Method for checking conformity of an engineering project to a mockup

Abstract

The application relates to a kind of engineering project real model consistency checking methods, comprising the following steps: scanning engineering project entity, and obtaining entity scanning data;The entity scanning data is preprocessed;Reverse modeling is carried out using the preprocessed entity scanning data, and scanning model is obtained;The scanning model is aligned with the BIM model constructed in advance in the same coordinate system;The geometric properties of the two models after alignment are compared, and the overall deviation distance between the two models is obtained;Consistency evaluation is carried out according to the overall deviation distance.The application can accurately check and mark the deviation between the two models by scanning the entity and carrying out reverse modeling, and comparing the entity model with the pre-created BIM model by combining algorithm.

Description

Technical Field

[0001] This invention relates to the field of engineering management, and in particular to a method for checking the consistency of the actual model of an engineering project. Background Technology

[0002] In engineering construction, the realism and accuracy of the physical model have a significant impact on the entire project implementation. Therefore, effectively checking the consistency between the physical engineering model and the design drawings is crucial for the quality control of the engineering project.

[0003] A common method for inspecting engineering structures is to conduct manual on-site measurements and then compare them with the design drawings. However, manual measurement is labor-intensive, inefficient, and prone to errors, especially for large or complex construction projects, where manual inspection proves inadequate. Summary of the Invention

[0004] To address the aforementioned issues, this invention provides a method for checking the consistency of physical models in engineering projects. By scanning the entity and performing reverse modeling, combined with an algorithm, the physical model and a pre-created BIM model are compared, which can accurately check and mark the deviations between the two models.

[0005] This invention is achieved through the following scheme: a method for checking the consistency of the actual model in an engineering project, comprising the following steps:

[0006] S1. Scan the physical entity of the project and obtain the entity scan data;

[0007] S2. Preprocess the entity scan data;

[0008] S3. Use the preprocessed entity scan data to perform reverse modeling to obtain the scan model;

[0009] S4. Align the scanned model with the pre-built BIM model in the same coordinate system;

[0010] S5. Compare the geometric properties of the two aligned models to obtain the overall deviation distance between the two models;

[0011] S6. Perform a consistency assessment based on the overall deviation distance.

[0012] A further improvement of the method for checking the consistency of the actual model of the project in this invention is that: when performing step S1, a radar device is used to scan along a predetermined route so that the acquired entity scan data is point cloud data covering the entire entity of the project.

[0013] A further improvement to the method for checking the consistency of the actual model in engineering projects according to the present invention is that, when performing step S2:

[0014] S2.1, Perform noise removal on the entity scan data;

[0015] S2.2 Sampling is performed on the noise-removed data to optimize the amount of data and ensure that the optimized data retains the features of the original data.

[0016] S2.3. Use the NDT registration algorithm to register the optimized point cloud data.

[0017] A further improvement to the method for checking the consistency of the actual model in engineering projects according to the present invention is that, when performing step S2.3:

[0018] Perform initial global alignment of the source point cloud and target point cloud data corresponding to the building characteristics;

[0019] After initial global alignment, all point cloud data are decomposed and distributed into the cells of the 3D mesh system;

[0020] Calculate the mean and variance for each point in the cell and create the corresponding Gaussian distribution;

[0021] The optimal rigid body transformation is found through iterative calculation and comparison, which maximizes the cumulative match between the probability distribution of the source point cloud and the target point cloud across all cells.

[0022] A further improvement to the method for checking the consistency of the actual model in engineering projects according to the present invention is that, when performing step S4:

[0023] S4.1 Load the data of the scanned model and the BIM model in the same coordinate system and perform initial alignment;

[0024] S4.2. Use the ICP algorithm to iterate the nearest point registration. Each iteration optimizes the rotation and translation parameters based on finding the corresponding point pair or facing each other to minimize the distance between matching points. Stop the iteration when the maximum number of iterations is reached or the change of the transformation matrix is ​​less than the preset threshold.

[0025] S4.3 Validate and optimize the registration results, and combine the RANSAC algorithm to improve the robustness of the registration results.

[0026] A further improvement of the method for checking the consistency of the actual model in the present invention is that the Hausdorff distance algorithm is used to compare the geometric attributes in step S5, and the Hausdorff distance between the scanned model set points and the BIM model set points is obtained as the overall deviation distance.

[0027] A further improvement to the method for checking the consistency of the actual model in engineering projects according to the present invention is as follows:

[0028] Before executing step S6, the overall deviation distance is divided into different standard accuracy levels according to the gradient, and corresponding processing opinions are configured for each standard accuracy level.

[0029] When performing step S6, the standard accuracy level to which the deviation should belong is determined based on the overall deviation distance calculated in step S5, and corresponding processing opinions are obtained. Then, a consistency assessment report is generated.

[0030] A further improvement to the method for checking the consistency of the actual model in engineering projects according to the present invention is that the different standard accuracy levels include:

[0031] For first-class accuracy with an overall deviation within ±3mm, the corresponding processing opinion is "perfect match, no processing required";

[0032] For Class II precision with an overall deviation within ±5mm, the corresponding handling opinion is "partial deviation, it is recommended to re-check the design or construction";

[0033] For Class III precision with an overall deviation within ±10mm, the corresponding handling opinion is "significant deviation exists, requiring design adjustment or reconstruction"; and

[0034] The overall deviation exceeds the level 4 accuracy of ±10mm, and the corresponding handling opinion is "there is a serious deviation, and engineering improvement measures must be implemented immediately".

[0035] This invention uses entity scanning and reverse modeling, combined with algorithms to compare the entity model with a pre-created BIM model, to accurately check and mark the deviations between the two models. This provides a data foundation for project managers, enabling them to intuitively understand the actual situation of the project through the model, monitor the project progress in real time, anticipate potential delays or deviations in advance, promptly identify and correct errors, reduce the additional costs and time caused by error correction, and improve the efficiency of project schedule and quality management. Attached Figure Description

[0036] Figure 1 A flowchart illustrating the overall process of the method of the present invention is shown.

[0037] Figure 2 A flowchart illustrating the preprocessing of entity scan data in this invention is shown.

[0038] Figure 3 The flowchart illustrates the process of aligning two models in the same coordinate system according to the present invention. Detailed Implementation

[0039] To address the problems of high labor intensity, low efficiency, and susceptibility to errors in traditional manual model comparison methods, this invention provides a method for checking the consistency of physical models in engineering projects. By scanning the entity and performing reverse modeling, combined with algorithms, the physical model is compared with a pre-created BIM model, enabling precise detection and identification of deviations between the two models. The following detailed embodiments, in conjunction with accompanying drawings, further illustrate this method for checking the consistency of physical models in engineering projects.

[0040] See Figures 1 to 3 As shown, a method for checking the consistency of the actual model in an engineering project includes the following steps:

[0041] Step S1: Scan the project entity and acquire entity scan data. Specifically: First, determine the specific area to be scanned and set the scan path; then, select the most suitable radar equipment based on the actual project requirements and environmental conditions; before starting the scan, calibrate the radar equipment and set the scan parameters, including scan resolution, scan depth, and scan frequency, etc. These scan parameters should be set according to the actual project requirements; during radar scanning, ensure that the radar equipment operates continuously and stably without sudden movement or jumping. The radar equipment should scan along the predetermined route to ensure the scan coverage (covering the entire project entity) and scan accuracy. After scanning, save the point cloud data generated by the radar equipment.

[0042] Step S2: Preprocess the entity scan data. Specifically, such as... Figure 2 As shown, the preprocessing steps include:

[0043] Step S2.1: Remove noise from the entity scan data. Due to environmental factors and equipment performance during the scanning process, point cloud data may contain noise, such as outliers and duplicate points. Therefore, in preprocessing, this noise is removed from the original data by setting thresholds or using statistical methods.

[0044] Step S2.2: Sample from the noise-removed data to optimize the data volume and ensure that the optimized data retains the features of the original data. For large-scale point cloud data, direct processing is not only computationally expensive but may also lead to data redundancy. Therefore, strategies such as uniform sampling or centroid selection are used to sample the point cloud data, optimize the data volume, and retain the features of the original data as much as possible.

[0045] Step S2.3: Register the optimized point cloud data using the NDT (Normal Distributions Transform) registration algorithm. This registration process includes: first, performing initial global alignment between the source and target point cloud data corresponding to the building characteristics; after initial global alignment, decomposing all point cloud data and distributing them into cells of the 3D mesh system; then, calculating the mean and variance of points in each cell and creating a corresponding Gaussian distribution; and finally, finding the optimal rigid body transformation (i.e., a combination of rotation and translation) through iterative calculation and comparison, ensuring the highest cumulative match in probability distributions between the source and target point clouds across all cells, thus achieving the registration effect. This process iterates until a predetermined accuracy or maximum number of iterations is reached. Using this NDT registration algorithm ensures accurate registration of the building's point cloud data, providing reliable input for subsequent data processing and modeling.

[0046] Step S3: Use the preprocessed entity scan data to perform reverse modeling to obtain the scan model.

[0047] Step S4: Align the scanned model with the pre-built BIM model in the same coordinate system. Specifically, as follows... Figure 3 As shown, this step specifically includes:

[0048] Step S4.1: Load the data of the scanned model and the BIM model in the same coordinate system, perform noise reduction, filtering and downsampling preprocessing operations on the data to optimize data quality and improve processing efficiency, and then perform initial alignment on the optimized data.

[0049] Step S4.2: After initial alignment to ensure that the two models are in the same coordinate system, run the ICP (Iterative Closest Point) algorithm to iterate the registration of the closest point. Each iteration optimizes the rotation and translation parameters based on finding corresponding point pairs or facing points to minimize the distance between matching points. The optimization process will continue until certain stopping criteria are met, such as reaching the maximum number of iterations or the change of the transformation matrix being less than a preset threshold, at which point the iteration stops.

[0050] Step S4.3: Finally, the registration results are verified and optimized, and the robustness of the registration results is improved by combining the RANSAC algorithm. Through step S4, the scanned model and the BIM model can be effectively aligned to the same coordinate system to ensure the consistency of the comparison and the effectiveness of further operations.

[0051] Step S5: Compare the geometric properties of the two aligned models to obtain the overall deviation distance between them. Specifically, the Hausdorff distance algorithm is used to compare the geometric properties of the two models. The smaller the Hausdorff distance value, the more similar the geometric properties are between the scanned model and the BIM model. The specific principle is as follows:

[0052] Suppose we have two sets A = {a1, ..., ap} and B = {b1, ..., bq}, then the Hausdorff distance between these two sets is defined as...

[0053] H(A,B)=max(h(A,B),h(B,A)) (1)

[0054] in,

[0055]

[0056]

[0057] ||·|| is the distance paradigm between point sets A and B.

[0058] Equation (1) is called the two-way Hausdorff distance, which is the most basic form of the Hausdorff distance;

[0059] In equation (2), h(A, B) and h(B, A) are called the one-way Hausdorff distances from set A to set B and from set B to set A, respectively. That is, h(A, B) is actually obtained by first sorting the distances ||ai-bj|| between each point ai in point set A and the nearest point bj in set B, and then taking the maximum value of these distances as the value of h(A, B). h(B, A) can be obtained in the same way.

[0060] The bidirectional Hausdorff distance H(A, B) is the larger of the one-way distances h(A, B) and h(B, A), and measures the maximum degree of mismatch between two point sets.

[0061] Steps S4 and S5 compare the scanned model obtained from reverse modeling with the pre-built BIM model. The comparison process can be carried out by specialized comparison software, such as Navisworks. These specialized software programs can easily identify the differences and deviations between the two models.

[0062] Step S6: Perform a consistency assessment based on the overall deviation distance. Specifically: Before performing the consistency assessment, first divide the overall deviation distance into different standard accuracy levels according to the gradient, and configure corresponding processing opinions for each standard accuracy level. For example, the different standard accuracy levels are divided into:

[0063] Level 1 accuracy (i.e., absolute high accuracy), with an overall deviation within ±3mm, corresponds to the processing opinion of "perfect match, no processing required".

[0064] Level 2 accuracy (i.e., higher accuracy), with an overall deviation within ±5mm, the corresponding handling opinion is "partial deviation, it is recommended to re-check the design or construction".

[0065] Level 3 accuracy (acceptable accuracy), with an overall deviation within ±10mm, the corresponding handling opinion is "there is a large deviation, and the design needs to be adjusted or the construction needs to be redone".

[0066] Level 4 accuracy (i.e., the overall deviation distance exceeds Level 3 accuracy), the overall deviation distance exceeds ±10mm, the corresponding handling opinion is "there is a serious deviation, and engineering improvement measures must be implemented immediately".

[0067] The above accuracy level classification is only one example. In actual applications, it can be set according to the actual situation. The overall deviation distance range and corresponding processing opinions corresponding to different accuracy levels can be customized.

[0068] During the consistency assessment, based on the actual overall deviation distance calculated in step S5, the standard accuracy level to which it should belong is determined, and corresponding handling opinions are obtained. Then, a consistency assessment report is generated. The report may include Hausdorf distance, standard accuracy level, handling opinions, etc., and is presented in the form of a report to facilitate project managers' understanding of the current status and progress of the project and to facilitate the next step of work arrangements.

[0069] This invention uses entity scanning and reverse modeling, combined with algorithms to compare the entity model with a pre-created BIM model, to accurately check and mark the deviations between the two models. This provides a data foundation for project managers, enabling them to intuitively understand the actual situation of the project through the model, monitor the project progress in real time, anticipate potential delays or deviations in advance, promptly identify and correct errors, reduce the additional costs and time caused by error correction, and improve the efficiency of project schedule and quality management.

[0070] The present invention has been described in detail above with reference to the accompanying drawings and embodiments. Those skilled in the art can make various modifications to the present invention based on the above description. Therefore, certain details in the embodiments should not be construed as limiting the present invention, and the scope of protection of the present invention shall be defined by the appended claims.

Claims

1. A method for checking the consistency of the actual model in an engineering project, characterized in that, Including the following steps: S1. Scan the physical entity of the project and obtain the entity scan data; S2. Preprocess the entity scan data: S2.1, Perform noise removal on the entity scan data; S2.2 Sampling is performed on the noise-removed data to optimize the amount of data and ensure that the optimized data retains the features of the original data. S2.

3. Register the optimized point cloud data using the NDT registration algorithm: Perform initial global alignment between the source point cloud and the target point cloud data corresponding to the building characteristics; After initial global alignment, all point cloud data is decomposed and distributed into cells of the 3D mesh system; the mean and variance of the points in each cell are calculated, and a corresponding Gaussian distribution is created. The optimal rigid body transformation is found through iterative calculation and comparison, which maximizes the cumulative match between the probability distribution of the source point cloud and the target point cloud across all cells. S3. Use the preprocessed entity scan data to perform reverse modeling to obtain the scan model; S4. Align the scanned model with the pre-built BIM model in the same coordinate system: S4.1 Load the data of the scanned model and the BIM model in the same coordinate system and perform initial alignment; S4.

2. Use the ICP algorithm to iterate the nearest point registration. Each iteration optimizes the rotation and translation parameters based on finding the corresponding point pair or facing each other to minimize the distance between matching points. Stop the iteration when the maximum number of iterations is reached or the change of the transformation matrix is ​​less than the preset threshold. S4.3 Validate and optimize the registration results, and combine the RANSAC algorithm to improve the robustness of the registration results; S5. Compare the geometric properties of the two aligned models to obtain the overall deviation distance between the two models; S6. Perform a consistency assessment based on the overall deviation distance: Before executing step S6, the overall deviation distance is divided into different standard accuracy levels according to the gradient, and corresponding processing opinions are configured for each standard accuracy level. When performing step S6, the standard accuracy level to which the deviation should belong is determined based on the overall deviation distance calculated in step S5, and corresponding processing opinions are obtained. Then, a consistency assessment report is generated.

2. The method for checking the consistency of the actual model of an engineering project as described in claim 1, characterized in that: When performing step S1, the radar equipment is used to scan along a predetermined route so that the acquired entity scan data is point cloud data covering the entire project entity.

3. The method for checking the consistency of the actual model of an engineering project as described in claim 1, characterized in that, The Hausdorff distance algorithm is used to compare the geometric attributes in step S5, and the bidirectional Hausdorff distance between the scanned model set points and the BIM model set points is obtained as the overall deviation distance.

4. The method for checking the consistency of the actual model of an engineering project as described in claim 1, characterized in that, The different standard accuracy levels include: For first-class accuracy with an overall deviation within ±3mm, the corresponding processing opinion is "perfect match, no processing required"; For Class II precision with an overall deviation within ±5mm, the corresponding handling opinion is "partial deviation, it is recommended to re-check the design or construction"; For Class III precision with an overall deviation within ±10mm, the corresponding handling opinion is "significant deviation exists, requiring design adjustment or reconstruction"; and The overall deviation exceeds the level 4 accuracy of ±10mm, and the corresponding handling opinion is "there is a serious deviation, and engineering improvement measures must be implemented immediately".

Images