Engineering pay-off process collaborative management system based on digital twinning

Abstract

The invention discloses an engineering pay-off process collaborative management system based on digital twinning, and the system comprises a scene construction module which is used for obtaining engineering information data, and generating a digital twinning three-dimensional engineering scene; the environment updating module is used for completing data preprocessing to obtain an updated three-dimensional engineering scene; the path simulation module is used for generating a target pay-off path and performing simulation deduction to obtain a simulation execution track; the execution control module is used for generating a pay-off execution instruction, performing track processing and generating a real pay-off path; the deviation re-planning module is used for performing deviation analysis on the real pay-off path to generate a local adjustment result; the collaborative scheduling module is used for executing Max-Sum distributed collaborative scheduling on the regional layering factor graph and generating a multi-main-body collaborative scheduling instruction; and the data link management module is used for recording a real pay-off path and a multi-main-body cooperative scheduling instruction to form an engineering pay-off data link. According to the invention, collaborative management of the engineering pay-off process is realized.

Description

Technical Field

[0001] This invention relates to the field of engineering management technology, and in particular to a collaborative management system for engineering layout process based on digital twins. Background Technology

[0002] Currently, most on-site layout work still relies on surveying equipment such as total stations and laser rangefinders in conjunction with manual operation. Some projects have introduced basic 3D modeling or BIM technology to display design drawings in 3D space to assist in layout positioning. However, these technologies are mostly limited to graphical visualization and single-point measurement annotation. There is a lack of unified structured management and integrated coordinate expression between engineering information data, component attribute data, and on-site measurement data. This makes it difficult to form a continuously updatable digital twin 3D engineering scene, resulting in the layout preparation work relying on manual integration of multi-source data, which is inefficient and prone to errors.

[0003] With the emergence of equipment such as intelligent line-laying robots, some existing technologies attempt to combine line-laying path planning with robot control. By using simple path generation algorithms and preset construction area information, a single line-laying robot can be driven to complete automatic line-laying operations. However, existing solutions often assume that the construction environment is relatively static, and the collection and processing of on-site environmental data are relatively crude. They are mostly corrected through local manual intervention or simple safety distance rules, and cannot make full use of digital twin models for refined control.

[0004] In scenarios involving multiple robots and multiple processes in parallel construction, existing collaborative management technologies for line laying mainly rely on human experience or simple rule scheduling. They do not adequately consider the mapping relationship between line laying tasks and material resources. The task allocation methods are mostly static planning or first-come-first-served strategies, lacking quantitative evaluation of the matching benefits between tasks and resources. This can easily lead to problems such as local congestion, uneven resource utilization, and interference between robots.

[0005] Therefore, how to provide a collaborative management system for engineering layout processes based on digital twins is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] One objective of this invention is to propose a collaborative management system for engineering layout processes based on digital twins. This invention achieves dynamic optimization allocation of tasks and material resources by introducing a local replanning mechanism and a Max-Sum distributed collaborative scheduling algorithm based on factor graphs, thereby improving the accuracy of layout operations and the efficiency of construction collaboration.

[0007] According to an embodiment of the present invention, a collaborative management system for engineering layout process based on digital twins includes:

[0008] The scene construction module is used to acquire engineering information data, perform structured encoding and coordinate unification, and generate a digital twin 3D engineering scene.

[0009] The environment update module is used to collect environmental status data, complete data preprocessing, and obtain the updated 3D engineering scene;

[0010] The path simulation module is used to generate the target layout path in the updated 3D engineering scene according to the on-site constraints and layout requirements, and to perform simulation and deduction to obtain the simulation execution trajectory.

[0011] The execution control module is used to generate wire laying execution instructions based on the simulated execution trajectory and issue them to the wire laying robot execution unit to perform trajectory processing and attitude correction, and generate the actual wire laying path.

[0012] The deviation replanning module is used to analyze the deviation between the actual surveying path and the target surveying path, and to perform local replanning to generate local adjustment results for dynamic correction.

[0013] The collaborative scheduling module is used to determine the candidate relationships for task allocation through a market bidding mechanism, execute Max-Sum distributed collaborative scheduling on the regional hierarchical factor graph, and generate multi-entity collaborative scheduling instructions.

[0014] The data chain management module is used to record the actual layout path and multi-entity collaborative scheduling instructions, forming an engineering layout data chain for quality acceptance and construction traceability.

[0015] Optionally, modules can be integrated using the following methods:

[0016] Acquire engineering information data, perform structuring, coordinate unification, and scene construction to form a digital twin three-dimensional engineering scene;

[0017] Based on the 3D engineering scene, environmental status data is collected, and data preprocessing including time synchronization, spatial registration and data cleaning is performed to obtain the updated 3D engineering scene.

[0018] In the updated 3D engineering scene, a target layout path is generated based on the on-site constraints, and the target layout path is simulated and deduced to obtain the simulated execution trajectory.

[0019] The wire laying execution command is generated based on the simulated execution trajectory and issued to the wire laying robot execution unit;

[0020] During the laying process, the position information and execution status of the laying robot's execution unit are collected in real time, and trajectory processing and attitude correction are performed to generate a real laying path and synchronize it to the updated 3D engineering scene.

[0021] The deviation between the actual and target survey paths is analyzed. When the deviation exceeds the preset range, a conflict-aware local replanning mechanism is triggered to locally adjust the survey execution command for the corresponding area, so as to achieve dynamic correction of the survey process and obtain the local adjustment result.

[0022] Based on the updated 3D engineering scene and local adjustment results, the layout task and material resources are mapped to the corresponding area. The candidate relationship of task allocation is determined through market bidding. The Max-Sum distributed collaborative scheduling is executed on the factor graph with the candidate relationship as a constraint to obtain multi-entity collaborative scheduling instructions.

[0023] The actual layout paths and multi-entity collaborative scheduling instructions generated during the layout process are recorded and archived to form an engineering layout data chain for quality acceptance and construction traceability.

[0024] Optionally, the formation of the digital twin's three-dimensional engineering scene specifically includes:

[0025] The process involves acquiring engineering information data, reading and importing design drawings, component attributes, and on-site measurements, classifying and organizing data from different sources according to component category, floor location, and project number, and screening for missing fields, duplicate records, and abnormal data to form a raw dataset of engineering information that can be used for structured processing.

[0026] The original dataset of engineering information is structured and encoded according to a unified field format. The coordinate and elevation references used by different drawings and measurement systems are uniformly converted so that all component information is expressed under a unified engineering coordinate system and an engineering information data table with a unified expression format is formed.

[0027] Based on the engineering information data table, the three-dimensional geometric shape of the component is generated according to the planar coordinates and elevation coordinates of the component. The connection, intersection and subordination relationships between components are constructed to form a digital twin three-dimensional engineering scene containing the geometric units of the components and the spatial topological relationships.

[0028] Optionally, obtaining the updated 3D engineering scene specifically includes:

[0029] Environmental status data is collected under a unified coordinate system in a three-dimensional engineering scene. The unified coordinate system includes planar coordinates and elevation coordinates. Environmental status data from different field sensing devices are recorded with time information according to the time of collection and are classified according to device type, collection area and time order. Data that cannot be identified, has missing key fields or abnormal format is removed. Each valid environmental status data has a clear collection time description and spatial location description, forming a set of original environmental status data with time and space markers.

[0030] The raw environmental condition data set is processed in a unified manner, and the spatial location description is converted into a spatial coordinate method consistent with the 3D engineering scene. Environmental condition data belonging to the same time period and in the same spatial area are sampled and integrated. Data with location jumps, time jumps or abnormal collection intervals are corrected or removed to form a preprocessed environmental condition data set that has undergone time synchronization, spatial alignment and abnormal data processing.

[0031] The environmental state preprocessing data set is matched with the engineering information and spatial location in the 3D engineering scene. According to the time and spatial description of each environmental state data, it is written into the state field of the corresponding scene node. The relevant state parameters in the 3D engineering scene are updated so that the scene reflects the real environmental state of the current construction area on the basis of the original geometric structure, and the updated 3D engineering scene is obtained.

[0032] Optionally, obtaining the simulation execution trajectory specifically includes:

[0033] In the updated 3D engineering scene, the construction area information used for line laying is read, the spatial position of components, the outer boundary of components and the location of obstacles in the construction area are extracted and marked, and the space range that the line laying robot can pass through is delineated in the 3D coordinate system. The inaccessible component area, obstacle area and safety isolation area are removed from the passage space to obtain the path planning space composed of multiple passable space units.

[0034] In the path planning space, based on the site constraints and specified layout requirements, the spatial locations of the starting and ending points of the layout are determined. Target locations within the passable space are selected sequentially according to the layout order. Each target location is associated with its corresponding spatial location and adjacent component relationships in the 3D engineering scene. The component relationships include spatial topological relationships, geometric relationships, and engineering attribute relationships between components. Spatial topological relationships include adjacent, connected, intersecting, and subordinate relationships between components. Geometric relationships include the relationship between the component's external boundary, occupied space, and passable space. Engineering attribute relationships include the component's floor affiliation, area affiliation, and system subordinate relationships. The target location points are combined into a continuous layout route according to the layout progress order, generating path data to describe the target layout path.

[0035] In the updated 3D engineering scene, the movement process of the wire-laying robot is simulated sequentially according to the path data. The arrival position and motion state of each target location are simulated, and the position and posture changes of the wire-laying robot during the simulation process are recorded to form a simulation execution trajectory arranged in chronological order.

[0036] Optionally, the process of generating the line-laying execution instruction specifically includes:

[0037] The simulation execution trajectory is read, and the trajectory points recorded during the simulation are reorganized in chronological order. The spatial position description and direction of motion of each trajectory point are analyzed, and a trajectory point sequence is generated based on the continuity and chronological relationship of the trajectory points, which serves as the input data set required for generating the line laying execution command.

[0038] Based on the trajectory point sequence, the spatial position description of each trajectory point is converted into the control coordinate system adopted by the wire-laying robot execution unit. The posture and motion direction of the trajectory points are encoded. The position data, posture parameters and motion direction parameters are combined according to the control field format of the execution unit to form a set of control parameters that can be directly used to control the behavior of the wire-laying robot execution unit.

[0039] The control parameter set is assembled according to the motion sequence of the trajectory points. An execution sequence identifier, execution condition field, and running time field are added to each set of control parameters. The assembled control parameters are integrated to generate a complete line laying execution instruction. The line laying execution instruction is written into the robot control communication structure and issued through the instruction interface of the execution unit, so that the line laying robot can perform the line laying operation according to the instruction.

[0040] Optionally, the generation of the actual laying path specifically includes:

[0041] During the line laying process, the operation data of the line laying robot execution unit is collected in real time. Each collection record is assigned a time description according to the collection time, and the corresponding spatial location and execution status parameters are recorded. At the same time, each record is labeled with the construction area and corresponding component domain identifier during the collection process. Records with missing fields, contradictory status or abnormal format are removed and re-collected to form a set of original operation data records with time description, component domain identifier and spatial location description.

[0042] Based on the original set of running data records, the spatial locations in the continuously collected records are combined into a trajectory point sequence in chronological order. Component domain constraints are applied to the trajectory point sequence according to the component domain identifier. The spatial location of each trajectory point is mapped to the traversable path domain of the corresponding component. Attitude fusion processing is performed on the trajectory point attitude based on the fusion of component surface normal vector, path tangential vector and device attitude vector. Digital twin inverse projection correction is performed on trajectory points that have position jumps, attitude reversals or inconsistent states, so that the trajectory point sequence satisfies the component domain constraints in spatial location and the path travel constraints in attitude direction, resulting in a real trajectory data sequence containing attitude fusion information and position correction information.

[0043] The real trajectory data sequence is reconstructed according to the time sequence of the laying out, generating a real laying out path with continuous trajectory segments, fused posture description and component domain correspondence. The spatial position of each trajectory point in the real laying out path is converted into a unified coordinate system of the updated 3D engineering scene. The component nodes, spatial region nodes and status fields to which the trajectory points belong are updated synchronously, so that the updated 3D engineering scene can fully express the actual movement path, posture changes and area coverage of the laying out robot on the construction site, thereby generating a real laying out path and completing real-time synchronization with the 3D engineering scene.

[0044] Optionally, obtaining the local adjustment result specifically includes:

[0045] The actual laying path and the trajectory points in the target laying path are paired point by point in time sequence. The spatial position description and attitude description of each trajectory point are extracted. The position deviation value composed of position deviation and attitude deviation is calculated. The position deviation value is used to describe the spatial distance difference between trajectory points, and the attitude deviation value is used to describe the angle difference between the attitude directions of trajectory points. All position deviation values ​​are combined into a deviation sequence according to the trajectory sequence for deviation identification processing.

[0046] The positional deviation value in the deviation sequence is compared with the preset deviation threshold. The spatial region corresponding to the trajectory point where the deviation exceeds the limit is located. The component distribution, obstacle layout and passable space in the current region are read. Regions with component encroachment or path failure are marked as conflict regions to form conflict identification results.

[0047] In the conflict area, local replanning is performed, the component constraints and passable space units in the conflict area are loaded, alternative location points are selected and local path segments are constructed, the connection relationship between the local path segments and the original path segments is corrected, and the attitude description of the local path segments is adjusted to meet the laying direction constraints and component spatial relationship constraints, thus forming a local adjusted path.

[0048] The local adjustment path is converted into a local adjustment instruction. The spatial position, attitude and execution order are encoded by control fields. The corresponding fields in the original wire laying execution instruction are replaced to generate a local adjustment result for driving the wire laying robot execution unit to perform dynamic correction actions.

[0049] Optionally, obtaining the multi-entity collaborative scheduling instruction specifically includes:

[0050] In the updated 3D engineering scene, the construction area involved in the local adjustment results is analyzed, and a mapping relationship is established between each layout task and the area and available material resources, forming a task and material resource area mapping set indexed by the construction area.

[0051] Based on the mapping set of tasks and material resources, the utility of each line laying task and the candidate material resources in its area is evaluated. A hybrid scheduling strategy combining market bidding mechanism is adopted, which comprehensively considers the matching degree between tasks and material resources, spatial distance, current load level of the area and expected load level. The matching degree is recorded as a score, the spatial distance is recorded as a distance, and the deviation between the current load and the expected load of the area is included in the load penalty in a weighted manner. The score, distance and load penalty are weighted and synthesized by pre-set weight parameters and attenuation parameters to obtain the bidding utility value between each task and the candidate material resources. The candidate relationships are sorted and filtered according to the size of the utility value to form a candidate relationship set for task allocation.

[0052] Using the candidate relationship set of task allocation as constraints, a regional hierarchical factor graph structure is constructed based on the division of construction areas. In the hierarchical factor graph structure, the material resource allocation of each layout task in each construction area is represented as an allocation variable. The task demand constraints, material resource capacity constraints, and regional construction restrictions are represented as task factors, resource factors, and regional factors, respectively. The results of local adjustments are written into the corresponding regional factors in the form of additional constraints. This allows the regional hierarchical factor graph structure to simultaneously reflect the candidate matching relationship between tasks and material resources, the resource and construction constraints of each construction area, and the path adjustment constraints brought about by the aforementioned local replanning.

[0053] Max-Sum distributed collaborative scheduling is performed on the hierarchical factor graph structure. The messages sent by each factor node to its connected allocation variable nodes are iteratively updated. During each update, the bidding utility value corresponding to the candidate relationship between the current task and material resources is combined with the candidate relationship, local adjustment constraints, and hierarchical relationship of the hierarchical factor graph structure as the comprehensive cost. The comprehensive cost of different task allocation combinations is compared and selected. By repeatedly passing and updating messages between factor nodes and variable nodes until the values ​​of each allocation variable converge, the optimal allocation combination of task and material resources that satisfies the overall utility maximization and regional constraints is obtained. This is then converted into multi-subject collaborative scheduling instructions for different wire-laying robot execution units and material supply entities.

[0054] Optionally, the formation of the engineering layout data link specifically includes:

[0055] The spatial location, attitude, and motion state descriptions of the actual cable laying path are continuously collected. The collected path data are organized according to the time description, each record is written into a path record entry in a unified coordinate format, and all record entries are organized in chronological order to form a time-series path record set of the actual cable laying path.

[0056] The optimal allocation combination generated by multi-entity collaborative scheduling is read and written into the scheduling record entry. The message update content from the factor node to the variable node during the scheduling iteration is also recorded. The scheduling record entry and message update content are combined to form a multi-entity collaborative scheduling instruction record set.

[0057] The time-series path record set of the actual layout path is associated with the multi-subject collaborative scheduling instruction record set according to the task number and time description quantity. The spatial location description in the path record is compared with the allocation description in the scheduling record, and archived according to the execution batch number to form an engineering layout data chain for quality acceptance and construction traceability.

[0058] The beneficial effects of this invention are:

[0059] This invention enables unified management of engineering information data and on-site environmental status data, allowing the layout process to be carried out in a real, continuous, and synchronous virtual-real mapping environment, fundamentally reducing modeling errors and path deviations caused by fragmented multi-source data.

[0060] This invention establishes a complete closed-loop link between path planning, simulation and execution control. Through real-time acquisition of the actual laying path, attitude fusion processing and digital twin inverse projection correction, the laying accuracy and execution stability are greatly improved.

[0061] The local replanning mechanism based on deviation analysis and conflict perception proposed in this invention can generate local adjustment paths in a timely manner when path failure, component encroachment, or abnormal posture occurs, thereby reducing manual intervention and improving construction continuity.

[0062] This invention obtains candidate relationships between tasks and material resources through a market bidding mechanism, and runs the Max-Sum distributed collaborative scheduling algorithm based on a regional hierarchical factor graph, thereby achieving efficient collaboration among multiple robots, multiple tasks, and multiple resources. Attached Figure Description

[0063] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0064] Figure 1 This is a flowchart of a collaborative management system for engineering layout process based on digital twins proposed in this invention;

[0065] Figure 2 This is a schematic diagram of the algorithm structure of a collaborative management system for engineering layout process based on digital twin proposed in this invention. Detailed Implementation

[0066] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0067] refer to Figure 1-2 A collaborative management system for engineering layout processes based on digital twins, comprising:

[0068] The scene construction module is used to acquire engineering information data, perform structured encoding and coordinate unification, and generate a digital twin 3D engineering scene.

[0069] The environment update module is used to collect environmental status data, complete data preprocessing, and obtain the updated 3D engineering scene;

[0070] The path simulation module is used to generate the target layout path in the updated 3D engineering scene according to the on-site constraints and layout requirements, and to perform simulation and deduction to obtain the simulation execution trajectory.

[0071] The execution control module is used to generate wire laying execution instructions based on the simulated execution trajectory and issue them to the wire laying robot execution unit to perform trajectory processing and attitude correction, and generate the actual wire laying path.

[0072] The deviation replanning module is used to analyze the deviation between the actual surveying path and the target surveying path, and to perform local replanning to generate local adjustment results for dynamic correction.

[0073] The collaborative scheduling module is used to determine the candidate relationships for task allocation through a market bidding mechanism, execute Max-Sum distributed collaborative scheduling on the regional hierarchical factor graph, and generate multi-entity collaborative scheduling instructions.

[0074] The data chain management module is used to record the actual layout path and multi-entity collaborative scheduling instructions, forming an engineering layout data chain for quality acceptance and construction traceability.

[0075] In this embodiment, the modules are interconnected using the following method:

[0076] Acquire engineering information data, perform structuring, coordinate unification, and scene construction to form a digital twin three-dimensional engineering scene;

[0077] Based on the 3D engineering scene, environmental status data is collected, and data preprocessing including time synchronization, spatial registration and data cleaning is performed to obtain the updated 3D engineering scene.

[0078] In the updated 3D engineering scene, a target layout path is generated based on the on-site constraints, and the target layout path is simulated and deduced to obtain the simulated execution trajectory.

[0079] The wire laying execution command is generated based on the simulated execution trajectory and issued to the wire laying robot execution unit;

[0080] During the laying process, the position information and execution status of the laying robot's execution unit are collected in real time, and trajectory processing and attitude correction are performed to generate a real laying path and synchronize it to the updated 3D engineering scene.

[0081] The deviation between the actual and target survey paths is analyzed. When the deviation exceeds the preset range, a conflict-aware local replanning mechanism is triggered to locally adjust the survey execution command for the corresponding area, so as to achieve dynamic correction of the survey process and obtain the local adjustment result.

[0082] Based on the updated 3D engineering scene and local adjustment results, the layout task and material resources are mapped to the corresponding area. The candidate relationship of task allocation is determined through market bidding. The Max-Sum distributed collaborative scheduling is executed on the factor graph with the candidate relationship as a constraint to obtain multi-entity collaborative scheduling instructions.

[0083] The actual layout paths and multi-entity collaborative scheduling instructions generated during the layout process are recorded and archived to form an engineering layout data chain for quality acceptance and construction traceability.

[0084] In this embodiment, the formation of the digital twin's three-dimensional engineering scene specifically includes:

[0085] The process involves acquiring engineering information data, reading and importing design drawings, component attributes, and on-site measurements, classifying and organizing data from different sources according to component category, floor location, and project number, and screening for missing fields, duplicate records, and abnormal data to form a raw dataset of engineering information that can be used for structured processing.

[0086] The original dataset of engineering information is structured and encoded according to a unified field format. The coordinate and elevation references used by different drawings and measurement systems are uniformly converted so that all component information is expressed under a unified engineering coordinate system and an engineering information data table with a unified expression format is formed.

[0087] Based on the engineering information data table, the three-dimensional geometric shape of the component is generated according to the planar coordinates and elevation coordinates of the component. The connection, intersection and subordination relationships between components are constructed to form a digital twin three-dimensional engineering scene containing the geometric units of the components and the spatial topological relationships.

[0088] In this embodiment, obtaining the updated 3D engineering scene specifically includes:

[0089] Environmental status data is collected under a unified coordinate system in a three-dimensional engineering scene. The unified coordinate system includes planar coordinates and elevation coordinates. Environmental status data from different field sensing devices are recorded with time information according to the time of collection and are classified according to device type, collection area and time order. Data that cannot be identified, has missing key fields or abnormal format is removed. Each valid environmental status data has a clear collection time description and spatial location description, forming a set of original environmental status data with time and space markers.

[0090] The raw environmental condition data set is processed in a unified manner, and the spatial location description is converted into a spatial coordinate method consistent with the 3D engineering scene. Environmental condition data belonging to the same time period and in the same spatial area are sampled and integrated. Data with location jumps, time jumps or abnormal collection intervals are corrected or removed to form a preprocessed environmental condition data set that has undergone time synchronization, spatial alignment and abnormal data processing.

[0091] The environmental state preprocessing data set is matched with the engineering information and spatial location in the 3D engineering scene. According to the time and spatial description of each environmental state data, it is written into the state field of the corresponding scene node. The relevant state parameters in the 3D engineering scene are updated so that the scene reflects the real environmental state of the current construction area on the basis of the original geometric structure, and the updated 3D engineering scene is obtained.

[0092] In this embodiment, obtaining the simulation execution trajectory specifically includes:

[0093] In the updated 3D engineering scene, the construction area information used for line laying is read, the spatial position of components, the outer boundary of components and the location of obstacles in the construction area are extracted and marked, and the space range that the line laying robot can pass through is delineated in the 3D coordinate system. The inaccessible component area, obstacle area and safety isolation area are removed from the passage space to obtain the path planning space composed of multiple passable space units.

[0094] In the path planning space, based on the site constraints and specified layout requirements, the spatial locations of the starting and ending points of the layout are determined. Target locations within the passable space are selected sequentially according to the layout order. Each target location is associated with its corresponding spatial location and adjacent component relationships in the 3D engineering scene. The component relationships include spatial topological relationships, geometric relationships, and engineering attribute relationships between components. Spatial topological relationships include adjacent, connected, intersecting, and subordinate relationships between components. Geometric relationships include the relationship between the component's external boundary, occupied space, and passable space. Engineering attribute relationships include the component's floor affiliation, area affiliation, and system subordinate relationships. The target location points are combined into a continuous layout route according to the layout progress order, generating path data to describe the target layout path.

[0095] In the updated 3D engineering scene, the movement process of the wire-laying robot is simulated sequentially according to the path data. The arrival position and motion state of each target location are simulated, and the position and posture changes of the wire-laying robot during the simulation process are recorded to form a simulation execution trajectory arranged in chronological order.

[0096] In this embodiment, the process of generating the line-laying execution instruction specifically includes:

[0097] The simulation execution trajectory is read, and the trajectory points recorded during the simulation are reorganized in chronological order. The spatial position description and direction of motion of each trajectory point are analyzed, and a trajectory point sequence is generated based on the continuity and chronological relationship of the trajectory points, which serves as the input data set required for generating the line laying execution command.

[0098] Based on the trajectory point sequence, the spatial position description of each trajectory point is converted into the control coordinate system adopted by the wire-laying robot execution unit. The posture and motion direction of the trajectory points are encoded. The position data, posture parameters and motion direction parameters are combined according to the control field format of the execution unit to form a set of control parameters that can be directly used to control the behavior of the wire-laying robot execution unit.

[0099] The control parameter set is assembled according to the motion sequence of the trajectory points. An execution sequence identifier, execution condition field, and running time field are added to each set of control parameters. The assembled control parameters are integrated to generate a complete line laying execution instruction. The line laying execution instruction is written into the robot control communication structure and issued through the instruction interface of the execution unit, so that the line laying robot can perform the line laying operation according to the instruction.

[0100] In this embodiment, the generation of the actual laying path specifically includes:

[0101] During the line laying process, the operation data of the line laying robot execution unit is collected in real time. Each collection record is assigned a time description according to the collection time, and the corresponding spatial location and execution status parameters are recorded. At the same time, each record is labeled with the construction area and corresponding component domain identifier during the collection process. Records with missing fields, contradictory status or abnormal format are removed and re-collected to form a set of original operation data records with time description, component domain identifier and spatial location description.

[0102] Based on the original set of running data records, the spatial locations in the continuously collected records are combined into a trajectory point sequence in chronological order. Component domain constraints are applied to the trajectory point sequence according to the component domain identifier. The spatial location of each trajectory point is mapped to the traversable path domain of the corresponding component. Attitude fusion processing is performed on the trajectory point attitude based on the fusion of component surface normal vector, path tangential vector and device attitude vector. Digital twin inverse projection correction is performed on trajectory points that have position jumps, attitude reversals or inconsistent states, so that the trajectory point sequence satisfies the component domain constraints in spatial location and the path travel constraints in attitude direction, resulting in a real trajectory data sequence containing attitude fusion information and position correction information.

[0103] The real trajectory data sequence is reconstructed according to the time sequence of the laying out, generating a real laying out path with continuous trajectory segments, fused posture description and component domain correspondence. The spatial position of each trajectory point in the real laying out path is converted into a unified coordinate system of the updated 3D engineering scene. The component nodes, spatial region nodes and status fields to which the trajectory points belong are updated synchronously, so that the updated 3D engineering scene can fully express the actual movement path, posture changes and area coverage of the laying out robot on the construction site, thereby generating a real laying out path and completing real-time synchronization with the 3D engineering scene.

[0104] In this embodiment, obtaining the local adjustment result specifically includes:

[0105] The actual laying path and the trajectory points in the target laying path are paired point by point in time sequence. The spatial position description and attitude description of each trajectory point are extracted. The position deviation value composed of position deviation and attitude deviation is calculated. The position deviation value is used to describe the spatial distance difference between trajectory points, and the attitude deviation value is used to describe the angle difference between the attitude directions of trajectory points. All position deviation values ​​are combined into a deviation sequence according to the trajectory sequence for deviation identification processing.

[0106] The positional deviation value in the deviation sequence is compared with the preset deviation threshold. The spatial region corresponding to the trajectory point where the deviation exceeds the limit is located. The component distribution, obstacle layout and passable space in the current region are read. Regions with component encroachment or path failure are marked as conflict regions to form conflict identification results.

[0107] In the conflict area, local replanning is performed, the component constraints and passable space units in the conflict area are loaded, alternative location points are selected and local path segments are constructed, the connection relationship between the local path segments and the original path segments is corrected, and the attitude description of the local path segments is adjusted to meet the laying direction constraints and component spatial relationship constraints, thus forming a local adjusted path.

[0108] The local adjustment path is converted into a local adjustment instruction. The spatial position, attitude and execution order are encoded by control fields. The corresponding fields in the original wire laying execution instruction are replaced to generate a local adjustment result for driving the wire laying robot execution unit to perform dynamic correction actions.

[0109] This implementation method achieves precise identification of deviations and regional conflict localization by quantifying the point-by-point deviations between the actual laying-out path and the target path, enabling the system to accurately determine the specific cause and location of path deviations. Combined with digital twin information such as component distribution, obstacle layout, and passable space, local replanning can quickly generate new paths within conflict areas that meet component constraints and movement direction requirements, achieving automated correction of execution deviations.

[0110] In this embodiment, obtaining the multi-entity collaborative scheduling instruction specifically includes:

[0111] In the updated 3D engineering scene, the construction area involved in the local adjustment results is analyzed, and a mapping relationship is established between each layout task and the area and available material resources, forming a task and material resource area mapping set indexed by the construction area.

[0112] Based on the mapping set of tasks and material resources, the utility of each line laying task and the candidate material resources in its area is evaluated. A hybrid scheduling strategy combining market bidding mechanism is adopted, which comprehensively considers the matching degree between tasks and material resources, spatial distance, current load level of the area and expected load level. The matching degree is recorded as a score, the spatial distance is recorded as a distance, and the deviation between the current load and the expected load of the area is included in the load penalty in a weighted manner. The score, distance and load penalty are weighted and synthesized by pre-set weight parameters and attenuation parameters to obtain the bidding utility value between each task and the candidate material resources. The candidate relationships are sorted and filtered according to the size of the utility value to form a candidate relationship set for task allocation.

[0113] Using the candidate relationship set of task allocation as constraints, a regional hierarchical factor graph structure is constructed based on the division of construction areas. In the hierarchical factor graph structure, the material resource allocation of each layout task in each construction area is represented as an allocation variable. The task demand constraints, material resource capacity constraints, and regional construction restrictions are represented as task factors, resource factors, and regional factors, respectively. The results of local adjustments are written into the corresponding regional factors in the form of additional constraints. This allows the regional hierarchical factor graph structure to simultaneously reflect the candidate matching relationship between tasks and material resources, the resource and construction constraints of each construction area, and the path adjustment constraints brought about by the aforementioned local replanning.

[0114] Max-Sum distributed collaborative scheduling is performed on the hierarchical factor graph structure. The messages sent by each factor node to its connected allocation variable nodes are iteratively updated. During each update, the bidding utility value corresponding to the candidate relationship between the current task and material resources is combined with the candidate relationship, local adjustment constraints, and hierarchical relationship of the hierarchical factor graph structure as the comprehensive cost. The comprehensive cost of different task allocation combinations is compared and selected. By repeatedly passing and updating messages between factor nodes and variable nodes until the values ​​of each allocation variable converge, the optimal allocation combination of task and material resources that satisfies the overall utility maximization and regional constraints is obtained. This is then converted into multi-subject collaborative scheduling instructions for different wire-laying robot execution units and material supply entities.

[0115] This implementation establishes a precise mapping between tasks and material resources in a digital twin scenario, and uses bidding utility values ​​as a quantitative basis to achieve refined evaluation and dynamic ranking of resource matching. Combined with a market bidding mechanism, it can adaptively reflect changes in regional load, making task allocation more reasonable and efficient. A regional hierarchical factor graph structure uniformly expresses task requirements, resource capacity, construction limitations, and local adjustment constraints, ensuring clear hierarchical relationships and global consistency in multi-agent scheduling. Based on this, Max-Sum distributed collaborative scheduling is executed, achieving a globally optimal task and resource allocation scheme in a distributed computing environment, improving the coordination, stability, and overall efficiency of multi-robot line-laying operations.

[0116] In this embodiment, the formation of the engineering layout data link specifically includes:

[0117] The spatial location, attitude, and motion state descriptions of the actual cable laying path are continuously collected. The collected path data are organized according to the time description, each record is written into a path record entry in a unified coordinate format, and all record entries are organized in chronological order to form a time-series path record set of the actual cable laying path.

[0118] The optimal allocation combination generated by multi-entity collaborative scheduling is read and written into the scheduling record entry. The message update content from the factor node to the variable node during the scheduling iteration is also recorded. The scheduling record entry and message update content are combined to form a multi-entity collaborative scheduling instruction record set.

[0119] The time-series path record set of the actual layout path is associated with the multi-subject collaborative scheduling instruction record set according to the task number and time description quantity. The spatial location description in the path record is compared with the allocation description in the scheduling record, and archived according to the execution batch number to form an engineering layout data chain for quality acceptance and construction traceability.

[0120] Example 1:

[0121] To verify the feasibility of this invention in practice, it was applied to a construction scenario involving complex component layouts and multi-robot collaborative operations. This scenario included various component types such as walls, beams, columns, floor slabs, and pipeline pre-installation. The site was dynamic, with construction equipment, material stacks, and personnel movement, representing a typical scenario characterized by complex multi-source data, frequent environmental changes, and high accuracy requirements for layout. In traditional layout methods, workers rely on total stations and handheld devices for measurement and positioning. Path planning depends on manual experience, and repeated measurements and manual adjustments are often necessary when encountering obstacles, component obstructions, or congested spaces, leading to low construction efficiency, high deviation rates, and chaotic resource scheduling.

[0122] After applying this invention in this scenario, the design drawings, component attributes, and on-site measurement data are first uniformly encoded using the scene construction module to generate a complete digital twin 3D engineering scene. This scene can accurately present the geometric shape, spatial location, and topological relationship of each component, providing a unified data foundation for subsequent path planning and scheduling. Environmental changes at the construction site are collected in real time through the environment update module and mapped onto the digital twin scene, enabling the virtual scene to dynamically reflect the actual construction status. For example, newly added equipment, temporary scaffolding, and impassable areas can all be updated in real time.

[0123] During the layout path planning stage, the system automatically generates layout paths based on the 3D engineering scene and identifies potential component interference, spatial conflicts, or passage restrictions through simulation. For example, in an area with insufficient space height under a beam, traditionally manually planned paths often result in limited robot posture or even impassable passage. However, this invention identifies this problem in a timely manner by simulating motion trajectories and avoids it in advance during the simulation stage, generating feasible alternative paths.

[0124] During the line laying execution phase, this invention acquires the real-time position and attitude data of the line laying robot and maps the acquired actual trajectory back to the digital twin scene. When the actual path deviates due to material obstructions or area congestion, the deviation replanning module of this invention can immediately identify trajectory points with deviations exceeding a threshold and perform local replanning for the corresponding area. For example, in a densely piped area, if the robot deviates due to obstacle avoidance, this invention automatically generates a local adjustment path and converts it into control commands, allowing the robot to return to the correct route without human intervention.

[0125] In multi-robot operations, traditional scheduling often leads to congestion due to multiple robots concentrating in the same narrow area. This invention calculates the utility value of tasks and resources through a market bidding mechanism and runs the Max-Sum cooperative scheduling algorithm on a regional hierarchical factor graph, effectively avoiding path conflicts, resource contention, and execution delays among multiple entities.

[0126] To verify the beneficial effects of the present invention, it was compared with the traditional manual line laying method and the single robot fixed path method. The specific data are shown in Table 1:

[0127] Table 1 Comparison of the Effectiveness of Collaborative Management in Engineering Layout

[0128] Comparison indicators Traditional manual layout Single robot fixed path laying This invention system Improvement effect Average layout deviation (mm) 12.4 8.7 3.1 Accuracy improved by 64%~75% Construction time per batch (minutes) 73 58 31 Time reduced by 47%~57% Number of multi-robot conflicts (times / day) None (manual stand-alone machine) 6.2 0.8 Conflict reduced by 87% Number of partial reworks (times / 50 paths) 11 7 2 Rework has decreased by 71% to 82%. Material resource utilization rate (%) 63 71 89 Utilization rate increased by 18% to 41%. Scheduling decision time (milliseconds) not applicable 410 36 Scheduling speed increased by more than 10 times

[0129] As shown in Table 1, the system of this invention significantly outperforms traditional manual line-laying methods and single-robot fixed-path methods in several key indicators. Regarding line-laying accuracy, the average deviation of this invention is only 3.1 mm, a reduction of approximately 75% compared to traditional methods, effectively avoiding rework and accumulated errors caused by inaccurate measurements. In terms of construction efficiency, the construction time for a single batch of work is approximately 31 minutes, nearly half that of traditional methods, making the overall construction rhythm more compact and controllable. Regarding multi-entity collaboration, this invention reduces the number of multi-robot conflicts to 0.8 times / day, and combined with an efficient resource scheduling mechanism, achieves a material utilization rate of 89%, significantly higher than other methods. Furthermore, the scheduling decision-making time of this invention is only 36 milliseconds, more than ten times faster than the fixed-path method, ensuring that the scheduling results can be applied to the construction process in real time.

[0130] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A collaborative management system for engineering layout process based on digital twins, characterized in that, include: The scene construction module is used to acquire engineering information data, perform structured encoding and coordinate unification, and generate a digital twin 3D engineering scene. The environment update module is used to collect environmental status data, complete data preprocessing, and obtain the updated 3D engineering scene; The path simulation module is used to generate the target layout path in the updated 3D engineering scene according to the on-site constraints and layout requirements, and to perform simulation and deduction to obtain the simulation execution trajectory. The execution control module is used to generate wire laying execution instructions based on the simulated execution trajectory and issue them to the wire laying robot execution unit to perform trajectory processing and attitude correction, and generate the actual wire laying path. The deviation replanning module is used to analyze the deviation between the actual surveying path and the target surveying path, and to perform local replanning to generate local adjustment results for dynamic correction. The collaborative scheduling module is used to determine the candidate relationships for task allocation through a market bidding mechanism, execute Max-Sum distributed collaborative scheduling on the regional hierarchical factor graph, and generate multi-entity collaborative scheduling instructions. The data chain management module is used to record the actual layout path and multi-entity collaborative scheduling instructions, forming an engineering layout data chain for quality acceptance and construction traceability.

2. The collaborative management system for engineering layout process based on digital twin as described in claim 1, characterized in that, The modules are connected in the following way: Acquire engineering information data, perform structuring, coordinate unification, and scene construction to form a digital twin three-dimensional engineering scene; Based on the 3D engineering scene, environmental status data is collected, data preprocessing is performed, and an updated 3D engineering scene is obtained. In the updated 3D engineering scene, the target layout path is generated according to the on-site constraints, and simulation is performed to obtain the simulation execution trajectory. The wire laying execution command is generated based on the simulated execution trajectory and issued to the wire laying robot execution unit; During the wire laying process, the position information and execution status of the wire laying robot's execution unit are collected in real time, and trajectory processing and attitude correction are performed to generate a real wire laying path; The deviation between the actual laying path and the target laying path is analyzed, and the laying execution command for the corresponding area is locally adjusted to obtain the local adjustment result. The line laying task is mapped to the corresponding area with the material resources. The candidate relationship for task allocation is determined by market bidding. Max-Sum distributed collaborative scheduling is executed on the factor graph to obtain multi-entity collaborative scheduling instructions. The actual layout paths and multi-entity collaborative scheduling instructions generated during the layout process are recorded and archived to form an engineering layout data chain.

3. The collaborative management system for engineering layout process based on digital twins according to claim 2, characterized in that, The formation of the digital twin's 3D engineering scene specifically includes: Acquire engineering information data, read and import design drawings, component attributes and on-site measurements, classify and organize data from different sources, and screen and process them to form the original dataset of engineering information; The raw dataset of engineering information is structured, encoded according to a unified field format, and the coordinate datum and elevation datum are uniformly converted to form an engineering information data table with a unified expression format. Based on the engineering information data table, the three-dimensional geometric shape of the component is generated according to the planar coordinates and elevation coordinates of the component, and the connection, intersection and subordination relationship between the components is constructed to form a digital twin three-dimensional engineering scene.

4. The collaborative management system for engineering layout process based on digital twin as described in claim 2, characterized in that, The updated 3D engineering scene is obtained specifically through: Environmental status data is collected under a unified coordinate system in a 3D engineering scene. The environmental status data is recorded with time information according to the collection time and classified. Data with abnormal format is removed. Each valid environmental status data has a clear collection time description and spatial location description, forming a set of original environmental status data. The raw environmental state data set is processed in a unified manner to form a preprocessed environmental state data set. The environmental state preprocessing data set is matched with the engineering information in the 3D engineering scene, and the relevant state parameters in the 3D engineering scene are updated to obtain the updated 3D engineering scene.

5. The collaborative management system for engineering layout process based on digital twins according to claim 2, characterized in that, The specific steps involved in obtaining the simulation execution trajectory are as follows: In the updated 3D engineering scene, the construction area information used for line laying is read, the spatial position of components, the outer boundary of components and the location of obstacles in the construction area are extracted and marked, and the space range that the line laying robot can pass through is delineated in the 3D coordinate system to obtain the path planning space. In the path planning space, based on the site constraints and specified layout requirements, the spatial positions of the starting and ending points of the layout are determined. Target positions are selected sequentially according to the layout order. Each target position is associated with its corresponding spatial position in the 3D engineering scene and its relationship with adjacent components. The target positions are combined into a continuous layout route according to the layout progress order to generate path data. In the updated 3D engineering scene, the movement process of the wire-laying robot is simulated sequentially according to the path data. The arrival position and motion state of each target location are simulated, and the position and posture changes of the wire-laying robot during the simulation process are recorded to form the simulation execution trajectory.

6. The collaborative management system for engineering layout process based on digital twins according to claim 2, characterized in that, The process of generating the line-laying execution instruction specifically includes: The simulation execution trajectory is read, and the trajectory points recorded during the simulation are reorganized in chronological order. A trajectory point sequence is generated based on the continuity and chronological relationship of the trajectory points. Based on the trajectory point sequence, the spatial position description of each trajectory point is converted into the control coordinate system adopted by the wire-laying robot execution unit. The posture and motion direction of the trajectory points are encoded. The position data, posture parameters and motion direction parameters are combined according to the control field format of the execution unit to form a set of control parameters. The control parameter set is assembled according to the motion sequence of the trajectory points. The assembled control parameters are integrated to generate a complete line laying execution command, which is then written into the robot control communication structure and issued through the command interface of the execution unit.

7. The collaborative management system for engineering layout process based on digital twin as described in claim 2, characterized in that, The generation of the actual laying path specifically includes: During the line laying process, the operation data of the line laying robot execution unit is collected in real time. Each collection record is assigned a time description according to the collection time. At the same time, each record is marked with the construction area and corresponding component domain identifier during the collection process, forming a set of original operation data records. Based on the original set of running data records, the spatial locations in the continuously collected records are combined into a trajectory point sequence in chronological order. Component domain constraints are applied to the trajectory point sequence according to the component domain identifier. The spatial location of each trajectory point is mapped to the traversable path domain of the corresponding component. Attitude fusion processing is performed on the trajectory point attitude based on the fusion of component surface normal vector, path tangential vector and device attitude vector. Digital twin inverse projection correction is performed on trajectory points that have position jumps, attitude reversals or inconsistent states, so that the trajectory point sequence satisfies the component domain constraints in spatial location and the path travel constraints in attitude direction, thus obtaining the real trajectory data sequence. The real trajectory data sequence is reconstructed according to the time sequence of the laying out, generating a real laying out path with continuous trajectory segments, fused attitude description and component domain correspondence. The spatial position of each trajectory point in the real laying out path is converted into a unified coordinate system of the updated 3D engineering scene, and the component nodes, spatial region nodes and status fields to which the trajectory points belong are updated synchronously.

8. The collaborative management system for engineering layout process based on digital twins according to claim 2, characterized in that, The results of the local adjustments are obtained specifically in the following ways: The actual laying path and the trajectory points in the target laying path are paired point by point in time sequence. The position deviation value composed of position deviation and attitude deviation is calculated, and all position deviation values ​​are combined into a deviation sequence in the trajectory sequence. The positional deviation values ​​in the deviation sequence are compared with the preset deviation threshold, and areas where there is component encroachment or path failure are marked as conflict areas, thus forming a conflict identification result. Within the conflict area, perform local replanning, load the component constraints and passable space units of the conflict area, select alternative location points and construct local path segments, and perform continuity correction on the connection relationship between the local path segments and the original path segments to form a local adjusted path. The local adjustment path is converted into a local adjustment instruction. The spatial position, attitude and execution order are encoded using control fields. The corresponding fields in the original line laying execution instruction are replaced to generate the local adjustment result.

9. A collaborative management system for engineering layout process based on digital twins according to claim 2, characterized in that, The acquisition of the multi-entity collaborative scheduling instruction specifically includes: In the updated 3D engineering scene, the construction area involved in the local adjustment results is analyzed, and a mapping relationship is established between each layout task and the area and available material resources, forming a task and material resource area mapping set indexed by the construction area. Based on the mapping set of tasks and material resources, the utility of each line laying task and the candidate material resources in its area is evaluated. A hybrid scheduling strategy combining market bidding mechanism is adopted, which comprehensively considers the matching degree between tasks and material resources, spatial distance, current load level of the area and expected load level. The matching degree is recorded as a score, the spatial distance is recorded as a distance, and the deviation between the current load and the expected load of the area is included in the load penalty in a weighted manner. The score, distance and load penalty are weighted and synthesized by pre-set weight parameters and attenuation parameters to obtain the bidding utility value between each task and the candidate material resources. The candidate relationships are sorted and filtered according to the size of the utility value to form a candidate relationship set for task allocation. Using the candidate relationship set for task allocation as a constraint, a regional hierarchical factor graph structure is constructed based on the division of construction areas. In the hierarchical factor graph structure, the material resource allocation of each layout task in each construction area is represented as an allocation variable. The task requirement constraint, material resource capacity constraint, and regional construction restriction are represented as task factors, resource factors, and regional factors, respectively. The local adjustment results are written into the corresponding regional factors in the form of additional constraints. Max-Sum distributed collaborative scheduling is performed on the hierarchical factor graph structure. The messages sent by each factor node to its connected allocation variable nodes are iteratively updated. During each update, the bidding utility value corresponding to the candidate relationship between the current task and material resources is combined with the candidate relationship, local adjustment constraints and hierarchical relationship of the hierarchical factor graph structure as the comprehensive cost. The comprehensive cost of different task allocation combinations is compared and selected to obtain the optimal allocation combination, and then converted into multi-agent collaborative scheduling instructions.

10. A collaborative management system for engineering layout process based on digital twins according to claim 2, characterized in that, The formation of the engineering layout data link specifically includes: The spatial location, attitude, and motion state descriptions of the actual cable laying path are continuously collected, organized according to the time description amount, written into the path record entries, and all record entries are organized in chronological order to form a time-series path record set of the actual cable laying path. The optimal allocation combination generated by multi-entity collaborative scheduling is read and written into the scheduling record entry. The message update content from the factor node to the variable node during the scheduling iteration is also recorded. The scheduling record entry and message update content are combined to form a multi-entity collaborative scheduling instruction record set. The time-series path record set of the actual surveying path is associated with the multi-entity collaborative scheduling instruction record set according to the task number and time description quantity, and archived according to the execution batch number to form an engineering surveying data chain.

Images