BIM-based precast component splicing displacement deviation monitoring system

Abstract

This application relates to the field of point cloud matching technology, specifically to a prefabricated component splicing displacement deviation monitoring system based on BIM technology. The system includes: acquiring a standard BIM model of the prefabricated component and acquiring its 3D point cloud data in real time; obtaining the detail saliency coefficient of each point cloud based on the directional diversity and surface roughness of each point cloud location at each time point; obtaining the matching influence coefficient at each time point based on the difference in detail saliency coefficients at different locations in the overall point cloud data at each time point, the detail saliency coefficients of all point clouds, and the pose difference between each time point and the previous time point; and then obtaining the leaf_size parameter at each time point, thereby obtaining the registration result between the prefabricated component and the standard BIM model, and ultimately obtaining the displacement deviation result of the prefabricated component at each time point. This application improves the accuracy of prefabricated component displacement deviation monitoring by adaptively acquiring the leaf_size parameter at each time point.

Description

Technical Field

[0001] This application relates to the field of point cloud matching technology, specifically to a prefabricated component splicing displacement deviation monitoring system based on BIM technology. Background Technology

[0002] Prefabricated buildings utilize an industrialized production model, where a large number of prefabricated components are manufactured in factories and then transported to the construction site for assembly, significantly reducing on-site workload. In the construction of prefabricated buildings, workers only need to assemble the prefabricated walls, floor slabs, and other components according to design requirements, increasing construction speed several times compared to traditional on-site casting.

[0003] Accurate tracking and positioning of prefabricated components enables better management of the entire prefabricated building process. Against this backdrop, digital technologies, represented by Building Information Modeling (BIM), have become key to solving construction quality control challenges due to their advantages in 3D visualization modeling and end-to-end collaborative management. By deploying a total station to collect 3D point cloud data of prefabricated components during assembly, and performing point cloud data registration and automated comparison calculations based on the measured point cloud data and the point cloud data in the BIM model during construction, dynamic installation deviation monitoring of prefabricated components can be achieved. During the assembly of large components, prefabricated components are subject to stress deformation or swaying vibration due to gravity and airflow factors, which can easily cause changes in posture during hoisting and movement. Existing methods do not fully consider the impact of gravity and airflow factors on the accuracy of point cloud data registration during component assembly, resulting in insufficient accuracy in monitoring prefabricated component assembly displacement deviation. Summary of the Invention

[0004] To address the aforementioned technical problems, the purpose of this application is to provide a BIM-based system for monitoring displacement deviations in prefabricated component splicing. The specific technical solution adopted is as follows:

[0005] This application proposes a prefabricated component splicing displacement deviation monitoring system based on BIM technology, the system comprising:

[0006] Data acquisition module: acquires the standard BIM model of prefabricated components and acquires the 3D point cloud data of prefabricated components in real time during the splicing process;

[0007] Matching accuracy evaluation module: Based on the similarity of the directions of the normal vectors at different positions in the fitted surface corresponding to the nearest neighbor point clouds of each point cloud at each time step, the dispersion of the curvature at different positions, and the dispersion of the shortest distance between all the nearest neighbor point clouds of each point cloud and their corresponding fitted surface, the detail significance coefficient of each point cloud is obtained; The point cloud data at each time step is segmented, and the diversity index at each time step is obtained based on the distance between the detail significance coefficients of all points in any two point cloud data clusters after segmentation, and the detail significance coefficients of all point clouds at the same time step.

[0008] Displacement deviation calculation module: Based on the difference in the axial direction of the precast component at each moment compared to the previous moment, the attitude difference coefficient of the precast component at each moment is obtained. Combined with the diversity index at each moment, the matching influence coefficient of the precast component at each moment is obtained. Then, the leaf_size parameter at each moment is obtained. This allows for the registration of the 3D point cloud data of the precast component at each moment with the point cloud data in the standard BIM model, thereby obtaining the displacement deviation result of the precast component at each moment.

[0009] Preferably, the formula for calculating the detail saliency coefficient of each point cloud is: In the formula, Let be the detail saliency coefficient of the i-th point cloud. Let i be the orientation consistency of the i-th point cloud. Let be the coefficient of variation of curvature at the mapping points of the i-th point cloud and all its nearest neighbor point clouds, and exp() be an exponential function with the natural constant e as the base. Let be the standard deviation of the Euclidean distance between all neighboring point clouds and their mapping points of the i-th point cloud.

[0010] Preferably, the mapping point of each point cloud refers to the point on the fitted surface corresponding to each point cloud that has the smallest Euclidean distance to the corresponding point cloud.

[0011] Preferably, the process for obtaining the orientation consistency of each point cloud is as follows: obtain the unit normal vector corresponding to the mapping point of each point cloud on the fitted surface; and take the mean of the cosine similarity coefficient between each point cloud and the unit normal vectors corresponding to all its nearest neighbor point clouds as the orientation consistency of each point cloud.

[0012] Preferably, the process of obtaining the diversity index at each time point is as follows: arranging all detail saliency coefficients corresponding to each point cloud data cluster in ascending order to obtain the detail saliency sequence of each point cloud data cluster; calculating the DTW distance between the detail saliency sequences of any two point cloud data clusters at each time point; and using the product of the mean of all DTW distances and the mean of the detail saliency coefficients of all point clouds in the entire point cloud data at the same time point as the diversity index at each time point.

[0013] Preferably, the calculation process of the attitude difference coefficient of the prefabricated component at each moment is as follows: extract the axis of the three-dimensional point cloud data of the prefabricated component at each moment, and take the angle between the axis and the direction of gravity as the verticality of the prefabricated component at each moment; take the absolute difference between the verticality of the prefabricated component at each moment and the corresponding verticality of the previous adjacent moment as the attitude difference coefficient of the prefabricated component at each moment.

[0014] Preferably, the matching influence coefficient of the prefabricated component at each time moment refers to the product of the attitude difference coefficient and the diversity index of the prefabricated component at each time moment.

[0015] Preferably, the formula for calculating the leaf_size parameter at each time point is: In the formula, The leaf_size parameter at time j. This is the preset minimum value for the leaf_size parameter. This is the preset maximum value for the leaf_size parameter. For normalization function, Let be the matching influence coefficient of the precast component at time j.

[0016] Preferably, in the process of registering the three-dimensional point cloud data of the prefabricated component at each time with the point cloud data in the standard BIM model, the leaf_size parameter calculated at each time is used as the leaf_size parameter of the KD-Tree in the corresponding time of the ICP algorithm processing.

[0017] Preferably, the process of obtaining the displacement deviation results at each moment is as follows: using the Cloud Compare tool to compare and analyze the point cloud registration results, and obtaining the displacement deviation results between the three-dimensional point cloud data and the point cloud data in the standard BIM model at each moment during the splicing process of the prefabricated component.

[0018] This application has the following beneficial effects:

[0019] This application proposes a BIM-based system for monitoring displacement deviation in prefabricated component splicing. By deeply analyzing the differences in the normal vector direction and surface roughness of the prefabricated component surface, detailed saliency coefficients of each point cloud are constructed. This allows for the evaluation of detailed features at various locations on the prefabricated component surface, which is beneficial for subsequent evaluation of the matching accuracy of the prefabricated component. By analyzing the overall complexity of the prefabricated component and its posture and swaying state during splicing, as well as the matching influence coefficient of the component at each moment, the influence of the shape and posture of the prefabricated component on the point cloud registration at each moment can be determined. Based on the matching influence coefficient at each moment, the parameters for point cloud matching processing are optimized. The advantage of this system is that it can reduce the impact of swaying and vibration of the prefabricated component caused by airflow, improve the real-time matching accuracy of the point cloud data and the BIM model during the splicing process of the prefabricated component, and thus improve the accuracy of the displacement deviation monitoring results of the prefabricated component. Attached Figure Description

[0020] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a block diagram of a prefabricated component splicing displacement deviation monitoring system based on BIM technology provided in one embodiment of this application;

[0022] Figure 2 This is a flowchart illustrating the process of obtaining the displacement deviation results of a prefabricated component at various times, as provided in one embodiment of this application. Detailed Implementation

[0023] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the BIM-based prefabricated component splicing displacement deviation monitoring system proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0025] The following description, in conjunction with the accompanying drawings, details the specific scheme of the BIM-based prefabricated component splicing displacement deviation monitoring system provided in this application.

[0026] Please see Figure 1 The diagram illustrates a block diagram of a BIM-based prefabricated component splicing displacement deviation monitoring system according to an embodiment of this application. The system includes: a data acquisition module, a matching accuracy evaluation module, and a displacement deviation calculation module.

[0027] Data acquisition module: Acquires standard BIM models of prefabricated components and acquires 3D point cloud data of prefabricated components in real time during the splicing process.

[0028] In the early stages of construction, this application first constructs a standard BIM model of the entire building based on the construction drawings. This standard BIM model contains detailed information about prefabricated components, such as dimensions, shapes, and connection methods. Through systematic integration of models from multiple disciplines, including structural, MEP, and pipeline systems, precise positioning and logical connections of components across disciplines are achieved in three-dimensional space. Simultaneously, detailed modeling is conducted for critical areas prone to quality issues, such as beam-column joints and waterproofing joints, fully incorporating details of joint construction, material properties, and connection methods into the model. This ultimately yields a standard BIM model of the prefabricated components in an ideal state, providing accurate spatial geometric parameter references for subsequent construction.

[0029] In the manufacturing stage, this application completes mass production in the factory according to the unified and finalized detailed design drawings, and integrates and transmits various data information during the production process through the open data interface of the BIM model components. Then, the transportation of prefabricated components is organized according to the construction sequence, so as to orderly stack the prefabricated components in the limited space on site.

[0030] During the on-site assembly phase, prefabricated components are assembled according to the construction plan. The prefabricated components are hoisted using tower cranes. This application uses an intelligent total station to acquire real-time 3D point cloud data of the prefabricated components during the assembly process and transmits it to the BIM software model. Based on the 3D information corresponding to the 3D point cloud data, the system analyzes whether displacement deviations occur during the assembly process. The intelligent total station, in addition to measuring angles and distances to targets, also has automatic target recognition and aiming functions, thus automatically completing the identification, aiming, and measurement of multiple targets. In this embodiment, the data acquisition interval of the intelligent total station is set to 5 seconds. To improve the accuracy of data acquisition, this application uses Gaussian filtering technology to perform preliminary noise reduction processing on the point cloud data. Thus, the 3D point cloud data of the prefabricated components at each moment is obtained.

[0031] Matching accuracy evaluation module: Based on the similarity of the directions of the normal vectors at different positions in the fitted surface corresponding to the nearest neighbor point clouds of each point cloud at each time step, the dispersion of the curvature at different positions, and the dispersion of the shortest distance between all the nearest neighbor point clouds of each point cloud and their corresponding fitted surface, the detail significance coefficient of each point cloud is obtained; The point cloud data at each time step is segmented, and the diversity index at each time step is obtained based on the distance between the detail significance coefficients of all points in any two point cloud data clusters after segmentation, and the detail significance coefficients of all point clouds at the same time step.

[0032] In actual construction, significant ambient wind speeds may cause continuous swaying and vibration during the hoisting and assembly of precast components. Large precast components are prone to stress deformation under gravity, and moisture and floating particles in the air can affect the accuracy of data registration between the measured point cloud data of the precast components and the BIM model, thereby reducing the accuracy of displacement deviation monitoring. Therefore, this application conducts an in-depth analysis of the aforementioned influencing factors during the assembly process and optimizes the point cloud data registration process in real time to improve the reliability of displacement deviation monitoring.

[0033] The surfaces of prefabricated building components are typically rough, and the detailed information contained in different prefabricated components also varies. For example, the structure at the connection of components is usually more complex and detailed, while the structure of other relatively flat parts is simpler and contains less detailed information. The greater the curvature of the surface corresponding to a part of the prefabricated component and the greater the difference between the directions of the surface normals, the more complex the corresponding structure. To obtain the features of the detailed information contained in its surface, the following processing is performed. Taking the i-th point cloud in the point cloud data of any prefabricated component at time j as an example, the K-nearest neighbor search algorithm is used to obtain the M nearest point clouds to the i-th point cloud as the nearest neighbor point clouds of the i-th point cloud data, where M is an integer in the range [25, 35], and is taken as 30 in this embodiment. The least squares method is used to obtain the fitted surface of the nearest neighbor point clouds of the i-th point cloud. The point on the fitted surface with the smallest Euclidean distance to each point cloud is recorded as the mapping point of each point cloud. The unit normal vector corresponding to the mapping point of each point cloud on the fitted surface is obtained. The cosine similarity coefficient between the i-th point cloud and the unit normal vectors of all its nearest neighbor point clouds is calculated. The mean of all cosine similarity coefficients is taken as the orientation consistency of the i-th point cloud, denoted as . The result The smaller the value, the greater the difference in the direction of the normal vector between the i-th point cloud and its nearest neighbor point clouds. Furthermore, curvature reflects the degree of bending of the precast component surface at that point; generally, the more complex the curvature, the greater the complexity of the corresponding part, and the greater the distance between the point cloud data and the obtained fitted surface due to the influence of surface roughness. Therefore, the coefficient of variation of curvature at the mapping points of the i-th point cloud and all its nearest neighbor point clouds is denoted as... The result This reflects the complex characteristics of the local bending variation of the precast component surface at point i. Then, the standard deviation of the Euclidean distance between all nearest neighbor point clouds and their mapped points at point i is denoted as... The result The larger the value, the more uneven the distance distribution between the fitted surface and the point cloud data, indicating that the surface of the building component at the location of the i-th point cloud data may be rougher.

[0034] As a preferred implementation, the detail saliency coefficients of each point cloud are obtained based on the similarity of the directions of the normal vectors at different positions in the fitted surface corresponding to the nearest neighbor point clouds of each point cloud at each time point, the dispersion of the curvature at different positions, and the dispersion of the shortest distance between all the nearest neighbor point clouds of each point cloud and its corresponding fitted surface. These coefficients are used to characterize the saliency of the detail features in the local region where each point cloud is located.

[0035] In this embodiment, the detail saliency coefficient of the i-th point cloud is denoted as... Its expression is: In the formula, Let be the detail saliency coefficient of the i-th point cloud. Let i be the orientation consistency of the i-th point cloud. Let be the coefficient of variation of curvature at the mapping points of the i-th point cloud and all its nearest neighbor point clouds, and exp() be an exponential function with the natural constant e as the base. Let be the standard deviation of the Euclidean distance between all nearest neighbor point clouds and their mapped points of the i-th point cloud. The resulting... The larger the value, the more significant the difference in the surface normal vector direction and the degree of curvature change of the i-th point cloud within its local region.

[0036] Furthermore, the more complex the structure of the precast component, the higher the accuracy required during registration. Detail saliency coefficients can more accurately reflect the complex characteristics of the structure. This application first employs a point cloud segmentation method based on principal component analysis to segment the point cloud data of the precast component at time j, outputting multiple point cloud data clusters. Some structures in the precast component are relatively simple, while others are relatively complex. By dividing the point cloud data clusters, simple and complex structures in the precast component can be distinguished. By analyzing the morphological differences of each structure, the overall structural complexity characteristics of the precast component are evaluated. For example, if the structure of a certain part is relatively simple, the distribution of detail saliency coefficients in its corresponding point cloud data cluster is relatively concentrated, and the overall mean of all detail saliency coefficients is relatively small. In contrast, in complex structures, the distribution of detail saliency coefficients corresponding to their point cloud data clusters is generally more dispersed, and the overall mean of all detail saliency coefficients is relatively large. Therefore, the detail saliency coefficients corresponding to each point cloud data cluster are arranged in ascending order to obtain the detail saliency sequence of each point cloud data cluster. Then, the DTW distance between the detail saliency sequences of any two point cloud data clusters in the entire point cloud data at time j is calculated. The product of the mean of all DTW distances and the mean of the detail saliency coefficients of all point clouds in the entire point cloud data is used as the diversity index of the prefabricated component at time j, denoted as . The result The larger the value, the more complex the overall structure of the prefabricated component at time j.

[0037] Displacement deviation calculation module: Based on the difference in the axial direction of the precast component at each moment compared to the previous moment, the attitude difference coefficient of the precast component at each moment is obtained. Combined with the diversity index at each moment, the matching influence coefficient of the precast component at each moment is obtained. Then, the leaf_size parameter at each moment is obtained. This allows for the registration of the 3D point cloud data of the precast component at each moment with the point cloud data in the standard BIM model, thereby obtaining the displacement deviation result of the precast component at each moment.

[0038] Furthermore, the precast components experience swaying and vibration during hoisting due to airflow, resulting in varying attitudes in the air at different times. The greater the degree of attitude variation, the more prone the component is to causing misalignment during assembly. The verticality of the precast component can be obtained from the point cloud data collected at each time point. This verticality is obtained as follows: first, the axis of the precast component's 3D point cloud data at each time point is extracted using a point cloud slicing method. The angle between the axis and the direction of gravity is then used as the verticality of the precast component at each time point. Therefore, the absolute difference between the verticality of the precast component at time j and its preceding adjacent time point is used as the attitude difference coefficient of the precast component at time j, denoted as [equation missing]. The result The larger the value, the more pronounced the swaying vibration of the precast component at time j, and the higher the accuracy requirement for point cloud matching.

[0039] As a preferred embodiment, the matching influence coefficient of the prefabricated component at each time moment is obtained based on the attitude difference coefficient and diversity index of the prefabricated component at each time moment, which is used to characterize the influence of the shape and attitude of the prefabricated component at each time moment on the point cloud registration.

[0040] In this embodiment, the matching influence coefficient of the prefabricated component at time j is denoted as... Its expression is: In the formula, Let be the matching influence coefficient of the precast component at time j. Let be the diversity index of the prefabricated component at time j. Let be the attitude difference coefficient of the precast component at time j. The obtained... The larger the value, the more complex the structure of the precast component and the greater the degree of swaying vibration at time j. Therefore, the registration accuracy of the precast component should be improved at time j.

[0041] Furthermore, the more complex the shape and the greater the attitude difference of the prefabricated component, the higher the accuracy requirement in the point cloud matching process. This application uses the ICP (Iterative Closest Point) algorithm for point cloud registration. As can be seen from the above analysis, the larger the matching influence coefficient at each time point, the higher the registration accuracy should be. Therefore, in the ICP algorithm processing at the corresponding time point, when using KD-Tree for fast search, a smaller leaf_size parameter should be set to improve the registration accuracy; conversely, a larger leaf_size parameter should be set to improve the registration efficiency. In this application, the value range of the leaf_size parameter is set to [a1, a2]. Since the value of the leaf_size parameter is usually within 20, in this embodiment, a1 is 5 and a2 is 20.

[0042] Then, the leaf_size parameter at each time point is optimized in real time based on the matching influence coefficient at each time point. The specific formula for calculating the leaf_size parameter at each time point is as follows: In the formula, The leaf_size parameter at time j. This is the preset minimum value for the leaf_size parameter. This is the preset maximum value for the leaf_size parameter. As the normalization function, this embodiment uses the tanh function for normalization. Let be the matching influence coefficient of the precast component at time j.

[0043] Furthermore, the calculated leaf_size parameters at each time step are used as the leaf_size parameters of the KD-Tree during the corresponding ICP algorithm processing. This is used to register the point cloud data in the prefabricated component's 3D point cloud data pre-BIM model, obtaining the point cloud registration results at each time step. Then, the Cloud Compare tool is used to compare and analyze the point cloud registration results, obtaining the displacement deviation results between the 3D point cloud data of the prefabricated component at each time step during the assembly process and the point cloud data in the standard BIM model. The flowchart for obtaining the displacement deviation results of the prefabricated component at each time step is shown below. Figure 2 As shown.

[0044] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0045] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0046] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the principles of this application should be included within the protection scope of this application.

Claims

1. A prefabricated component splicing displacement deviation monitoring system based on BIM technology, characterized in that, The system includes: Data acquisition module: acquires the standard BIM model of prefabricated components and acquires the 3D point cloud data of prefabricated components in real time during the splicing process; Matching accuracy evaluation module: Obtains the detail saliency coefficients of each point cloud, expressed as: In the formula, Let be the detail saliency coefficient of the i-th point cloud. Let i be the orientation consistency of the i-th point cloud. Let be the coefficient of variation of curvature at the mapping points of the i-th point cloud and all its nearest neighbor point clouds, and exp() be an exponential function with the natural constant e as the base. Let be the standard deviation of the Euclidean distance between all neighboring point clouds and their mapping points of the i-th point cloud; arrange all detail saliency coefficients corresponding to each point cloud data cluster in ascending order to obtain the detail saliency sequence of each point cloud data cluster; calculate the DTW distance between the detail saliency sequences of any two point cloud data clusters at each time step, and use the product of the mean of all DTW distances and the mean of the detail saliency coefficients of all point clouds in the entire point cloud data at the same time step as the diversity index at each time step; Displacement deviation calculation module: Extracts the axis of the precast component's 3D point cloud data at each moment, and uses the angle between the axis and the gravity direction as the verticality of the precast component at each moment; uses the absolute difference between the verticality of the precast component at each moment and the corresponding value of the previous adjacent moment as the attitude difference coefficient of the precast component at each moment; uses the product of the attitude difference coefficient of the precast component at each moment and the diversity index as the matching influence coefficient of the precast component at each moment, and then obtains the leaf_size parameter at each moment, with the expression as follows: In the formula, The leaf_size parameter at time j. This is the preset minimum value for the leaf_size parameter. This is the preset maximum value for the leaf_size parameter. For normalization function, Let be the matching influence coefficient of the prefabricated component at time j; then, the three-dimensional point cloud data of the prefabricated component at each time moment is registered with the point cloud data in the standard BIM model, and the displacement deviation results of the prefabricated component at each time moment are obtained.

2. The prefabricated component splicing displacement deviation monitoring system based on BIM technology as described in claim 1, characterized in that, The mapping point of each point cloud refers to the point on the fitted surface corresponding to each point cloud that has the smallest Euclidean distance to the corresponding point cloud.

3. The prefabricated component splicing displacement deviation monitoring system based on BIM technology as described in claim 2, characterized in that, The process of obtaining the orientation consistency of each point cloud is as follows: obtain the unit normal vector corresponding to the mapping point of each point cloud on the fitted surface; take the mean of the cosine similarity coefficient between each point cloud and the unit normal vectors corresponding to all its nearest neighbor point clouds as the orientation consistency of each point cloud.

4. The prefabricated component splicing displacement deviation monitoring system based on BIM technology as described in claim 1, characterized in that, In the process of registering the three-dimensional point cloud data of the prefabricated component at each time with the point cloud data in the standard BIM model, the leaf_size parameter calculated at each time is used as the leaf_size parameter of the KD-Tree in the corresponding time of the ICP algorithm processing.

5. The prefabricated component splicing displacement deviation monitoring system based on BIM technology as described in claim 1, characterized in that, The process of obtaining the displacement deviation results at each time point is as follows: the Cloud Compare tool is used to compare and analyze the point cloud registration results to obtain the displacement deviation results between the three-dimensional point cloud data and the point cloud data in the standard BIM model at each time point during the splicing process of the prefabricated component.

Images