Intelligent construction progress management and control system based on digital twinning

Abstract

This invention relates to the field of construction management technology and discloses an intelligent construction progress control system based on digital twins, including modules for data mapping, voxel evolution, potential energy analysis, inverse decoupling, and command generation. The system receives heterogeneous data and extracts unified geometric and state information to generate voxel mesh data containing rheological stiffness attributes; it calculates global potential energy field data to quantify spatiotemporal conflict risks based on occupancy status and rheological stiffness attributes; it constructs a disturbance inverse transmission map based on the voxel mesh and potential energy field, and calculates the optimal spatiotemporal displacement parameters for mitigating progress deviations by searching for and adjusting the path with the minimum energy; finally, it maps the parameters back to the model to generate spatiotemporal operation commands and dynamic electronic fence signals. This invention, by introducing rheological stiffness and potential energy field mechanisms, effectively quantifies physical adjustment resistance and environmental risks, achieving dynamic and precise correction and safety control of construction progress in complex environments.

Description

Technical Field

[0001] This invention relates to the field of construction management technology, specifically to an intelligent control system for construction progress based on digital twins. Background Technology

[0002] Currently, large-scale construction projects face significant temporal and spatial complexity, and precise control of construction progress directly impacts project delivery efficiency. Digital twin technology is increasingly being introduced into construction management, aiming to achieve digital tracking of the construction process by establishing a real-time mapping between the physical site and a virtual model. Managers rely on data feedback to attempt to maintain established construction plans in a dynamically changing environment, hoping to solve common problems such as schedule delays.

[0003] Regarding the application of digital twins in construction, existing management systems are typically based on the association between Building Information Modeling (BIM) and Work Schedule (WBS). The system collects point cloud data or image information from the site and performs a geometric comparison with the design model. Identified schedule discrepancies are marked on the 4D model, using color to distinguish components that have not yet started, are in progress, or have been completed. Based on this visualization information, schedulers use the Critical Path Method (CPM) to logically extrapolate and adjust the start time or duration of subsequent work processes.

[0004] However, existing technologies rely solely on the linear relationship between geometric models and time parameters, neglecting the state evolution characteristics of physical entities. During construction, the difficulty of adjusting the work surface is not constant. For example, the physical stiffness of concrete changes dramatically over time as it transitions from a fluid to a solid state. Existing models cannot quantify this increasing adjustment resistance as the process progresses, leading to adjustment schemes that often violate physical laws. Furthermore, conventional collision detection is limited to hard collisions between components, lacking a field-specific description of non-contact environmental risks such as high temperatures, harmful gases, or vibrations, and failing to automatically identify hidden spatial exclusion zones formed by environmental repulsion. When dealing with schedule deviations, traditional algorithms lack a global energy assessment of the cost of disturbance transmission, making it difficult to calculate the decoupling path with the least impact on the overall process, resulting in corrective measures that are often ineffective.

[0005] Therefore, the present invention provides an intelligent management and control system for building construction progress based on digital twins to address the shortcomings of existing technologies. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides an intelligent construction progress management and control system based on digital twins. This system solves the problem that in complex construction environments, the lack of quantification of the evolution of the physical stiffness of construction processes and non-contact environmental risks leads to progress correction strategies being detached from actual physical constraints and making it difficult to automatically avoid spatiotemporal conflicts.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an intelligent management and control system for building construction progress based on digital twins, comprising: The data mapping module is used to receive building information model and on-site measured data, and process them to generate standardized data containing geometric and state information under a unified coordinate system. The voxel evolution module is used to divide the construction site into continuous voxels based on the standardized data, construct a rheological stiffness evolution function, and calculate the rheological stiffness that characterizes the spatial physical state and resists scheduling adjustment resistance. The potential energy analysis module is used to combine the state information with the rheological stiffness to calculate the operational repulsive potential energy, and superimpose the environmental derived potential energy to generate a global potential energy field characterizing the on-site safety constraints. The reverse decoupling module is used to construct a disturbance reverse transmission map based on the construction logic when a construction delay is detected. It uses the rheological stiffness as a damping weight, searches for the reverse decoupling path with the minimum adjustment energy under the constraints of the global state energy field, and calculates the optimal spatiotemporal displacement parameters of the process to be adjusted. The instruction generation module is used to generate spatiotemporal operation instructions based on the optimal spatiotemporal displacement parameters, and to generate dynamic electronic fence boundary signals based on the global situational energy field.

[0008] By adopting the above technical solution, the discrete schedule scheduling problem is transformed into an energy minimization optimization problem within a continuous physical field by using rheological stiffness to quantify the physical space's resistance to process adjustments and constructing a global potential energy field to superimpose safety constraints and environmental influences. When schedule delays occur, the reverse decoupling module uses rheological stiffness as a damping weight to search for the path with the minimum adjustment energy under the constraint of the potential energy field, rather than simply compressing based on time logic. Therefore, deep coupling between physical space constraints and time schedule logic is achieved, ensuring that the generated schedule correction scheme can automatically avoid hardened structural areas and high-risk environmental areas, achieving the optimal adjustment effect that balances construction safety and on-site physical feasibility.

[0009] Preferably, when performing coordinate transformation processing, the data mapping module calculates the rigid body transformation matrix using the coordinates of preset control points, and maps the measured data on site to a unified coordinate system through this matrix; at the same time, it calculates the axis-aligned bounding box of the construction scene to define the global spatial domain range of voxels, and performs manifold checks and normal repairs on the building information model; the voxel evolution module establishes a multi-dimensional attribute vector for each voxel, including occupancy flag, rheological stiffness, environmentally derived potential energy value, and resource carrying capacity threshold.

[0010] By adopting the above technical solution, the spatial benchmark differences of multi-source heterogeneous data are eliminated, the spatial consistency of voxel segmentation is ensured, and an accurate geometric topological basis and data structure support are provided for physical field calculations.

[0011] Preferably, the rheological stiffness evolution function is generated as follows: for processes with physical phase transition characteristics, the rheological stiffness is controlled to increase exponentially over time until it reaches its maximum value based on the material physical hardening principle after the operation is completed; for processes with only spatial exclusivity, a constant high value is assigned during the operation and reset after the operation is completed; the potential energy analysis module obtains the environmental derived potential energy by calculating the propagation and attenuation distribution of the potential energy generated by the potential energy emission source to the surrounding space, and linearly superimposes it with the operation repulsion potential energy calculated based on the occupancy state and rheological stiffness to generate a global potential energy field.

[0012] By adopting the above technical solutions, the system can dynamically distinguish between physically irreversible and reversible operations and quantify the spatial distribution of environmental risks; by simulating the hardening process of materials such as concrete through an exponentially growing stiffness function, the scheduling algorithm can identify and adapt to the physical properties of materials; through multi-field coupling, complex safety hazards are transformed into a unified scalar field, realizing the digital marking of high-risk areas.

[0013] Preferably, the reverse decoupling module starts from the source of the progress disturbance and constructs a disturbance reverse propagation map along the reverse direction of the process logic and the lateral spatial direction; when searching for a path, it calculates the decoupling cost weight that is positively correlated with the process rheological stiffness, executes the minimum energy path search algorithm, and guides the adjustment path to the low rheological stiffness region; the spatiotemporal operation instructions generated by the instruction generation module include a description of the regional rheological stiffness state, and generate a virtual geofence based on the region exceeding the safety threshold in the global situational energy field.

[0014] By adopting the above technical solution and utilizing graph theory and the principle of minimum energy, soft connections are automatically found and adjusted in complex process networks. That is, processes with movable resources, unhardened spaces, and low safety risks are prioritized for adjustment. In theory, this ensures that the implementation cost of the schedule correction scheme is minimized and the disturbance to the physical state on site is minimized. It also realizes closed-loop control from calculation results to on-site execution instructions and safety boundaries.

[0015] This invention provides an intelligent management and control system for building construction progress based on digital twins. It has the following beneficial effects: 1. This invention introduces time-varying rheological stiffness properties through a voxel evolution module, discretizing the construction site into a voxel mesh containing physical state information. Compared to traditional schedule management based solely on time parameters, this scheme can dynamically quantify the adjustment resistance of the work surface according to the process characteristics, ensuring that the generated schedule control strategy fully considers the plasticity changes of physical entities, thus improving the practicality and scientific nature of construction plan adjustments.

[0016] 2. This invention utilizes a potential energy analysis module to construct a global potential energy field encompassing both operational repulsion and environmentally derived risks, quantifying both contact-based and non-contact-based risks during construction. This mechanism not only identifies geometrical spatial conflicts but also transforms environmental hazards such as high temperatures and harmful gases into high-potential-energy repulsion zones. Consequently, it automatically plans safe avoidance paths when calculating optimal spatiotemporal displacement parameters, effectively reducing the probability of spatiotemporal conflicts and the risk of safety accidents at the construction site.

[0017] 3. This invention constructs a disturbance reverse propagation map through a reverse decoupling module and executes a minimum energy path search algorithm, transforming the schedule deviation mitigation problem into an energy optimization problem. Utilizing the rheological stiffness integral as a damping weight, the system can distinguish between hard process constraints and soft logic constraints, quickly calculating the optimal spatiotemporal displacement parameters that minimize the impact on related processes and reduce adjustment costs. This achieves precise correction and efficient control of construction progress under complex disturbance environments. Attached Figure Description

[0018] Figure 1 This is a framework diagram of a digital twin-based intelligent management and control system for building construction progress according to an embodiment of the present invention. Figure 2 This is a flowchart of a digital twin-based intelligent management and control method for building construction progress according to an embodiment of the present invention.

[0019] Among them, 110 is the data mapping module; 120 is the voxel evolution module; 130 is the potential energy analysis module; 140 is the reverse decoupling module; and 150 is the instruction generation module. Detailed Implementation

[0020] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] See attached document Figure 1 This invention provides an intelligent construction progress management system based on digital twins, which operates on a computing device with data processing capabilities. The computing device includes at least one processor, a memory, and a communication interface. The memory stores computer program instructions, and the processor executes these instructions to implement system functions. The computing device communicates with data acquisition terminals at the construction site via the communication interface. Data acquisition terminals include, but are not limited to, laser scanners, oblique photography drones, RFID readers, and video surveillance equipment installed at key nodes on the construction site. The computing device is also connected to a display terminal to present a 3D visualization interface and management commands to the user.

[0022] like Figure 1 As shown, the intelligent construction progress control system based on digital twins includes a data mapping module 110, a voxel evolution module 120, a potential energy analysis module 130, a reverse decoupling module 140, and an instruction generation module 150. The modules transmit and interact with each other via a system bus or internal communication protocol.

[0023] The data mapping module 110 is configured as the system's input interface, connecting to both the database storing the Building Information Model (BIM) and the on-site data acquisition terminal. The data mapping module 110 receives BIM data generated during the building construction design phase. This BIM data includes the geometric dimensions, material properties, and preset schedules of building components. Simultaneously, the data mapping module 110 also receives on-site measured data from the data acquisition terminal, including point cloud data of the construction site, component location coordinates, and environmental monitoring data. The data mapping module 110 cleans, aligns, and transforms the received heterogeneous data, extracting geometric and state information in a unified coordinate system, and then transmits the processed, standardized data to the voxel evolution module 120.

[0024] The voxel evolution module 120 is connected to the data mapping module 110 and is used to receive standardized data and construct a spatiotemporal discretized model of the construction site. The voxel evolution module 120 divides the three-dimensional space of the construction site into continuous cubic units, i.e., voxels. The voxel evolution module 120 establishes a multi-dimensional attribute vector for each voxel, including occupancy status, rheological stiffness, environmental potential energy, and resource carrying capacity limit. Based on a preset process characteristic library, the voxel evolution module 120 generates an evolution function of the rheological stiffness of each voxel over time. This rheological stiffness characterizes the physical state of the corresponding spatial region at a specific time point and the magnitude of its resistance to scheduling adjustments. The voxel evolution module 120 transmits the voxel mesh data containing time-varying attributes to the potential energy analysis module 130 and the reverse decoupling module 140.

[0025] The potential energy analysis module 130 is connected to the voxel evolution module 120 and is used to calculate the overall potential energy field distribution at the construction site in real time. Based on the voxel occupancy state and rheological stiffness properties, the potential energy analysis module 130 calculates the repulsive potential energy generated when different work processes compete for the same spatiotemporal resources. Furthermore, the potential energy analysis module 130 also incorporates environmental monitoring data to calculate the environmentally derived potential energy generated by construction activities, quantifying non-contact spatial conflict risks. The potential energy analysis module 130 feeds back the calculated overall potential energy field data to the reverse decoupling module 140 to assist in path search and decision-making.

[0026] The reverse decoupling module 140 is connected to the voxel evolution module 120 and the potential energy analysis module 130, serving as the core computing unit of the system. The reverse decoupling module 140 triggers a rescheduling mechanism when the actual construction progress lags behind the preset plan. Based on the construction logic, the reverse decoupling module 140 constructs a disturbance reverse propagation graph and uses the rheological stiffness integral of the voxels as the damping weight of the edge connections in the graph. Under the premise of satisfying the overall potential energy field safety constraints, the reverse decoupling module 140 searches for the reverse decoupling path with the minimum adjustment energy in the disturbance reverse propagation graph, calculates the optimal spatiotemporal displacement parameters of each process to be adjusted, and sends these displacement parameters to the instruction generation module 150.

[0027] The instruction generation module 150 is connected to the reverse decoupling module 140 and is used to receive optimal spatiotemporal displacement parameters and convert them into executable control instructions. The instruction generation module 150 maps abstract time and spatial coordinates back to the building information model, generating spatiotemporal operation instructions that include specific work areas and work time windows. Simultaneously, the instruction generation module 150 generates dynamic electronic fence boundary signals based on high potential energy area data calculated by the potential energy analysis module 130. The instruction generation module 150 sends the spatiotemporal operation instructions and dynamic electronic fence boundary signals to a display terminal for visualization, or to a handheld terminal for on-site operations to guide construction.

[0028] See attached document Figure 1 The data mapping module 110 is configured as the data input and preprocessing front end of the intelligent construction progress management system based on digital twins. The data mapping module 110 establishes a communication connection with the server storing the building information model (BIM) via a high-speed data interface, reading the BIM file conforming to industrial basic standards. The data mapping module 110 parses the BIM file, extracting the globally unique identifier, geometric mesh data, design material parameters, and preset schedule time nodes for each building component. The data mapping module 110 parses the geometric mesh data into vertex coordinate arrays and face index arrays, clarifying the precise spatial form of each component in the design coordinate system. Simultaneously, the data mapping module 110 extracts the theoretical start and end times of each component from the schedule, establishing a mapping table between component identifiers and the time dimension.

[0029] The data mapping module 110 also connects to data acquisition terminals distributed across the construction site via a wireless sensor network or a wired local area network. The data mapping module 110 receives in real-time point cloud data streams from a laser scanner, 3D mesh update data from an oblique photogrammetry drone, and component arrival status signals from an RFID reader. For the point cloud data acquired by the laser scanner, the data mapping module 110 performs data decompression and format conversion operations, converting the raw binary data stream into a 3D point set with intensity information and spatial coordinates. For the RFID signals, the data mapping module 110 parses the tag ID and matches it with the unique identifier of the component in the building information model, thereby obtaining the current physical location status of the component, such as whether it has arrived on site, been hoisted, or been fixed.

[0030] After acquiring heterogeneous data from multiple sources, the data mapping module 110 performs coordinate system registration and unification processing. Since the Building Information Model (BIM) uses a design coordinate system, while field acquisition equipment is typically based on a local measurement coordinate system or a geodetic coordinate system, the data mapping module 110 calculates a rigid body transformation matrix using preset control point coordinates or feature matching algorithms. The data mapping module 110 applies this rigid body transformation matrix to map the coordinates of all field measured data to the design coordinate system of the BIM, ensuring strict spatial alignment between the virtual model and the physical entity. This step eliminates spatial deviations between data sources, providing a unified geometric benchmark for subsequent spatiotemporal conflict calculations.

[0031] To accommodate the discretization requirements of the subsequent voxel evolution module 120, the data mapping module 110 preprocesses the extracted geometric information. The data mapping module 110 calculates the axis-aligned bounding box of the entire construction scene, determines the extreme values ​​of the three-dimensional spatial boundaries of the construction area, and thus defines the global spatial domain range for voxel mesh generation. The data mapping module 110 also performs manifold checks and normal repair on the complex triangular meshes extracted from the building information model, ensuring that the geometric topology of each component is closed and the normal direction faces outwards, preventing internal voids or incorrect occupancy determinations during voxelization. For point cloud data collected on-site, the data mapping module 110 performs statistical filtering and outlier removal operations, eliminating noise caused by airborne dust, birds, or measurement noise, retaining valid point cloud data representing the entity surface. The cleaned, registered, and repaired standard geometric data packets and state data packets are sent by the data mapping module 110 to the voxel evolution module 120.

[0032] See attached document Figure 1The voxel evolution module 120, connected to the data mapping module 110, is configured as the system's spatiotemporal discretization modeling and state evolution calculation unit. The voxel evolution module 120 first receives cleaned and registered standard geometric data packets and performs voxelization operations based on preset spatial resolution parameters. The voxel evolution module 120 discretizes the continuous three-dimensional Euclidean space of the construction site into a set of regularly arranged cubic units, i.e., a voxel mesh. The voxel evolution module 120 assigns a unique spatial index coordinate to each voxel and establishes a four-dimensional state vector storage structure. This four-dimensional state vector contains not only the geometric position information of the voxel in three-dimensional space but also the state sequence information in the time dimension. Specifically, the data structure of each voxel stores an occupancy flag, a rheological stiffness value, an environmentally derived potential energy value, and a resource carrying capacity threshold. The occupancy flag indicates whether the voxel is occupied by a physical component at a specific moment; the rheological stiffness value quantifies the resistance of the space to external disturbances or scheduling adjustments at a specific moment; the environmentally derived potential energy value stores the intensity of non-contact influences derived from surrounding operations; and the resource carrying capacity threshold defines the maximum physical load or work density that the space can withstand.

[0033] The core function of the voxel evolution module 120 is to generate and maintain the rheological stiffness evolution model for each voxel. The voxel evolution module 120 has a built-in construction process attribute library, which defines the physical state change patterns corresponding to different types of construction procedures. Based on the procedure type information extracted from the building information model, the voxel evolution module 120 maps the corresponding rheological stiffness evolution function to the relevant voxel set. Rheological stiffness is defined here as a scalar attribute that evolves unidirectionally over time, and its value reflects the strength of the physical phase transition process and logical constraints of the working surface.

[0034] For processes involving physical phase transitions, such as concrete pouring, the rheological stiffness function generated by the voxel evolution module 120 exhibits a phased variation. During the pre-pouring rebar tying and formwork erection phase, the voxel evolution module 120 sets the rheological stiffness of the corresponding voxels to a low baseline value, indicating that the work surface is physically reversible, allowing for fine-tuning of the process sequence or the interleaving of workers. Upon entering the concrete pouring phase, the voxel evolution module 120 increases the rheological stiffness to a work-locked value, indicating that the area is under high-intensity, exclusive operation, strictly prohibiting the intervention of other unrelated processes. During the curing phase after pouring, the voxel evolution module 120, based on the hydration reaction principle of concrete, controls the rheological stiffness value to increase exponentially over time. This growth process simulates the physical process of concrete transforming from a fluid to a solid state, indicating that as time progresses, the difficulty and cost of dismantling or modifying the component increase sharply until it reaches a fully solidified state, at which point the rheological stiffness reaches its maximum extreme value, logically forming an insurmountable rigid barrier.

[0035] For processes that lack physical phase transition characteristics but are spatially exclusive, such as steel structure hoisting or welding, the voxel evolution module 120 employs linear or step-type rheological stiffness functions. During the operation, the voxel evolution module 120 assigns constant high rheological stiffness values ​​to relevant voxels to prevent spatial conflicts. During work breaks or after completion, if the space is no longer occupied by entities, the voxel evolution module 120 resets the rheological stiffness values ​​to zero or the environmental background value, releasing the space resource for reuse by subsequent processes. Through this voxel-by-voxel, time-by-time calculation, the voxel evolution module 120 generates a dynamic four-dimensional scalar field covering the entire construction cycle. This scalar field precisely describes the physical stiffness and logical availability of every cubic centimeter of space on the construction site every second. The voxel evolution module 120 sends the generated spatiotemporal voxel dataset containing rheological stiffness evolution information to the potential energy analysis module 130, providing a physical constraint basis for subsequent conflict detection and path planning.

[0036] See attached document Figure 1 The potential energy analysis module 130 is communicatively connected to the voxel evolution module 120 and is configured as a processing unit to perform global spatiotemporal conflict quantification calculations. The potential energy analysis module 130 receives a spatiotemporal voxel dataset containing rheological stiffness properties output by the voxel evolution module 120 and constructs a dynamic potential energy field covering the entire construction cycle based on the principles of physical field theory. The potential energy analysis module 130 transforms the space occupancy competition and environmental safety risks during construction into scalar potential energy values, and characterizes the accessibility and safety of spatiotemporal resources through the gradient distribution of the potential energy field. The global potential energy field calculated by the potential energy analysis module 130 is composed of a work-repulsive potential energy field and an environmentally derived potential energy field, used to accurately identify physical collision risks between physical components and non-contact process compatibility conflicts.

[0037] The potential energy analysis module 130 first calculates the repulsive potential energy field, which mainly reflects the resistance of the physical component or work space to external intrusion. The module 130 traverses each voxel in the spatiotemporal voxel grid, calculating its inherent repulsive potential energy intensity based on the voxel's rheological stiffness value and occupancy status at the current time step. When a voxel is marked as occupied and its rheological stiffness value is high, such as during the concrete hardening period or the steel structure fixing period, the module 130 generates an extremely high-intensity repulsive potential energy peak at that voxel location and generates a rapidly decaying potential energy gradient band in its neighboring voxels. This high-gradient potential energy distribution logically constitutes an absolute spatial no-entry zone; any subsequent process attempting to plan entry into this zone within the same time period will be judged as a serious conflict by the system because the calculated potential energy cost exceeds a threshold. Conversely, when the rheological stiffness value of the voxel is low, such as during the rebar binding or temporary storage stage, the potential energy analysis module 130 generates a repulsive potential energy field with low intensity and slow decay, allowing other non-rigid dependent processes to perform spatiotemporal compression or adjacent operations at the cost of certain decoupling.

[0038] Based on the calculation of the repulsive potential energy of the operation, the potential energy analysis module 130 further calculates the environmentally derived potential energy field to resolve implicit safety conflicts that traditional geometric collision detection cannot identify. The potential energy analysis module 130 identifies operation sources with environmental impact attributes in the construction site, including but not limited to welding operation points that generate high-temperature sparks, painting operation surfaces that emit volatile gases, and the locations of mechanical equipment that generate high-frequency vibrations. The potential energy analysis module 130 defines these operation sources as potential energy emission sources and calculates the propagation and attenuation distribution of potential energy into the surrounding space based on a preset physical diffusion model. For example, for welding operations, the potential energy analysis module 130 establishes a thermal radiation potential energy field centered on the operation point, whose potential energy value decreases nonlinearly with increasing spatial distance.

[0039] The potential energy analysis module 130 performs multi-field coupling operations, linearly superimposing the job-rejecting potential energy field with the environmentally derived potential energy field to generate the final global total potential energy field. During the superposition process, the potential energy analysis module 130 detects the overlap of peaks in different potential energy fields. If a certain area is not occupied by physical components, i.e., the job-rejecting potential energy is zero, but the environmentally derived potential energy exceeds a preset safety threshold due to the influence of nearby high-risk operations, the potential energy analysis module 130 marks this area as a high-potential-energy hazard zone. For example, when the voxel of the flammable material storage area overlaps with the high-potential-energy area of ​​the welding operation's heat radiation, even though the two are not geometrically in contact, the total potential energy value calculated by the potential energy analysis module 130 will trigger an alarm. The potential energy analysis module 130 sends the calculated global total potential energy field data matrix to the reverse decoupling module 140 as the numerical basis for subsequent calculation of the rescheduling path cost function.

[0040] See attached document Figure 1The reverse decoupling module 140 establishes communication connections with both the voxel evolution module 120 and the potential energy analysis module 130, and is configured as the core decision-making and computational unit and adaptive rescheduling engine of the system. The reverse decoupling module 140 receives real-time construction progress status data and compares it with a preset baseline plan. When the actual progress lag on the critical path exceeds a preset tolerance threshold, or when a high-risk potential energy warning is received from the potential energy analysis module 130, the reverse decoupling module 140 automatically triggers the spatiotemporal rescheduling calculation process. The goal of the reverse decoupling module 140 is to absorb the generated progress deviations by adjusting the spatiotemporal distribution of unfinished processes without violating physical safety constraints or adding extra time.

[0041] The reverse decoupling module 140 first constructs a disturbance reverse propagation graph based on the current remaining workload and process logic relationships. In this graph structure, each node represents an independent process or work unit to be executed, and nodes are connected by directed edges, representing the process precedence relationship, resource sharing relationship, and spatial adjacency relationship between processes. The reverse decoupling module 140 not only constructs this graph as a time-dimensional network, but also extends it into a spatiotemporal topology network containing spatial location information. Starting from the source node of the progress disturbance, the reverse decoupling module 140 performs a breadth-first search along the reverse direction of the logical connection line and the lateral spatial direction to identify all related process nodes that may be affected logically or spatially, thereby defining the computational domain boundary for rescheduling.

[0042] To achieve intelligent decision-making based on physical properties, the reverse decoupling module 140 introduces a stiffness damping mechanism to weight the disturbance reverse propagation spectrum. The reverse decoupling module 140 calls the rheological stiffness data provided by the voxel evolution module 120 to calculate the comprehensive stiffness characteristic value of each associated process within a specific time window. The reverse decoupling module 140 maps this comprehensive stiffness characteristic value to the damping weight of the corresponding connection edge in the spectrum. Specifically, for processes in a high rheological stiffness state, such as the final setting process of concrete or the loading process of stressed components, the reverse decoupling module 140 assigns a higher damping weight, defining it as a rigid connection at the algorithm level. This means that adjusting the time or spatial location of this process requires a significant physical cost or safety risk. For processes in a low rheological stiffness state, such as decoration and finishing or material transportation on non-critical routes, the reverse decoupling module 140 assigns a lower damping weight, defining it as a flexible connection. This means that the process possesses high spatiotemporal compressive elasticity or portability.

[0043] Based on the weighted perturbation reverse propagation map, the reverse decoupling module 140 executes a minimum energy path search algorithm. This algorithm aims to find an optimal combination of process adjustments that minimizes the total damping energy required to absorb the schedule lag. The reverse decoupling module 140 simulates the propagation process of perturbation energy in the map, automatically avoiding hard-connected paths with high damping weights and guiding the adjustment flow to soft-connected regions with low damping weights. In this way, the reverse decoupling module 140 can identify time gaps in a low-stiffness state and calculate how to safely insert subsequent processes into these gaps. The reverse decoupling module 140 finally calculates the optimal spatiotemporal displacement parameters for each relevant process. These parameters precisely define the start time, end time, and workspace coordinates of each adjusted process, and send the results to the instruction generation module 150.

[0044] See attached document Figure 1 The instruction generation module 150 establishes a communication connection with the reverse decoupling module 140 and the potential energy analysis module 130, and is configured as the system's output execution and human-machine interface. The main function of the instruction generation module 150 is to transform the abstract numerical results output by the reverse decoupling module 140 into specific operational instructions that can be directly understood and executed by on-site construction personnel and managers. The instruction generation module 150 receives optimized and calculated process spatiotemporal displacement parameters, which include the new start time, new end time, and precise spatial operating range for each adjustment process. Using a built-in coordinate reverse mapping algorithm, the instruction generation module 150 converts the voxel space index coordinates in these parameters back to the engineering axis coordinate system and floor elevation data of the construction site, ensuring that the spatial orientation of the instructions is consistent with the on-site construction language.

[0045] The instruction generation module 150 generates a standardized spatiotemporal operation ticket based on the converted spatiotemporal parameters. This spatiotemporal operation ticket is a digital instruction package containing multi-dimensional work information, including the work process name, work area boundary description, work time window accurate to the minute, required resource configuration list, and description of the current rheological stiffness status of the area. The instruction generation module 150 encodes the spatiotemporal operation ticket into a data format conforming to the engineering management software interface standard and pushes it to the construction management server and on-site handheld mobile terminals. For on-site workers, the instruction generation module 150 displays their assigned work tasks on the mobile terminal in the form of a Gantt chart or a 3D highlighted model, clearly indicating when to enter and where to perform what kind of work, and when to evacuate the area to avoid conflicts with other high-stiffness processes.

[0046] In addition, the instruction generation module 150 is also responsible for generating a dynamic electronic fence based on a safety potential energy field. The instruction generation module 150 reads the overall potential energy field data output by the potential energy analysis module 130 in real time and identifies high-potential energy areas where the potential energy value exceeds a preset safety warning threshold. These areas correspond to spaces where high-risk operations are underway or during the physical solidification period. Based on the edge voxel coordinates of these high-potential energy areas, the instruction generation module 150 generates virtual geofence boundary data. The instruction generation module 150 synchronizes this dynamic electronic fence data to the on-site positioning monitoring system or augmented reality device. When personnel or machinery carrying positioning tags approach the boundary of the dynamic electronic fence, the instruction generation module 150 triggers an early warning signal, prompting relevant personnel to immediately stop entering through on-site broadcasting or terminal vibration, thereby achieving proactive defense against construction conflicts at the physical level.

[0047] The instruction generation module 150 also drives the display terminal to perform four-dimensional visualization rendering. The instruction generation module 150 dynamically binds the adjusted construction schedule to the building information model, simulating the entire construction process after rescheduling on the screen. The instruction generation module 150 uses different color codes to distinguish the rheological stiffness states of different processes; for example, red marks high-stiffness incompressible processes, green marks low-stiffness interleaved processes, and semi-transparent blocks mark temporary restricted areas formed due to the superposition of potential energy fields. Through this intuitive visualization, the instruction generation module 150 assists project managers in verifying the feasibility of the rescheduling scheme and making a final confirmation of potential remaining risks.

[0048] See attached document Figure 2 This invention provides an intelligent management and control method for building construction progress based on digital twins. In step S1, the system performs voxel mesh initialization and attribute mapping operations. This step aims to integrate the continuous geometric space of the building information model with the discrete construction progress timeline to construct a basic data structure for subsequent calculations. The voxel evolution module first determines the boundary of the three-dimensional computational domain based on the global bounding box coordinates of the construction site provided by the data mapping module. The voxel evolution module uses a spatial octree partitioning algorithm or a uniform mesh partitioning algorithm to divide the computational domain into uniformly sized cubic units, i.e., voxels. The side length of the voxels is set according to the accuracy requirements of engineering management, usually set to the length that meets the minimum work surface requirement of the process. The voxel evolution module traverses each component mesh in the building information model and detects the spatial overlap relationship between the component geometry and each voxel through the separating axis theorem or ray casting algorithm. When the geometric mesh of a certain component has a volume intersection with a certain voxel, the system marks the voxel as occupied and writes the globally unique identifier of the component into the attribute list of the voxel, thereby completing the discretization mapping of geometric information.

[0049] After completing the spatial geometric mapping, the voxel evolution module further performs attribute mapping in the time dimension. The module reads the schedule data associated with the components, obtaining the planned start and end times for each component. The system assigns these time parameters to all voxels constituting the component, ensuring that each voxel possesses not only spatial coordinates but also a clear time span attribute. Based on this, the system identifies the physical attribute category of the process to which the component belongs. For processes involving physical transformation, such as cast-in-place concrete construction, the system identifies phase transition processes involving fluid, gel, and solid states; for prefabricated processes, such as steel structure hoisting, the system identifies constraint processes involving free, temporarily fixed, and permanently connected states. Based on these physical attribute categories, the system loads the corresponding rheological stiffness evolution model for each voxel.

[0050] Rheological stiffness is a core physical quantity introduced in this embodiment, used to quantify the resistance of a specific space to external disturbances or schedule adjustments at a specific point in time. A larger rheological stiffness value means a higher degree of locking of the spatiotemporal unit, making rescheduling or spatial replacement costly or even infeasible; a smaller rheological stiffness value means the spatiotemporal unit is in a flexible state, allowing for process overlap or time compression. The voxel evolution module calculates the stiffness change curve of each voxel throughout its entire lifecycle based on a preset rheological stiffness function. This rheological stiffness function uses a piecewise defined mathematical form to accurately describe the nonlinear evolution of the working surface state during construction.

[0051] The system uses the following formula to calculate the rheological stiffness value of any voxel at any time: ; in, Indicates the target voxel at time... The rheological stiffness value; This indicates the planned start time of the process corresponding to this voxel; This indicates the planned end time of the process corresponding to this voxel; It represents the basic environmental stiffness constant before the start of the process, and represents the minimum spatial constraint level when the site is not occupied. It indicates the process stiffness during the execution of a work procedure, representing the exclusivity strength of the space occupied by workers and machinery; This represents a stiffness correction factor related to the type of work process, used to distinguish the stiffness differences between high-risk and ordinary work processes. It represents the final-state ultimate stiffness value of a component after it has been fully formed or solidified, and represents the irreversibility of a physical entity after it has been formed. The time constant representing the material hardening rate or the fastening rate of a connection is used to describe how quickly the stiffness transitions from its process value to its final value.

[0052] Through the above calculations, the system transforms the original static building model into a dynamically changing spatiotemporal voxel field of rheological stiffness. In this field, unconstructed areas appear as a low-stiffness background, areas under construction appear as high-stiffness bright areas, and completed areas appear as stable regions where stiffness gradually increases over time until it is completely solidified. This data structure based on physical evolution provides a precise quantitative benchmark for subsequent identification of construction conflicts and calculation of rescheduling costs, enabling the system to distinguish between soft logical constraints and hard physical constraints, thereby achieving flexible control of the schedule while ensuring structural safety.

[0053] See attached document Figure 2 In step S2, the system performs real-time monitoring and calculation of the multidimensional potential energy field. This step is configured to transform discrete voxel states into continuous risk quantification indicators, thereby providing numerical basis for subsequent path planning. The potential energy analysis module first initiates a periodic global scan process, the time step of which is consistent with the minimum control unit of construction progress management. Within each time step, the potential energy analysis module traverses all voxel indices in the storage unit, reads the occupancy status bit of each voxel at the current moment, and the rheological stiffness value calculated in step S1. Using this basic data, the potential energy analysis module calculates the repulsive potential energy based on physical contact and the environmentally derived potential energy based on spatial proximity effect, and superimposes the two to form the global total potential energy field.

[0054] The potential energy analysis module first calculates the repulsive potential energy, an index used to quantify the direct physical resistance of a physical component to external intrusion. When a voxel is marked as occupied, the physical components or work activities within it constitute a high-energy potential barrier, preventing other processes from entering the space simultaneously. The magnitude of the repulsive potential energy directly depends on the rheological stiffness of the current process in that space; that is, the harder the work surface, the greater the repulsive force, and the higher the risk of collision or damage caused by forcibly entering the area.

[0055] The system uses the following formula to calculate the repulsive potential energy of any voxel at a specific moment: ; in, Indicates spatial index as voxels at all times The repulsive potential energy value of the operation; This represents the global collision penalty coefficient, which is a preset large numerical constant used to ensure that the potential energy value of the physical collision zone is significantly higher than that of free space; This represents the binary occupancy state variable of a voxel. It takes the value 1 when the voxel is occupied by an entity or is in a job-locked state, and otherwise takes the value 0. This indicates that the voxel is at time [time]. The rheological stiffness value is obtained by directly calling the calculation result in step S1. Through this formula, the system can map the high-stiffness concrete pouring area as an insurmountable infinite potential energy wall, and the low-stiffness material stacking area as a finite high potential energy slope that can be climbed by paying a certain cost.

[0056] While calculating the repulsive potential energy, the potential energy analysis module simultaneously calculates the environmentally derived potential energy, an indicator used to quantify the radioactive impact of non-contact hazards on the surrounding space. The system first identifies all active potential energy emission sources within the construction site, including welding points that generate high-temperature spatter, spraying points that produce harmful gases, and areas below hoisting paths where there is a risk of falling objects. For each unoccupied void, the system calculates the intensity of its cumulative impact from all active potential energy emission sources, which decreases with distance.

[0057] The system uses the following formula to calculate the environmentally derived potential energy value of any voxel at a specific moment: ; in, Indicates spatial index as voxels at all times Environmentally derived potential energy value; This represents the total number of potential energy emission sources that are currently active. Indicates the index number of the potential energy emission source; Indicates the first The inherent intensity value of each potential energy emission source is preset according to the type and level of the hazard source; This represents the media propagation attenuation factor, used to describe the diffusion efficiency of hazardous effects in air or other media; Represents the target voxel The central three-dimensional coordinate vector; Indicates the first The three-dimensional coordinate vector of the center of a potential energy emission source; Represents the square of the Euclidean distance between the target voxel and the hazard source; This represents the distance smoothing constant, which is a very small positive number used to prevent mathematical singularities in the calculation results when the distance approaches zero.

[0058] After calculating the above two components, the potential energy analysis module linearly superimposes the work-related repulsive potential energy and the environmentally derived potential energy to generate the final global total potential energy field matrix. Each element in this matrix represents the comprehensive resistance cost of construction work at the corresponding time and space point. The system transmits this total potential energy field matrix to the reverse decoupling module in real time, serving as a navigation map for subsequently finding the minimum energy path. In high potential energy regions, the gradient descent algorithm automatically guides rescheduling paths, while in low potential energy valleys, the system determines them as safe time-space windows suitable for staggered construction.

[0059] See attached document Figure 2 In step S3, the system performs a reverse decoupling and rescheduling operation based on stiffness damping. The reverse decoupling module receives actual progress data from the progress monitoring system in real time and compares it with a preset baseline plan. The system sets a progress lag threshold, which is a specific time difference. When the actual completion time of any critical path operation is detected to be later than the planned completion time and the lag exceeds the progress lag threshold, or when the potential energy value of a local area calculated by the potential energy analysis module exceeds the safety warning line, the system automatically triggers the rescheduling calculation process. The system first constructs a disturbance reverse propagation map, which starts from the lagging operation and searches backward along the logical dependencies for all affected upstream and parallel operation nodes.

[0060] In the perturbation reverse propagation graph, the system calculates the decoupling cost weight for each edge connecting process nodes. This weight represents the physical cost and risk associated with adjusting the logical relationship or reducing the process time. Based on the rheological stiffness data provided by the voxel evolution module and the repulsive potential energy data provided by the potential energy analysis module, the system calculates the weight value of each edge using a stiffness damping function.

[0061] The system uses the following formula to calculate the connection process. With process Decoupling cost weights: ; in, Indicate process With process Decoupling cost weights between them; Indicate process The current duration; and Each represents a process. The start and end times; Indicate process The corresponding voxel set at time The average rheological stiffness value reflects the physical incompressibility of the process itself; Indicate process At the proposed adjustment time The repulsive potential energy value of the target space is reflected in the conflict cost of occupying space resources; Indicate process and The coefficient representing the tightness of the process logic relationship between processes is set to a maximum value for processes that must be strictly serial and a minimum value for processes that can be parallelized. , , These are the normalized weighting coefficients for the stiffness factor, potential energy factor, and logic factor, respectively.

[0062] Based on the calculated decoupling cost weights, the system constructs a global rescheduling objective function. The reverse decoupling module employs heuristic search algorithms or dynamic programming algorithms to find the optimal path in the spatiotemporal solution space, aiming to minimize the total damping energy caused by schedule adjustments. This total damping energy includes not only the time-dimensional rush cost but also the spatial-dimensional conflict risk cost.

[0063] The system uses the following formula to calculate the objective function value of the rescheduling scheme: ; in, This represents the total adjustment cost energy of the rescheduling scheme, and the system seeks a solution that minimizes this value. This represents the set of time translations and spatial displacements for all related processes; This represents the set of all affected process connection edges in the perturbation reverse propagation graph; The decoupling cost weights calculated using the aforementioned formula; Indicate process and The absolute value of the adjustment range of the time interval between them; Indicate process Adjusted to new spatial location Additional environmental potential energy risk value introduced later; This represents the risk penalty coefficient.

[0064] By minimizing the objective function, the system automatically identifies and distinguishes between soft and hard connections at the algorithmic level. For processes with high rheological stiffness, in a solidification phase, or in a high-risk operational state, the algorithm automatically avoids time compression or spatial movement of these processes when searching for the minimum energy path due to their extremely large decoupling cost weights, thus identifying them as untouchable hard connections. Conversely, for processes with low rheological stiffness, in decorative or non-critical paths, their weight values ​​are smaller, and the algorithm tends to absorb schedule lags by adjusting the start and end times of these processes or by using low-potential-energy idle areas for spatial displacement, thus identifying them as flexible soft connections. The system uses low-potential-energy valley regions calculated by the potential energy analysis module as safe time and space windows, inserting the processes to be adjusted into these windows, thereby achieving high-density overlapping construction of multiple processes without violating physical safety constraints.

[0065] See attached document Figure 2 In step S4, the system performs closed-loop control and dynamic command output operations. This step aims to transform the abstract mathematical solution obtained through inverse decoupling calculations into specific executable control signals and operational credentials in the physical world, and to establish a complete closed loop from command issuance to execution feedback. The command generation module first performs coordinate inverse mapping on the optimal spatiotemporal path solution output in step S3. Since the preceding steps are performed in a discretized voxel index space, the command generation module uses a pre-stored spatial mapping matrix to convert the index coordinates of each voxel unit involved in the adjustment back to the axis grid coordinate system and absolute elevation data commonly used in construction sites. The command generation module also converts the discrete step length of the time dimension into a standard calendar time format, accurate to the specific year, month, day, hour, and minute.

[0066] The instruction generation module generates standardized spatiotemporal operation tickets based on the converted data. This spatiotemporal operation ticket is a digital instruction package containing multi-dimensional constraint information. Unlike traditional two-dimensional construction schedules, it defines a precise four-dimensional spatiotemporal bounding box for each work team. The data structure of the spatiotemporal operation ticket includes a unique identifier for the work unit, the absolute start time for permitted entry, the absolute end time for mandatory evacuation, and the set of vertex coordinates of the three-dimensional spatial work boundary. The instruction generation module explicitly marks the rheological stiffness state of the surrounding work area in the spatiotemporal operation ticket, marking adjacent high rheological stiffness areas as untouchable hard constraint boundaries and low rheological stiffness areas as flexible buffer boundaries. The system distributes the generated spatiotemporal operation tickets to the handheld terminals of construction site managers and the vehicle-mounted control terminals of construction machinery via a wireless communication network. The permitted work area is rendered in the form of a three-dimensional visualization model on the terminal screen, intuitively instructing construction personnel on when, where, and what operations to perform.

[0067] In parallel, the instruction generation module generates a dynamic electronic fence based on real-time global potential energy field data output by the potential energy analysis module. The system sets a potential energy safety threshold, representing the maximum risk exposure level that a human or device can withstand. The instruction generation module traverses the global voxel field, filtering out all voxel units whose total potential energy exceeds the safety threshold. The system extracts the spatial outer edge coordinates of these high-potential-energy voxel units and generates closed or semi-closed virtual boundary lines using a geometric envelope algorithm. These virtual boundary lines constitute the geometric description data of the dynamic electronic fence, and their extent is dynamically updated in real-time according to changes in the rheological stiffness field at the construction site and the movement of hazard sources. For example, as the concrete pouring process progresses, the high-potential-energy area moves accordingly, and the dynamic electronic fence also shifts in the digital model.

[0068] The system synchronizes dynamic electronic fence data to the on-site IoT positioning system. Smart safety helmets worn by on-site workers and positioning tags installed on construction machinery report their location coordinates to the system in real time. The system compares the relative positions of these coordinates with the dynamic electronic fence in real time on the backend server. When the system detects that the location coordinates of a person or piece of equipment are about to cross the boundary of the dynamic electronic fence and enter a high-potential-energy area, it immediately triggers a tiered early warning mechanism. The system sends vibration or audible and visual alarm signals to the corresponding on-site terminals, forcibly requiring the relevant objects to stop moving or change their path. Simultaneously, the system monitors the deviation between the actual construction progress and the issued time-space operation ticket. If the actual working time exceeds the time window specified in the operation ticket, or the actual working space crosses the specified voxel boundary, the system will determine this as a new progress disturbance or safety violation event and feed this status back to the data mapping module, thereby triggering a new round of voxel evolution and potential energy analysis, achieving full-cycle adaptive closed-loop control of the construction process.

[0069] To verify the practical application effect of the technical solution of this invention under complex working conditions, this embodiment constructs a construction scenario of a high-rise building core tube. In this scenario, the construction area is set as the fifteenth-floor core tube area, involving two key processes with a spatiotemporal interrelationship: the core tube shear wall concrete pouring process and the core tube internal electromechanical pipeline support installation process. In the initial baseline schedule, considering the narrow construction space and operational safety, the traditional management model sets these two processes with a strict end-start logic relationship, that is, requiring the electromechanical installation team to enter the site for construction only after the shear wall concrete has been poured and completed its initial setting. Under the interference conditions set in this embodiment, it is assumed that the concrete pumping equipment experiences a mechanical failure, causing the actual start time of the shear wall concrete pouring process to be delayed by four hours compared to the planned time.

[0070] In the traditional Critical Path Method (CPM) management model, because the installation of electromechanical pipeline supports is the immediate successor to the shear wall concrete pouring process and is located on the critical path, the system will directly pass on the four-hour delay to all subsequent processes, resulting in a corresponding four-hour extension of the overall project duration. The traditional CPM method treats the fifteenth-floor core as a single resource node, failing to identify subtle state differences within that space or assess the differences in the physical properties of the processes themselves. Therefore, it can only adopt a passive delay strategy.

[0071] The adaptive construction progress control system proposed in this invention generates distinctly different handling schemes through the collaborative work of various modules. The data mapping module 110 captures the real-time shutdown status and estimated maintenance time of the concrete pumping equipment and inputs this progress lag information into the system. The voxel evolution module 120 immediately recalculates the spatiotemporal voxel properties of the area. The voxel evolution module 120 identifies that the concrete pouring process has a high rheological stiffness value, belonging to an incompressible and strongly exclusive hard constraint process, while the electromechanical pipeline support installation process involves lightweight metal components with a low rheological stiffness value, belonging to a soft constraint process with high elasticity in both space and time.

[0072] The potential energy analysis module 130 further calculated that although the shear wall of the core tube was in a high potential energy warning state due to the waiting for concrete pouring, the corridor area and elevator lobby area inside the core tube were not currently undergoing high-risk operations, and their operational repulsion potential energy and environmentally derived potential energy were both at low, safe levels. The reverse decoupling module 140 constructed a disturbance reverse transmission map based on the above data and performed a minimum energy path search. The algorithm calculation results showed that although the shear wall pouring process must be executed according to the original logical sequence, some prefabrication and assembly work in the electromechanical pipeline support installation process could be spatially decoupled. The reverse decoupling module 140 found that temporarily adjusting the work location of the electromechanical installation team to the low-potential-energy elevator lobby area and performing ground assembly and rust prevention treatment of the supports in advance resulted in a lower decoupling cost weight.

[0073] Based on this calculation, the instruction generation module 150 generated a new spatiotemporal operation ticket. This instruction did not postpone the start time of the electromechanical installation process; instead, it instructed the electromechanical installation team to immediately enter the 15th-floor elevator lobby area (low-potential voxel area) to perform pre-assembly of the supports during the concrete equipment maintenance period. Simultaneously, the instruction generation module 150 generated a dynamic electronic fence around the shear wall to prevent installation personnel from accidentally entering the high-risk waiting area. Through this interleaved construction scheme based on rheological stiffness differences, the system successfully utilized the downtime during equipment maintenance to complete the preparatory work for subsequent processes. Once the concrete pouring was completed and reached its initial setting strength, the electromechanical installation team only needed to move the pre-assembled supports to the wall position for rapid fixing. The final implementation results showed that, through the system's scheduling, the original four-hour delay was completely absorbed, and the actual start time of subsequent processes remained consistent with the original baseline plan, achieving automatic schedule correction without additional resource input.

Claims

1. A digital twin-based intelligent control system for building construction progress, characterized in that, include: The data mapping module (110) is used to receive building information model and on-site measured data, and process them to generate standardized data containing geometric and state information under a unified coordinate system; The voxel evolution module (120) is used to divide the construction site into continuous voxels based on the standardized data, construct the rheological stiffness evolution function, and calculate the rheological stiffness that characterizes the spatial physical state and resists scheduling adjustment resistance. The potential energy analysis module (130) is used to combine the state information and the rheological stiffness to calculate the operation repulsion potential energy, and superimpose the environmental derived potential energy to generate a global potential energy field characterizing the on-site safety constraints. The reverse decoupling module (140) is used to construct a disturbance reverse transmission map based on the construction logic when a construction progress delay is detected, and to use the rheological stiffness as a damping weight to search for the reverse decoupling path with the minimum adjustment energy under the constraints of the global situational energy field, and to calculate the optimal spatiotemporal displacement parameters of the process to be adjusted. The instruction generation module (150) is used to generate spatiotemporal operation instructions based on the optimal spatiotemporal displacement parameters, and to generate dynamic electronic fence boundary signals based on the global situational energy field.

2. The intelligent construction progress control system based on digital twin as described in claim 1, characterized in that, The coordinate transformation process performed by the data mapping module (110) when generating the standardized data includes: The rigid body transformation matrix is ​​calculated using the coordinates of preset control points, and the coordinates of the field measured data are mapped to the unified coordinate system using the rigid body transformation matrix. The data mapping module (110) is also used to calculate the axis-aligned bounding box of the construction scene, define the global spatial domain range of the voxel generation, and perform manifold checks and normal repairs on the geometric topology of the components in the building information model.

3. The intelligent construction progress control system based on digital twin as described in claim 1, characterized in that, The voxel evolution module (120) is used to establish a multidimensional attribute vector for each voxel, the multidimensional attribute vector including: An occupancy flag indicating whether the voxel is occupied by a physical component at a given time, the rheological stiffness, an environmentally derived potential energy value for storing the intensity of non-contact influences derived from surrounding operations, and a resource carrying capacity threshold for defining the maximum physical load that the space can withstand.

4. The intelligent construction progress control system based on digital twin as described in claim 1, characterized in that, The voxel evolution module (120) generates the rheological stiffness evolution function in the following ways: For processes with physical phase change characteristics, the rheological stiffness is set to change in stages: it is set as a baseline value in the work preparation stage, increased to the work lock value in the work process stage, and after the work is completed, the rheological stiffness is controlled to increase exponentially with time according to the principle of material physical hardening until it reaches the maximum extreme value. For processes that do not have physical phase change characteristics but are spatially exclusive, a constant high rheological stiffness value is assigned during the operation. During the operation interval or after completion and when the space is not occupied, the rheological stiffness value is reset to zero or the environmental background value.

5. The intelligent construction progress control system based on digital twin as described in claim 1, characterized in that, The potential energy analysis module (130) calculates the repulsive potential energy of the operation in the following ways: Traverse the voxels and calculate the repulsive potential energy intensity based on the rheological stiffness of the voxels at the current time step and the occupancy state in the state information. When the occupancy status is displayed as occupied and the rheological stiffness is at a high level, a repulsive potential energy peak is generated at the voxel position, and a decaying potential energy gradient band is generated in the voxel neighborhood of the repulsive potential energy peak.

6. The intelligent construction progress control system based on digital twin as described in claim 1, characterized in that, The potential energy analysis module (130) calculates the environmentally derived potential energy in the following ways: Identify work sources with environmental impact attributes in the construction site and define them as potential energy emission sources. Calculate the propagation and attenuation distribution of the potential energy generated by the potential energy emission sources into the surrounding space based on a preset physical diffusion model. The potential energy emission sources include welding work points that generate high-temperature sparks, coating work surfaces that emit volatile gases, and mechanical equipment locations that generate high-frequency vibrations.

7. The intelligent construction progress control system based on digital twin as described in claim 6, characterized in that, The potential energy analysis module (130) generates the global potential energy field in the following ways: Perform multi-field coupling operations to linearly superimpose the job repulsion potential energy with the environment-derived potential energy; During the superposition process, if the repulsive potential energy of the operation in a certain area is zero but the total potential energy value after superposition exceeds the preset safety threshold, the area will be marked as a high potential energy hazard area.

8. The intelligent construction progress control system based on digital twin as described in claim 1, characterized in that, The reverse decoupling module (140) constructs the perturbation reverse propagation map in the following ways: Starting from the source node where the progress disturbance occurs, a breadth-first search is performed along the reverse direction and lateral spatial direction of the process logic connection line to identify all process nodes that are logically or spatially related. The nodes are connected by directed edges, which represent the process precedence relationship, resource sharing relationship and spatial adjacency relationship between processes.

9. A construction progress intelligent control system based on digital twins according to claim 8, characterized in that, The reverse decoupling module (140) utilizes the rheological stiffness as a damping weight and searches for the reverse decoupling path with the minimum adjustment energy in the following ways: Calculate the decoupling cost weight of each edge in the disturbance reverse propagation graph, wherein the decoupling cost weight is positively correlated with the rheological stiffness of the corresponding process; The minimum energy path search algorithm is executed to find the process adjustment combination that minimizes the sum of the total decoupling cost weights required to eliminate the lag in the progress. Specifically, processes in a high rheological stiffness state are assigned high damping weights, while processes in a low rheological stiffness state are assigned low damping weights, guiding the adjustment path towards the region with the low damping weight.

10. A construction progress intelligent control system based on digital twins according to claim 1, characterized in that, The spatiotemporal operation instruction generated by the instruction generation module (150) is a spatiotemporal operation ticket, which includes the process name, work area boundary description, work time window, required resource configuration list and regional rheological stiffness state description. The dynamic electronic fence boundary signal generated by the instruction generation module (150) is virtual geographic fence boundary data generated based on the voxel coordinates of the edge of the high potential energy area in the global potential energy field where the potential energy value exceeds the preset safety warning threshold.

Images