Construction project whole-process management system and method based on BIM3D-6D technology

Abstract

The invention discloses a construction project whole-process management system and method with a BIM3D-6D technology as a base, and relates to the technical field of construction project management, and the method comprises the steps: loading a 3D building information model through a BIM lightweight calculation server, extracting the spatial topological relation and IFC semantic attributes of geometric components, and carrying out the calculation of the 3D building information model through a BIM lightweight calculation server; spatial topology analysis and semantic attribute mapping are carried out on the spatial topology relation of the geometric components and the IFC semantic attributes, and a three-dimensional data topology network packet based on a graph structure is generated; performing multi-dimensional deduction operation on the three-dimensional data topology network packet based on the graph structure to obtain a deviation compensation strategy containing an optimal cost control path; and carrying out dimension expansion and data fusion on the 3D building information model to generate a 5D cost dimension data set. Discrete decisions are converted into a continuous field model through mathematical transformation, so that a resource flow path has spatial relevance and time continuity, and the beneficial effect of dynamically optimizing a construction process is achieved.

Description

Technical Field

[0001] This invention relates to the field of construction project management technology, and in particular to a construction project full-process management system and method based on BIM3D-6D technology. Background Technology

[0002] In Building Information Modeling (BIM)-driven project management, conventional methods primarily rely on the static correlation between 3D geometric models and schedule and cost data to achieve full-process control. A typical workflow includes: integrating the BIM model with the construction plan and bill of quantities through a BIM platform to create a visual control interface; using a rule engine to parse model component attributes and generate resource allocation schemes; and combining this with IoT devices to provide feedback on site status, enabling dynamic monitoring of construction progress and costs. This approach, through structured data mapping and visual interaction, significantly improves construction collaboration efficiency and provides a digital foundation for quality and safety management, becoming a core technological support for modern intelligent construction.

[0003] However, existing methods still have two limitations: Insufficient coupling of multidimensional data: The association between cost, schedule and spatial topology relies on manual rule configuration, making it difficult to dynamically capture the spatiotemporal constraints of the process logic chain, resulting in the risk of decoupling between resource scheduling strategy and actual on-site working conditions; Dynamic response lag: Decision-making based on static models lacks a real-time disturbance correction mechanism. When on-site resource consumption deviates from the plan, parameters need to be manually readjusted, making it difficult to ensure the global optimality of cost control path. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a project construction process management system and method based on BIM3D-6D technology, which solves the problems of insufficient dynamic coupling of multi-dimensional data and lagging global optimization of cost control path.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: Firstly, this invention provides a method for full-process management of construction projects based on BIM 3D-6D technology, comprising: The 3D building information model is loaded by a BIM lightweight computing server, the spatial topological relationships and IFC semantic attributes of geometric components are extracted, and the spatial topological relationships and IFC semantic attributes of geometric components are analyzed and mapped to generate a three-dimensional data topology network package based on graph structure. Multi-dimensional extrapolation is performed on graph-based 3D data topology network packets to obtain deviation compensation strategies containing the optimal cost control path; By expanding the dimensions and fusing data from 3D building information models, a 5D cost dimension dataset is generated. The resource adjustment rules in the deviation compensation strategy containing the optimal cost control path are analyzed, a resource allocation optimization model is constructed, and the gradient descent algorithm is used to optimize the resource allocation optimization model. Based on the 5D cost dimension dataset, a construction scheduling instruction set that minimizes the cost disturbance factor is generated. By using cognitive mirror synchronization, the construction scheduling instruction set that minimizes cost disturbance factors is subjected to tensor convolution operation to generate a reconstructed 4D spatiotemporal progress topology. Based on the reconstructed 4D spatiotemporal progress topology, the facility status record is updated, and combined with the readings of IoT sensing devices, a 6D full lifecycle data record table is generated.

[0007] As a preferred embodiment of the project whole-process management method based on BIM3D-6D technology described in this invention, the steps of loading a 3D building information model through a BIM lightweight computing server, extracting the spatial topological relationships and IFC semantic attributes of geometric components, and performing spatial topological analysis and semantic attribute mapping on the spatial topological relationships and IFC semantic attributes of the geometric components to generate a graph-based three-dimensional data topology network packet are as follows: By loading the 3D building information model through the BIM lightweight computing server, the non-critical structural mesh is simplified to generate the geometric deformation degree and topological relationship matrix, and the 3D building information model is scanned based on the geometric deformation degree to obtain the geometric element set; The semantic attributes of IFC in the set of geometric elements are analyzed, and the topological relation matrix is ​​converted into a topological graph with weighted attributes. Based on the spatial overlap threshold, adjacent 3D nodes of the topology graph with weighted attributes are merged to construct an R-tree 3D spatial index, and the R-tree 3D spatial index is converted into a 3D topology data network. Verify the 3D data topology network package, generate a connectivity graph of building structural components and an attribute distribution report, and encapsulate it to obtain a graph-based 3D data topology network package.

[0008] As a preferred embodiment of the project whole-process management method based on BIM3D-6D technology described in this invention, the steps for performing multi-dimensional deduction operations on the graph-based three-dimensional data topology network package to obtain a deviation compensation strategy containing the optimal cost control path are as follows: Topological node and edge data are extracted from the 3D data topological network package of the graph structure. Based on the time window constraint, the spatial coordinates of the building structure components are extended into 4D spatiotemporal parameters. A preset resource cost parameter table is superimposed as a dynamic correction factor to generate an enhanced topological network with perturbation factor. Based on the enhanced topology network with perturbation factor, the perturbation effect is propagated along the topology edge, and the stability perturbation guarantee value is obtained through iteration. Then, multi-dimensional scenario simulation is performed on the stability perturbation guarantee value to generate a three-dimensional diagnostic report. A multi-objective optimization function is constructed based on the 3D diagnostic report. Then, the optimal non-dominated solution set is selected from the multi-objective optimization function to generate a deviation compensation strategy containing the optimal cost control path.

[0009] As a preferred embodiment of the project whole-process management method based on BIM3D-6D technology described in this invention, the preset resource cost parameter table includes a component type matching field, cost parameter values, and structured data related to parameter constraints.

[0010] As a preferred embodiment of the project whole-process management method based on BIM3D-6D technology described in this invention, the steps for generating a 5D cost dimension dataset by expanding the 3D building information model through dimensional expansion and data fusion are as follows: A replacement simplification method is used to process the geometric element set of the 3D building information model, and a lightweight geometric dataset is generated based on the geometric entities and attribute entities of the 3D building information model. By associating the construction schedule with the building structural component ID, the installation time nodes are bound to the building structural components in the lightweight geometry dataset. The external bill of quantities is parsed, the geometric quantities of the building structural components in the lightweight geometry dataset are matched, and the cost baseline matrix is ​​obtained. Dynamic cost correction is performed on the 4D spatiotemporal dataset and cost baseline matrix to generate a 5D cost dimension dataset.

[0011] As a preferred embodiment of the project whole-process management method based on BIM3D-6D technology described in this invention, the external bill of quantities is a structured resource quantification table constructed based on cost specifications, which accurately associates the geometric quantities, resource consumption rules and cost parameters of building structural components.

[0012] As a preferred embodiment of the project whole-process management method based on BIM3D-6D technology described in this invention, the steps of analyzing the resource adjustment rules in the deviation compensation strategy containing the optimal cost control path, constructing a resource allocation optimization model, optimizing the resource allocation optimization model using the gradient descent algorithm, and generating a construction scheduling instruction set that minimizes the cost disturbance factor based on the 5D cost dimension dataset are as follows: The resource type, adjustment direction, and adjustment amount in the deviation compensation strategy containing the optimal cost control path are analyzed, and the resource type, adjustment direction, and adjustment amount are converted into mathematical constraints to obtain a structured resource adjustment rule table. Using the resource adjustment rule table as a constraint and the dynamic weight parameters in the 5D cost dimension dataset as decision variables, a resource allocation optimization model is constructed with minimizing the cost disturbance factor as the objective function. The gradient descent algorithm is then used to solve the objective function and update the dynamic weight parameters to obtain the optimal combination of dynamic weight parameters and generate the optimized resource allocation optimization model. By matching the spatiotemporal coordinates in the 5D cost dimension dataset with the IDs of building structural components, the optimal dynamic weight parameter combination is injected into the spatiotemporal nodes. Subsequently, the task instructions are structured and coded, and cost disturbance warning instructions are added. After encapsulation, the construction scheduling instruction set that minimizes the cost disturbance factor is obtained.

[0013] As a preferred embodiment of the project whole-process management method based on BIM3D-6D technology described in this invention, the step of performing tensor convolution operation on the construction scheduling instruction set that minimizes cost disturbance factors through cognitive mirror synchronization to generate a reconstructed 4D spatiotemporal progress topology structure is as follows: The construction scheduling instruction set that minimizes cost disturbance factors is analyzed, and spatial location, time node, and optimal dynamic weight parameter combination are extracted. Then, based on the spatial location, time node, and optimal dynamic weight parameter combination, it is mapped into a spatiotemporal resource matrix. Based on the spatiotemporal resource matrix, a three-dimensional convolution kernel is constructed, and then tensor convolution operation is performed on the three-dimensional convolution kernel to generate a convolution-optimized spatiotemporal resource matrix. Each coordinate node in the convolution-optimized spatiotemporal resource matrix is ​​transformed into a topological vertex. The topological vertices are connected based on the resource flow path to form topological edges. Weight values ​​are assigned to the topological edges according to the resource consumption intensity to generate the reconstructed 4D spatiotemporal progress topology.

[0014] As a preferred embodiment of the project full-process management method based on BIM3D-6D technology described in this invention, the steps for updating facility status records based on the reconstructed 4D spatiotemporal progress topology and generating a 6D full lifecycle data record table by combining readings from IoT sensing devices are as follows: Based on the association of building structural component IDs with 4D spatiotemporal progress topology vertices and on-site physical facilities, the resource flow path of the topology vertex is extracted, and then the verified IoT device data is fused to the corresponding topology vertex to generate an enhanced topology with real-time status. Integrate construction period data, associate the topology vertices of the enhanced topology structure with real-time status with construction logs and quality inspection reports, and construct a 6D full life cycle data record table with building structural component ID as the core node.

[0015] Secondly, this invention provides a project construction process management system based on BIM3D-6D technology, comprising: The topology modeling module loads the 3D building information model through the BIM lightweight computing server, extracts the spatial topological relationships and IFC semantic attributes of geometric components, and performs spatial topology analysis and semantic attribute mapping on the spatial topological relationships and IFC semantic attributes of geometric components to generate a three-dimensional data topology network package based on graph structure. The path deduction module performs multi-dimensional deduction operations on the graph-based three-dimensional data topology network packet to obtain the deviation compensation strategy containing the optimal cost control path. The dimension fusion module generates a 5D cost dimension dataset by expanding the dimensions of the 3D building information model and fusing the data. The resource optimization module analyzes the resource adjustment rules in the deviation compensation strategy containing the optimal cost control path, constructs a resource allocation optimization model, and uses the gradient descent algorithm to optimize the resource allocation optimization model. Based on the 5D cost dimension dataset, it generates a construction scheduling instruction set that minimizes the cost disturbance factor. The structural reconstruction module performs tensor convolution operations on the construction scheduling instruction set that minimizes cost disturbance factors through cognitive mirror synchronization, generating a reconstructed 4D spatiotemporal progress topology. The closed-loop feedback module updates the facility status record based on the reconstructed 4D spatiotemporal progress topology and generates a 6D full lifecycle data record table by combining the readings of IoT sensing devices.

[0016] The beneficial effects of this invention are as follows: by performing tensor convolution operations on the construction scheduling instruction set through cognitive mirror synchronization, discrete scheduling instructions are mapped into a continuous spatiotemporal topology. The weight decay characteristics of the three-dimensional convolution kernel are used to achieve spatiotemporal smooth diffusion of resource paths, generating a reconstructed 4D spatiotemporal progress topology. Through mathematical transformation, discrete decisions are transformed into a continuous field model, enabling resource flow paths to have spatial correlation and temporal continuity, thus achieving the beneficial effect of dynamically optimizing the construction process. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a project management methodology based on BIM 3D-6D technology.

[0019] Figure 2 A flowchart for generating a 3D data topology network packet.

[0020] Figure 3A flowchart generated for the deviation compensation strategy.

[0021] Figure 4 A flowchart for reconstructing the 4D spatiotemporal progress topology. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figure 1 This is one embodiment of the present invention, which provides a method for full-process management of construction projects based on BIM3D-6D technology, including the following steps: S1: Load the 3D building information model through the BIM lightweight computing server, extract the spatial topological relationships and IFC semantic attributes of the geometric components, and perform spatial topological analysis and semantic attribute mapping on the spatial topological relationships and IFC semantic attributes of the geometric components to generate a three-dimensional data topology network package based on graph structure.

[0026] For details, please refer to Figure 2 The steps are as follows: S1.1: Load the 3D building information model through the BIM lightweight computing server, apply the dynamic level of detail selection algorithm to simplify the non-critical parts in the structural mesh, generate the geometric deformation degree and topological relationship matrix, and use the Octree spatial index to scan the 3D building information model to obtain the geometric element set based on the geometric deformation degree.

[0027] Specifically, a 3D building information model is loaded through a lightweight BIM computing server. A dynamic level of detail selection algorithm is executed to simplify the non-critical structural meshes in the 3D building information model while preserving the complete geometric accuracy of the critical structural meshes. The degree of geometric deformation of the 3D building information model is obtained, and iterative simplification is performed to generate a topological relationship matrix. Subsequently, using the degree of geometric deformation of the 3D building information model as weight, the Octree spatial index is used to perform a three-dimensional spatial scan of the 3D building information model to obtain the contact area and containment level between geometric elements, and integrate them to generate a set of geometric elements.

[0028] The formula for the dynamic level of detail selection algorithm is as follows: ; in, This represents a newly generated vertex in a 3D building information model. This represents the original mesh vertices in a 3D building information model. This represents the transpose coordinates of newly generated vertices in the 3D Building Information Model. Represents the original mesh vertices in a 3D building information model. The vertex error matrix, Represents the original mesh vertices in a 3D building information model. The vertex error matrix, This indicates the original mesh vertices in the 3D Building Information Model. and Merge into newly generated vertices The degree of geometric deformation introduced at that time.

[0029] It should be explained that 3D Building Information Modeling is based on the building industry standard IFC format and uses BIM construction software (such as Revit and Archicad) to perform structured modeling of geometric data (such as the size and location of beams and columns) and semantic attributes (such as material specifications and construction stages) in building structural components. The core of 3D Building Information Modeling is to achieve the automated generation and logical association of building elements through parametric modeling rules and engineering knowledge bases, and finally obtain a digital building twin containing geometric topology and engineering mathematics.

[0030] Among them, the parametric modeling rules are based on building structural codes, component size standards and construction process requirements. By analyzing the geometric characteristics and relationships of different types of building structural components, a parameter template that can be executed programmatically is formed. The engineering knowledge base is based on historical engineering data, construction experience and material performance information. It organizes component types, construction stages, resource consumption rules and cost parameters in a structured way, and establishes logical constraints and relationships between components.

[0031] It should be explained that the determination of critical and non-critical parts of the structural mesh is based on the structural importance and functional attributes of the building structural components. Specifically, critical structural meshes refer to the geometric data that directly affects the load-bearing performance of the building and the structural components. The determination criteria include the structural type mark in the IFC semantic attributes, mechanical analysis parameters, and mandatory requirements in the structural code. Non-critical structural meshes refer to decorative, auxiliary, or replaceable building structural components, which are usually automatically identified through the building structural component classification code, non-load-bearing attribute mark, and LOD (Level of Detail) parameters.

[0032] Among them, IFC semantic attributes are a digital attribute system defined by the IFC standard, used to accurately describe the engineering characteristics, functional classification and physical behavior of building structural components; IFC semantic attributes realize the standardized storage and exchange of component information through structured data, and the specific content is divided into the following two categories: IFC entity type and IFC attribute set.

[0033] S1.2: Parse the IFC semantic attributes in the set of geometric elements, convert the topological relation matrix into a topological graph with weighted attributes, and accurately bind the IFC semantic attributes to the corresponding topological graph nodes in the topological graph with weighted attributes.

[0034] Specifically, the IFC entity type and IFC attribute set of each building structural component in the geometric element set are extracted. Then, the topology relation matrix is ​​traversed, and each non-zero element in the topology relation matrix is ​​mapped to an edge of the topology graph with weighted attributes. The weight value of the topology relation matrix is ​​used as the connection strength coefficient. The IFC entity type and IFC attribute set of the geometric element set are bound to the topology graph node corresponding to the topology graph with weighted attributes, generating a topology graph with weighted attributes containing complete IFC semantic annotations.

[0035] It should be explained that IFC entity types refer to the standard-defined classification of building components, specifically including: structural components, building envelope components, equipment components, and auxiliary components. Structural components include, but are not limited to: walls, structural columns, structural beams, and floor slabs; building envelope components include, but are not limited to: doors, windows, and curtain walls; equipment components include, but are not limited to: air ducts and cable trays; and auxiliary components include, but are not limited to: stairs, ramps, and railings. IFC attribute sets refer to the set of engineering parameters bound to IFC entity types, specifically including: material attributes, structural attributes, construction attributes, and custom attribute sets. Material attributes include, but are not limited to: strength grade, fire rating, and thermal conductivity; structural attributes include, but are not limited to: cross-sectional dimensions, span, and load rating; construction attributes include, but are not limited to: construction stage and prefabrication markings; and custom attribute sets refer to extended attributes, including, but not limited to: unit price and supplier information.

[0036] It needs to be explained that the non-zero element mapping transformation process converts the non-zero elements in the topological relation matrix into directed edges connecting nodes in the topological graph with weighted attributes. The data values ​​of the non-zero elements are directly assigned as the weight attributes of the edges, and the engineering meaning of the original matrix is ​​recorded in the weight attributes of the edges, thus completing the transformation of the topological graph with weighted attributes.

[0037] It needs to be explained that complete IFC semantic annotation refers to the engineering information system that conforms to the IFC standard carried by each node and edge in the weighted attribute topology graph. Specifically, it includes node annotation and edge annotation. Node annotation represents the engineering characteristics that define the building structural components and determines the functional classification of the engineering characteristics in the construction process, including but not limited to: IFC entity type, IFC attribute set, and component unique identifier. Edge annotation represents the digital definition of the engineering relationship between components in the weighted attribute topology graph. The core is to accurately describe the interaction between building elements through standardized data structure, including but not limited to: connection type, connection strength coefficient, and engineering basis.

[0038] S1.3: Based on the spatial overlap threshold, merge adjacent 3D nodes of the topology graph with weighted attributes, construct an R-tree 3D spatial index, and convert the R-tree 3D spatial index into a GraphML format 3D topology data network.

[0039] Specifically, all 3D nodes in the topology graph with weighted attributes are traversed, and the spatial bounding box overlap of each pair of adjacent 3D nodes is calculated. If the spatial bounding box overlap of adjacent 3D nodes exceeds a preset spatial overlap threshold, the 3D nodes are merged into new 3D nodes. The weight of the new 3D node is the average of the weights of the merged adjacent 3D nodes. After merging, the topology edge relationships are updated. Then, an R-tree 3D spatial index is constructed based on the spatial distribution of the merged 3D nodes. The minimum bounding cube of each new 3D node is stored as the leaf node of the R-tree 3D spatial index. Finally, the hierarchical structure and node spatial coordinates of the R-tree 3D spatial index are converted into a 3D topology data network in GraphML format.

[0040] The formula for calculating the overlap of spatial bounding boxes is as follows: ; in, as well as These represent the axis-aligned bounding boxes of the 3D objects that are the 3D nodes. Represents two 3D nodes and The overlap of the bounding boxes between the axis-aligned bounding boxes. Represents spatial dimension variables. Indicates the length direction of the corresponding engineering structure. Indicates the width direction of the corresponding building structure. Indicates the vertical direction of the corresponding engineering structure. Indicates axis-aligned bounding box In spatial dimension Maximum boundary coordinates Indicates axis-aligned bounding box In spatial dimension Maximum boundary coordinates Indicates axis-aligned bounding box In spatial dimension The minimum boundary coordinates, Indicates axis-aligned bounding box In spatial dimension The minimum boundary coordinates.

[0041] It should be explained that the preset spatial overlap threshold is a critical value used to determine whether adjacent 3D nodes need to be merged. When the calculated spatial bounding box overlap of two 3D nodes exceeds the preset spatial overlap threshold, a merging operation is triggered. The preset spatial overlap threshold is constructed based on engineering accuracy requirements, hardware computing resources, and industry standards.

[0042] It needs to be explained that the transformation process of the 3D topology data network involves converting the leaf nodes of the R-tree 3D spatial index into GraphML node elements, maintaining the uniqueness of node IDs, and converting the cube range coordinates into "bbox" attribute key-value pairs. Subsequently, all structural nodes of the R-tree 3D spatial index are converted into nested graph structural elements of GraphML, forming a hierarchical relationship. The spatial containment relationships between R-tree nodes are then converted into edge elements of GraphML, where the source node and target node of the GraphML edge element point to the parent graph structure and the child node, respectively. At the same time, all attributes (such as IFC type and material parameters) of the topology graph nodes corresponding to the weighted attribute topology graph are converted into additional data elements of GraphML. Finally, GraphML standard header information is added to complete the format conversion of the 3D topology data network.

[0043] S1.4: Verify the node connectivity and attribute integrity of the 3D data topology network, generate a connectivity graph and attribute distribution report of building structural components, and encapsulate it to obtain a 3D data topology network package based on graph structure.

[0044] Specifically, the process iterates through all 3D nodes in the 3D data topology network, checks the number of connecting edges for each 3D node and verifies the integrity of attribute fields, generates verified 3D node and edge data, and maps each 3D node to a vertex of a connected graph based on the verified 3D node and edge data, retaining the node ID and IFC entity type. At the same time, it extracts 3D node attributes as additional vertex data, converts the verified edge data into undirected edges of the connected graph, constructs a connected graph of building structural components, and finally encapsulates the connected graph of building structural components, the attribute distribution report, and the 3D topology data network into a graph-based 3D data topology network package conforming to the GraphML specification based on graph structure, based on the attribute distribution report.

[0045] It needs to be explained that the specific verification of attribute field integrity is to check whether the attribute fields of each 3D node in the 3D data topology network contain the required IFC semantic attributes, and at the same time verify whether the effective engineering parameter range of each 3D node in the 3D data topology network conforms to the multidimensional engineering constraints. 3D nodes with missing fields or engineering parameter ranges that exceed the limits are marked as verification failures.

[0046] Among them, multidimensional engineering constraints include: engineering geometric parameter constraints, engineering structure type constraints, engineering material property constraints, engineering construction stage constraints, engineering mechanics constraints, and engineering safety constraints.

[0047] It should be explained that the attribute distribution report is a document that statistically analyzes the attribute values ​​of all verified nodes in the 3D topology data network, recording the frequency and classification ratio of the numerical distribution of different IFC semantic attributes; the attribute distribution report includes the completeness rate of required fields, such as the completeness rate of the cross-sectional dimension field, as well as the interval distribution of engineering quantitative parameters with units (such as interface dimensions, connection strength, and load level).

[0048] S2: Perform multi-dimensional deduction operations on the graph-based three-dimensional data topology network package to obtain the deviation compensation strategy containing the optimal cost control path.

[0049] For details, please refer to Figure 3 The steps are as follows: S2.1: Extract topology node and topology edge data from the 3D data topology network package of the graph structure, and then combine time window constraints to expand the spatial coordinates of the topology nodes of the building structure components into 4D spatiotemporal parameters. The preset resource cost parameter table is used as a dynamic correction factor to obtain an enhanced topology network with perturbation factors.

[0050] Specifically, topology node data and topology edge data are parsed from the 3D data topology network package of the graph structure. The spatial coordinates and IFC semantic attributes of the topology nodes, as well as the connection relationships and weights of the topology edges, are extracted. Combined with time window constraints, the spatial coordinates of the topology nodes are expanded into 4D spatiotemporal parameters, while the connection relationships of the topology edge data are preserved. Then, the preset resource cost parameter table is converted into a dynamic correction factor, which is matched to the corresponding topology node and topology edge data according to the component type. Through the superposition operation of spatiotemporal parameters and dynamic correction factors, an enhanced topology network with perturbation factors is generated.

[0051] It should be explained that time window constraint refers to a discrete time node defined based on the construction schedule or project phase division, used to expand three-dimensional spatial coordinates into 4D spatiotemporal parameters. Time window constraint exists in standard time format or relative construction period form, and forcibly limits the validity of the spatial state of building components within a specific time interval.

[0052] It should be explained that the preset resource cost parameter table is a structured data table. Its core function is to store resource cost parameter information associated with different building structural component types, ensuring accurate matching during the generation of enhanced topology networks with perturbation factors. The preset resource cost parameter table is constructed based on the breakdown items of the bill of quantities, resource consumption quota standards, and supplementary clauses of the engineering contract. The preset resource cost parameter table contains structured data related to component type matching fields, cost parameter values, and parameter constraints.

[0053] S2.2: Based on the enhanced topology network with perturbation factor, the perturbation effect is propagated along the topology edge, and the stability perturbation guarantee value is obtained through iteration. Then, multi-dimensional scenario simulation is performed on the stability perturbation guarantee value to generate a three-dimensional diagnostic report.

[0054] Specifically, based on all topological edges in the enhanced topological network with perturbation factors, the perturbation factors of the initial three-dimensional network are passed to the target three-dimensional nodes along the direction of the topological edges. Through weighted accumulation operations, the perturbation values ​​of all nodes are iteratively updated until convergence, and the stability perturbation guarantee value is obtained. Then, multi-dimensional scenario simulations are performed on the stability perturbation guarantee value, the perturbation response of the topological network under different scenarios is recorded, and a three-dimensional diagnostic report containing indicators such as node displacement and edge weight change rate is generated.

[0055] It needs to be explained that the determination of the target 3D node is defined by the directionality of the topological edge. Each topological edge in the enhanced topological network has a clearly designated start node and target node. All topological edges are traversed, and the start node and target node are extracted according to the preset direction of each topological edge. The perturbation factor of the start node is weighted by the edge weight and added to the current perturbation value of the target node, and the perturbation values ​​of all nodes are updated.

[0056] It should be explained that the multi-dimensional scenario simulation operation generates multiple disturbance scenarios by adjusting and combining disturbance parameters based on multi-dimensional engineering constraints and historical engineering data. The stability disturbance guarantee value of the topology network under each scenario is then re-obtained, and key response indicators of nodes and edges are recorded. Specifically, the operation involves selecting core parameters affecting disturbance factors (e.g., material cost fluctuations and schedule compression rates), generating orthogonal experimental combinations, iterating disturbance propagation for each orthogonal experimental combination to obtain new stability disturbance guarantee values, extracting quantitative data such as node displacement and edge weight change rates from the new stability disturbance guarantee values, integrating indicator values ​​according to scenario coding, and generating a three-dimensional diagnostic report.

[0057] S2.3: Construct a multi-objective optimization function based on the three-dimensional diagnostic report, and then use the Pareto front algorithm to select the optimal non-dominated solution set from the multi-objective optimization function to generate a deviation compensation strategy containing the optimal cost control path.

[0058] Specifically, based on parameters such as node displacement and edge weight change rate in the 3D diagnostic report, a multi-objective optimization function is constructed, which includes cost control objectives, structural safety objectives, and schedule objectives. The Pareto front algorithm is then used to traverse the solution space and select the optimal non-dominated solution set that satisfies all constraints and is mutually non-dominated. Finally, cost control paths are extracted from the optimal non-dominated solution set to generate a deviation compensation strategy containing resource adjustment rules and compensation amounts.

[0059] The formula for the Pareto front algorithm is as follows: ; in, Denotes the optimal non-dominated solution set. This represents the combination of control parameters for the deviation compensation strategy. Indicates a candidate solution. This represents the solution space defined by the multi-objective optimization function. This represents the cost control objective function for comparing solutions. This represents the structural safety objective function of the compared solutions. This represents the progress objective function for comparing solutions. The objective function for cost control represents the combination of control parameters. The structural safety objective function represents the combination of control parameters. This represents the progress objective function of the combination of control parameters.

[0060] It should be explained that satisfying the constraints means that the candidate solution must simultaneously meet the following engineering boundary restrictions: the node displacement does not exceed the upper limit allowed by the project, the edge weight change rate is not lower than the safety threshold, and the resource adjustment amount is within the scope of the contract bill of quantities; satisfying the non-dominant condition means that when comparing the candidate solutions, under the premise of satisfying the constraints, if there is no other candidate solution that is better than the current candidate solution in all objectives, the current candidate solution is regarded as the non-dominant solution.

[0061] The scope of the contract bill of quantities refers to a structured resource quantification table formed based on the quantity of building structural components, material consumption, and construction procedures listed in the tender documents, construction contract, and cost specifications. After structuring and associating with BIM component IDs, it is accurately mapped to each building structural component in the three-dimensional building information model, thereby limiting the acceptable range of each resource adjustment.

[0062] S3: Generate a 5D cost dimension dataset by expanding the dimensions of the 3D building information model and fusing the data.

[0063] Specifically, the steps are as follows: S3.1: The geometric element set of the 3D Building Information Model is processed using a replacement simplification method, and a lightweight geometric dataset is generated based on the geometric entities and attribute entities of the 3D Building Information Model.

[0064] Specifically, a simplified method based on component type and bounding box replacement is used to process the geometric element set of the 3D Building Information Model. The original geometric accuracy of key structural components is preserved, and axis-aligned bounding boxes are extracted from non-key structural components and replaced with the original geometry. At the same time, non-standard annotations are filtered according to the geometric entities and attribute entities of the 3D Building Information Model, and finally a lightweight geometric dataset is generated.

[0065] It should be explained that the key determination of structural components is achieved by analyzing the structural type markers in the IFC semantic attributes, mechanical analysis parameters (such as bending moment, shear force, axial force, etc.), and the requirements of building structural codes for building structural components. This process automatically identifies components that play a decisive role in the overall safety, stability, and load-bearing capacity of the building structure, such as load-bearing beams, columns, and core tube walls, and defines the components that play a decisive role as key structural components.

[0066] It should be explained that the simplified method based on component type and bounding box replacement achieves lightweighting of 3D building information models through engineering semantic-driven geometric dimensionality reduction technology. The structural importance is determined based on the component IFC entity type. The original geometric accuracy of key structural components is preserved, while spatial bounding boxes are extracted for non-key structural components. The original complex geometry is replaced with the minimum bounding cube. At the same time, annotation information in geometric entities that does not conform to the IFC standard is filtered out, and finally a lightweight geometric dataset is generated.

[0067] It should be explained that geometric entities refer to structured data in 3D building information models that carry spatial topology and geometric form, accurately describing the spatial location, shape and size relationship of building components, specifically including: basic geometric elements, spatial relationships and accuracy levels; attribute entities refer to the engineering semantic data system bound to geometric entities, which establishes a structured description of component characteristics, functions and behaviors based on IFC standards, specifically including: classification labels, engineering parameter sets and constraint rules.

[0068] S3.2: Associate the construction schedule with the building structural component ID, bind the installation time node to the building structural component in the lightweight geometry dataset, parse the external bill of quantities, match the geometric quantities of the building structural components in the lightweight geometry dataset, and obtain the cost baseline matrix.

[0069] Specifically, the installation time nodes in the construction schedule are associated with the building structural component IDs. The installation time nodes are then bound to the attribute fields of the corresponding building structural components in the lightweight geometry dataset. Subsequently, the external bill of quantities is parsed to match the geometric quantities of the building structural components in the lightweight geometry dataset. A two-dimensional index matrix is ​​then established based on the component IDs and time nodes. The matched quantities and unit costs are then filled into the two-dimensional index matrix to generate a cost benchmark matrix.

[0070] It should be explained that a construction schedule is a construction task execution plan based on a time dimension in an engineering project. Its core is to precisely control the installation sequence of building structural components and the division of engineering stages through discrete time nodes.

[0071] It should be explained that the external bill of quantities is a structured resource quantification table built on cost specifications. It accurately links the geometric quantities, resource consumption rules, and cost parameters of building structural components. The construction of the external bill of quantities uses component ID as the link to integrate geometric quantities, resource consumption rules, and cost parameters into a structured measurement system, supporting the accurate generation of the cost benchmark matrix.

[0072] S3.3: The gradient descent algorithm is used to dynamically correct the cost of the 4D spatiotemporal dataset and the cost baseline matrix, generating a 5D cost dimension dataset.

[0073] Specifically, the gradient descent algorithm is used to dynamically correct the cost of the 4D spatiotemporal dataset and the cost benchmark matrix. The cost correction factor is initialized as the benchmark ratio parameter, and the overall difference between the cost benchmark matrix value and the predicted cost of the 4D spatiotemporal dataset is quantified by defining a loss function. Then, based on the gradient direction of the loss function and combined with the preset learning rate parameter, the value of the correction factor is gradually adjusted. After each iteration, the loss change is re-evaluated. Finally, the converged correction factor is merged with the original 4D spatiotemporal dataset to expand and form a 5D cost dimension dataset containing spatial three-dimensional coordinates, time dimension and dynamically corrected cost values.

[0074] The formula for the gradient descent algorithm is as follows: ; in, Indicates the iteration count index. Discrete variables representing component traversal. This represents the learning rate parameter. Indicates the total number of structural components of a building. Indicates the first Cost adjustment factor for the next iteration Indicates the first Cost adjustment factor for the next iteration Indicates the first The predicted cost value of each component, Indicates the first The baseline cost value of each component.

[0075] It should be explained that the preset learning rate parameter is set based on the magnitude characteristics of the engineering cost data, the convergence stability requirements, and hardware resource limitations, and is determined through the following core principles: cost data dispersion analysis, convergence stability constraints, calculation of real-time resource boundaries, and engineering experience rules.

[0076] It should be explained that the 5D cost dimension dataset refers to a 5D structured data matrix formed by dynamically adjusting and integrating cost quantification parameters on the basis of the 4D spatiotemporal dataset. This matrix dynamically links the static cost benchmark with the spatiotemporal progress, thereby achieving a precise spatiotemporal mapping of the resource costs of construction projects.

[0077] S4: Analyze the resource adjustment rules in the deviation compensation strategy containing the optimal cost control path, construct a resource allocation optimization model, and use the gradient descent algorithm to optimize the resource allocation optimization model. Generate a construction scheduling instruction set that minimizes the cost disturbance factor based on the 5D cost dimension dataset.

[0078] Specifically, the steps are as follows: S4.1: Analyze the resource type, adjustment direction, and adjustment amount in the deviation compensation strategy containing the optimal cost control path, convert the resource type, adjustment direction, and adjustment amount into mathematical constraints, and obtain a structured resource adjustment rule table.

[0079] Specifically, the resource type, adjustment direction, and adjustment amount in the deviation compensation strategy containing the optimal cost control path are analyzed. The resource type name, adjustment direction, and adjustment amount value are extracted. Then, the resource type name is converted into a mathematical variable through resource type standardization mapping. According to the adjustment direction to operator conversion rule, the adjustment direction word is mapped to a mathematical operator. According to the adjustment amount to mathematical coefficient conversion rule, the adjustment amount value is converted into a decimal coefficient. Finally, the results are integrated to generate mathematical constraints, which are stored in a structured table to generate a structured resource adjustment rule table.

[0080] It's important to explain that resource type standardization is a rule-based process of converting engineering resource names into mathematical variable symbols, and its core is establishing a unique correspondence between engineering terms and numerical variables. For example, if the engineering resource name is "concrete," the numerical variable is determined to be "C" based on the first letter of "concrete" in English.

[0081] It needs to be explained that the adjustment direction conversion operator rule is a standardized conversion that maps natural language instructions to mathematical operators, and the essence of the adjustment direction conversion operator rule is to define the logical equivalence relationship between instructions and operators. For example, the adjustment direction word is "append", and the mathematical operator is defined as "+".

[0082] It needs to be explained that the rule of converting adjustment quantities to mathematical coefficients refers to the quantification process of converting engineering numerical descriptions into mathematical coefficients or constants. The core of the rule of converting adjustment quantities to mathematical coefficients is to unify the mathematical expression of engineering numerical values. For example, if the adjustment quantity is described as a "percentage", it is converted into a "decimal" mathematical coefficient; if the adjustment quantity is described as an "absolute value", it is converted into a "constant" mathematical coefficient.

[0083] S4.2: Using the resource adjustment rule table as a constraint and the dynamic weight parameters in the 5D cost dimension dataset as decision variables, construct a resource allocation optimization model with minimizing the cost disturbance factor as the objective function, and use the gradient descent algorithm to solve the objective function and update the dynamic weight parameters to obtain the optimized resource allocation optimization model.

[0084] Specifically, the mathematical constraints in the resource adjustment rule table are bound to the dynamic weight parameters in the 5D cost dimension dataset to construct an objective function that minimizes the cost disturbance factor. The gradient descent algorithm is used to iteratively solve this objective function. The weight parameters are initialized to baseline values. The direction of change of the objective function is calculated, and the dynamic weight parameters are adjusted accordingly. This calculation and adjustment process is repeated. When the gradient magnitude of the dynamic weight parameters during repeated calculations is less than the convergence threshold, the direction of change of the objective function tends to stabilize (based on data precision requirements and floating-point precision limitations). The optimal combination of dynamic weight parameters that satisfies all constraints is obtained, generating the optimized resource allocation optimization model. The expression for the objective function is: ; in, The objective function is an instruction that declares that the objective of the entire mathematical expression is to minimize it. Represent decision variables; Indicates the first The constructed predicted cost values ​​are derived from the 5D cost dimension dataset; Indicates the first The baseline cost value for each construction.

[0085] It needs to be explained that the constraint satisfaction determination is carried out during the gradient descent algorithm iteration. After each update of the dynamic weight parameters, the validity of all mathematical constraints must be verified simultaneously. The determination criteria are divided into three categories: equality constraint determination criteria, inequality constraint boundary determination criteria, and engineering physical constraint determination criteria. Among them, the equality constraint determination criteria are: the equality relationship in the resource adjustment rules (e.g., concrete addition ratio) must satisfy the condition that the difference between the values ​​on the left and right sides is less than the engineering-specified error threshold; the inequality constraint boundary determination criteria are: the resource boundary restrictions (e.g., the lower limit of steel reinforcement usage) must satisfy the condition that the adjusted value is not lower than the constraint boundary threshold; the engineering physical constraint determination criteria are: the weight parameters must meet the basic requirements of the engineering feasible region (e.g., the resource allocation amount is negative). During the determination, all constraints are traversed, and only when all types of constraints are satisfied simultaneously is the combination of dynamic weight parameters accepted as a valid solution.

[0086] It should be explained that the constraint boundary threshold refers to the upper and lower limits of material consumption and component quantity specified in the engineering cost specifications, construction technical specifications, and construction contract for each type of engineering resources, and is corrected by combining the geometric quantity of the component and construction feasibility parameters.

[0087] S4.3: Match the spatiotemporal coordinates in the 5D cost dimension dataset by the ID of the building structural components, inject the optimal dynamic weight parameter combination into the spatiotemporal node, then structure the task instructions and add cost disturbance warning instructions, and encapsulate them to obtain the construction scheduling instruction set that minimizes the cost disturbance factor.

[0088] Specifically, the target spatiotemporal node is located by accurately matching the spatiotemporal coordinates in the 5D cost dimension dataset with the ID of the building structural components. The optimal dynamic weight parameter combination is then injected into the attribute field of the corresponding spatiotemporal node as a new attribute. Subsequently, task instructions are generated using structured coding rules, and cost disturbance warning instructions are added. Cost deviation thresholds and response actions are defined. Finally, the target spatiotemporal node, task instructions, and cost disturbance warning instructions are encapsulated to generate a construction scheduling instruction set that minimizes the cost disturbance factor.

[0089] It should be explained that the cost deviation threshold is based on the contract bill of quantities and cost specifications. The benchmark cost and constraint boundary threshold for each type of engineering resource are determined. Combined with historical engineering data and market price fluctuations, typical ranges of cost disturbances are statistically analyzed and an acceptable risk tolerance range is extracted. Then, according to the characteristics of the engineering project (e.g., the tightness of the construction period and the stability of resource supply), a dynamic monitoring range is set at the intersection of the benchmark cost and the risk tolerance range as the cost deviation threshold.

[0090] It should be explained that the structured coding rules are a standardized conversion mechanism for construction task instructions based on preset field formats. By using mandatory delimiters, the task type code, component ID, spatiotemporal coordinates, and resource parameters are converted into machine-readable instruction strings. Value domain constraints (such as coordinates retaining three decimal places and timestamps in ISO8601 format) and fault tolerance rules (missing fields trigger error codes and illegal values ​​are replaced with default values) are added to achieve lossless parsing of construction instructions and integration of cost disturbance early warning.

[0091] S5: By using cognitive mirror synchronization, the construction scheduling instruction set that minimizes cost disturbance factors is subjected to tensor convolution operation to generate the reconstructed 4D spatiotemporal progress topology.

[0092] For details, please refer to Figure 4 The steps are as follows: S5.1: Analyze the construction scheduling instruction set that minimizes cost disturbance factors, extract spatial location, time nodes, and optimal dynamic weight parameter combinations, and then map them into a spatiotemporal resource matrix based on spatial location, time nodes, and optimal dynamic weight parameter combinations.

[0093] Specifically, the construction scheduling instruction set that minimizes cost disturbance factors is analyzed, and the spatial location coordinates, time nodes, and optimal dynamic weight parameter combinations of each instruction item are extracted. Based on the spatial location coordinates and time nodes, the row and column indices of the spatiotemporal resource matrix are constructed. The optimal dynamic weight parameter combinations are split into independent vectors according to resource type and filled into the corresponding cells of the spatiotemporal resource matrix to obtain quantized allocation values. Finally, based on the row and column indices of the spatiotemporal resource matrix and the quantized allocation values, the spatiotemporal resource matrix is ​​generated using the discretized grid mapping method.

[0094] It needs to be explained that the discretized grid mapping method is an engineering method that converts the row and column indices of the spatiotemporal resource matrix and the quantized allocation values ​​into discrete row and column indices, and stores the resource allocation data based on a grid structure. The operation steps of the discretized grid mapping method are as follows: the row and column indices of the spatiotemporal resource matrix are used as unique row indices, then an independent two-dimensional table is created according to each resource type, and the quantized allocation values ​​are filled into the row and column intersection cells of the corresponding independent two-dimensional table to obtain the spatiotemporal resource matrix.

[0095] S5.2: Based on the spatiotemporal resource matrix, construct a three-dimensional convolution kernel, and then perform tensor convolution operation on the three-dimensional convolution kernel to generate a convolution-optimized spatiotemporal resource matrix.

[0096] Specifically, the convolution kernel size is determined based on the spatial-temporal-hardware three-dimensional coupling, covering the spatial three-dimensional coordinates and temporal dimension in the spatiotemporal resource matrix. Weights are allocated according to the distribution of the spatial three-dimensional coordinates and temporal dimension by the center diffusion decay to construct a three-dimensional convolution kernel. Then, the three-dimensional convolution kernel and the spatiotemporal resource matrix are subjected to a full-dimensional convolution operation to obtain the convolution-optimized spatiotemporal resource matrix.

[0097] It should be explained that the three-dimensional coupling of space-time-hardware includes three factors: engineering structural characteristics, construction stage granularity, and computing resource constraints. Among them, engineering structural characteristics are set according to the structural geometric density and resource interaction range, construction stage granularity is set according to the construction rhythm and process continuity requirements, and computing resource constraints are set according to the real-time response time and memory limit.

[0098] It should be explained that full-dimensional convolution operation is a sliding window weighted operation performed on a 4D tensor (three spatial dimensions and time dimension). The core of full-dimensional convolution operation is to achieve spatiotemporal continuity optimization of resource allocation by synchronously sliding the three-dimensional convolution kernel across all dimensions of the spatiotemporal resource matrix.

[0099] S5.3: Transform each coordinate node in the spatiotemporal resource matrix after convolution optimization into a topological vertex, connect the topological vertices based on the resource flow path to form topological edges, and assign weight values ​​to the topological edges according to the resource consumption intensity to generate the reconstructed 4D spatiotemporal progress topology.

[0100] Specifically, each coordinate node in the spatiotemporal resource matrix after convolution optimization is directly mapped to a topological vertex. The attributes of the topological vertex include the original spatial coordinates and timestamp. Based on the preset resource flow path rules, topological vertices with resource flow relationships are connected to form directed topological edges. Then, the weight values ​​of the topological edges are calculated according to the resource consumption intensity and flow distance. Finally, the reconstructed 4D spatiotemporal progress topology structure is generated through the vertex set, directed topological edges, and the attributes of the topological vertices.

[0101] The formula for calculating the weight of topological edges is as follows: ; in, This represents the starting vertex of each coordinate node in the convolution-optimized spatiotemporal resource matrix. This represents the endpoint vertex of each coordinate node in the convolution-optimized spatiotemporal resource matrix. This represents the resource intensity weighting coefficient. This represents the distance decay weighting coefficient. Represents the starting vertex To the final vertex Topological edge weights, Represents the starting vertex To the final vertex The flow distance, Indicates the maximum permissible flow distance. Indicates the starting vertex Resource consumption per unit time Represents the endpoint vertex Resource consumption per unit of time.

[0102] It should be explained that the preset resource flow path rules are topological connection criteria based on engineering constraints and process logic. By quantifying spatial reachability, temporal continuity, and process dependencies, the rules determine whether there are resource flow paths between topological vertices and generate directed edges. The preset resource flow path rules determine the resource flow paths between topological vertices through triple constraints of engineering equipment capabilities, construction rhythm, and process specifications.

[0103] It needs to be explained that the 4D spatiotemporal schedule topology refers to a dynamic graph network that integrates spatial three-dimensional coordinates and time dimension, generated by tensor convolution operations. The core of the 4D spatiotemporal schedule topology consists of topological vertices and directed edges. Furthermore, the 4D spatiotemporal schedule topology transforms the discrete parameters of the construction scheduling instruction set into a spatiotemporal resource matrix after convolution optimization. By constraining the spatial proximity and temporal continuity of resource flow paths, it dynamically optimizes the construction process.

[0104] S6: Update facility status records based on the reconstructed 4D spatiotemporal progress topology, and generate a 6D full lifecycle data record table by combining readings from IoT sensing devices.

[0105] Specifically, the steps are as follows: S6.1: Based on the building structure component ID, accurately associate the topological vertices in the reconstructed 4D spatiotemporal progress topology with the on-site physical facilities, extract the resource flow paths corresponding to the topological vertices, then receive and verify IoT device data in real time, map the IoT device data to the corresponding topological vertices, and generate an enhanced topology with real-time status.

[0106] Specifically, based on the ID of the building structural components, the topological vertices in the reconstructed 4D spatiotemporal progress topology are precisely associated with the on-site physical facilities, establishing a correspondence between the topological vertices and the on-site physical facilities. The resource flow paths associated with the topological vertices are extracted, and the spatial coordinate sequence of the resource flow paths is recorded. Subsequently, IoT device data is received in real time. When verifying the validity of the IoT device data, the IoT device data values ​​are verified to be within the preset engineering threshold range. At the same time, it is verified whether the IoT device data timestamp is included in the vertex time dimension interval. The verified IoT device data is mapped to the real-time status attributes of the corresponding topological vertex, the vertex attribute fields are updated, and finally, an enhanced topology with real-time status is generated.

[0107] It needs to be explained that the correspondence between topological vertices and on-site physical facilities is established through a two-way mapping between topological vertices and physical facilities via building structural component IDs. Specifically, it includes the following three dimensions of precise association: unique identifier binding, space-time state synchronization, and materialization of resource flow paths.

[0108] It should be explained that the preset engineering threshold range is set based on building structure specifications, construction process requirements, and mathematical constraints. It sets the allowable range of engineering parameter set and corrects it by combining historical engineering data and empirical statistical results of similar building structural components. Ultimately, the preset engineering threshold range not only ensures the engineering rationality of IoT device data, but also serves as a judgment standard for real-time status verification.

[0109] It should be explained that IoT device data refers to engineering physical quantities and equipment status data collected in real time through field sensors and controllers. The core is the real-time information flow connecting physical facilities and digital topology. Among them, IoT device data includes: resource consumption data, environmental and status data, and spatial positioning data.

[0110] S6.2: Integrate construction period data, associate the topology vertices of the enhanced topology structure with real-time status with construction logs and quality inspection reports, and construct a 6D full life cycle data record table with building structural component ID as the core node.

[0111] Specifically, by integrating construction period data, the topological vertices of the enhanced topology with real-time status are accurately associated with construction log entries and quality inspection reports through the building structural component ID. A mapping relationship between component ID, log entry, and report number is established. Using the building structural component ID as the core node, the real-time status attributes of the topological vertices, construction log operation records, and quality inspection report parameters are extracted. Then, the real-time status attributes of the topological vertices are converted into real-time status data, the construction log operation records are converted into construction log entry indexes, and the quality inspection report parameters are converted into quality inspection report numbers, thus constructing a 6D full life cycle data record table.

[0112] It should be explained that a construction log entry is a digital construction operation record unit with the building structural component ID as a unique index. It is created in real time when the process starts and dynamically updated as the construction progresses. The construction log entry is a structured record automatically generated when the construction operation begins. It specifically includes spatial coordinates, timestamp, operation type (such as pouring), resource consumption, equipment parameters and responsible person information. It is bound to the topological vertex corresponding to the component ID and time window through the component ID and time window, and finally converted into an entry index.

[0113] It should be explained that the quality inspection report is a digital quality inspection result file with the building structural component ID as the unique index. Physical parameters are collected in real time by on-site sensors, and the results are automatically compared with technical specifications and standards to generate a judgment result. The report is also bound to the construction log entry index and the supervisor's electronic signature to form a structured record.

[0114] It should be explained that the 6D full life cycle data record table uses the building structural component ID as the core index, is based on the 5D cost dimension dataset, and combines historical engineering data with real-time status data. By integrating resource consumption records and the long-term status evolution of facilities, it realizes continuous tracking of resource flow paths in terms of spatial correlation and temporal continuity.

[0115] This embodiment also provides a project construction process management system based on BIM3D-6D technology, including: The topology modeling module loads the 3D building information model through the BIM lightweight computing server, extracts the spatial topological relationships and IFC semantic attributes of geometric components, and performs spatial topology analysis and semantic attribute mapping on the spatial topological relationships and IFC semantic attributes of geometric components to generate a three-dimensional data topology network package based on graph structure. The path deduction module performs multi-dimensional deduction operations on the graph-based three-dimensional data topology network packet to obtain the deviation compensation strategy containing the optimal cost control path. The dimension fusion module generates a 5D cost dimension dataset by expanding the dimensions of the 3D building information model and fusing the data. The resource optimization module analyzes the resource adjustment rules in the deviation compensation strategy containing the optimal cost control path, constructs a resource allocation optimization model, and uses the gradient descent algorithm to optimize the resource allocation optimization model. Based on the 5D cost dimension dataset, it generates a construction scheduling instruction set that minimizes the cost disturbance factor. The structural reconstruction module performs tensor convolution operations on the construction scheduling instruction set that minimizes cost disturbance factors through cognitive mirror synchronization, generating a reconstructed 4D spatiotemporal progress topology. The closed-loop feedback module updates the facility status record based on the reconstructed 4D spatiotemporal progress topology and generates a 6D full lifecycle data record table by combining the readings of IoT sensing devices.

[0116] This embodiment also provides a computer device, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the full-process management method for construction projects based on BIM3D-6D technology as proposed in the above embodiment.

[0117] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0118] This embodiment also provides a storage medium storing a computer program. When executed by a processor, the program implements the full-process management method for construction projects based on BIM3D-6D technology as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0119] In summary, this invention performs tensor convolution operations on the construction scheduling instruction set through cognitive mirror synchronization, mapping discrete scheduling instructions into a continuous spatiotemporal topology. It utilizes the weight decay characteristics of the three-dimensional convolution kernel to achieve spatiotemporal smooth diffusion of resource paths, generating a reconstructed 4D spatiotemporal progress topology. Through mathematical transformation, discrete decisions are converted into a continuous field model, enabling resource flow paths to possess spatial correlation and temporal continuity, thus achieving the beneficial effect of dynamically optimizing the construction process.

[0120] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for managing the entire construction project process based on BIM 3D-6D technology, characterized in that: include: The 3D building information model is loaded by a BIM lightweight computing server, the spatial topological relationships and IFC semantic attributes of geometric components are extracted, and the spatial topological relationships and IFC semantic attributes of geometric components are analyzed and mapped to generate a three-dimensional data topology network package based on graph structure. Multi-dimensional extrapolation is performed on graph-based 3D data topology network packets to obtain deviation compensation strategies containing the optimal cost control path; By expanding the dimensions and fusing data from 3D building information models, a 5D cost dimension dataset is generated. The resource adjustment rules in the deviation compensation strategy containing the optimal cost control path are analyzed, a resource allocation optimization model is constructed, and the gradient descent algorithm is used to optimize the resource allocation optimization model. Based on the 5D cost dimension dataset, a construction scheduling instruction set that minimizes the cost disturbance factor is generated. By using cognitive mirror synchronization, the construction scheduling instruction set that minimizes cost disturbance factors is subjected to tensor convolution operation to generate a reconstructed 4D spatiotemporal progress topology. Based on the reconstructed 4D spatiotemporal progress topology, the facility status record is updated, and combined with the readings of IoT sensing devices, a 6D full lifecycle data record table is generated.

2. The project construction process management method based on BIM 3D-6D technology as described in claim 1, characterized in that: The steps are as follows: Loading a 3D building information model using a lightweight BIM computing server, extracting the spatial topological relationships and IFC semantic attributes of geometric components, performing spatial topological analysis and semantic attribute mapping on the spatial topological relationships and IFC semantic attributes of the geometric components, and generating a graph-based 3D data topology network packet. By loading the 3D building information model through the BIM lightweight computing server, the non-critical structural mesh is simplified to generate the geometric deformation degree and topological relationship matrix, and the 3D building information model is scanned based on the geometric deformation degree to obtain the geometric element set; The semantic attributes of IFC in the set of geometric elements are analyzed, and the topological relation matrix is ​​converted into a topological graph with weighted attributes. Based on the spatial overlap threshold, adjacent 3D nodes of the topology graph with weighted attributes are merged to construct an R-tree 3D spatial index, and the R-tree 3D spatial index is converted into a 3D topology data network. Verify the 3D data topology network, generate a connectivity graph of building structural components and an attribute distribution report, and encapsulate it to obtain a graph-based 3D data topology network package.

3. The project construction process management method based on BIM3D-6D technology as described in claim 1, characterized in that: The steps for performing multi-dimensional deduction on graph-based 3D data topology network packets to obtain a deviation compensation strategy containing the optimal cost control path are as follows: Topological node and edge data are extracted from the 3D data topological network package of the graph structure. Based on the time window constraint, the spatial coordinates of the building structure components are extended into 4D spatiotemporal parameters. A preset resource cost parameter table is superimposed as a dynamic correction factor to generate an enhanced topological network with perturbation factor. Based on the enhanced topology network with perturbation factor, the perturbation effect is propagated along the topology edge, and the stability perturbation guarantee value is obtained through iteration. Then, multi-dimensional scenario simulation is performed on the stability perturbation guarantee value to generate a three-dimensional diagnostic report. A multi-objective optimization function is constructed based on the 3D diagnostic report. Then, the optimal non-dominated solution set is selected from the multi-objective optimization function to generate a deviation compensation strategy containing the optimal cost control path.

4. The project construction process management method based on BIM3D-6D technology as described in claim 3, characterized in that: The preset resource cost parameter table includes a component type matching field, cost parameter values, and structured data related to parameter constraints.

5. The project construction process management method based on BIM 3D-6D technology as described in claim 1, characterized in that: The steps for generating a 5D cost dimension dataset from a 3D building information model through dimensional expansion and data fusion are as follows: A replacement simplification method is used to process the geometric element set of the 3D building information model, and a lightweight geometric dataset is generated based on the geometric entities and attribute entities of the 3D building information model. By associating the construction schedule with the building structural component ID, the installation time nodes are bound to the building structural components in the lightweight geometry dataset. The external bill of quantities is parsed, the geometric quantities of the building structural components in the lightweight geometry dataset are matched, and the cost baseline matrix is ​​obtained. Dynamic cost correction is performed on the 4D spatiotemporal dataset and cost baseline matrix to generate a 5D cost dimension dataset.

6. The project construction process management method based on BIM 3D-6D technology as described in claim 5, characterized in that: The external bill of quantities is a structured resource quantification table built on cost standards, which accurately links the geometric quantities, resource consumption rules, and cost parameters of building structural components.

7. The project construction process management method based on BIM 3D-6D technology as described in claim 1, characterized in that: The resource adjustment rules in the deviation compensation strategy containing the optimal cost control path are analyzed, a resource allocation optimization model is constructed, and the gradient descent algorithm is used to optimize the resource allocation optimization model. Based on the 5D cost dimension dataset, a construction scheduling instruction set that minimizes the cost disturbance factor is generated. The steps are as follows: The resource type, adjustment direction, and adjustment amount in the deviation compensation strategy containing the optimal cost control path are analyzed, and the resource type, adjustment direction, and adjustment amount are converted into mathematical constraints to obtain a structured resource adjustment rule table. Using the resource adjustment rule table as a constraint and the dynamic weight parameters in the 5D cost dimension dataset as decision variables, a resource allocation optimization model is constructed with minimizing the cost disturbance factor as the objective function. The gradient descent algorithm is then used to solve the objective function and update the dynamic weight parameters to obtain the optimal combination of dynamic weight parameters and generate the optimized resource allocation optimization model. By matching the spatiotemporal coordinates in the 5D cost dimension dataset with the IDs of building structural components, the optimal dynamic weight parameter combination is injected into the spatiotemporal nodes. Subsequently, the task instructions are structured and coded, and cost disturbance warning instructions are added. After encapsulation, the construction scheduling instruction set that minimizes the cost disturbance factor is obtained.

8. The project construction process management method based on BIM 3D-6D technology as described in claim 1, characterized in that: The steps for generating a reconstructed 4D spatiotemporal progress topology by performing tensor convolution operations on the construction scheduling instruction set that minimizes cost disturbance factors through cognitive mirror synchronization are as follows: The construction scheduling instruction set that minimizes cost disturbance factors is analyzed, and spatial location, time node, and optimal dynamic weight parameter combination are extracted. Then, based on the spatial location, time node, and optimal dynamic weight parameter combination, it is mapped into a spatiotemporal resource matrix. Based on the spatiotemporal resource matrix, a three-dimensional convolution kernel is constructed, and then tensor convolution operation is performed on the three-dimensional convolution kernel to generate a convolution-optimized spatiotemporal resource matrix. Each coordinate node in the convolution-optimized spatiotemporal resource matrix is ​​transformed into a topological vertex. The topological vertices are connected based on the resource flow path to form topological edges. Weight values ​​are assigned to the topological edges according to the resource consumption intensity to generate the reconstructed 4D spatiotemporal progress topology.

9. The project construction process management method based on BIM 3D-6D technology as described in claim 1, characterized in that: The process of updating facility status records based on the reconstructed 4D spatiotemporal progress topology, combined with readings from IoT sensing devices, to generate a 6D full lifecycle data record table involves the following steps: Based on the association of building structural component IDs with 4D spatiotemporal progress topology vertices and on-site physical facilities, the resource flow path of the topology vertex is extracted, and then the verified IoT device data is fused to the corresponding topology vertex to generate an enhanced topology with real-time status. Integrate construction period data, associate the topology vertices of the enhanced topology structure with real-time status with construction logs and quality inspection reports, and construct a 6D full life cycle data record table with building structural component ID as the core node.

10. A project construction process management system based on BIM3D-6D technology, comprising the project construction process management method based on BIM3D-6D technology as described in any one of claims 1 to 9, characterized in that: include: The topology modeling module loads the 3D building information model through the BIM lightweight computing server, extracts the spatial topological relationships and IFC semantic attributes of geometric components, and performs spatial topology analysis and semantic attribute mapping on the spatial topological relationships and IFC semantic attributes of geometric components to generate a three-dimensional data topology network package based on graph structure. The path deduction module performs multi-dimensional deduction operations on the graph-based three-dimensional data topology network packet to obtain the deviation compensation strategy containing the optimal cost control path. The dimension fusion module generates a 5D cost dimension dataset by expanding the dimensions of the 3D building information model and fusing the data. The resource optimization module analyzes the resource adjustment rules in the deviation compensation strategy containing the optimal cost control path, constructs a resource allocation optimization model, and uses the gradient descent algorithm to optimize the resource allocation optimization model. Based on the 5D cost dimension dataset, it generates a construction scheduling instruction set that minimizes the cost disturbance factor. The structural reconstruction module performs tensor convolution operations on the construction scheduling instruction set that minimizes cost disturbance factors through cognitive mirror synchronization, generating a reconstructed 4D spatiotemporal progress topology. The closed-loop feedback module updates the facility status record based on the reconstructed 4D spatiotemporal progress topology and generates a 6D full lifecycle data record table by combining the readings of IoT sensing devices.

Images