vector t that minimize the error function. In actual calculations, the data in the feature data set are substituted into the error function, and the matrix singular value decomposition algorithm is used to perform complex mathematical calculations, and the values ​​of R and t are gradually adjusted until the combination of R and t that minimizes the error function is found.

[0081] S3.6.3. Based on the obtained rotation matrix R and translation vector t, the eigenvalue x of the source point cloud p After transformation, we get x p ,h is the updated eigenvalue;

[0082] S3.6.4, based on x p ,h updated feature value, calculate the average distance between all matching feature point pairs;

[0083] The eigenvalues ​​of the source point cloud are transformed based on the solved rotation matrix R and translation vector t. For the eigenvalues ​​of the source point cloud in the feature data set, the updated eigenvalue xp,c' is calculated by ‖xp,c′=Rpc+t‖ (wherein it is the coordinate of the source point cloud in the dimension related to the eigenvalue c). Then, the average distance between all matching feature point pairs is calculated. The specific calculation method is to calculate the distance between each pair of matching feature point pairs (xp,c', xp,c) (such as the Euclidean distance ‖xp,c'-xp,c‖), and then sum all these distances and divide them by the number of matching feature point pairs a to obtain the average distance.

[0084] S3.6.5. If the average distance is less than the set threshold, the optimization goal has been achieved and the calculation process stops; otherwise, return to S3.6.1 to continue iterative calculation until convergence, and output the optimal transformation matrix T best , thereby completing the precise registration processing of the point cloud real scene model and the point cloud design model.

[0085] Set a threshold (for example, 0.05 meters) and compare the calculated average distance with the threshold. If the average distance is less than the set threshold, it means that the optimization goal has been achieved. At this time, the transformation matrix composed of the rotation matrix R and the translation vector t can make the two point cloud models achieve a high registration accuracy, and the calculation process stops. If the average distance is not less than the threshold, return to the above step of screening feature point clouds to construct a data set, continue iterative calculations, re-solve the error function, update the source point cloud eigenvalues ​​and calculate the average distance until the average distance is less than the threshold, and then output the optimal transformation matrix T best, thereby completing the precise registration processing of the point cloud models of the stone arch bridges in two different periods, so that the overlap accuracy of the two models in space reaches a high level, providing an accurate basis for the subsequent posture difference detection.

[0086] The advantages of Example 6 are: by screening feature point pairs to construct a feature data set, using the error function to represent the error between the eigenvalues ​​of the source point cloud and the target point cloud, and using the matrix singular value decomposition method to solve the rotation matrix and translation vector to minimize the error, the average distance is calculated iteratively and compared with the set threshold, and finally the precise alignment of the point cloud real-scene model and the point cloud design model is achieved, ensuring the precise alignment of the model in space, providing a high-precision model foundation for subsequent posture difference detection, and effectively improving the accuracy of virtual acceptance.

[0087] Embodiment 7: In S3.7, the KNN neighbor algorithm is used to detect the posture difference of the point cloud real scene model and the point cloud design model after accurate registration to obtain a second appearance difference result, including:

[0088] S3.7.1. Calculate the normal vectors of all source point clouds and all target point clouds in the accurately registered point cloud real scene model and the point cloud design model, and use the calculated normal vectors as the feature descriptors of each point cloud; for example, for the source point cloud p1 in the point cloud real scene model, its normal vector is For the target point cloud q1 in the point cloud design model, its normal vector is

[0089] S3.7.2. For each feature descriptor of the source point cloud, the distance between the feature descriptor of each source point cloud and the feature descriptors of all points in the target point cloud is calculated using the Euclidean distance formula to obtain the difference between the feature descriptors:

[0090] d(p i ,q j )=||f(p i )-f(q j )||,

[0091] Among them, d(p i ,q j ) is the difference value, f(p i ) and f(q j ) are the source point clouds p i ∈P and target point cloud q j ∈Q feature descriptor,

[0092] Select the nearest K target points q j , that is, select iThe most similar K points; for example, select the K target points with the smallest difference value, assuming that K = 3 here, and select the target points q1, q2, and q3 as the points most similar to the source point cloud p1

[0093] S3.7.3. For each source point cloud p i and its corresponding K target points q j , calculate the source point cloud p by the Euclidean distance formula i With K target points q j Straight-line distance in space:

[0094] δ(p i -q j )=‖p i -q j ‖,

[0095] Among them, δ(p i -q j ) is the straight-line distance;

[0096] S3.7.4. Calculate each source point cloud p separately by using the mean square error algorithm i With K target points q j The mean square error value and average deviation value of the straight-line distance are calculated to obtain the second shape difference result. For example, the straight-line distances between the source point cloud p1 and the target points q1, q2, and q3 are calculated to be d1 = 0.05m, d2 = 0.06m, and d3 = 0.04m, respectively. The mean square error value is: The average deviation value is: (d1+d2+d3) / 3.

[0097] Through the above process, the appearance difference information between the models can be obtained comprehensively and meticulously, providing rich data for accurately evaluating the manufacturing accuracy of prefabricated steel tube arch ribs, which helps to more accurately determine whether the arch ribs meet the design and construction requirements. At the same time, the difference detection can be carried out by calculating the steel tube diameter deviation, fitting the arch axis function curve comparison and other methods to further enhance the comprehensiveness and accuracy of the model difference evaluation. For example, feature points that can represent the diameter of the steel tube are selected from the point cloud, and the diameter deviation of the steel tube in the point cloud real scene model and the point cloud design model is calculated; or the vertical coordinate data of the centroid points of each cross section of the real point cloud model and the design point cloud model under the same horizontal coordinate are extracted, and the arch axis function curve is fitted, and the two curves are compared to judge the difference in overall shape between the design model and the actual model.

[0098] The advantages of Example 7 are: by calculating the normal vectors of all source point clouds and target point clouds in the model after accurate registration as feature descriptors, using Euclidean distance to calculate the difference value between feature descriptors, the straight-line distance between the source point cloud and the target point cloud, and further calculating the mean square error value and the average deviation value, the posture difference of the model is detected from multiple dimensions, and the appearance difference information between the models can be obtained comprehensively and meticulously, providing rich data for accurately evaluating the manufacturing accuracy of prefabricated steel pipe arch ribs, which helps to more accurately determine whether the arch ribs meet the design and construction requirements. At the same time, this method is not limited to the KNN neighbor algorithm, but can also perform difference detection by calculating the steel pipe diameter deviation, fitting the arch axis function curve comparison and other methods, providing multiple analysis angles, and further enhancing the comprehensiveness and accuracy of the model difference evaluation.

[0099] This embodiment is not limited to performing posture difference detection on the point cloud real scene model and the point cloud design model after accurate registration by using the KNN neighbor algorithm, and the difference detection can also be performed by the following methods:

[0100] For example, to further analyze the differences between the models, feature points that can help calculate the diameter of the steel pipe are selected from the point cloud. Through these points, the diameter deviation of the steel pipe in the point cloud real scene model and the point cloud design model can be calculated. This process can evaluate the difference in specific dimensions between the design and the actual model, especially for the accuracy requirements of structural parts such as steel pipes.

[0101] Another example: by extracting the ordinate data of the centroid points of each cross section of the actual point cloud model and the design point cloud model under the same horizontal coordinate, a more detailed comparison can be made. The ordinate data of these centroid points are fitted using the least squares method to obtain the arch axis function curve of each model. These two fitted arch axis function curves can be used for comparison and acceptance to determine the difference in overall shape between the design model and the actual model. The least squares method is a commonly used mathematical method that minimizes the error between the fitting curve and the data points to obtain the curve that best fits the data trend.

[0102] Example 8: Figure 2 As shown, the embodiment of the present invention further provides a virtual acceptance device for a steel tube arch rib, comprising:

[0103] A data import module, used for importing multi-directional arch rib point cloud data, wherein the multi-directional arch rib point cloud data is obtained by scanning the prefabricated steel pipe arch rib in multiple directions by a three-dimensional laser scanner pre-installed in the site;

[0104] A model construction module, used to construct a BIM real-scene model and a point cloud real-scene model according to the multi-directional arch rib point cloud data, and to construct a BIM design model and a point cloud design model according to the prefabricated steel pipe arch rib drawing data, and to merge the BIM real-scene model and the BIM design model into the same file, and to merge the point cloud real-scene model and the point cloud design model into the same file;

[0105] The virtual acceptance module is used to perform a qualitative analysis on the BIM real-scene model and the BIM design model in the same file to obtain a first appearance difference result, and to perform a quantitative analysis on the point cloud real-scene model and the point cloud design model in the same file to obtain a second appearance difference result, thereby completing the virtual acceptance of the steel tube arch rib.

[0106] Embodiment 9: In the virtual acceptance module, the BIM real scene model and the BIM design model in the same file are qualitatively analyzed to obtain a first appearance difference result, including:

[0107] Importing the BIM real-scene model and the BIM design model in the same file into a BIM design tool as the same BIM file, overlapping the BIM real-scene model and the BIM design model, and obtaining a common spatial coordinate system;

[0108] Based on the spatial coordinate system, the arch rib bottom plane is selected in the BIM real scene model and the BIM design model respectively, the arch rib bottom plane includes two upper and lower circles and a middle connecting part of the two circles, and the center of the upper circle in the arch rib bottom plane is taken as the centroid and the length of the three sides is set as a fixed value to construct a characteristic triangle corresponding to the BIM real scene model and a characteristic triangle corresponding to the BIM design model;

[0109] The characteristic triangle corresponding to the BIM real scene model and the three corners of the characteristic triangle corresponding to the BIM design model are used as three characteristic point pairs, and based on the principle that three points that are not in the same straight line determine the position of a plane in space, the positions of the end planes of the BIM real scene model and the BIM design model in space are determined and overlapped through the three characteristic point pairs;

[0110] The difference between the overlapped BIM real-scene model and the BIM design model is calculated by using a Boolean operation method, and a first appearance difference result is obtained according to the difference.

[0111] Example 10: The embodiment of the present invention also provides a virtual acceptance device for steel tube arch ribs, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the virtual acceptance method for steel tube arch ribs as described above is implemented.

[0112] The advantages of the present invention are: 1) In terms of model construction, the BIM real-scene model and the point cloud real-scene model are constructed based on the multi-directional arch rib point cloud data, which can highly restore the real state of the prefabricated steel tube arch rib. At the same time, the prefabricated steel tube arch rib drawing data is imported to construct the BIM design model and the point cloud design model, so that the actual constructed model and the design model have a basis for direct comparison. The model constructed in this way is not only accurate in geometric shape, but also can achieve a complete mapping from design to reality at the information level, which is helpful to comprehensively evaluate the status of the prefabricated steel tube arch rib at various stages;

[0113] 2) In terms of acceptance analysis, qualitative and quantitative analysis are combined, and the BIM real-scene model and the BIM design model are qualitatively analyzed. The difference between the two is calculated through Boolean operation to obtain the first appearance difference result, which can quickly and intuitively judge the approximate differences between the design model and the actual model in terms of shape, size, etc., such as whether there are obvious omissions, deformations, etc., and grasp the appearance quality of the prefabricated steel tube arch ribs as a whole;

[0114] The point cloud real scene model and the point cloud design model are quantitatively analyzed, and the precise deviation between the models is calculated using algorithms such as RANSAC, ICP and KNN, including the spatial position deviation between the point cloud pairs, the steel pipe diameter deviation, the arch axis function curve comparison, etc., to obtain the second shape difference result. This quantitative analysis can provide specific values ​​and accurately evaluate the manufacturing accuracy of the prefabricated steel pipe arch ribs, such as the deviation size of each position, etc., providing detailed data support for quality control. The combination of qualitative and quantitative analysis realizes a comprehensive and in-depth evaluation of the prefabricated steel pipe arch ribs, making the acceptance results more scientific and reliable.

[0115] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0116] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0117] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0118] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present invention.

[0119] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0120] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A virtual acceptance method for steel tube arch ribs, characterized in that: The steps include: S1. Perform multi-directional scanning on the prefabricated steel tube arch ribs by using a 3D laser scanner pre-installed in the site to obtain multi-directional arch rib point cloud data; S2. Construct a BIM real-scene model and a point cloud real-scene model according to the multi-directional arch rib point cloud data, and construct a BIM design model and a point cloud design model according to the prefabricated steel pipe arch rib drawing data, and merge the BIM real-scene model and the BIM design model into the same file, and merge the point cloud real-scene model and the point cloud design model into the same file; S3. Performing a qualitative analysis on the BIM real-scene model and the BIM design model in the same file to obtain a first appearance difference result, and performing a quantitative analysis on the point cloud real-scene model and the point cloud design model in the same file to obtain a second appearance difference result, thereby completing the virtual acceptance of the steel tube arch rib.

2. The virtual acceptance method according to claim 1, characterized in that: In S2, a BIM real-scene model and a point cloud real-scene model are constructed according to the multi-directional arch rib point cloud data, and a BIM design model and a point cloud design model are constructed according to the prefabricated steel pipe arch rib drawing data, and the BIM real-scene model and the BIM design model are merged into the same file, and the point cloud real-scene model and the point cloud design model are merged into the same file, including: S2.

1. Importing the multi-directional arch rib point cloud data into a point cloud analysis tool for data splicing to obtain point cloud models of multiple prefabricated arch rib segments, and virtually assembling the point cloud models of the multiple prefabricated arch rib segments to obtain a point cloud real-scene model of the prefabricated steel tube arch rib, and converting the point cloud real-scene model into a BIM real-scene model through a BIM design tool; S2.2, importing the prefabricated steel tube arch rib drawing data, inputting the prefabricated steel tube arch rib drawing data into the BIM design tool to generate a BIM design model, and converting the BIM design model into a point cloud design model through the point cloud analysis tool; S2.

3. Import the BIM real-scene model and the BIM design model into a BIM design tool as the same BIM file to merge them into the same file; import the point cloud real-scene model and the point cloud design model into a point cloud analysis tool as the same point cloud file to merge them into the same file.

3. The virtual acceptance method according to claim 1, characterized in that: In S3, the BIM real scene model and the BIM design model in the same file are qualitatively analyzed to obtain a first appearance difference result, including: S3.

1. Overlap the BIM real-scene model and the BIM design model in the same file to obtain a common spatial coordinate system; S3.2, based on the spatial coordinate system, respectively select the arch rib bottom plane in the BIM real scene model and the BIM design model, the arch rib bottom plane includes two upper and lower circles and the middle connecting part of the two circles, take the center of the upper circle in the arch rib bottom plane as the centroid and set the lengths of the three sides as fixed values ​​to construct the characteristic triangle corresponding to the BIM real scene model and the characteristic triangle corresponding to the BIM design model; S3.

3. Taking the characteristic triangle corresponding to the BIM real scene model and the three corners of the characteristic triangle corresponding to the BIM design model as three characteristic point pairs, and based on the principle of determining the position of a plane in space by three points that are not in the same straight line, determine the positions of the end planes of the BIM real scene model and the BIM design model in space through the three characteristic point pairs and make them coincide with each other; S3.

4. Calculate the difference between the overlapped BIM real-scene model and the BIM design model using a Boolean operation method, and obtain a first appearance difference result according to the difference.

4. The virtual acceptance method according to claim 1, characterized in that: In S3, the point cloud real scene model and the point cloud design model in the same file are quantitatively analyzed to obtain a second appearance difference result, including: S3.5, selecting a plurality of point cloud pairs corresponding to the same position from the point cloud real scene model and the point cloud design model in the same file, and performing preliminary registration processing on the selected plurality of point cloud pairs by using the RANSAC random sampling consensus algorithm; S3.6, perform precise registration processing on the point cloud real scene model and the point cloud design model after preliminary registration through ICP iterative closest point algorithm; S3.

7. Perform posture difference detection on the point cloud real scene model and the point cloud design model after precise registration through the KNN neighbor algorithm to obtain a second appearance difference result.

5. The virtual acceptance method according to claim 4, characterized in that: In S3.5, a plurality of point cloud pairs corresponding to the same position are selected from the point cloud real scene model and the point cloud design model, and a preliminary registration process is performed on the selected plurality of point cloud pairs using a RANSAC random sampling consensus algorithm, including: S3.5.

1. Taking each point cloud in the point cloud design model as a source point cloud, collecting all the source point clouds to obtain a source point cloud set P; taking each point cloud in the point cloud real scene model as a target point cloud, collecting all the target point clouds to obtain a target point cloud set Q; selecting multiple source point clouds and target point clouds corresponding to the same position to obtain multiple groups of point cloud pairs; S3.5.2, calculating normal vectors of point clouds in the plurality of point clouds to obtain eigenvalues ​​of each point cloud pair; S3.5.

3. Randomly select a set of point cloud pairs from multiple sets of point cloud pairs, denoted as Among them, p i is the source point cloud, p i ∈P,q i is the target point cloud, q i ∈Q, K is the number of point cloud pairs; S3.5.

4. Construct a rigid transformation matrix T according to the eigenvalues ​​corresponding to the selected point cloud pair, wherein the rigid transformation matrix T includes a rotation matrix R and a translation vector t, such that: T(P i )=R·p i +t≈q i ; S3.5.5, based on the rigid transformation matrix T applied to all source point clouds in the source point cloud set P for transformation processing, a transformed source point cloud set P' is obtained, wherein the source point cloud set P' is composed of the transformed source point clouds, and the distance difference between each transformed source point cloud in the source point cloud set P' and the target point cloud belonging to the same point cloud pair in the target point cloud set Q is calculated, each distance difference is compared with a preset difference, and point cloud pairs corresponding to distance differences less than the preset difference are obtained, and the number of point cloud pairs obtained by comparison is counted, and the rigid transformation matrix T is updated according to the number of point cloud pairs obtained by comparison; S3.5.6, repeat S3.5.3 to S3.5.5 according to a preset number of iterations, until the preset number of iterations is reached, thereby obtaining an updated rigid transformation matrix T; S3.5.

7. Apply the updated rigid transformation matrix T to all source point clouds in the source point cloud set P for transformation processing, and complete the preliminary registration processing of the point cloud real scene model and the point cloud design model.

6. The virtual acceptance method according to claim 5, characterized in that: In S3.6, the point cloud real scene model and the point cloud design model after preliminary registration are accurately registered by using the ICP iterative closest point algorithm, including: S3.6.

1. Select point cloud pairs containing eigenvalue c from all point cloud pairs of the point cloud real scene model and the point cloud design model after preliminary registration, and construct a feature data set based on the eigenvalues ​​corresponding to each point cloud in the selected point cloud pairs. The feature data set includes x p ,c∈x p and x q ,c∈x q , where x p ,c and x q ,c is the matching feature point pair; S3.6.

2. The error between the source point cloud feature value and the target point cloud feature value in the feature data set is represented by an error function, wherein the error function is: Among them, R is the rotation matrix of the source point cloud during transformation, t is the translation vector of the source point cloud during transformation, and x p,h and x q,h are the hth eigenvalues ​​in the matching feature point pairs, a is the number of matching feature point pairs, Solving the error function by matrix singular value decomposition method to obtain rotation matrix R and translation vector t, so as to minimize the error function; S3.6.

3. Based on the obtained rotation matrix R and translation vector t, the eigenvalue x of the source point cloud p After transformation, we get x p ,h is the updated eigenvalue; S3.6.4, based on x p ,h updated feature value, calculate the average distance between all matching feature point pairs; S3.6.

5. If the average distance is less than the set threshold, the optimization goal has been achieved and the calculation process stops; otherwise, return to S3.6.1 to continue the iterative calculation until convergence, and output the optimal transformation matrix, thereby completing the precise registration processing of the point cloud real scene model and the point cloud design model.

7. The virtual acceptance method according to claim 6, characterized in that: In S3.7, the KNN neighbor algorithm is used to detect the posture difference of the point cloud real scene model and the point cloud design model after accurate registration, and a second appearance difference result is obtained, including: S3.7.

1. Calculate normal vectors for all source point clouds and all target point clouds in the accurately registered point cloud real scene model and the point cloud design model, and use the calculated normal vectors as feature descriptors of each point cloud; S3.7.

2. For each feature descriptor of the source point cloud, the distance between the feature descriptor of each source point cloud and the feature descriptors of all points in the target point cloud is calculated using the Euclidean distance formula to obtain the difference between the feature descriptors: d(p i ,q j )=||f(p i )-f(q j )||, Among them, d(p i ,q j ) is the difference value, f(p i ) and f(q j ) are the source point clouds p i ∈P and target point cloud q j ∈Q feature descriptor, Select the nearest K target points q j , that is, select i The most similar K points; S3.7.

3. For each source point cloud p i and its corresponding K target points q j , calculate the source point cloud p by the Euclidean distance formula i With K target points q j Straight-line distance in space: δ(p i -q j )=||p i -q j ||, Among them, δ(p i -q j ) is the straight-line distance; S3.7.

4. Calculate each source point cloud p separately by using the mean square error algorithm i With K target points q j The mean square error value and the average deviation value of the straight-line distance are calculated to obtain the second appearance difference result.

8. A virtual acceptance device for steel tube arch ribs, characterized in that: include: A data import module, used for importing multi-directional arch rib point cloud data, wherein the multi-directional arch rib point cloud data is obtained by scanning the prefabricated steel pipe arch rib in multiple directions by a three-dimensional laser scanner pre-installed in the site; A model construction module, used to construct a BIM real-scene model and a point cloud real-scene model according to the multi-directional arch rib point cloud data, and to construct a BIM design model and a point cloud design model according to the prefabricated steel pipe arch rib drawing data, and to merge the BIM real-scene model and the BIM design model into the same file, and to merge the point cloud real-scene model and the point cloud design model into the same file; The virtual acceptance module is used to perform a qualitative analysis on the BIM real-scene model and the BIM design model in the same file to obtain a first appearance difference result, and to perform a quantitative analysis on the point cloud real-scene model and the point cloud design model in the same file to obtain a second appearance difference result, thereby completing the virtual acceptance of the steel tube arch rib.

9. The virtual acceptance device according to claim 8, characterized in that: In the virtual acceptance module, the BIM real scene model and the BIM design model in the same file are qualitatively analyzed to obtain a first appearance difference result, including: Importing the BIM real-scene model and the BIM design model in the same file into a BIM design tool as the same BIM file, overlapping the BIM real-scene model and the BIM design model, and obtaining a common spatial coordinate system; Based on the spatial coordinate system, the arch rib bottom plane is selected in the BIM real scene model and the BIM design model respectively, the arch rib bottom plane includes two upper and lower circles and a middle connecting part of the two circles, and the center of the upper circle in the arch rib bottom plane is taken as the centroid and the length of the three sides is set as a fixed value to construct a characteristic triangle corresponding to the BIM real scene model and a characteristic triangle corresponding to the BIM design model; The characteristic triangle corresponding to the BIM real scene model and the three corners of the characteristic triangle corresponding to the BIM design model are used as three characteristic point pairs, and based on the principle that three points that are not in the same straight line determine the position of a plane in space, the positions of the end planes of the BIM real scene model and the BIM design model in space are determined and overlapped through the three characteristic point pairs; The difference between the overlapped BIM real-scene model and the BIM design model is calculated by using a Boolean operation method, and a first appearance difference result is obtained according to the difference.

10. A virtual acceptance device for steel tube arch ribs, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the virtual acceptance method for the steel tube arch rib as described in any one of claims 1 to 7 is implemented.

Images