Intelligent tracking method and system for construction progress based on point cloud and BIM

Abstract

The present disclosure provides an intelligent tracking method and system for construction progress based on point cloud and BIM, which comprises: obtaining a design BIM of a target building and actual point cloud data of the actual building at a construction site; inputting the actual point cloud data into a generated model of a preset BIM to output an actual BIM of the actual building; taking mutually matched design components and actual components as target component pairs; for each target component pair, obtaining a gap value of a plurality of evaluation indexes of each target component pair based on a corresponding design sub-BIM and a corresponding actual sub-BIM; obtaining a comprehensive gap value of each target component pair based on the gap value; and intelligently tracking the construction progress of the target building based on the comprehensive gap value of each target component pair. The present disclosure compares two BIMs, comprehensively considers a plurality of evaluation indexes, and tracks the construction progress of the building, thereby improving the accuracy and reliability of the determination of the construction progress.

Description

Technical Field

[0001] This disclosure relates to the field of construction progress tracking technology, and in particular to an intelligent method and system for tracking construction progress based on point cloud and BIM (Building Information Model). Background Technology

[0002] Building construction refers to the entire process of constructing various building products on a designated site by using various building materials, components, and equipment, through a series of technical means and organizational management, in accordance with design drawings and relevant specifications.

[0003] Construction tracking refers to the continuous and systematic recording, monitoring, evaluation, and management of the entire process of a construction project from commencement to completion, in order to ensure that the project proceeds smoothly in accordance with the contract, drawings, budget, and schedule.

[0004] Current construction project progress tracking largely relies on manual on-site comparison and general labeling in visualization software. Manually judging the construction progress of building components through point clouds makes it difficult to achieve comparability and quantification of component parameters, resulting in inaccurate and unreliable determination of construction progress. Summary of the Invention

[0005] The technical problem to be solved by this disclosure is to overcome the defects of the existing technology, which uses point clouds to judge the construction progress of building components, such as inaccuracy and unreliability in judging the construction progress. This disclosure provides an intelligent tracking method and system for building construction progress based on point clouds and BIM.

[0006] This disclosure solves the above-mentioned technical problems through the following technical solution:

[0007] This disclosure provides an intelligent tracking method for building construction progress based on point cloud and BIM, the intelligent tracking method including:

[0008] Acquire the BIM design data of the target building and the actual point cloud data of the building at the construction site;

[0009] The actual building is the building constructed according to the target building.

[0010] The actual point cloud data is input into the preset BIM generation model to output the actual BIM of the actual building;

[0011] The target building includes several design components, and the design BIM includes a design sub-BIM corresponding to each design component; the actual building includes several actual components, and the actual BIM includes an actual sub-BIM corresponding to each actual component.

[0012] The mutually matching design components and actual components are taken as target component pairs;

[0013] For each target component pair, based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component, the difference values ​​of several evaluation indicators for each target component pair are obtained.

[0014] Based on the aforementioned gap value, a comprehensive gap value is obtained for each pair of target components;

[0015] The comprehensive gap value is used to characterize the current construction progress of the corresponding component;

[0016] The construction progress of the target building is intelligently tracked based on the comprehensive gap value of each target component pair.

[0017] Optionally, the design components include regular components and irregularly shaped components;

[0018] The design sub-BIM corresponding to the rule component is represented using the first parameter set;

[0019] For the irregularly shaped component and the design file of the irregularly shaped component contains the parameters required for NURBS (a modeling technique), the design sub-BIM corresponding to the irregularly shaped component is represented by NURBS, and the NURBS corresponds to the second parameter set;

[0020] For the irregularly shaped component and the design file of the irregularly shaped component does not contain the parameters required for NURBS representation, the design sub-BIM corresponding to the irregularly shaped component is represented by Mesh (a modeling technique), and the Mesh corresponds to a third parameter set;

[0021] And / or,

[0022] The actual components include regular components and irregularly shaped components;

[0023] The actual sub-BIM corresponding to the rule component is represented using the first parameter set;

[0024] For the irregular component and the actual confidence level of the NURBS representation of the irregular component output by the preset BIM generation model is not less than the preset confidence level, the actual sub-BIM corresponding to the irregular component is represented by NURBS, and the NURBS corresponds to the second parameter set.

[0025] For the irregularly shaped component, and the actual confidence level corresponding to the NURBS representation of the irregularly shaped component output by the preset BIM generation model is less than the preset confidence level, the actual sub-BIM corresponding to the irregularly shaped component is represented by a Mesh, and the Mesh corresponds to the third parameter set.

[0026] Optionally, the preset BIM generation model is obtained based on several sets of sample training data;

[0027] The training data for each set of samples includes sample point cloud data and sample BIM corresponding to the sample buildings.

[0028] Optionally, for the regular component, the evaluation index includes at least one of the regular component's position, orientation, size, cross-section, and installation time;

[0029] For the irregularly shaped component, and both the design sub-BIM and the actual sub-BIM corresponding to the irregularly shaped component are represented by NURBS, the evaluation index includes at least one of the following: the position, orientation, installation time, shape, control grid, and weight of the control grid of the irregularly shaped component.

[0030] For the irregularly shaped component, and at least one of the design sub-BIM and the actual sub-BIM corresponding to the irregularly shaped component is represented by a Mesh, the evaluation index includes at least one of the following: position, orientation, installation time, shape, main direction, and size of the outer bounding box of the irregularly shaped component.

[0031] Optionally, before the step of using the mutually matching design component and the actual component as a target component pair, the method further includes:

[0032] Obtain the design parameter values ​​and actual parameter values ​​of the designed component and the actual component under several parameter information;

[0033] In response to the fact that the design parameter value and the actual parameter value under each parameter information meet the preset requirements, it is determined that the design component and the actual component are mutually matched.

[0034] Optionally, the parameter information includes at least one of the component's type, location, size, and shape.

[0035] Optionally, the step of obtaining the difference values ​​of several evaluation indicators for each target component pair based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component includes:

[0036] For each target component pair, based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component, the design index value and actual index value corresponding to each evaluation index are obtained.

[0037] The gap value is obtained based on the design index value and the actual index value;

[0038] And / or,

[0039] The step of obtaining the comprehensive gap value for each pair of target components based on the gap value includes:

[0040] Obtain the weight corresponding to each of the aforementioned gap values;

[0041] Based on each of the gap values ​​and the corresponding weights, the comprehensive gap value of each of the target component pairs is obtained;

[0042] And / or,

[0043] The step of intelligently tracking the construction progress of the target building based on the comprehensive gap value of each target component pair includes:

[0044] In response to the comprehensive gap value falling within a first preset range, the construction progress of the corresponding design component is determined to be in a normal state;

[0045] In response to the comprehensive gap value falling within the second preset range, the construction progress of the corresponding design component is determined to be in an early warning state;

[0046] In response to the comprehensive gap value falling within a third preset range, the construction progress of the corresponding design component is determined to be in an alarm state;

[0047] Wherein, the upper limit of the first preset range is less than the lower limit of the second preset range, and the upper limit of the second preset range is less than the lower limit of the third preset range.

[0048] This disclosure also provides an intelligent tracking system for building construction progress based on point cloud and BIM, the intelligent tracking system comprising:

[0049] The design BIM acquisition module is used to acquire the design BIM of the target building.

[0050] The point cloud data acquisition module is used to acquire actual point cloud data of the actual building at the construction site;

[0051] The actual building is the building constructed according to the target building.

[0052] The actual BIM output module is used to input the actual point cloud data into the preset BIM generation model to output the actual BIM of the actual building;

[0053] The target building includes several design components, and the design BIM includes a design sub-BIM corresponding to each design component; the actual building includes several actual components, and the actual BIM includes an actual sub-BIM corresponding to each actual component.

[0054] The component pair determination module is used to identify mutually matching design components and actual components as target component pairs.

[0055] The gap value acquisition module is used to obtain the gap value of several evaluation indicators for each target component pair based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component.

[0056] The comprehensive gap value acquisition module is used to obtain the comprehensive gap value for each of the target component pairs based on the gap value;

[0057] The comprehensive gap value is used to characterize the current construction progress of the corresponding component;

[0058] The construction progress tracking module is used to intelligently track the construction progress of the target building based on the comprehensive gap value of each target component pair.

[0059] Optionally, the target component includes regular components and irregularly shaped components;

[0060] The design sub-BIM corresponding to the rule component is represented using the first parameter set;

[0061] For the irregularly shaped component and the design file of the irregularly shaped component contains the parameters required for NURBS representation, the design sub-BIM corresponding to the irregularly shaped component is represented by NURBS, and the NURBS corresponds to the second parameter set;

[0062] For the irregularly shaped component and the design file of the irregularly shaped component does not contain the parameters required for NURBS representation, the design sub-BIM corresponding to the irregularly shaped component is represented by Mesh, and the Mesh corresponds to the third parameter set;

[0063] And / or,

[0064] The actual sub-BIM corresponding to the rule component is represented using the first parameter set;

[0065] For the irregular component and the actual confidence level of the NURBS representation of the irregular component output by the preset BIM generation model is not less than the preset confidence level, the actual sub-BIM corresponding to the irregular component is represented by NURBS, and the NURBS corresponds to the second parameter set.

[0066] For the irregularly shaped component, and the actual confidence level corresponding to the NURBS representation of the irregularly shaped component output by the preset BIM generation model is less than the preset confidence level, the actual sub-BIM corresponding to the irregularly shaped component is represented by a Mesh, and the Mesh corresponds to the third parameter set.

[0067] Optionally, the preset BIM generation model is obtained based on several sets of sample training data;

[0068] The training data for each set of samples includes sample point cloud data and sample BIM corresponding to the sample buildings.

[0069] Optionally, for the regular component, the evaluation index includes at least one of the regular component's position, orientation, size, cross-section, and installation time;

[0070] For the irregularly shaped component, and both the design sub-BIM and the actual sub-BIM corresponding to the irregularly shaped component are represented by NURBS, the evaluation index includes at least one of the following: the position, orientation, installation time, shape, control grid, and weight of the control grid of the irregularly shaped component.

[0071] For the irregularly shaped component, and at least one of the design sub-BIM and the actual sub-BIM corresponding to the irregularly shaped component is represented by a Mesh, the evaluation index includes at least one of the following: position, orientation, installation time, shape, main direction, and size of the outer bounding box of the irregularly shaped component.

[0072] Optionally, the intelligent tracking system further includes:

[0073] The parameter value acquisition module is used to acquire the design parameter values ​​and actual parameter values ​​of the design component and the actual component respectively under several parameter information.

[0074] The matching determination module is used to determine that the design component and the actual component are mutually matched in response to the fact that the design parameter value and the actual parameter value under each parameter information meet the preset requirements.

[0075] Optionally, the parameter information includes at least one of the component's type, location, size, and shape.

[0076] Optionally, the difference value acquisition module includes:

[0077] The index value acquisition unit is used to obtain, for each of the target component pairs, the design index value and the actual index value corresponding to each of the evaluation indexes based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component.

[0078] The gap value acquisition unit is used to obtain the gap value based on the design index value and the actual index value;

[0079] And / or,

[0080] The comprehensive difference value acquisition module includes:

[0081] A weight acquisition unit is used to acquire the weight corresponding to each of the difference values;

[0082] A comprehensive gap value acquisition unit is used to obtain the comprehensive gap value of each target component pair based on each gap value and the corresponding weight;

[0083] And / or,

[0084] The construction progress tracking module is also used to determine that the construction progress of the corresponding design component is in a normal state in response to the comprehensive gap value falling within a first preset range.

[0085] In response to the comprehensive gap value falling within the second preset range, the construction progress of the corresponding design component is determined to be in an early warning state;

[0086] In response to the comprehensive gap value falling within a third preset range, the construction progress of the corresponding design component is determined to be in an alarm state;

[0087] Wherein, the upper limit of the first preset range is less than the lower limit of the second preset range, and the upper limit of the second preset range is less than the lower limit of the third preset range.

[0088] This disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and for running on the processor, wherein the processor executes the computer program to implement the above-described intelligent tracking method for building construction progress based on point cloud and BIM.

[0089] This disclosure also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described intelligent tracking method for building construction progress based on point cloud and BIM.

[0090] This disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements the intelligent tracking method for building construction progress based on point cloud and BIM as described above.

[0091] Based on common knowledge in the field, the above-mentioned preferred conditions can be combined arbitrarily to obtain various preferred embodiments of this disclosure.

[0092] The positive and progressive effects of this disclosure are as follows:

[0093] This disclosure obtains the difference values ​​of several evaluation indicators of the matching design components and actual components by comparing the design BIM of the target building and the actual BIM of the building. Then, a comprehensive difference value is obtained to intelligently track the construction progress of the target building. That is, by comparing the two BIMs and comprehensively considering multiple evaluation indicators, a comprehensive difference value is obtained to track the construction progress of the building, thereby improving the accuracy and reliability of the determination of the construction progress. Attached Figure Description

[0094] Figure 1 This is a first flowchart of the intelligent tracking method for building construction progress based on point cloud and BIM according to Embodiment 1 of this disclosure;

[0095] Figure 2 This is a second flowchart of the intelligent tracking method for building construction progress based on point cloud and BIM according to Embodiment 1 of this disclosure;

[0096] Figure 3 This is a flowchart of step S14 in the intelligent tracking method for building construction progress based on point cloud and BIM according to Embodiment 1 of this disclosure;

[0097] Figure 4 This is a flowchart of step S15 in the intelligent tracking method for building construction progress based on point cloud and BIM according to Embodiment 1 of this disclosure;

[0098] Figure 5 This is a schematic diagram of the first module of the intelligent tracking system for building construction progress based on point cloud and BIM according to Embodiment 2 of this disclosure;

[0099] Figure 6 This is a schematic diagram of the second module of the intelligent tracking system for building construction progress based on point cloud and BIM according to Embodiment 2 of this disclosure;

[0100] Figure 7 This is a schematic diagram of the structure of the electronic device according to Embodiment 3 of this disclosure. Detailed Implementation

[0101] The present disclosure is further illustrated below by way of embodiments, but the present disclosure is not limited to the scope of the embodiments described herein.

[0102] The prefixes such as "first" and "second" used in this disclosure are merely for distinguishing different descriptive objects and do not limit the position, order, priority, quantity, or content of the described objects. The use of ordinal numbers and other prefixes used to distinguish descriptive objects in this disclosure does not constitute a limitation on the described objects. The description of the described objects is given in the context of the embodiments, and the use of such prefixes should not constitute unnecessary restrictions. Furthermore, in the description of this embodiment, unless otherwise stated, "multiple" means two or more.

[0103] Example 1

[0104] This embodiment provides an intelligent tracking method for building construction progress based on point cloud and BIM, such as... Figure 1 As shown, the intelligent tracking method includes:

[0105] S11. Obtain the design BIM of the target building and the actual point cloud data of the actual building at the construction site;

[0106] The actual building is the building constructed according to the target building.

[0107] S12. Input the actual point cloud data into the preset BIM generation model to output the actual BIM of the actual building;

[0108] The target building includes several design components, and the design BIM includes a design sub-BIM corresponding to each design component; the actual building includes several actual components, and the actual BIM includes an actual sub-BIM corresponding to each actual component.

[0109] S13. Take the mutually matching design components and actual components as target component pairs;

[0110] S14. For each target component pair, based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component, obtain the difference values ​​of several evaluation indicators for each target component pair.

[0111] S15. Based on the gap value, obtain the comprehensive gap value for each target component pair;

[0112] Among them, the comprehensive gap value is used to characterize the current construction progress of the corresponding component;

[0113] S16. Based on the comprehensive gap value of each target component pair, intelligently track the construction progress of the target building.

[0114] Specifically, the actual building is obtained by constructing according to the target building. The actual point cloud data of the actual building at the construction site includes laser point cloud data and / or visual point cloud data.

[0115] Ground-based 3D laser scanners were used to acquire laser point cloud data, while high-resolution industrial cameras were used to acquire visual point cloud data. The actual point cloud data acquisition intervals were set according to the project construction milestones, covering key process stages such as structural construction, electromechanical installation, and decoration, ensuring that the data is representative of each stage. For example, when installing glass, data was collected once before installation and once after installation.

[0116] For actual point cloud data, preprocessing such as denoising, registration, and fusion can be performed to generate high-quality point cloud data of the construction site, providing a reliable data foundation for the training and comparison of the subsequent pre-set BIM generation model.

[0117] The visual point cloud data consists of multi-view image data. The image data undergoes distortion correction and exposure equalization. Camera pose and sparse point cloud are recovered using SfM (Structure from Motion), followed by MVS (Multi-View Stereo) to generate a dense visual point cloud. Camera pose is optimized using Bundle Adjustment (BA), and the visual point cloud is transformed to a unified engineering reference coordinate system through extrinsic parameter calibration or reference feature matching. The visual point cloud data is represented as P1={(x1,y1,z1,A1,N1)}, where x1, y1, and z1 represent a point P in the visual point cloud data P1. 11 The three-dimensional coordinates of point P are given by A1=(R,G,B). 11 In the RGB color space, N1 represents point P. 11 The direction of the law.

[0118] For laser point cloud data, FPFH (Fast Point Feature Histograms) features and RANSAC (Random Sample Consensus) are used to achieve coarse registration between ground-based 3D laser scanner stations. Then, ICP (Iterative Closest Point) is used for fine registration to unify the laser point cloud data to the engineering reference coordinate system. The laser point cloud data is represented as P2={(x2,y2,z2,I,N2)}, where x2, y2, and z2 represent a point P in the visual point cloud data P2. 21 The three-dimensional coordinates, where I represents point P. 21 The reflection intensity, N2 represents the reflection intensity at point P. 21 The direction of the law.

[0119] For actual point cloud data, adaptive resampling can be performed based on the geometric complexity of the components. The design components of the target building and the actual components of the actual building are collectively referred to as components, which include regular components and irregular components. High-density point clouds are maintained for complex areas such as irregular components, node connection areas, and curved curtain walls, while voxel mesh downsampling is performed for regular areas such as large-area flat panels and walls. This reduces storage and subsequent computation costs while ensuring detail fidelity and not affecting boundary recognition.

[0120] Laser point cloud data and visual point cloud data are fused. Geometrically, optimization is performed through ICP iterative registration and global error minimization to ensure that the position, size, and boundary contour of the same component are consistent in both types of point cloud data, achieving spatial alignment. Attribute-wise, the colors of the visual point cloud data are converted from the RGB color space to the Lab color space. The converted visual point cloud data is represented as follows: ={(x1,y1,z1,A2,N1)}, where A2=(L,a,b), representing point P. 11 In the Lab color space, L represents luminance, and a and b represent chromaticity. The reflection intensity I of the laser point cloud data is normalized to the [0,1] interval, and copied into three channels to align with the shape of the Lab color space. The converted laser point cloud data is represented as follows: ={(x2,y2,z2, ,N2)}, where, =(I norm ,I norm ,I norm ), representing point P 21 Aligned reflection intensity, I norm This represents the normalized reflection intensity, with a value range of [0,1]. Through projection consistency constraints, the color and intensity distributions of the two types of point cloud data at the same geometric location are kept consistent, thereby unifying the attributes of the point cloud data. The final output is high-precision multi-source fused point cloud data in a unified engineering reference coordinate system. The fused point cloud data is represented as P. fuse ={(x,y,z,A,N,s)}, where x, y, and z represent the fused point cloud data P. fuse The three-dimensional coordinates of a point P are given, where A represents the attribute value of point P corresponding to color or intensity. N represents the normal vector of point P, s represents the source identifier of point P, s∈{visual,lidar}, where visual indicates that point P originates from visual point cloud data, and lidar indicates that point P originates from laser point cloud data. This indicates that point P is derived from visual point cloud data in terms of brightness and chromaticity. This indicates the brightness of point P when it originates from laser point cloud data. =g(Inorm ), g() represents the mapping function from intensity to brightness obtained by calibration based on projection consistency constraints.

[0121] By using an adaptive cleaning and registration method based on actual buildings, the spatial geometric accuracy of actual point cloud data is guaranteed while achieving the unification of color and intensity attributes. Combined with adaptive resampling based on the geometric complexity of components, the detail restoration accuracy and comparability of actual building components, especially irregular components and complex nodes, are significantly improved.

[0122] In this solution, by comparing the design BIM of the target building with the actual BIM of the building, the difference values ​​of several evaluation indicators of the matching design components and actual components are obtained, and then a comprehensive difference value is obtained to intelligently track the construction progress of the target building. That is, by comparing the two BIMs and comprehensively considering multiple evaluation indicators, a comprehensive difference value is obtained to track the construction progress, thereby improving the accuracy and reliability of the determination of the construction progress.

[0123] In a feasible solution, the design components include regular components and irregular components;

[0124] The design sub-BIM corresponding to the regular component is represented using the first parameter set;

[0125] For irregularly shaped components and whose design files contain the parameters required for NURBS representation, the corresponding design sub-BIM of the irregularly shaped component is represented using NURBS, with NURBS corresponding to the second parameter set.

[0126] For irregularly shaped components whose design files do not contain the parameters required for NURBS representation, the corresponding design sub-BIM of the irregularly shaped component is represented by a Mesh, which corresponds to the third parameter set.

[0127] Specifically, the design BIM of the target building in the design phase is parameterized to unify its geometric parameters such as size information, shape features, connection relationships, and installation location, so as to ensure that the design BIM is comparable to the actual BIM generated later.

[0128] The parameter fields of the design BIM are standardized. The design BIM is divided into design sub-BIMs corresponding to regular components and design sub-BIMs corresponding to irregular components, i.e., BIMs. design =BIM design_regular ∪BIM design_unregular Among them, BIM design BIM design, BIM design_regular This represents the design sub-BIM corresponding to the regular component. design_unregular This represents the design sub-BIM corresponding to the irregularly shaped component.

[0129] Regular components include walls, beams, columns, slabs, etc., and the parameter fields of regular components are standardized as follows: , That is, the first parameter set.

[0130] Where ID represents the identifier of the component, Type represents the type of the component, L, W, and H represent the length, width, and height of the main dimensions of the regular component, respectively, and c represents the center of gravity or design reference point of the regular component, c=(x c ,y c ,z c )∈R 3 x c y c z c R represents the three-dimensional coordinates of the centroid or design reference point c of a regular component. 3 Let f represent the orientation quaternion of the regular component, which is used to characterize the orientation of the regular component's local coordinate system relative to the engineering reference coordinate system. f = (f... x ,f y ,f z ,f w )∈S 3 , satisfying ||f||=1, f x f y f z f w S represents the four components of a quaternion. 3 Representing a three-dimensional sphere, the direction cosine matrix R(f) corresponding to f is as follows:

[0131] ;

[0132] R(f) is an orthogonal matrix, and its determinant det(R) = 1. Section represents the cross-sectional parameters of the regular member, and the specific parameters are as follows:

[0133] ;

[0134] Where t represents the thickness of the regular member along the local normal direction, which is the direction facing the outside of the member; b and h represent the two principal dimensions of the local section of the regular member, namely the width and height; Wall indicates that the type of the regular member is a wall; Slab indicates that the type of the regular member is a slab; Beam indicates that the type of the regular member is a beam; and Column indicates that the type of the regular member is a column.

[0135] When the type of regular component is a wall or slab, the section parameter of the regular component is its thickness; when the type of regular component is a beam or column, the section parameters of the regular component are its width and height.

[0136] Irregularly shaped components include curved curtain walls, curved beams, and irregularly shaped panels. The parameter fields of irregularly shaped components are standardized as follows: Where ID represents the identifier of the component, Type represents the type of the component, and Rep represents the set of parameters other than ID and Type for the irregular component, where Rep∈{NURBS representation, Mesh representation}.

[0137] For irregularly shaped components, NURBS is preferred for representation. In this case, the design file of the irregularly shaped component, i.e., the source BIM file, contains the parametric definitions required for NURBS representation. When the design file of the irregularly shaped component only provides a mesh representation and does not contain the parametric definitions required for NURBS representation, Mesh representation is used.

[0138] When irregularly shaped components are represented using NURBS, Rep = NURBS(C,w,U,V,p,q). That is, the second parameter set.

[0139] Where C represents the set of control points for the irregularly shaped component, and the set of control points is a two-dimensional control grid or matrix, C={P ij}, P ij =(x ij ,y ij, z ij ), x ij y ij、 z ij Represent a control point P ij The three-dimensional coordinates of P; ij In this context, i and j represent the row and column indices of the control point in the control network, respectively; the total number of control points is (m+1)×(n+1); m and n represent the maximum index values ​​of the control network in the row and column, respectively, and m and n define the size of the control grid, which is directly related to the complexity of the NURBS surface, i.e., its order; w represents the set of weights for the control points, w={w ij}, w ij Represent a control point P ij The weights; U and V represent node vectors, U=[U0,…,U m+p+1 ],V=[V0,…,V n+q+1 The lengths of the node vectors are m+p+2 and n+q+2, respectively; the endpoint multiplicity is p+1 and q+1; the internal multiplicity is ≤p and ≤q; and the effective parameter domain is [u]. p ,u m+1 ]×[v q ,v n+1 ]; p and q represent the orders of the two-directional polynomials, satisfying p and q∈N, taking values ​​from 2 to 5, where N represents an integer.

[0140] When irregularly shaped components are represented using a mesh, Rep = Mesh(V, F, BB, PCA, Curv). That is, the third parameter set.

[0141] Where V and F represent the vertices and faces of the triangular mesh, respectively, V = {v k}∈R 3 F={△( , , )},v k Represents a vertex of a triangular mesh. , , This represents the three vertices of a facet of a triangular mesh; BB represents the minimum bounding box of the irregular component, used to provide the approximate dimensions and orientation of the irregular component. c BB、 R BB、 d represents the center, three axes, and three side lengths of the minimum bounding box, respectively; d1, d2, and d3 represent the length, width, and height of the minimum bounding box, respectively; Curv represents curvature, a numerical value used to characterize the degree of curvature of the surface of the irregular component at a certain point, and is a parameter expressing the shape characteristics of the irregular component; PCA represents the principal directions and eigenvalues ​​of the irregular component, used to characterize anisotropy. e1, e2, and e3 represent three eigenvectors calculated through principal component analysis, representing the three main directions of the irregular component's geometry, i.e., the principal axes, such as the longest, second longest, and shortest directions. These are used to align two components with the same shape but different orientations. e1 typically points to the longest direction of the irregular component. λ1, λ2, and λ3 represent the eigenvalues ​​corresponding to e1, e2, and e3, respectively, used to quantify the dimensions of the irregular component along these three principal axes to determine shape similarity. For example, for a slender rod, λ1 will be much larger than λ2 and λ3.

[0142] Standardize the connection relationships between components. Record the topological relationships between components, such as support, connection, overlap, and embedding, in the form of a relationship table, including the topological connection type (RelType), interface description (InterfaceNote), and contact area (InterfaceAreaA) between components. ij Boundary length between components ij It supports one-to-many and many-to-many scenarios, facilitating cross-software read and write operations.

[0143] Standardize the semantics and classification of components. Standardize component category coding, floor and area identifiers, and component sequence label T. planIt is associated with the work processes in the construction plan for type consistency constraints, floor area aggregation, and schedule alignment, without changing the plan data itself.

[0144] Establish a cross-platform mapping mechanism. Create a mapping table F between various modeling software such as Revit (a building information modeling software), Rhino (a freeform surface modeling software), and SketchUp (a 3D sketching software) and standardized parameter fields. in and inverse mapping table F out This clarifies field alignment, value specifications, and default value rules, enabling reversible conversion and version management, and ensuring that design BIMs from different sources can directly participate in subsequent mapping and comparison under a unified parameter system.

[0145] Perform consistency checks on parameter fields. Perform basic checks on parameters imported from the modeling software. Basic check rules include positive dimensions, valid orientation representations, self-consistent relationship tables, and unique IDs. Automatically generate a checklist and optional rule-based corrections without altering the design intent. Rule-based corrections include, for example, supplementing missing orientation parameters for components and standardizing the types of connection relationships between components.

[0146] For regular components, the first parameter set is used for representation; for irregular components, NURBS is preferred for representation. When the design file of an irregular component only provides a mesh representation and does not include the parametric definitions required for NURBS representation, Mesh is used for representation; lightweight relational tables are used to record the connection relationships between components; a mapping relationship is established between various modeling software and standardized parameter fields; design BIMs from different sources can be directly used for subsequent parameter mapping and multidimensional comparison under a unified semantic, improving the comparability of parameters and the efficiency of project implementation.

[0147] In this scheme, the design sub-BIM corresponding to regular and irregular components are represented in different ways according to the actual conditions met by the components, thus ensuring the accuracy and reliability of the design sub-BIM.

[0148] In a feasible solution, the actual components include regular components and irregularly shaped components;

[0149] The actual sub-BIM corresponding to the regular component is represented using the first parameter set;

[0150] For irregularly shaped components, the actual confidence level of the NURBS representation of the irregularly shaped component output by the pre-set BIM is not less than the pre-set confidence level. The actual sub-BIM corresponding to the irregularly shaped component is represented by NURBS, and NURBS corresponds to the second parameter set.

[0151] For irregularly shaped components, and the actual confidence level of the NURBS representation of the irregularly shaped component output by the preset BIM is less than the preset confidence level, the actual sub-BIM corresponding to the irregularly shaped component is represented by a Mesh, and the Mesh corresponds to the third parameter set.

[0152] Specifically, the pre-defined BIM generated model outputs an actual BIM with construction time sequence labels, which is represented as BIM. gen BIM gen =BIM gen_regular ∪BIM gen_unregular BIM gen_regular This represents the actual sub-BIM corresponding to the rule component. gen_unregular This represents the actual sub-BIM corresponding to the irregularly shaped component.

[0153] During the training process of the pre-defined BIM generative model, the sample BIM adopts the same modeling rules as the design BIM. Therefore, during the application of the pre-defined BIM generative model, the modeling rules of the actual BIM and the design BIM are also consistent. The design BIM includes a design sub-BIM corresponding to each component, and the actual BIM includes an actual sub-BIM corresponding to each component. Thus, the semantics of the geometric parameters such as dimensions, shape features, connection relationships, and installation positions of the actual sub-BIM and the design sub-BIM are consistent.

[0154] For irregularly shaped components, the improved PointTransformer will output a specific parameter NURBS representing the corresponding actual confidence level s. nurbs Normalizing this value to between 0 and 1 represents the generator network's confidence in whether its generated NURBS representation accurately represents the actual input point cloud data slice. This confidence level (s) can be adjusted by... nurbs With a preset, configurable threshold, i.e., a preset confidence level τ nurbs Comparisons are used to quantitatively define situations where confidence levels are insufficient. τ nurbs This is a hyperparameter of the generator network, which can be set according to accuracy requirements, for example, to 0.85. τ nurbs =0.85 indicates that if the generator network has a certain confidence level s corresponding to its generated NURBS representation, then the confidence level is 0.85. nurbs If the confidence level is below 85%, then the confidence level is insufficient, and a more robust Mesh representation will be selected for the irregular component.

[0155] Actual confidence level s nurbs The corresponding calculation formula is as follows:

[0156] ;

[0157] Where exp(-x) is the exponential decay function, with an output range of (0,1). x represents the penalty term; when the penalty term is 0, the fit is perfect. nurbs =exp(0)=1, indicating a confidence level of 100%; the larger the penalty term, the worse the fit, s nurbs The closer the value is to 0, the lower the confidence level. This structure ensures that the actual confidence level is always within a standardized and easily understood range. `E_fit_norm` represents the normalized fitting error, used to measure the geometric deviation between the fitted NURBS representation and the original point cloud (i.e., the sample BIM). `V_curv_norm` represents the normalized curvature variance, used to measure the smoothness of the component surface. `C_param_norm` represents the normalized parameter complexity, used to penalize cases where too many control points are used to barely achieve a fit, encouraging the model to express geometry using the simplest NURBS. `w_e` represents the weight of the fitting error. `w_c` represents the weight of the curvature variance. `w_p` represents the weight of the parameter complexity. The three hyperparameters `w_e`, `w_c`, and `w_p` are set according to the needs of the actual application scenario, satisfying `w_e + w_c + w_p = 1`. If the application scenario has the highest requirement for geometric accuracy, a higher `w_e` should be set, for example, `w_e = 0.6`, `w_c = 0.2`, `w_p = 0.2`. If you want to prioritize identifying naturally smooth components best suited for NURBS representation, you can appropriately increase w_c. If you want the model to be as concise and efficient as possible, you can increase w_p.

[0158] The calculation process for E_fit_norm is as follows:

[0159] First, calculate the original fitting error E_fit, which is usually calculated using the root mean square error. The corresponding calculation formula is as follows:

[0160] E_fit=sqrt((1 / N)*Σ||P_i- P'_i||²);

[0161] Where N represents the number of points of the component in the original point cloud, P_i represents the i-th point in the original point cloud, and P'_i represents the nearest projection point of P_i onto the fitted NURBS representation.

[0162] Then, E_fit is normalized and scaled to an interval to eliminate the influence of dimensions. The corresponding calculation formula is as follows:

[0163] E_fit_norm=min(1,E_fit / E_max);

[0164] Here, E_max represents a preset maximum permissible error threshold, for example, set to 50mm according to engineering accuracy requirements. When the actual error exceeds this threshold, the normalized error is capped at 1.

[0165] The calculation process of V_curv_norm is as follows:

[0166] First, for each point P_i belonging to the component in the original point cloud, its principal curvature is estimated through its neighboring points to obtain k_i.

[0167] Then calculate the variance V_curv of the curvature at all points, using the following formula:

[0168] V_curv=Variance({k_1,k_2……k_N});

[0169] A flat or smoothly curved surface, such as a glass curtain wall, has roughly the same curvature value at all points, so the variance V_curv is very low; a surface full of bumps and edges, such as a steel cage, has drastic changes in curvature value, so the variance V_curv is very high.

[0170] Finally, V_curv is normalized, and the corresponding calculation formula is as follows:

[0171] V_curv_norm=min(1,V_curv / V_max);

[0172] Where V_max represents a preset maximum allowable curvature variance threshold.

[0173] The calculation process for C_param_norm is as follows:

[0174] First, calculate the original complexity C_param. This is typically done using the logarithm of the total number of control points to reflect the diminishing marginal utility. The corresponding formula is as follows:

[0175] C_param=log[(m+1)*(n+1)];

[0176] Where m and n represent the order of the NURBS surface in two directions, and (m+1)*(n+1) represents the total number of control points.

[0177] Then, C_param is normalized, and the corresponding calculation formula is as follows:

[0178] C_param_norm=min(1,C_param / C_max);

[0179] Where C_max represents a preset maximum allowed complexity threshold, for example, log(100*100).

[0180] In this scheme, the actual sub-BIM corresponding to regular and irregular components are represented in different ways according to the actual conditions met by the components, thus ensuring the accuracy and reliability of the actual sub-BIM.

[0181] In one feasible approach, the pre-defined BIM generative model is obtained based on several sets of sample training data;

[0182] Each set of sample training data includes sample point cloud data and sample BIM corresponding to the sample building.

[0183] Specifically, a training library is constructed using measured point cloud data from multiple engineering projects and construction nodes, along with their reverse-engineered BIM models, with parameters standardized during the library's construction. Based on this training library, an improved PointTransformer (a deep learning model) is used to train a point cloud-to-BIM generative network, i.e., a pre-defined BIM generative model, to generate the actual BIM of the construction site. Construction timeline labels are embedded to achieve dynamic correlation with the construction schedule.

[0184] The training library includes several sets of sample training data. Each set of sample training data includes sample point cloud data and sample BIM corresponding to the sample building. The sample point cloud data is the measured point cloud data, which can be used for fusion point cloud data. Sample BIM refers to BIM generated from measured point cloud data, and is represented as BIM. As_built BIM As_built During modeling, the same modeling rules as the design BIM are adopted, that is, the parameters of the sample BIM are standardized to unify its geometric parameters such as size information, shape features, connection relationships, and installation location, so as to ensure the semantic consistency of the parameters.

[0185] An improved PointTransformer network architecture is adopted, introducing anisotropic weights guided by local covariance in each attention layer. The encoder extracts multi-scale neighborhood features, and the decoder outputs a 128-dimensional parameter vector, which includes parameter branches for regular components, parameter branches for irregular components, and the actual confidence s corresponding to the NURBS representation of the irregular components. nurbs Output of connection relationships between components, output of construction sequence labels, etc.

[0186] Model training is performed using PyTorch (an open-source machine learning framework) and CUDA (a parallel computing platform and programming model). The training process seeks to minimize the value of the total loss function. The formula for calculating the total loss function is as follows:

[0187] ;

[0188] ;

[0189] ;

[0190] ;

[0191] Among them, L total L represents the total loss function; geo d represents the geometric loss function, used to measure the geometric similarity between the actual BIM output by the pre-defined BIM generative model and the sample BIM. It considers both point position and surface orientation. Point position is represented by chamfer distance, and surface orientation is represented by normal difference; chamfer The chamfer distance is primarily used to measure the spatial positional error of point sets between the actual BIM and the sample BIM. This distance comprehensively considers the most classic bidirectional average distance between the actual BIM and the sample BIM, effectively assessing macroscopic differences in location, size, and shape; d normal The normal difference is represented by the coefficient μ, which constrains the surface orientation of the model by calculating the difference between the normal vectors of corresponding points in the actual BIM and the sample BIM. This constrains inconsistencies in surface orientation and ensures the accuracy of the orientation. For example, it calculates the cosine deviation of the angle between the normal vectors of the closest point pairs in the actual BIM and the sample BIM. quality This represents the parametric quality loss function, expressed through the actual BIM parameter vector s. i Compared with the actual parameters s in the sample BIM i The mean square error between * is used to calculate and force the network to output numerically correct parameter values, such as length, width, and control point coordinates; s i s represents the i-th parameter of the actual BIM. i * indicates the i-th parameter of the sample BIM; K represents the total number of parameters in the actual BIM or sample BIM. For example, if a regular component is defined by three scalars (length, width, and height), three scalars (centroid coordinates), and four scalars (orientation quaternions), then K = 10; for complex NURBS surfaces, K is the sum of all control point coordinates and weight values; L time The time-series loss function is expressed as the normalized actual BIM installation time T and the normalized planned installation time T. plan The squared L2 norm of the distance between the two points is calculated, i.e., the squared Euclidean distance, to ensure that the model can accurately predict the construction sequence; φ() represents the normalization function, used to normalize the installation time to [0,1]; λ g , λ q , λ t λ represents the weight hyperparameter, used to control the relative importance of geometry, parameter quality, and temporal loss function in the total loss function. g , λ q , λ tAll are greater than 0.

[0192] During training monitoring, if L is found quality It has decreased significantly, but L geo The parameter values ​​remain persistently high, meaning the model outputs seemingly reasonable values, but the instantiated 3D model and the point cloud are severely mismatched in space. This indicates that the model may be stuck in rote memorization of parameter distributions without truly understanding geometric constraints. In this case, λ needs to be increased. g To correct the direction of learning.

[0193] When the training set contains a large number of irregularly shaped components such as curved curtain walls and irregularly shaped beams, its core value lies in the accuracy of its geometric form. For such components, geometric fidelity should be given the highest priority, therefore a high λ value needs to be set. g .

[0194] In the initial stage of model training, λ can be appropriately increased. g The guided model first learns to roughly locate and shape-fit point clouds, establishing a macroscopic understanding of spatial geometry and laying the foundation for subsequent refined parameter regression.

[0195] When the input point cloud contains noise, occlusion, or uneven density, a higher λ g This can enhance the model's robustness to the overall shape. Because d chamfer These indicators evaluate the overall distribution and can effectively suppress overfitting of the model to local noise points.

[0196] In geometric loss function L geo In the middle, relying solely on the chamfer distance d chamfer Insufficient to penalize incorrect surface orientation. For example, an actual BIM image of a surface with wrinkles or completely incorrect orientation may still result in a lower chamfer distance. Normal difference d normal The introduction of this feature can compensate for this deficiency. By minimizing this loss term, the generative network is not only explicitly guided to learn "where the surface is", but also explicitly guided to learn "how the surface should be oriented in various places".

[0197] During model training, data augmentation and class equalization sampling are employed, including rotation, scaling, noise, and occlusion adjustments; construction time series labels are embedded as conditional vectors to participate in decoding.

[0198] By constructing standardized parameters for actual BIM and employing an improved PointTransformer (a deep learning model) network architecture, the parameter vectors of regular and irregular components are decoded. Combined with the joint loss function of geometry, parameter quality, and temporal sequence, the integrated parameter-level reconstruction of regular and irregular components is achieved.

[0199] In this solution, a pre-defined BIM generation model is trained using sample point cloud data corresponding to the sample building and sample BIM training data. This ensures the accuracy and reliability of the pre-defined BIM generation model and guarantees the accurate output of the actual BIM corresponding to the actual point cloud data of the subsequent construction site.

[0200] In a feasible solution, for regular components, the evaluation metrics include at least one of the following: the position, orientation, size, cross-section, and installation time of the regular component.

[0201] For irregularly shaped components, both the design sub-BIM and the actual sub-BIM corresponding to the irregularly shaped components are represented by NURBS. The evaluation indicators include at least one of the following: the location, orientation, installation time, shape, control grid, and weight of the control grid of the irregularly shaped components.

[0202] For irregularly shaped components, at least one of the design sub-BIM and the actual sub-BIM corresponding to the irregularly shaped component is represented by a Mesh. The evaluation indicators include at least one of the following: the location, orientation, installation time, shape, main direction, and size of the outer bounding box of the irregularly shaped component.

[0203] Specifically, the design BIM is matched with the actual BIM. A parameter mapping relationship is established between the design BIM and the actual BIM, and a multi-dimensional comparative analysis is conducted to generate quantitative results of differences in location, size, orientation, shape, and installation time.

[0204] Under the standard parameter system, the parameter θ of the design BIM (design) Compared with the actual BIM parameter θ (gen) Perform a one-to-one mapping and output the difference between fields with the same name and the necessary unified evaluation metrics across types.

[0205] For regular components, the evaluation index includes at least one of the component's position, orientation, dimensions, cross-section, and installation time. The comparable parameter difference corresponding to the evaluation index, i.e., the difference value, includes at least one of the following: position difference, orientation difference, dimension difference, cross-section difference, and installation time difference. The corresponding calculation formula is as follows:

[0206] Δc=c (gen) -c (design) ;

[0207] Where Δc represents the position difference, c (gen) c represents the centroid or design reference point of the actual sub-BIM. (design) Indicates the center of gravity or design reference point of the sub-BIM;

[0208] Δf=∠(f (gen) ,f (design) );

[0209] Where Δf represents the attitude difference: f (gen) f represents the orientation quaternion of the actual sub-BIM. (design) The quaternion representing the orientation of the design sub-BIM;

[0210] ΔL=L (gen) -L (design) ;

[0211] Where ΔL represents the length difference in the size difference, L (gen) L represents the actual length of the sub-BIM. (design) Indicates the length of the design sub-BIM;

[0212] ΔW=W (gen) -W (design) ;

[0213] Where ΔW represents the width difference in the dimensional difference, W (gen) W represents the width of the actual sub-BIM. (design) Indicates the width of the design sub-BIM;

[0214] ΔH=H (gen) -H (design) ;

[0215] Where ΔH represents the height difference in the dimensional difference, H (gen) H represents the actual height of the sub-BIM. (design) Indicates the height of the design sub-BIM;

[0216] Type∈{Wall,Slab},Δt=t (gen) -t (design) ;

[0217] Type∈{Beam,Column}, Δb=b (gen) -b (design) , Δh=h (gen) -h (design) ;

[0218] Where Δt represents the thickness difference in the cross-sectional difference, t (gen) The thickness of the actual sub-BIM is represented by t. (design) The thickness of the sub-BIM design is represented by Δb; Δb represents the width difference in the cross-sectional differences, b (gen) Indicates the width of the actual sub-BIM, b (design) The width of the sub-BIM design is represented by Δh; Δh represents the height difference in the section difference, h (gen) h represents the actual height of the sub-BIM. (design) Indicates the height of the design sub-BIM;

[0219] ΔT=T (gen) -T (design) ;

[0220] Where ΔT represents the installation time difference, T (gen) T represents the actual installation time of the sub-BIM. (design) This indicates the installation time of the sub-BIM design.

[0221] For irregularly shaped components, the evaluation index includes at least one of the component's position, orientation, installation time, and shape. The comparable parameter difference corresponding to the evaluation index, i.e., the difference value, includes at least one of the position difference, orientation difference, installation time difference, and shape difference. The calculation formulas for position difference, orientation difference, and installation time difference are similar to those for regular components, and will not be repeated here.

[0222] The shape difference can be calculated in two ways, one of which is the 95th quantile of the symmetric Hausdorff distance. This is a robust shape similarity metric used to measure the distance between two sets of points, while being insensitive to a small number of outliers; another approach is the average point-to-surface distance between two grids. It provides a measure of the average deviation between two shapes.

[0223] For irregularly shaped components where both the design sub-BIM and the actual sub-BIM are represented using NURBS, the surfaces corresponding to both the design and actual sub-BIMs are first subjected to degree lifting and node refinement down to a common node vector. This standardizes the two NURBS surfaces to the same mathematical framework, enabling precise mathematical calculations. Evaluation metrics also include at least one of the following: the control mesh of the irregularly shaped component and the weights of the control mesh. The comparable parameter difference corresponding to the evaluation metric, i.e., the difference value, includes at least one of the root mean square difference of the control mesh and the difference in the weights of the control mesh. The corresponding calculation formula is as follows:

[0224] ;

[0225] in, The root mean square error between the control points of the actual sub-BIM and the design sub-BIM is expressed as the root mean square error of the control grid, which measures the average geometric deviation of the control grid. Represents the control points of the actual sub-BIM. These represent the control points in the design sub-BIM. For curved surfaces, the control points are connected to form a control grid, which outlines the basic shape of the surface.

[0226] ;

[0227] Where Δw represents the weight difference between the actual sub-BIM and the design sub-BIM; w (gen)Represents the set of weights for the actual sub-BIM; w (design) This represents the set of weights for the design sub-BIM; This represents the infinite norm or maximum norm, which is the maximum absolute value of the difference between corresponding elements of two weight matrices.

[0228] For irregularly shaped components, where at least one of the design sub-BIM and the actual sub-BIM corresponding to the irregularly shaped component is represented by a Mesh, the evaluation indicators also include at least one of the main direction of the irregularly shaped component and the size of the outer bounding box. The comparable parameter difference corresponding to the evaluation indicators, i.e., the difference value includes at least one of the main direction angle difference and the minimum outer bounding box size difference, is calculated using the following formula:

[0229] ;

[0230] ;

[0231] ;

[0232] ;

[0233] Where ΔBB represents the size difference between the actual sub-BIM and the design sub-BIM's minimum bounding box, and Δd1 represents the length difference within the size difference of the minimum bounding box. This indicates the length of the smallest outer bounding box of the actual sub-BIM. Δd2 represents the length of the minimum bounding box of the design sub-BIM, and Δd2 represents the width difference within the dimensional differences of the minimum bounding box. This represents the width of the minimum outer bounding box of the actual sub-BIM. Δd3 represents the width of the minimum bounding box of the design sub-BIM, and Δd3 represents the height difference within the dimensional differences of the minimum bounding box. This indicates the height of the minimum outer bounding box of the actual sub-BIM. This indicates the height of the minimum outer bounding box of the design sub-BIM;

[0234] ΔPCA=∠(PCA (design) PCA (gen) );

[0235] Where ΔPCA represents the difference in principal directions between the actual sub-BIM and the design sub-BIM, PCA (gen) PCA represents the main direction of the actual sub-BIM. (design) This indicates the main direction of the design sub-BIM.

[0236] For all components, a difference quantification result table can be output. The table includes the following fields: `genid`: the ID of the actual sub-BIM of the component; `designid`: the ID of the design sub-BIM of the component; `type`: the type of the component; `match_status`: the matching result between the actual sub-BIM and the design sub-BIM of the component, including both matched and non-matching results; `Δc`: position difference; `Δf`: orientation difference; `ΔL`: length difference; `ΔW`: width difference; `ΔH`: height difference; `Δt`: thickness difference in section differences; `Δb`: width difference in section differences; `Δh`: height difference in section differences. : Represents the 95th percentile of the symmetric Hausdorff distance. : Represents the average distance from a point to a surface. : represents the root mean square difference of the control network, Δw: represents the weight difference, ΔBB: represents the size difference of the smallest outer bounding box, ΔPCA: represents the difference in the included angle of the principal direction, ΔT: represents the installation time difference.

[0237] All components have the fields genid, designid, type, match_status, Δc, Δf, and ΔT.

[0238] For regular components, when the type is wall or slab, it has fields ΔL, ΔW, ΔH, and Δt, and the remaining fields in the difference quantification result table are left empty; when the type is beam or column, it has fields ΔL, ΔW, ΔH, Δb, and Δh.

[0239] For irregularly shaped components, when both the design sub-BIM and the actual sub-BIM are represented using NURBS, they have fields. , , Δw; When at least one of the design sub-BIM and the actual sub-BIM is represented by a Mesh, it has fields , , ΔBB, ΔPCA.

[0240] In this scheme, the evaluation indicators for regular and irregular components are determined based on the actual conditions met by the components, thus ensuring the accuracy and reliability of the evaluation indicators.

[0241] In a feasible solution, such as Figure 2 As shown, before step S13, the following steps are also included:

[0242] S131. Obtain the design parameter values ​​and actual parameter values ​​of the designed component and the actual component under several parameter information;

[0243] S132. In response to the fact that the design parameter value and the actual parameter value under each parameter information meet the preset requirements, it is determined that the design component and the actual component are matched.

[0244] Specifically, in the process of matching design components with actual components, a candidate set is first selected based on the type and location constraints of the components. Within the candidate set, a small number of alternative sub-BIM pairs are obtained based on their centroid proximity, namely the actual sub-BIM and the design sub-BIM. The centroid proximity can be achieved using the k-nearest neighbor algorithm, where k can be 3, 4, or 5.

[0245] The candidate sub-BIM pairs undergo a fine screening process. This fine screening consists of three steps, with each pair passing the screening process to be considered a successful match: location verification, dimension verification, and shape verification.

[0246] Position verification: ||Δc||≤δ c δ c =10mm;

[0247] Dimensional verification: For regular components, ||ΔL||≤δ L , ||ΔW||≤δ W , ||ΔH||≤δ H δ L =δ W =δ H =max(0.1*L) （design） (10mm)

[0248] For irregularly shaped components, take the local axis after alignment. ,|Δd k |≤10mm, k=1,2,3;

[0249] Shape verification: Shape verification is only applied to irregularly shaped components. When both the design sub-BIM and the actual sub-BIM corresponding to the irregularly shaped component are represented using NURBS, the surfaces corresponding to the design sub-BIM and the actual sub-BIM are first subjected to degree boosting and node refinement to common node vectors before judgment. and ;

[0250] If the above formula is not satisfied, then judge ;

[0251] Where, ε ctrl =5mm, ε w =0.1, δ H =10mm, =5mm.

[0252] If any of the above fine screening steps are not passed, the matching result of the corresponding component will be marked as not matching in the difference quantification result table; if all are passed, they will be marked as matching.

[0253] The above values ​​are just an example and can be adjusted according to the actual situation.

[0254] In this scheme, when the design parameter values ​​and actual parameter values ​​of the designed component and the actual component under each parameter information meet the preset requirements, the designed component and the actual component are determined to be matched with each other, thus ensuring the accuracy and reliability of the matching process between the designed component and the actual component.

[0255] In an feasible solution, the parameter information includes at least one of the component's type, location, size, and shape.

[0256] In this scheme, the design components and actual components are matched using a variety of parameters to ensure the accuracy and reliability of the matching determination.

[0257] In a feasible solution, such as Figure 3 As shown, step S14 includes:

[0258] S141. For each target component pair, based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component, the design index value and actual index value corresponding to each evaluation index are obtained.

[0259] S142. Based on the design index value and the actual index value, obtain the gap value.

[0260] Specifically, design BIM is represented as BIM. design In practice, BIM is represented as BIM. gen Regarding BIM design Corresponding design components and BIM gen After matching the corresponding actual components, the difference value is calculated for each matched design component and actual component under each evaluation index. The design index value is the parameter value of the design sub-BIM under the evaluation index; the actual index value is the parameter value of the actual sub-BIM under the evaluation index. For example, if the evaluation index is attitude, the design index value is the orientation quaternion of the design sub-BIM, and the actual index value is the orientation quaternion of the actual sub-BIM. The relevant content of the difference value has been explained above and will not be repeated here. For mismatched design components and actual components, their respective parameter information is recorded in the difference quantification result table.

[0261] In this scheme, the gap value is calculated for each evaluation index for the matching design components and actual components, ensuring the accuracy and reliability of the gap value.

[0262] In a feasible solution, such as Figure 4 As shown, step S15 includes:

[0263] S151. Obtain the weight corresponding to each difference value;

[0264] S152. Based on each gap value and its corresponding weight, obtain the comprehensive gap value for each target component pair.

[0265] Specifically, based on the gap values ​​of several evaluation indicators of the target component pair, namely the parameter-level difference vector and the installation time difference, a unified comprehensive gap value R is constructed and a configurable threshold is set. The component status is automatically determined and the construction plan is written back, generating a deviation heat map and a progress Gantt chart, realizing an integrated closed-loop management of judgment, early warning, feedback and visualization driven by difference quantification.

[0266] Construction progress includes construction time schedule and construction deviation status.

[0267] The construction progress is determined, and a time threshold δ is set. t =1 day, if |ΔT|<δ t If ΔT≤-δ, then construction is considered to be proceeding according to the planned schedule; t If Δt ≥ δ, then it is considered advanced construction; t If so, it is considered that the construction is delayed.

[0268] To determine the state of construction deviation, a comprehensive gap value R needs to be constructed, and the corresponding calculation formula is as follows:

[0269] ;

[0270] ;

[0271] Where R represents the overall difference value; w c w f w size w shape w T These represent the weights corresponding to the position difference, orientation difference, size difference, shape difference, and installation time difference, respectively. c +w f +w size +w shape +w T =1; Indicates positional tolerance. Indicates attitude tolerance, Indicates the installation time tolerance. Indicates length tolerance. Indicates width tolerance. Indicates height tolerance, Indicates the cross-sectional tolerance. =10mm, =2°, = = =10mm, =5mm.

[0272] For irregularly shaped components S=0, w size =0; for regular component D=0, w shape =0.

[0273] The values ​​for the various tolerances mentioned above are just examples and can be set or adjusted according to the actual situation.

[0274] The weights w corresponding to position difference, orientation difference, size difference, shape difference, and installation time difference c w f w size w shape w T The configuration should be a dynamic, multi-dimensional decision-making process, and these weights can be adjusted according to the construction stage, component function, and importance.

[0275] During the foundation and main structure construction phase, to ensure the geometric accuracy and safety of the structural framework and to establish precise benchmarks for all subsequent construction work, the weights are set as follows:

[0276] The Δc of load-bearing components is decisive; any deviation will be amplified layer by layer, making it the primary indicator of structural safety. Therefore, compared to other weights, a larger w is set. c .

[0277] The verticality and horizontality of the associated components, represented by Δf, are crucial for structural stability. Therefore, a larger value is assigned to w compared to other weights. f .

[0278] Dimensional deviation S is relatively important, but its priority is lower than position and orientation. Provided that positioning and orientation are correct, minor dimensional deviations, such as beam cross-sectional dimensions, are usually within the tolerances allowed by the specifications. Therefore, a moderately large w is set. size .

[0279] Compared to permanent and irreversible deviations such as structural position and attitude, catch-up ΔT carries a lower risk level. At this stage, doing things correctly is far more important than doing them quickly; therefore, a smaller w is set. T .

[0280] At this stage, the components are mostly regular components, and their shape difference D and weight w are defined accordingly. shape All are 0.

[0281] During the secondary structure and electromechanical installation phase, it is necessary to ensure spatial coordination and assembly accuracy under the cross-disciplinary operations. The weights for each are set as follows:

[0282] Δc determines whether the equipment foundation, pipeline supports, and reserved holes can be accurately positioned, and is the primary factor in avoiding spatial conflicts. Therefore, compared to other weights, a larger w is set. c .

[0283] In this stage, the processes are tightly coupled, and a delay in ΔT will directly block the construction of multiple subsequent work surfaces, causing a chain reaction of delays. This is the core of the schedule risk in this stage. Therefore, a larger w is set. T .

[0284] Dimensional deviations (S) of components and pre-drilled holes require close monitoring to ensure smooth installation. However, their risk priority is generally lower than catastrophic spatial errors or critical path delays. Therefore, setting a medium-sized w is acceptable. size .

[0285] Pipeline slope, equipment orientation, etc., are important, but their adjustment flexibility and impact are usually smaller than that of location deviation. Therefore, setting a moderate w is sufficient. f .

[0286] Unless it involves irregularly shaped ducts or customized components, most electromechanical components have relatively fixed shapes, and the weighting term w shape Lower, the lowest setting can be 0.

[0287] During the curtain wall and interior decoration phase of construction, it is necessary to ensure spatial coordination and assembly accuracy under the cross-disciplinary work. The weights for each are set as follows:

[0288] Δc determines whether the seams between panels are uniform and whether the lines are aligned, directly reflecting the appearance quality. Therefore, compared to other weights, a larger w is set. c .

[0289] Dimensional deviation S directly affects the size of the seam and physical properties such as sealing and waterproofing. Therefore, compared to other weights, a larger w is set. size .

[0290] For curved curtain walls and irregularly shaped ceilings, shape deviation D is the core indicator for evaluating their design fidelity and aesthetic effect. Therefore, compared to other weights, a larger w is set. shape .

[0291] The flatness of the panel, the installation angle, and other factors (Δf) will affect the lighting effects and functionality, but their priority is slightly lower than the decisive factors of position, size, and shape. Therefore, setting a medium-sized w... f .

[0292] In this final stage, delivering quality is the primary objective. To ensure the final result, tolerating a certain ΔT lag is often necessary; therefore, a smaller w is set. T .

[0293] The weights are adjusted based on the function and importance of the components. For major load-bearing components, such as columns and beams, a larger weight (w) is assigned. c and w f medium-sized wsize smaller w T w shape Setting it to 0 here is consistent with the weight configuration logic of the foundation and main structure construction phase, emphasizing geometric accuracy far more than schedule.

[0294] For prefabricated components, such as PC wall panels, a larger w is configured c w f w size medium-sized w T smaller w shape w shape The minimum configurable value is 0. Positional difference, orientation difference, and dimensional difference together define assembly tolerances, and all three are equally important. ΔT is also critical because the production and transportation schedules for prefabricated components are usually very tight.

[0295] For irregularly shaped / external enclosure components, such as curved curtain wall units, a larger w is configured. c w size、 w shape medium-sized w f smaller w T This configuration aligns with the logic of the curtain wall and interior decoration construction phase, emphasizing the final visual and physical quality.

[0296] For critical construction components or long-cycle equipment, such as chillers, a larger W is required. c w T medium-sized w size w f smaller w shape w shape The minimum configurability is 0. ΔT is the most significant source of risk. Δc must be precise to ensure proper installation on the equipment foundation. Dimension S and orientation difference Δf are secondary risks, while shape is generally not a major risk factor for this type of equipment.

[0297] In this scheme, based on each gap value and its corresponding weight, the comprehensive gap value of each component pair is obtained, which ensures the accuracy and reliability of the comprehensive gap value and improves the accuracy and reliability of the judgment of construction deviation status.

[0298] In one feasible embodiment, step S16 includes:

[0299] In response to the fact that the comprehensive gap value falls within the first preset range, the construction progress of the corresponding actual component is determined to be in a normal state;

[0300] In response to the comprehensive gap value falling within the second preset range, the construction progress of the corresponding design component is determined to be in an early warning state;

[0301] In response to the comprehensive gap value falling within the third preset range, the construction progress of the corresponding design component is set to an alarm state;

[0302] The upper limit of the first preset range is less than the lower limit of the second preset range, and the upper limit of the second preset range is less than the lower limit of the third preset range.

[0303] Specifically, if the overall difference value R < 1, the construction deviation status is determined to be normal; if 1.0 ≤ R < 1.5, the construction deviation status is determined to be warning status; if R ≥ 1.5, the construction deviation status is determined to be alarm status. R < 1 is the first preset range, 1.0 ≤ R < 1.5 is the second preset range, and R ≥ 1.5 is the third preset range.

[0304] Furthermore, if the construction progress of a designed component is in a warning or alarm state, status information feedback and resource optimization can be implemented. Information about the designed component in a warning or alarm state is written back to the construction planning system, automatically adjusting the subsequent process sequence, optimizing resource allocation, and forming a closed-loop log according to adjustment rules. This facilitates review and parameter adaptation. Adjustment rules include, for example, prioritizing lagging processes, prioritizing advanced processes, or freezing dependent processes; optimizing resource allocation includes, for example, adding work teams / hoisting crews, inserting retesting and re-registration, and adding quality inspection batches; forming a closed-loop log includes, for example, recording the current round's thresholds, weights, adjustment actions, and execution results.

[0305] It can also build a visual management platform and provide decision support. It can generate deviation heatmaps and progress Gantt charts in conjunction with each other, support component-level alarms and export reports in CSV / JSON (comma-separated value file format / lightweight data exchange format), and realize real-time monitoring and decision support for key components.

[0306] By using a parameter-level difference vector, including the installation time difference, as a unified input, a comprehensive gap value with configurable thresholds is constructed and linked bidirectionally with the construction plan to achieve a single-step closed loop of judgment, early warning, feedback, and visualization, ensuring intelligent tracking and early warning of the construction progress at the component level of the target building.

[0307] In this scheme, the construction progress of the corresponding design components is determined according to the preset range to which the comprehensive gap value belongs, thus ensuring the accuracy and reliability of the component construction progress judgment.

[0308] In this embodiment, by comparing the design BIM of the target building with the actual BIM of the building, the difference values ​​of several evaluation indicators of the matching design components and actual components are obtained, and then a comprehensive difference value is obtained to intelligently track the construction progress of the target building. That is, by comparing the two BIMs and comprehensively considering multiple evaluation indicators, a comprehensive difference value is obtained to track the construction progress of the building, thereby improving the accuracy and reliability of the determination of the construction progress.

[0309] Example 2

[0310] Corresponding to the aforementioned embodiments of the intelligent tracking method for building construction progress based on point cloud and BIM, this disclosure also provides embodiments of an intelligent tracking system for building construction progress based on point cloud and BIM.

[0311] like Figure 5 As shown, the intelligent tracking system includes:

[0312] Design BIM Acquisition Module 1 is used to acquire the design BIM of the target building;

[0313] Point cloud data acquisition module 2 is used to acquire actual point cloud data of the actual building at the construction site;

[0314] The actual building is the building constructed according to the target building.

[0315] Actual BIM output module 3 is used to input actual point cloud data into the preset BIM generation model to output the actual BIM of the actual building;

[0316] The target building includes several design components, and the design BIM includes a design sub-BIM corresponding to each design component; the actual building includes several actual components, and the actual BIM includes an actual sub-BIM corresponding to each actual component.

[0317] Component pair determination module 4 is used to identify mutually matching design components and actual components as target component pairs;

[0318] The gap value acquisition module 5 is used to obtain the gap value of several evaluation indicators for each target component pair based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component.

[0319] The comprehensive gap value acquisition module 6 is used to obtain the comprehensive gap value of each target component pair based on the gap value;

[0320] Among them, the comprehensive gap value is used to characterize the current construction progress of the corresponding component;

[0321] The construction progress tracking module 7 is used to intelligently track the construction progress of the target building based on the comprehensive gap value of each target component pair.

[0322] In a feasible solution, the target components include regular components and irregularly shaped components;

[0323] The design sub-BIM corresponding to the regular component is represented using the first parameter set;

[0324] For irregularly shaped components and whose design files contain the parameters required for NURBS representation, the corresponding design sub-BIM of the irregularly shaped component is represented using NURBS, with NURBS corresponding to the second parameter set.

[0325] For irregularly shaped components and whose design files do not contain the parameters required for NURBS representation, the design sub-BIM corresponding to the irregularly shaped component is represented by a Mesh, which corresponds to the third parameter set.

[0326] And / or,

[0327] The actual sub-BIM corresponding to the regular component is represented using the first parameter set;

[0328] For irregularly shaped components, the actual confidence level of the NURBS representation of the irregularly shaped component output by the pre-set BIM is not less than the pre-set confidence level. The actual sub-BIM corresponding to the irregularly shaped component is represented by NURBS, and NURBS corresponds to the second parameter set.

[0329] For irregularly shaped components, and the actual confidence level of the NURBS representation of the irregularly shaped component output by the preset BIM is less than the preset confidence level, the actual sub-BIM corresponding to the irregularly shaped component is represented by a Mesh, and the Mesh corresponds to the third parameter set.

[0330] In one feasible approach, the pre-defined BIM generative model is obtained based on several sets of sample training data;

[0331] Each set of sample training data includes sample point cloud data and sample BIM corresponding to the sample building.

[0332] In a feasible solution, for regular components, the evaluation metrics include at least one of the following: the position, orientation, size, cross-section, and installation time of the regular component.

[0333] For irregularly shaped components, both the design sub-BIM and the actual sub-BIM corresponding to the irregularly shaped components are represented using NURBS. The evaluation indicators include at least one of the following: the location, orientation, installation time, shape, control grid, and weight of the control grid of the irregularly shaped components.

[0334] For irregularly shaped components, at least one of the design sub-BIM and the actual sub-BIM corresponding to the irregularly shaped component is represented by a Mesh. The evaluation indicators include at least one of the following: the location, orientation, installation time, shape, main direction, and size of the outer bounding box of the irregularly shaped component.

[0335] In a feasible solution, such as Figure 6 As shown, the intelligent tracking system also includes:

[0336] The parameter value acquisition module 8 is used to acquire the design parameter values ​​and actual parameter values ​​of the design component and the actual component under several parameter information.

[0337] The matching determination module 9 is used to determine the mutual matching between the design component and the actual component in response to the fact that the design parameter value and the actual parameter value under each parameter information meet the preset requirements.

[0338] In an feasible solution, the parameter information includes at least one of the component's type, location, size, and shape.

[0339] In one feasible solution, the gap value acquisition module 5 includes:

[0340] The index value acquisition unit 51 is used to obtain the design index value and actual index value corresponding to each evaluation index for each target component pair, based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component.

[0341] The gap value acquisition unit 52 is used to obtain the gap value based on the design index value and the actual index value;

[0342] And / or,

[0343] The comprehensive gap value acquisition module 6 includes:

[0344] Weight acquisition unit 61 is used to acquire the weight corresponding to each difference value;

[0345] The comprehensive gap value acquisition unit 62 is used to obtain the comprehensive gap value of each target component pair based on each gap value and its corresponding weight;

[0346] And / or,

[0347] The construction progress tracking module 7 is also used to determine the construction progress of the corresponding design component as normal in response to the comprehensive gap value falling within the first preset range.

[0348] In response to the comprehensive gap value falling within the second preset range, the construction progress of the corresponding design component is determined to be in an early warning state;

[0349] In response to the comprehensive gap value falling within the third preset range, the construction progress of the corresponding design component is set to an alarm state;

[0350] The upper limit of the first preset range is less than the lower limit of the second preset range, and the upper limit of the second preset range is less than the lower limit of the third preset range.

[0351] In this embodiment, by comparing the design BIM of the target building with the actual BIM of the building, the difference values ​​of several evaluation indicators of the matching design components and actual components are obtained, and then a comprehensive difference value is obtained to intelligently track the construction progress of the target building. That is, by comparing the two BIMs and comprehensively considering multiple evaluation indicators, a comprehensive difference value is obtained to track the construction progress of the building, thereby improving the accuracy and reliability of the determination of the construction progress.

[0352] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The system embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs.

[0353] Example 3

[0354] Figure 7 This is a schematic diagram of the structure of an electronic device according to an example embodiment of the present disclosure. The electronic device includes a memory, a processor, and a computer program stored in the memory and used to run on the processor. When the processor executes the computer program, it implements the intelligent tracking method for building construction progress based on point cloud and BIM as described in any of the above embodiments. Figure 7 The electronic device 90 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments disclosed herein.

[0355] like Figure 7 As shown, the electronic device 90 can be manifested as a general-purpose computing device, such as a server device. The components of the electronic device 90 may include, but are not limited to: at least one processor 91, at least one memory 92, and a bus 93 connecting different system components (including memory 92 and processor 91).

[0356] Bus 93 includes a data bus, an address bus, and a control bus.

[0357] The memory 92 may include volatile memory, such as random access memory (RAM) 921 and / or cache memory 922, and may further include read-only memory (ROM) 923.

[0358] The memory 92 may also include a program tool 925 (or utility) having a set (at least one) program module 924, such program module 924 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.

[0359] The processor 91 executes various functional applications and data processing by running computer programs stored in the memory 92, such as the intelligent tracking method for building construction progress based on point cloud and BIM provided in any of the above embodiments.

[0360] Electronic device 90 can also communicate with one or more external devices 94 (e.g., keyboard, pointing device, etc.). This communication can be performed through input / output (I / O) interface 95. Furthermore, electronic device 90 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public network, such as the Internet) via network adapter 96. As shown, network adapter 96 communicates with other modules of electronic device 90 via bus 93. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with electronic device 90, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems.

[0361] It should be noted that although several units / modules or sub-units / modules of the electronic device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided and embodied by multiple units / modules.

[0362] Example 4

[0363] This disclosure also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the intelligent tracking method for building construction progress based on point cloud and BIM provided in any of the above embodiments.

[0364] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0365] Example 5

[0366] This disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements the intelligent tracking method for building construction progress based on point cloud and BIM as described above.

[0367] The program code for executing the computer program product of this disclosure can be written in any combination of one or more programming languages, and the program code can be executed entirely on a user device, partially on a user device, as a stand-alone software package, partially on a user device and partially on a remote device, or entirely on a remote device.

[0368] While specific embodiments of this disclosure have been described above, those skilled in the art should understand that these are merely illustrative examples, and the scope of protection of this disclosure is defined by the appended claims. Those skilled in the art can make various changes or modifications to these embodiments without departing from the principles and essence of this disclosure, but all such changes and modifications fall within the scope of protection of this disclosure.

Claims

1. A method for intelligently tracking building construction progress based on point cloud and BIM, characterized in that, The intelligent tracking method includes: Acquire the BIM design data of the target building and the actual point cloud data of the building at the construction site; The actual building is the building constructed according to the target building. The actual point cloud data is input into the preset BIM generation model to output the actual BIM of the actual building; The target building includes several design components, and the design BIM includes a design sub-BIM corresponding to each design component; the actual building includes several actual components, and the actual BIM includes an actual sub-BIM corresponding to each actual component. The mutually matching design components and actual components are taken as target component pairs; For each target component pair, based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component, the difference values ​​of several evaluation indicators for each target component pair are obtained. Based on the aforementioned gap value, a comprehensive gap value is obtained for each pair of target components; The comprehensive gap value is used to characterize the current construction progress of the corresponding component; Based on the comprehensive gap value of each target component pair, the construction progress of the target building is intelligently tracked; The design components include regular components and irregularly shaped components; The design sub-BIM corresponding to the rule component is represented using the first parameter set; For the irregularly shaped component and the design file of the irregularly shaped component contains the parameters required for NURBS representation, the design sub-BIM corresponding to the irregularly shaped component is represented by NURBS, and the NURBS corresponds to the second parameter set; For the irregularly shaped component and the design file of the irregularly shaped component does not contain the parameters required for NURBS representation, the design sub-BIM corresponding to the irregularly shaped component is represented by Mesh, and the Mesh corresponds to the third parameter set; The actual components include regular components and irregularly shaped components; The actual sub-BIM corresponding to the rule component is represented using the first parameter set; For the irregular component and the actual confidence level of the NURBS representation of the irregular component output by the preset BIM generation model is not less than the preset confidence level, the actual sub-BIM corresponding to the irregular component is represented by NURBS, and the NURBS corresponds to the second parameter set. For the irregular component and the actual confidence level of the NURBS representation of the irregular component output by the preset BIM generation model is less than the preset confidence level, the actual sub-BIM corresponding to the irregular component is represented by a Mesh, and the Mesh corresponds to the third parameter set; The formula for calculating the actual confidence level is as follows: ； E_fit_norm=min(1,E_fit / E_max); E_fit=sqrt((1 / N)*Σ||P_i-P'_i||²); V_curv_norm=min(1,V_curv / V_max); V_curv=Variance({k_1,k_2……k_N}); C_param_norm=min(1,C_param / C_max); C_param=log[(m+1)*(n+1)]; Among them, s nurbs The values ​​represent the actual confidence level, E_fit_norm represents the normalized fitting error, V_curv_norm represents the normalized curvature variance, C_param_norm represents the normalized parameter complexity, w_e represents the weight of the fitting error, w_c represents the weight of the curvature variance, w_p represents the weight of the parameter complexity, E_fit represents the original fitting error, E_max represents the preset maximum allowable error threshold, N represents the number of points of the irregular component in the corresponding original point cloud of NURBS, and P_ i represents the i-th point in the original point cloud, P'_i represents the nearest projection point of P_i onto the fitted NURBS representation, V_curv represents the variance of the curvature of all points of the irregular component in the original point cloud, V_max represents the preset maximum allowable curvature variance threshold, k_i represents the principal curvature of P_i, C_param represents the original complexity, C_max represents the preset maximum allowable complexity threshold, m and n represent the order of the NURBS surface in two directions, and (m+1)*(n+1) represents the total number of control points.

2. The intelligent tracking method for building construction progress based on point cloud and BIM as described in claim 1, characterized in that, The preset BIM generation model is obtained based on several sets of sample training data; The training data for each set of samples includes sample point cloud data and sample BIM corresponding to the sample buildings.

3. The intelligent tracking method for building construction progress based on point cloud and BIM as described in claim 1, characterized in that, For the regular component, the evaluation index includes at least one of the following: position, orientation, size, cross-section, and installation time. For the irregularly shaped component, and both the design sub-BIM and the actual sub-BIM corresponding to the irregularly shaped component are represented by NURBS, the evaluation index includes at least one of the following: the position, orientation, installation time, shape, control grid, and weight of the control grid of the irregularly shaped component. For the irregularly shaped component, and at least one of the design sub-BIM and the actual sub-BIM corresponding to the irregularly shaped component is represented by a Mesh, the evaluation index includes at least one of the following: position, orientation, installation time, shape, main direction, and size of the outer bounding box of the irregularly shaped component.

4. The intelligent tracking method for building construction progress based on point cloud and BIM as described in claim 1, characterized in that, Before the step of using the mutually matching design components and actual components as target component pairs, the method further includes: Obtain the design parameter values ​​and actual parameter values ​​of the designed component and the actual component under several parameter information; In response to the fact that the design parameter value and the actual parameter value under each parameter information meet the preset requirements, it is determined that the design component and the actual component are mutually matched.

5. The intelligent tracking method for building construction progress based on point cloud and BIM as described in claim 4, characterized in that, The parameter information includes at least one of the component's type, location, size, and shape.

6. The intelligent tracking method for building construction progress based on point cloud and BIM as described in any one of claims 1-5, characterized in that, The step of obtaining the difference values ​​of several evaluation indicators for each target component pair based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component includes: For each target component pair, based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component, the design index value and actual index value corresponding to each evaluation index are obtained. The gap value is obtained based on the design index value and the actual index value; And / or, The step of obtaining the comprehensive gap value for each pair of target components based on the gap value includes: Obtain the weight corresponding to each of the aforementioned gap values; Based on each of the gap values ​​and the corresponding weights, the comprehensive gap value of each of the target component pairs is obtained; And / or, The step of intelligently tracking the construction progress of the target building based on the comprehensive gap value of each target component pair includes: In response to the comprehensive gap value falling within a first preset range, the construction progress of the corresponding design component is determined to be in a normal state; In response to the comprehensive gap value falling within the second preset range, the construction progress of the corresponding design component is determined to be in an early warning state; In response to the comprehensive gap value falling within a third preset range, the construction progress of the corresponding design component is determined to be in an alarm state; Wherein, the upper limit of the first preset range is less than the lower limit of the second preset range, and the upper limit of the second preset range is less than the lower limit of the third preset range.

7. An intelligent tracking system for building construction progress based on point cloud and BIM, characterized in that, The intelligent tracking system includes: The design BIM acquisition module is used to acquire the design BIM of the target building. The point cloud data acquisition module is used to acquire actual point cloud data of the actual building at the construction site; The actual building is the building constructed according to the target building. The actual BIM output module is used to input the actual point cloud data into the preset BIM generation model to output the actual BIM of the actual building; The target building includes several design components, and the design BIM includes a design sub-BIM corresponding to each design component; the actual building includes several actual components, and the actual BIM includes an actual sub-BIM corresponding to each actual component. The component pair determination module is used to identify mutually matching design components and actual components as target component pairs. The gap value acquisition module is used to obtain the gap value of several evaluation indicators for each target component pair based on the design sub-BIM of the corresponding design component and the actual sub-BIM of the corresponding actual component. The comprehensive gap value acquisition module is used to obtain the comprehensive gap value for each of the target component pairs based on the gap value; The comprehensive gap value is used to characterize the current construction progress of the corresponding component; The construction progress tracking module is used to intelligently track the construction progress of the target building based on the comprehensive gap value of each target component pair. The design components include regular components and irregularly shaped components; The design sub-BIM corresponding to the rule component is represented using the first parameter set; For the irregularly shaped component and the design file of the irregularly shaped component contains the parameters required for NURBS representation, the design sub-BIM corresponding to the irregularly shaped component is represented by NURBS, and the NURBS corresponds to the second parameter set; For the irregularly shaped component and the design file of the irregularly shaped component does not contain the parameters required for NURBS representation, the design sub-BIM corresponding to the irregularly shaped component is represented by Mesh, and the Mesh corresponds to the third parameter set; The actual components include regular components and irregularly shaped components; The actual sub-BIM corresponding to the rule component is represented using the first parameter set; For the irregular component and the actual confidence level of the NURBS representation of the irregular component output by the preset BIM generation model is not less than the preset confidence level, the actual sub-BIM corresponding to the irregular component is represented by NURBS, and the NURBS corresponds to the second parameter set. For the irregular component and the actual confidence level of the NURBS representation of the irregular component output by the preset BIM generation model is less than the preset confidence level, the actual sub-BIM corresponding to the irregular component is represented by a Mesh, and the Mesh corresponds to the third parameter set; The formula for calculating the actual confidence level is as follows: ； E_fit_norm=min(1,E_fit / E_max); E_fit=sqrt((1 / N)*Σ||P_i-P'_i||²); V_curv_norm=min(1,V_curv / V_max); V_curv=Variance({k_1,k_2……k_N}); C_param_norm=min(1,C_param / C_max); C_param=log[(m+1)*(n+1)]; Among them, s nurbs The values ​​represent the actual confidence level, E_fit_norm represents the normalized fitting error, V_curv_norm represents the normalized curvature variance, C_param_norm represents the normalized parameter complexity, w_e represents the weight of the fitting error, w_c represents the weight of the curvature variance, w_p represents the weight of the parameter complexity, E_fit represents the original fitting error, E_max represents the preset maximum allowable error threshold, N represents the number of points of the irregular component in the corresponding original point cloud of NURBS, and P_ i represents the i-th point in the original point cloud, P'_i represents the nearest projection point of P_i onto the fitted NURBS representation, V_curv represents the variance of the curvature of all points of the irregular component in the original point cloud, V_max represents the preset maximum allowable curvature variance threshold, k_i represents the principal curvature of P_i, C_param represents the original complexity, C_max represents the preset maximum allowable complexity threshold, m and n represent the order of the NURBS surface in two directions, and (m+1)*(n+1) represents the total number of control points.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and for running on the processor, characterized in that, When the processor executes the computer program, it implements the intelligent tracking method for building construction progress based on point cloud and BIM as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the intelligent tracking method for building construction progress based on point cloud and BIM as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the intelligent tracking method for building construction progress based on point cloud and BIM as described in any one of claims 1-6.

Images