An intelligent municipal pipeline wiring method and system based on path optimization

Abstract

The application relates to the technical field of municipal pipeline wiring, and discloses an intelligent municipal pipeline wiring method and system based on path optimization, which comprises the following steps: preprocessing multi-source municipal pipeline related data to obtain a standardized data set, constructing a BIM and GIS model and fusing the two models to form a three-dimensional digital base; dividing three detection areas by hierarchical spatial indexing and establishing a pipeline detection unit, completing conflict checking and semantic topology analysis, and demarcating a constraint boundary; constructing a multi-constraint optimization model, adopting a multi-objective genetic algorithm and deep reinforcement learning hybrid optimization, checking conflicts in real time and adjusting parameters, and generating an optimal pipeline path scheme; generating construction instructions according to the scheme, completing field lofting through RTK and laser radar, collecting laying parameters in real time, comparing the laying parameters, and generating a correction scheme when the deviation exceeds the limit. The application can realize global optimization of a pipeline path under multi-constraint conditions and dynamic adaptation in a construction stage.

Description

Technical Field

[0001] This application relates to the technical field of municipal pipeline wiring, and in particular to an intelligent municipal pipeline wiring method based on path optimization. Background Technology

[0002] Urban municipal pipelines are crucial infrastructure for ensuring the normal operation of cities. With the continuous expansion of urban scale and the deepening development of underground space, pipelines from multiple disciplines, such as water supply, drainage, gas, electricity, and communications, are intertwined and laid out. Pipeline planning and layout involve various types of information, including geographical environment, existing pipeline networks, geological conditions, planning control, and construction requirements. The data sources are scattered and the formats are heterogeneous. Traditional municipal pipeline wiring mainly relies on two-dimensional drawings and manual experience, making it difficult to efficiently integrate and uniformly apply multi-source data, and also failing to achieve three-dimensional and visual management of underground space.

[0003] In the current process of urban construction and renovation, the layout of municipal pipelines needs to meet multiple requirements such as standards, geological safety, construction conditions, and engineering costs. Various constraints are mutually restrictive and difficult to balance. Furthermore, the planning and design schemes are difficult to link with the on-site construction conditions in real time, the schemes are lagging behind and lack adaptability. Overall, it still relies on manual segmented management and lacks systematic and intelligent path optimization support.

[0004] As can be seen from the above, existing municipal pipeline routing methods are difficult to achieve global optimization of pipeline paths and dynamic adaptation during the construction phase under multiple constraints. Summary of the Invention

[0005] To achieve global optimization of pipeline routes and dynamic adaptation during construction under multiple constraints, this application provides an intelligent municipal pipeline routing method based on route optimization.

[0006] Firstly, this application provides an intelligent municipal pipeline routing method based on path optimization, employing the following technical solution: A path optimization-based intelligent municipal pipeline routing method includes: The collected urban basic geography, existing pipeline network, geological survey, planning red line and pipeline layout standard data are preprocessed with format unification and cleaning correction to obtain a standardized multi-source dataset. Based on the standardized multi-source dataset, BIM parametric pipeline model and GIS geospatial model are constructed respectively. The BIM parametric pipeline model and GIS geospatial model are incorporated into the same coordinate system to complete spatial registration and fusion, so as to construct a three-dimensional digital base for integrated indoor and outdoor rendering. Based on the aforementioned three-dimensional digital base, the detection area is divided into three progressively more detailed levels: macroscopic, mesoscopic, and microscopic, using hierarchical spatial indexing technology. Corresponding pipeline detection units are established for each level. Through each pipeline detection unit, pipeline spatial conflict verification within the corresponding level is completed. Semantic topology analysis is performed in conjunction with pipeline functional attributes, ownership, and maintenance requirements to delineate the constraint boundaries for pipeline layout at each level. Based on the aforementioned constraint boundaries, a multi-constraint optimization model is constructed. Constraints such as regulatory requirements, geological environment, economic efficiency, and construction conditions are imported into pipeline detection units at each level to complete hierarchical weighting. Through a hybrid optimization method combining multi-objective genetic algorithm and deep reinforcement learning, in each generation of the genetic algorithm, the corresponding level pipeline detection unit is synchronously called to complete real-time conflict verification. The verification results are input into deep reinforcement learning to adjust the search direction and constraint weights, thereby generating the optimal pipeline layout path scheme. Based on the pipeline layout plan, construction instructions are generated. Using the three-dimensional digital base as a reference, the on-site pipeline is accurately laid out through RTK positioning and lidar point cloud matching. Pipeline laying parameters are collected in real time and compared with the design model. When the deviation exceeds the limit, an early warning is triggered and a correction plan is generated.

[0007] Optionally, the step of generating pipeline routing schemes using a hybrid optimization approach combining multi-objective genetic algorithms and deep reinforcement learning includes: Before each generation of the genetic algorithm, a dual pre-verification of constraint boundary compliance and pipeline detection unit availability is triggered, and the dual pre-verification is set as the only hard condition for individual admission in this iteration of the population. Perform constraint boundary compliance verification to confirm that the pipeline path corresponding to the individual in the population to be iterated is completely matched with the constraint boundary defined by the pipeline detection unit at the corresponding level. If the verification passes, proceed to the availability verification of the pipeline detection unit; if the verification fails, remove the corresponding invalid individual. Perform a pipeline detection unit availability check to confirm that there is a matching pipeline detection unit that can be called in real time for the path level corresponding to the individual in the population to be iterated. If the check passes, proceed to the iterative calculation stage; if the check fails, remove the corresponding invalid individual. Only valid individuals that have completed dual pre-verification and passed both verifications can enter the computation phase of this iteration, synchronously calling the pipeline detection unit at the corresponding level to complete real-time conflict verification.

[0008] Optionally, the step of iteratively calculating the valid individuals that have passed the double pre-verification includes: Set the constraint priority for iterative optimization, set the mandatory constraints of the standard as the highest priority unbreakable hard constraints, set the geological environment constraints and construction condition constraints as the second priority constraints, and set the economic constraints as the third priority constraints. The priority of each level of constraints is used as the core basis for fitness value calculation. The optimization objectives are weighted, with the lowest probability of pipeline space conflict as the core optimization objective and the lowest overall cost of pipeline engineering as the secondary optimization objective. The core optimization objective has a higher weight than the secondary optimization objective. For valid individuals that pass the double pre-validation, the fitness value is calculated according to the preset constraint priority and optimization target weight allocation. A preset number of individuals with the best fitness value are retained to enter the next iteration. At the same time, the conflict validation results of this iteration are input into deep reinforcement learning. When the preset maximum number of iterations is reached, or when the fitness value of the best individual does not improve after a preset number of generations, the iteration is terminated, and the pipeline layout path scheme corresponding to the individual with the best current fitness value is output.

[0009] Optionally, the step of inputting the iterative conflict verification results into deep reinforcement learning to adjust the search direction and constraint weights includes: The adjustment trigger logic of deep reinforcement learning is set, with the conflict verification result of each iteration as the core triggering basis. When the number of pipeline conflict points detected in the next iteration exceeds the preset threshold, the search direction and constraint weight are immediately adjusted. Set rules for adjusting the search direction, lock the path region with concentrated conflicts based on the conflict verification results, narrow the search range of the next generation of the genetic algorithm, and strengthen the global search power of the conflict-free path region; Set rules for adjusting constraint weights, and based on the constraint type corresponding to the conflict point, simultaneously increase the weight ratio of the corresponding constraint priority to ensure that the adjusted constraint weights match the risk level of the on-site deployment. The adjusted search direction and constraint weights are then simultaneously incorporated into the dual pre-verification and fitness value calculation stages of the next iteration, forming a closed-loop adjustment chain for iterative optimization.

[0010] Optionally, the steps of dividing the detection area, establishing pipeline detection units, and defining constraint boundaries using hierarchical spatial indexing technology include: The classification criteria are set at three levels: macro, meso, and micro. The spatial scope and detection accuracy corresponding to each level are clearly defined. The macro level corresponds to the entire planning red line scope, and the detection accuracy matches the requirements of the city's overall road network and planning control. The meso level corresponds to the construction scope of the block, and the detection accuracy matches the requirements of the topography and existing pipeline distribution within the block. The micro level corresponds to the pipeline laying point scope, and the detection accuracy matches the spatial layout requirements of single pipeline segments and fittings. Each level is matched with a unique pipeline detection unit. Each pipeline detection unit is assigned verification permissions and data access permissions that match the corresponding level. Macro-level pipeline detection units can only access global planning data, meso-level pipeline detection units can only access the existing pipeline network and geological data within the corresponding block, and micro-level pipeline detection units can only access the pipeline parameters and fitting size data of the corresponding point. Each pipeline detection unit can only perform pipeline spatial conflict verification within the corresponding spatial range within the scope of its corresponding level's authority. After verification, it outputs the spatial conflict verification result of the corresponding level for subsequent semantic topology analysis.

[0011] Optionally, the step of performing semantic topology analysis by combining pipeline functional attributes, ownership, and maintenance requirements, and delineating the constraint boundaries for pipeline layout at each level, includes: Pre-matching corresponding semantic analysis dimensions for pipeline detection units at each level: macro level matching the overall pipeline planning direction and cross-block ownership division semantic dimensions; meso level matching the semantic dimensions of pipeline function adaptation and maintenance channel reservation within the block; micro level matching the semantic dimensions of single pipeline safety distance and pipe fitting connection requirements. Based on the spatial conflict verification results output by the pipeline detection unit at the corresponding level, and combined with the matching semantic analysis dimension, the semantic information of pipeline functional attributes, ownership, and daily maintenance requirements is extracted to complete the pipeline semantic topology analysis and identify non-geometric layout restrictions between pipelines. By combining the spatial conflict verification results with the non-geometric layout restrictions output by semantic topology analysis, the compliant range of pipelines allowed to be laid out and the prohibited range of pipelines within the corresponding level are defined, thus forming the pipeline layout constraint boundary of the corresponding level. Constraint boundaries are set for pipeline layout at each level, and output synchronously to the corresponding pipeline detection unit in the subsequent path optimization step according to the hierarchical correspondence, which is determined as the hard constraint benchmark for path optimization.

[0012] Optionally, after the steps of real-time acquisition of pipeline laying parameters and comparison with the design model, triggering an early warning and generating a correction plan when the deviation exceeds the limit, the method further includes: The actual laying parameters of the laid pipelines and the real-time working conditions data of the construction site are collected in real time and updated to the three-dimensional digital base in a synchronous manner to complete the dynamic correction of the three-dimensional digital base and make the spatial data of the three-dimensional digital base consistent with the physical state of the construction site. Based on the corrected 3D digital base, the pipeline detection unit at the corresponding level is triggered to perform incremental spatial conflict verification. Combining the actual spatial location of the laid pipelines and the data of newly added obstacles on site, the spatial conflict verification results at the corresponding level are updated. Combined with the pipeline functional attributes, ownership, and maintenance requirements, supplementary semantic topology analysis is completed, and the layout constraint boundary of the unlaid pipeline segment in the corresponding level is dynamically adjusted. Based on the dynamically adjusted layout constraint boundary, a hybrid optimization method combining multi-objective genetic algorithm and deep reinforcement learning is called to perform path re-optimization only on the unlaid pipeline segments, lock the path scheme corresponding to the laid pipeline segments without modification, and generate a modified pipeline layout path scheme adapted to the actual working conditions of the construction site. The revised pipeline layout plan will be updated to the construction instructions, and the subsequent accurate layout and laying of pipelines will be completed based on the revised pipeline layout plan.

[0013] Secondly, this application provides an intelligent municipal pipeline cabling system based on path optimization, which adopts the following technical solution: An intelligent municipal pipeline cabling system based on path optimization includes: The 3D base construction module performs preprocessing, including format unification and cleaning correction, on the collected urban basic geography, existing pipeline network, geological survey, planning red line and pipeline layout specifications data to obtain a standardized multi-source dataset. Based on the standardized multi-source dataset, a BIM parametric pipeline model and a GIS geospatial model are constructed respectively. The BIM parametric pipeline model and the GIS geospatial model are incorporated into the same coordinate system to complete spatial registration and fusion, so as to construct a 3D digital base with integrated indoor and outdoor rendering. The hierarchical conflict detection module, based on the three-dimensional digital base, uses hierarchical spatial indexing technology to divide the detection area into three progressively refined levels: macroscopic, mesoscopic, and microscopic. A corresponding pipeline detection unit is established for each level. Through each pipeline detection unit, the spatial conflict verification of pipelines within the corresponding level is completed. Semantic topology analysis is completed in combination with pipeline functional attributes, ownership, and maintenance requirements to delineate the constraint boundaries of pipeline layout at each level. The intelligent path optimization module constructs a multi-constraint optimization model based on the aforementioned constraint boundaries. It imports the constraints of regulatory requirements, geological environment, economic efficiency, and construction conditions into the pipeline detection units at each level to complete hierarchical weighting. Through a hybrid optimization method combining multi-objective genetic algorithms and deep reinforcement learning, the corresponding pipeline detection units at each level are synchronously invoked in each generation of the genetic algorithm to complete real-time conflict verification. The verification results are input into deep reinforcement learning to adjust the search direction and constraint weights, thereby generating the optimal pipeline layout path scheme. The construction layout and correction module generates construction instructions based on the pipeline layout path scheme. Using the three-dimensional digital base as a reference, it completes the accurate layout of the pipeline on site through RTK positioning and lidar point cloud matching. It collects pipeline laying parameters in real time and compares them with the design model. When the deviation exceeds the limit, it triggers an early warning and generates a correction scheme.

[0014] Thirdly, this application provides an electronic device that adopts the following technical solution: An electronic device includes a processor running a program for the intelligent municipal pipeline wiring method based on path optimization as described in any one of the preceding claims.

[0015] Fourthly, this application provides a storage medium, which adopts the following technical solution: A storage medium storing a program for the intelligent municipal pipeline wiring method based on path optimization as described in any one of the above.

[0016] In summary, this application includes at least one of the following beneficial technical effects: By standardizing multi-source data, integrating BIM and GIS to construct a 3D digital foundation, and employing hierarchical spatial indexing and three-level pipeline detection units, multiple constraints such as geography, geology, pipeline network, planning, and standards are structurally decomposed and weighted hierarchically, achieving unified control of constraints across the entire domain from macro to micro. Combining multi-objective genetic algorithms with deep reinforcement learning for hybrid optimization, the detection units are invoked in real time during the iteration process to complete conflict verification, and the search direction and constraint weights are dynamically adjusted based on the conflict results. Under the parallel constraints of multiple factors such as mandatory standards, geological environment, construction conditions, and economic efficiency, the system can quickly converge to obtain the globally optimal pipeline route scheme, achieving global optimization of pipeline routes under complex constraints.

[0017] Based on a 3D digital base, data integration between design and construction is achieved. Precise on-site layout is completed through RTK positioning and lidar point cloud matching. Laying parameters are collected in real time for comparison with the design model, and deviation warnings are issued. At the same time, actual construction data is dynamically updated to the 3D digital base, triggering incremental conflict verification and constraint boundary adjustment. Oriented path re-optimization is performed on unlaid pipe sections while locking the laid sections unchanged. This allows the pipeline path to adapt to changes in construction site conditions in real time, maintaining the compliance, safety, and feasibility of the layout plan throughout the entire process, and achieving dynamic adaptation and closed-loop management of the entire construction phase. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating an intelligent municipal pipeline routing method based on path optimization, according to an exemplary embodiment.

[0019] Figure 2 This is a structural block diagram of an intelligent municipal pipeline cabling system based on path optimization, according to an exemplary embodiment. Detailed Implementation

[0020] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.

[0021] In the description of this specification, the references to "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" refer to specific features, structures, materials, or characteristics described in connection with the described embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0022] This application discloses an intelligent municipal pipeline routing method based on path optimization. This method is applicable to various scenarios, including comprehensive municipal pipeline layout in newly built urban areas, pipeline renovation and upgrading in old urban areas, and pipeline laying for road reconstruction and expansion. It pre-collects multi-source basic data for the target routing area, obtaining urban basic geographic data, existing pipeline network data, geological survey data, planning boundary data, and pipeline layout specification data for the target area, referencing... Figure 1 Specifically, it includes the following steps: S100 performs preprocessing, including format unification and cleaning correction, on the collected urban basic geography, existing pipeline network, geological survey, planning red line and pipeline layout specifications data to obtain a standardized multi-source dataset.

[0023] In this embodiment of the invention, the specific execution process of S100 is implemented through the following sub-steps: S101 completes the full collection and classification of multiple core data for the target cabling area, covering the management and control needs of the entire pipeline cabling process, with no data blind spots.

[0024] The collected urban basic geographic data includes digital elevation models, digital orthophotos, urban road network vector data, cadastral ownership data, surface building vector data, river systems, green vegetation and other spatial geographic data of the target area. The data accuracy meets the standards of large-scale urban topographic maps and covers the surface spatial information of the entire target area. The collected existing pipeline network data includes the completed data of multi-disciplinary pipelines already built in the target area, covering the spatial coordinates, pipe diameter, material, pressure rating, burial depth, laying year, ownership unit, connection point, manhole location and other geometric and attribute information of each pipeline segment. At the same time, the existing underground space data such as underground civil defense projects, underground structures, and underground obstacles are collected to clarify the spatial boundaries of the existing underground facilities. The collected geological survey data includes engineering geological survey reports, borehole data, soil layer distribution profile data, physical and mechanical parameters of each soil layer, groundwater level depth and variation patterns, and adverse geological distribution data for the target area, clarifying the spatial distribution and engineering characteristics of the underground geological environment; The collected planning red line and control data include various legal control boundary data of the target area, as well as planning control data such as control detailed planning, special planning for various professional pipelines, land use planning, and urban renewal planning, which clarify the legal control scope and planning requirements for pipeline layout; The collected pipeline layout standard data includes currently valid national and industry standards. Mandatory clauses and recommended requirements from these standards are extracted and converted into quantifiable constraint parameters. Simultaneously, on-site construction-related data are collected for the target area, including construction section divisions, construction enclosure areas, existing traffic diversion requirements, construction machinery operating parameters, distribution of surrounding sensitive buildings, and construction schedule requirements. After data collection, the data is categorized and aggregated according to data type, spatial hierarchy, and professional attributes. A multi-source data association index is established to achieve bidirectional mapping of various types of data within the same spatial location.

[0025] S102 involves standardizing the format and converting the coordinate system to a single system for the collected multi-source heterogeneous data, eliminating differences in data format and spatial coordinate system deviations. This process unifies the original data of different formats into a standard format compatible with both BIM and GIS, while converting unstructured specification clauses and survey descriptions into structured parametric data, establishing a one-to-one correspondence between attribute fields and spatial data. Furthermore, the coordinate system of all spatial data is converted to the national geodetic coordinate system, and the elevation datum is uniformly converted to the national elevation datum. For local coordinate system data existing in historical data, a corresponding conversion model is used to complete the precise coordinate system conversion, ensuring complete uniformity of the coordinate system for all spatial data.

[0026] S103 performs full data cleaning and error correction on the standardized multi-source data, removing invalid data and correcting biased data.

[0027] By using pre-defined verification rules, duplicate data, outlier coordinates, data with severe attribute deficiencies, and logically contradictory data in the dataset are removed. Key data with missing attributes are supplemented and improved, while invalid data that cannot be supplemented is removed. For spatial misalignment deviations in existing pipeline, terrain, and structure data, the same-name control point matching and iterative nearest-point algorithm are used to complete precise correction. Fixed points such as road intersections, building corners, existing manholes, and permanent survey markers measured on-site are used as same-name control points to complete coarse registration correction, and then the iterative nearest-point algorithm is used to complete fine registration, eliminating spatial misalignment and ensuring that the spatial data completely matches the actual situation on site. Logical verification is performed on standard data and planning control data, outdated and invalid standard provisions are removed, and control requirements that conflict with current statutory planning are corrected. At the same time, geological exploration data is layered and verified to correct logical contradictions in soil layer distribution in borehole data.

[0028] S104. After cleaning and correction, the standardized multi-source data is structured and stored to build a hierarchical spatial database. The data is stored in three layers: macro planning layer, meso block layer, and micro viewpoint layer. At the same time, it is divided into multiple sub-databases according to professional attributes. Each sub-database establishes a corresponding spatial index and attribute association relationship to realize the rapid retrieval, accurate matching and hierarchical calling of data.

[0029] S200, based on the standardized multi-source dataset, construct BIM parametric pipeline model and GIS geospatial model respectively, incorporate BIM parametric pipeline model and GIS geospatial model into the same coordinate system to complete spatial registration and fusion, so as to construct a three-dimensional digital base for integrated indoor and outdoor rendering.

[0030] In this embodiment of the invention, the specific execution process of S200 is implemented through the following sub-steps: S201, based on standardized multi-source datasets, uses parametric modeling to construct a full-discipline, full-element BIM parametric pipeline model.

[0031] A parametric BIM family library covering all disciplines of municipal pipelines is pre-built, including three major categories: pipeline families, fitting families, and ancillary facility families. Each family file not only includes a precise geometric model, but also has built-in semantic information such as corresponding functional attributes, ownership information, maintenance requirements, and specification constraint parameters, realizing deep binding between geometric models and semantic information.

[0032] Based on the planning requirements, existing pipeline data, and standards of standardized multi-source datasets, parametric BIM models are built for various professional pipelines in the target cabling area. During the model building process, each pipeline segment, each fitting, and each ancillary facility is generated through parametric driving. Parameters such as pipeline diameter, length, slope, burial depth, and corner radius can be adjusted in real time, and the geometric shape and attribute information of the model are updated synchronously after adjustment. At the same time, each component in the model is given a unique identity and associated with semantic information such as the corresponding ownership unit, design service life, maintenance cycle, and construction requirements.

[0033] After the initial BIM model is built, a preliminary compliance check is performed on the model based on the built-in standard constraint parameters. This check examines whether parameters such as pipe diameter, slope, and radius of curvature of the pipelines meet the corresponding professional standard requirements, and removes model components with non-compliant parameters.

[0034] S202, based on standardized multi-source datasets, constructs a GIS geospatial model covering the entire target wiring area.

[0035] Based on digital elevation models and digital orthophotos, an irregular triangular mesh algorithm is used to construct a 3D topographic model of the target area, accurately reproducing the topographic undulations and surface elevation changes. Simultaneously, the orthophoto texture is mapped onto the 3D topographic model, achieving realistic topographic reconstruction. Based on standardized geological exploration borehole data and soil layer distribution profile data, an interpolation method is used to construct a 3D geological body model of the target area, accurately reproducing the spatial distribution, thickness variations, and physical and mechanical parameters of different underground soil layers. The spatial extent and hazard level of adverse geological areas are also marked, achieving 3D visualization and parametric representation of the underground geological environment. Based on standardized multi-source datasets containing data on buildings, roads, rivers, green spaces, existing pipelines, and land... Data such as civil defense projects are used to construct corresponding GIS 3D vector models, accurately restoring the outline, height, and base coordinates of surface structures, the centerline, red line range, road surface elevation, and cross-sectional structure of roads, as well as the spatial coordinates, burial depth, geometric dimensions, and attribute information of underground structures and existing pipelines, achieving a GIS digital representation of all surface and underground elements. Based on planning red lines and control data, various legal control boundaries are converted into 3D spatial vector models, clarifying the spatial scope and control requirements of each control boundary. At the same time, attribute information such as land use nature, construction height restrictions, and routing control requirements of pipeline special plans in the control detailed plan are bound to the corresponding spatial models, realizing a spatialized and visual representation of planning control requirements.

[0036] S203 integrates the BIM parametric pipeline model and the GIS geospatial model into a unified national geodetic coordinate system and national elevation datum, completing the entire process of spatial alignment from coarse registration to fine registration, and achieving deep integration of the BIM model and the GIS model.

[0037] Fixed points such as permanent survey markers, road intersection center points, existing manhole coordinates, and building corner points within the target area are selected as corresponding control points for BIM and GIS model registration. These control points correspond one-to-one in both models and are evenly distributed across the entire target wiring area. Coordinate matching of these control points achieves coarse alignment of the coordinate systems of the two models, resulting in overall spatial position matching. After coarse registration, an iterative nearest-point algorithm is used to perform fine point cloud-level registration of corresponding entities in the BIM and GIS models. Surface point cloud data of existing pipelines, manholes, and structures in the BIM model are extracted and iteratively matched with the corresponding point cloud data of the corresponding entities in the GIS model. The rotation and translation matrices between the two models are continuously optimized until the point cloud matching error meets the preset accuracy requirements, eliminating subtle spatial deviations between the two models and achieving precise spatial coordinate alignment between the BIM and GIS models.

[0038] After spatial registration is completed, the attribute data of the BIM model and the GIS model are mapped and integrated in two directions. The refined parameter information and semantic information of the BIM model are mapped to the corresponding spatial location of the GIS model. At the same time, the geospatial information, geological environment information, planning and control information, and on-site construction condition information of the GIS model are mapped to the corresponding components of the BIM model, realizing the deep integration of the refined parameter capabilities of the BIM model and the large-scale spatial control capabilities of the GIS model.

[0039] S204, after completing the integration of BIM and GIS models, uses detailed rendering technology to construct an integrated, multi-level seamless 3D digital base for indoor and outdoor environments, while also encapsulating corresponding spatial analysis, conflict verification, and data retrieval function modules.

[0040] Based on level-of-detail (LMD) technology, multi-level LMDs are set for the fused 3D model, enabling seamless rendering from macroscopic global perspectives to microscopic viewpoints. This allows for integrated rendering of indoor underground spaces and outdoor surface spaces, as well as seamless switching from macroscopic urban scenes to microscopic scenes of individual pipe components. The 3D digital base encapsulates core functional modules such as spatial measurement, coordinate positioning, attribute querying, spatial conflict verification, semantic topology analysis, path visualization, and construction simulation. It also reserves data interfaces for RTK equipment, LiDAR, and on-site construction terminals, enabling real-time interaction between the 3D digital base and on-site construction data.

[0041] S300, based on the three-dimensional digital base, uses hierarchical spatial indexing technology to divide the detection area into three progressively refined levels: macroscopic, mesoscopic, and microscopic. A corresponding pipeline detection unit is established for each level. Through each pipeline detection unit, the spatial conflict verification of pipelines within the corresponding level is completed. Semantic topology analysis is completed in combination with pipeline functional attributes, ownership, and maintenance requirements to delineate the constraint boundaries of pipeline layout at each level.

[0042] In this embodiment of the invention, the specific execution process of S300 is implemented through the following sub-steps: S301 employs hierarchical spatial indexing technology to divide the target wiring detection area into three progressively more detailed levels: macroscopic, mesoscopic, and microscopic. It clarifies the spatial range, detection accuracy, and control priorities corresponding to each level, thus establishing a three-level hierarchical spatial indexing system.

[0043] The macro level corresponds to the entire planning red line range of the target cabling area, covering all urban blocks, road networks, and construction land areas involved in the project, including the entire space where all pipelines are laid. The detection accuracy matches the overall urban road network and planning control requirements. The core control focus is on the overall direction of the pipelines, cross-block route planning, and compliance with the overall urban plan and statutory control red line. The meso-level corresponds to the independent construction area divided within the entire region, and is divided according to road sections, construction land plots, functional areas, and ownership management areas. Each block corresponds to an independent construction section, geological unit, and ownership unit. The detection accuracy matches the control requirements of the topography, existing pipeline distribution, and geological conditions within the block. The core control focus is on the planar layout, vertical layering, maintenance channel reservation, connection with the existing pipeline network, avoidance with structures within the block, and compatibility with geological conditions of various professional pipelines within the block. The micro-level corresponds to the range of pipeline laying points within the block, including the laying path of a single pipeline segment, pipe fitting connection points, manhole locations, pipeline intersections, and points adjacent to structures, etc., which are subject to refined control. The detection accuracy matches the spatial layout requirements of a single pipeline segment and pipe fitting. The core control focus is on the pipe diameter, burial depth, slope, net distance from adjacent pipelines, pipe fitting connection accuracy, vertical avoidance at intersections, and detailed compliance with mandatory clauses of regulations for a single pipeline segment.

[0044] After completing the division into three levels, a corresponding hierarchical spatial index is established for each level based on the spatial indexing algorithm, enabling rapid positioning of spatial regions at different levels, accurate data retrieval, and seamless connection of spatial ranges.

[0045] S302 is divided into three levels: macroscopic, mesoscopic, and microscopic. Each level is matched with a unique pipeline inspection unit, and each pipeline inspection unit is assigned verification permissions and data access permissions that match the corresponding level.

[0046] A unique full-area pipeline detection unit is configured at the macro level. This full-area pipeline detection unit can only access full-area planning data, including planning red lines, legal control boundaries, urban master plans, special plans for various professional pipelines, cross-block ownership division data, and full-area mandatory constraint clauses in the regulations. It has no permission to access block-level detailed geological data and pipeline parameter data. The verification permissions assigned to this full-area pipeline detection unit are compliance verification of the overall pipeline route within the full area, planning compliance verification of cross-block routes, verification of avoidance with urban legal control red lines, and boundary compliance verification of cross-ownership pipelines. For each predefined mesoscopic block, a unique corresponding block pipeline detection unit is configured. The detection units of each block are independent of each other and can only access the existing pipeline network data, geological survey data, block control detailed planning, block topography data, block structure distribution data, and block construction condition data within the corresponding block. They do not have permission to access cross-block data at the overall planning level, nor do they have permission to access pipe size data at micro-viewpoints. The verification permissions assigned to the pipeline detection unit of this block are: spatial conflict verification of various professional pipelines within the corresponding block, compliance verification of connection with the existing pipeline network, verification of maintenance channel reservation, verification of avoidance with structures within the block, geological condition compatibility verification, and compliance verification of ownership boundaries within the block. For each micro-layout point within a meso-level block, a unique corresponding pipeline detection unit is configured. Each detection unit is independent of the others and can only access pipeline parameter data, fitting size data, mandatory clauses regarding pipeline clearance, burial depth, and slope in the specifications, geological borehole data, and detailed parameter data of adjacent pipelines for the corresponding point. It has no permission to access irrelevant data at the block level or the global level. The verification permissions assigned to the pipeline detection units are: pipeline spatial clearance verification, vertical conflict verification at intersection points, spatial matching verification of fitting connections, compliance verification of burial depth and slope, and safety distance verification with adjacent obstacles.

[0047] After configuring the pipeline inspection units at each level, each inspection unit is deployed to the 3D digital base and bound to the corresponding hierarchical spatial index, so as to achieve a one-to-one correspondence between the inspection unit and the corresponding spatial area, data permission, and verification permission.

[0048] S303, each pipeline detection unit can only complete the pipeline space conflict verification within the corresponding spatial range within the permission scope of the corresponding level, and output the space conflict verification result of the corresponding level after the verification is completed.

[0049] The macro-level pipeline detection unit, based on the accessed global planning data, performs a global conflict check on the preset overall pipeline routing scheme. After verification, it outputs the macro-level spatial conflict check results, including the coordinates of the conflict points, the conflict type, the violated planning control requirements, and the scope of the conflict's impact. The pipeline detection units for each meso-level block, based on the accessed block-level data, perform block-level conflict checks on the pipeline layout scheme within the block. After verification, it outputs the meso-level spatial conflict check results, including the coordinates of the conflict points, the pipeline disciplines involved in the conflict, the conflict type, the spatial scope of the conflict, and the violated constraints. The pipeline detection units for each micro-level location, based on the accessed refined location data, perform a precise conflict check on the pipeline laying scheme. After verification, it outputs the micro-level spatial conflict check results, including the precise coordinates of the conflict points, the pipeline segments involved in the conflict, the conflict type, the deviation value, and the violated mandatory regulatory clauses.

[0050] S304, based on the spatial conflict verification results output by the pipeline detection units at each level, combines the pipeline functional attributes, ownership, and maintenance requirements to complete semantic topology analysis, identify non-geometric layout constraints between pipelines, and delineate the constraint boundaries of pipeline layout at each level by combining the geometric conflict verification results.

[0051] Semantic analysis dimensions are pre-matched to pipeline detection units at each level. At the macro level, semantic dimensions related to overall pipeline planning and cross-block ownership are matched; at the meso level, semantic dimensions related to intra-block pipeline functional adaptation and maintenance channel reservation are matched; and at the micro level, semantic dimensions related to single-segment pipeline safety spacing and fitting connection requirements are matched. Based on the spatial conflict verification results output by the corresponding pipeline detection units, and combined with the matched semantic analysis dimensions, full semantic information of the pipeline is extracted to complete pipeline semantic topology analysis. For each pipeline segment and each professional system, three core semantic information categories—functional attributes, ownership, and maintenance requirements—are extracted. Based on the extracted semantic information and combined with the spatial conflict verification results, pipeline semantic topology analysis is completed, constructing a semantic topology relationship network for the pipelines and identifying non-geometric layout constraints between pipelines.

[0052] By combining the geometric constraints output by spatial conflict verification with the non-geometric layout restrictions output by semantic topology analysis, the compliant range and the prohibited range for pipeline layout within the corresponding level are defined, forming the pipeline layout constraint boundary of the corresponding level. The pipeline layout constraint boundary is divided into two categories: hard constraint boundary and flexible constraint boundary. The hard constraint boundary is an absolute constraint that cannot be broken, while the flexible constraint boundary is a constraint that can be optimized and adjusted within the compliant range.

[0053] For the macro, meso, and micro levels, corresponding constraint boundaries are defined respectively. The pipeline layout constraint boundaries at each level are set and synchronously output to the corresponding level pipeline detection unit in the subsequent path optimization step according to the hierarchical correspondence, which is determined as the hard constraint benchmark for path optimization.

[0054] S400: Based on the aforementioned constraint boundaries, a multi-constraint optimization model is constructed. Constraints related to regulatory requirements, geological environment, economic efficiency, and construction conditions are imported into pipeline detection units at each level for hierarchical weighting. A hybrid optimization approach combining multi-objective genetic algorithms and deep reinforcement learning is used. In each generation of the genetic algorithm, the corresponding level of pipeline detection units is synchronously invoked to perform real-time conflict verification. The verification results are input into deep reinforcement learning to adjust the search direction and constraint weights, generating the optimal pipeline layout path. Based on the pipeline layout path, construction instructions are generated. Using the three-dimensional digital base as a reference, precise on-site pipeline layout is completed through RTK positioning and lidar point cloud matching. Pipeline laying parameters are collected in real-time and compared with the design model. When deviations exceed limits, an early warning is triggered and a correction plan is generated.

[0055] In this embodiment of the invention, the specific execution process of S400 is implemented through the following sub-steps: S401, based on the defined constraint boundaries of pipeline layout at each level, constructs a multi-constraint, multi-objective optimization model, imports the four major categories of core constraints into the pipeline detection units at each level, completes the hierarchical weighted configuration, and clarifies the optimization objectives and constraint conditions.

[0056] All constraints related to pipeline layout are quantified into four categories of calculable and verifiable constraints. Each category is converted into a quantifiable mathematical expression and imported into the optimization model. These constraints are: mandatory regulatory constraints, geological environment constraints, construction condition constraints, and economic constraints. Mandatory regulatory constraints are inviolable hard constraints; violations render the entire plan invalid. These four categories of constraints are then mapped to pipeline monitoring units at three levels: macro, meso, and micro. Differentiated constraint weights are assigned to different levels based on their respective control priorities, achieving layered and precise control. Based on these constraints and the layered weighted configuration, a multi-objective optimization objective function is constructed, identifying two core optimization objectives: minimizing the probability of pipeline spatial conflict and minimizing the overall cost of pipeline engineering. The core objective has a higher weight than the secondary objective.

[0057] S402 employs a hybrid optimization approach combining multi-objective genetic algorithms and deep reinforcement learning for algorithm initialization. A dual pre-verification mechanism is implemented as the entry condition for population iteration. The hybrid optimization algorithm uses a multi-objective genetic algorithm as its basic iterative framework and a deep reinforcement learning agent as the core for adjusting the iterative direction. In each generation of the genetic algorithm, the corresponding level pipeline detection unit is synchronously invoked to perform real-time conflict verification. The verification results are input into the deep reinforcement learning agent, which adjusts the search direction and constraint weights of the genetic algorithm in real time, forming a closed-loop framework for iterative optimization. Before each generation of the genetic algorithm, dual pre-verification of constraint boundary compliance and pipeline detection unit availability is triggered. This dual pre-verification is set as the sole hard condition for individual entry into the population in this iteration; only individuals that pass both verifications can enter the computation phase of this iteration.

[0058] First, perform constraint boundary compliance verification to confirm that the pipeline path corresponding to the individual in the population to be iterated fully matches the constraint boundary defined by the pipeline detection unit at the corresponding level. If the verification passes, proceed to the availability verification of the pipeline detection unit; if the verification fails, the corresponding invalid individual is removed. Then, perform pipeline detection unit availability verification to confirm that there is a matching pipeline detection unit at the path level corresponding to the individual in the population to be iterated, which can be called in real time. If the verification passes, proceed to the iterative calculation stage; if the verification fails, the corresponding invalid individual is removed.

[0059] After setting up the verification mechanism, the parameters of the hybrid optimization algorithm are initialized, including the initialization of population parameters, constraint priorities, and deep reinforcement learning parameters. The constraint priorities for iterative optimization are set, with mandatory constraints set as the highest priority unbreakable hard constraints, geological environment constraints and construction condition constraints as secondary priorities, and economic constraints as third-level priorities. The priority of each constraint level is used synchronously as the core basis for fitness value calculation. The weight allocation of optimization objectives is also set, with the lowest probability of pipeline space conflict as the core optimization objective and the lowest overall cost of pipeline engineering as the secondary optimization objective. The core optimization objective has a higher weight than the secondary optimization objective.

[0060] In each generation of population iteration, S403 calculates the fitness value of valid individuals that have passed the double pre-verification according to the preset constraint priority and optimization target weight allocation, completes the selection, crossover and mutation operations of the population, generates the next generation of population, and outputs the conflict verification results of this iteration.

[0061] For valid individuals that pass the double pre-verification, the pipeline detection unit at the corresponding level is called synchronously to complete the real-time spatial conflict verification. Macro-level individuals call the global pipeline detection unit, meso-level individuals call the pipeline detection unit of the corresponding block, and micro-level individuals call the pipeline detection unit of the corresponding point. After the verification is completed, the number of conflict points, conflict type, conflict area, and violated constraint type of the path scheme corresponding to the individual are output, and the spatial conflict probability of the individual is calculated.

[0062] Based on the preset constraint priorities and optimization objective weights, the fitness value is calculated for each valid individual. First, it is checked whether the individual violates the mandatory constraints of the norm. If there is a violation, the fitness value is directly assigned to 0, and the individual is judged as invalid and removed from the population. If the constraints are fully satisfied, the deduction is calculated by combining the violation of the secondary priority constraints. The core fitness score is calculated by combining the weights of the two major optimization objectives, and finally the final fitness value of the individual is obtained.

[0063] Based on the final fitness value of each individual, a non-dominated sorting algorithm is used to rank the individuals in the population. Combined with crowding calculation, a predetermined number of individuals with the best fitness values ​​are selected as elite individuals and retained for the next generation. Simultaneously, through crossover and mutation operations, new offspring individuals are generated based on the elite individuals. The newly generated offspring individuals must pass a double pre-verification before they can be added to the next generation. After completing this population iteration, the full conflict verification results of this iteration are output synchronously, including the total number of conflict points in the population, the path regions where conflicts are concentrated, the constraint types corresponding to the conflict points, and the average number of conflicts per individual. This result is then input into the deep reinforcement learning agent.

[0064] S404, based on the conflict verification results of each iteration, uses a deep reinforcement learning agent to dynamically adjust the search direction and constraint weights, and then substitutes the adjusted parameters into the next iteration to form a closed-loop optimization chain.

[0065] A deep reinforcement learning adjustment trigger logic is established, using the conflict verification results of each iteration as the core trigger. When the number of pipeline conflict points detected in a given iteration exceeds a preset threshold, the search direction and constraint weights are immediately adjusted. Based on the conflict verification results, path regions with concentrated conflicts are identified, narrowing the search range of the next iteration of the genetic algorithm and strengthening the global search intensity of conflict-free path regions. Based on the constraint type corresponding to the conflict points, the weight ratio of the corresponding constraint priority is simultaneously increased to ensure that the adjusted constraint weights match the risk level deployed on-site. The adjusted search direction and constraint weights are then simultaneously substituted into the dual pre-verification and fitness value calculation stages of the next iteration, forming a closed-loop adjustment chain for iterative optimization.

[0066] S405: During the iteration process, the iteration status is monitored in real time. When the preset maximum number of iterations is reached, or the fitness value of the best individual in a consecutive preset number of generations does not improve, the iteration is terminated, and the pipeline layout path scheme corresponding to the individual with the best fitness value is output.

[0067] Before outputting the optimal solution, the solution is fully reviewed and verified again through pipeline inspection units at three levels: macro, meso, and micro. This confirms that the solution fully complies with all constraint boundary requirements, and has no spatial conflicts, code violations, or ownership boundary issues. The final output pipeline layout plan includes routing plans for each professional pipeline, laying parameters, pipe fitting and ancillary facility layout information, a 3D visualization model, a compliance verification report, a cost accounting report, and a construction feasibility analysis.

[0068] S406 generates standardized and executable construction instructions based on the optimal pipeline layout plan, completing the digital handover to the on-site construction end.

[0069] The optimal pipeline layout plan is broken down according to construction section, professional type, and construction procedure, and standardized construction instructions for the corresponding section are generated. The construction instructions include the coordinate parameters of pipeline laying, pipe diameter and material, burial depth and slope, connection requirements, layout location of pipe fittings and ancillary facilities, construction procedures, quality control standards, safety protection requirements, and acceptance specifications. It also includes a three-dimensional visualization model of the corresponding road section, conflict avoidance prompts, and descriptions of key control points.

[0070] The generated construction instructions, corresponding 3D digital base model, and pipeline layout BIM model are simultaneously distributed to the on-site construction terminal. On-site construction personnel, surveyors, and supervisors can view the 3D model of the pipeline, laying parameters, and construction requirements in real time through the construction terminal, and complete spatial coordinate queries, distance measurements, and construction simulations. Using the 3D digital base as a reference, precise pipeline layout is achieved on-site through RTK positioning and LiDAR point cloud matching. An RTK reference station is set up on the construction site, and the coordinate system of the reference station is completely unified with that of the 3D digital base. Joint calibration of the measuring equipment used on-site is performed to ensure that the coordinate systems of all measuring equipment are completely unified, eliminating equipment system errors. According to the pipeline design parameters in the construction instructions, precise layout of pipeline laying points is completed on-site using an RTK mobile station. During the layout process, the measured coordinates of the mobile station are compared with the design coordinates in the 3D digital base in real time. When the deviation exceeds a preset threshold, real-time adjustment prompts are issued until the layout points completely match the design coordinates, completing the initial layout.

[0071] After the initial layout is completed, the layout area is scanned on-site using LiDAR to generate a real-time 3D point cloud model of the layout area. The point cloud model measured on-site is matched with the design model in the 3D digital base using an iterative nearest point algorithm to complete fine correction, correct minor deviations in the initial layout, and ensure that the deviation between the final layout points and the design model meets the preset accuracy requirements, thus completing accurate layout. At the same time, the coordinates of the layout points are updated synchronously to the 3D digital base.

[0072] S407 collects the actual laying parameters of the pipeline in real time during the pipeline laying process and compares them with the design model in real time. When the deviation exceeds the limit, an early warning is triggered.

[0073] Throughout the entire pipeline laying process, actual pipeline laying parameters are collected in real time through various methods. During the pipeline trench excavation stage, RTK measurement is used to collect the bottom width, depth, slope, and centerline coordinates of the trench. During the pipeline laying stage, RTK measurement and lidar scanning are used to collect the actual spatial coordinates, burial depth, slope, net distance to adjacent pipelines, and pipe fitting connection positions of the pipeline. Before pipeline backfilling, downhole endoscopes and 3D laser scanning are used to complete the parameter collection and 3D modeling of the entire pipeline section.

[0074] The actual pipeline laying parameters collected in real time are compared synchronously with the design model parameters in the 3D digital base. The actual deviation values ​​of each parameter are calculated. When the actual deviation exceeds the preset threshold, an early warning is immediately triggered through the on-site construction terminal. At the same time, the deviation point, deviation value, deviation cause, and rectification requirements are marked in the 3D model, reminding the construction personnel to suspend construction and resume work after rectification. After each section of pipeline is laid, a construction acceptance report is automatically generated based on the collected actual parameters. The report verifies whether each parameter meets the design requirements and specifications. After acceptance, the actual pipeline laying parameters, acceptance report, and on-site image data are synchronously archived to the 3D digital base.

[0075] S408 synchronously updates the actual laying parameters of the laid pipelines and the real-time working conditions data of the construction site to the three-dimensional digital base, completing the dynamic correction of the three-dimensional digital base and ensuring that the spatial data of the three-dimensional digital base is consistent with the physical state of the construction site.

[0076] Based on the corrected 3D digital base, the pipeline detection unit at the corresponding level is triggered to perform incremental spatial conflict verification. Combining the actual spatial location of the laid pipelines and the data of newly added obstacles on site, the spatial conflict verification results at the corresponding level are updated. Supplementary semantic topology analysis is completed by combining pipeline functional attributes, ownership, and maintenance requirements, and the layout constraint boundary of the unlaid pipeline segments in the corresponding level is dynamically adjusted.

[0077] Based on dynamically adjusted layout constraints, a hybrid optimization approach combining multi-objective genetic algorithms and deep reinforcement learning is employed. Path re-optimization is performed only on unlaid pipeline segments, while the path schemes for already laid pipeline segments remain unchanged. This generates a revised pipeline layout path scheme adapted to the actual working conditions at the construction site. The revised pipeline layout path scheme is then synchronously updated to the construction instructions. Subsequent precise pipeline layout and laying operations are completed based on this revised scheme, forming a closed-loop management system encompassing design, construction, revision, and optimization.

[0078] This solution was implemented in a comprehensive municipal pipeline cabling project in a newly built core area of ​​a city. The area covers municipal roads, commercial land, residential land, and public facilities land, involving the simultaneous deployment of six major professional pipelines: water supply, drainage, gas, heating, electricity, and communications. The project faced multiple constraints, including numerous legally controlled boundaries, complex existing underground structures, a wide distribution of areas with adverse geological conditions, great difficulty in coordinating multiple professional pipelines, and limited construction conditions. During project implementation, firstly, comprehensive data including geographic, geological, planning, existing pipeline network, and standard data from multiple sources were integrated through standardized preprocessing. This constructed a three-dimensional digital foundation integrating BIM and GIS, enabling digital mapping of the entire underground space. Subsequently, through a three-level hierarchical spatial index and pipeline detection unit, the four major constraints—mandatory regulations, geological environment, construction conditions, and economic efficiency—were broken down into macro, meso, and micro levels. This completed hierarchical weighting and constraint boundary delineation, transforming all constraints into quantitative indicators that can be verified in real time and dynamically adjusted during path optimization. Finally, through a hybrid optimization approach combining multi-objective genetic algorithms and deep reinforcement learning, real-time conflict verification and dynamic parameter adjustment were performed simultaneously in each iteration. Within the boundaries of multiple constraints, the globally optimal pipeline path scheme was quickly converged, achieving collaborative layout of multi-disciplinary pipelines, avoiding spatial conflicts, and simultaneously considering the compliance and construction feasibility of the scheme.

[0079] During the construction phase of this project, this solution achieved seamless data integration between the design scheme and on-site construction through a 3D digital base. Standardized construction instructions generated based on the optimal path scheme were directly issued to the on-site construction terminal. High-precision on-site layout was achieved through RTK positioning and LiDAR point cloud matching, ensuring the accurate implementation of the design scheme. During construction, pipeline laying parameters and on-site working condition data were collected in real time and updated synchronously to the 3D digital base, achieving dynamic synchronization between the digital twin model and the on-site entity. For changes in on-site working conditions discovered during construction, such as unexplored underground obstacles, changes in geological conditions, and adjustments to traffic diversion schemes, incremental spatial conflict verification and supplementary semantic topology analysis were immediately triggered. The constraint boundaries of the unlaid pipeline segments were dynamically adjusted, and a hybrid optimization algorithm was invoked to re-optimize the directional paths only for the unlaid pipeline segments. The completed pipeline segments were locked without modification throughout the process, avoiding rework and schedule delays, and achieving dynamic adaptation during the construction phase.

[0080] This project, through the implementation of this solution, achieved global optimization of pipeline routes under multiple constraints and dynamic adaptation during the construction phase. It addressed industry pain points in traditional cabling methods, such as difficulties in multi-disciplinary collaboration, frequent spatial conflicts, disconnect between design and construction, and delayed response to on-site changes. After implementation, pipeline spatial conflicts were virtually eliminated during the design phase, rework during construction was significantly reduced, construction time was significantly shortened, and overall project costs were effectively controlled. Furthermore, it established a closed-loop management system encompassing design, construction, correction, and optimization, ultimately delivering a digital archive of the pipeline's entire lifecycle.

[0081] This application discloses an intelligent municipal pipeline cabling system based on path optimization, referring to... Figure 2 ,include: The 3D base construction module 001 performs preprocessing, including format unification and cleaning correction, on the collected urban basic geography, existing pipeline network, geological survey, planning red line and pipeline layout standard data to obtain a standardized multi-source dataset. Based on the standardized multi-source dataset, a BIM parametric pipeline model and a GIS geospatial model are constructed respectively. The BIM parametric pipeline model and the GIS geospatial model are incorporated into the same coordinate system to complete spatial registration and fusion, so as to construct a 3D digital base with integrated indoor and outdoor rendering. The hierarchical conflict detection module 002, based on the three-dimensional digital base, uses hierarchical spatial indexing technology to divide the detection area into three progressively refined levels: macroscopic, mesoscopic, and microscopic. A corresponding pipeline detection unit is established for each level. Through each pipeline detection unit, the spatial conflict verification of pipelines within the corresponding level is completed. Semantic topology analysis is completed in combination with pipeline functional attributes, ownership, and maintenance requirements to delineate the constraint boundaries of pipeline layout at each level. The intelligent path optimization module 003 constructs a multi-constraint optimization model based on the aforementioned constraint boundaries. It imports the constraints of regulatory requirements, geological environment, economic efficiency, and construction conditions into the pipeline detection units at each level to complete hierarchical weighting. Through a hybrid optimization method combining multi-objective genetic algorithm and deep reinforcement learning, the corresponding pipeline detection units at each level are synchronously called to complete real-time conflict verification in each generation of the genetic algorithm population iteration. The verification results are input into deep reinforcement learning to adjust the search direction and constraint weights, thereby generating the optimal pipeline layout path scheme. The construction layout and correction module 004 generates construction instructions based on the pipeline layout path scheme. Using the three-dimensional digital base as a reference, it completes the accurate layout of the pipeline on site through RTK positioning and lidar point cloud matching. It collects pipeline laying parameters in real time and compares them with the design model. When the deviation exceeds the limit, it triggers an early warning and generates a correction scheme.

[0082] This application also discloses an electronic device, including a processor, wherein the processor runs a program of any one of the above-described intelligent municipal pipeline wiring methods based on path optimization.

[0083] This application also discloses a storage medium storing a program for the intelligent municipal pipeline wiring method based on path optimization as described in any one of the above embodiments.

[0084] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. An intelligent municipal pipeline routing method based on path optimization, characterized in that, include: The collected urban basic geography, existing pipeline network, geological survey, planning red line and pipeline layout standard data are preprocessed with format unification and cleaning correction to obtain a standardized multi-source dataset. Based on the standardized multi-source dataset, BIM parametric pipeline model and GIS geospatial model are constructed respectively. The BIM parametric pipeline model and GIS geospatial model are incorporated into the same coordinate system to complete spatial registration and fusion, so as to construct a three-dimensional digital base for integrated indoor and outdoor rendering. Based on the aforementioned three-dimensional digital base, the detection area is divided into three progressively more detailed levels: macroscopic, mesoscopic, and microscopic, using hierarchical spatial indexing technology. Corresponding pipeline detection units are established for each level. Through each pipeline detection unit, pipeline spatial conflict verification within the corresponding level is completed. Semantic topology analysis is performed in conjunction with pipeline functional attributes, ownership, and maintenance requirements to delineate the constraint boundaries for pipeline layout at each level. Based on the aforementioned constraint boundaries, a multi-constraint optimization model is constructed. Constraints such as regulatory requirements, geological environment, economic efficiency, and construction conditions are imported into pipeline detection units at each level to complete hierarchical weighting. Through a hybrid optimization method combining multi-objective genetic algorithm and deep reinforcement learning, in each generation of the genetic algorithm, the corresponding level pipeline detection unit is synchronously called to complete real-time conflict verification. The verification results are input into deep reinforcement learning to adjust the search direction and constraint weights, thereby generating the optimal pipeline layout path scheme. Based on the pipeline layout path scheme, construction instructions are generated. Using the three-dimensional digital base as a reference, the precise on-site pipeline layout is completed through RTK positioning and lidar point cloud matching. The pipeline laying parameters are collected in real time and compared with the design model. When the deviation exceeds the limit, an early warning is triggered and a correction plan is generated.

2. The intelligent municipal pipeline routing method based on path optimization according to claim 1, characterized in that, The steps for generating pipeline routing schemes using a hybrid optimization approach combining multi-objective genetic algorithms and deep reinforcement learning include: Before each generation of the genetic algorithm, a dual pre-verification of constraint boundary compliance and pipeline detection unit availability is triggered, and the dual pre-verification is set as the only hard condition for individual admission in this iteration of the population. Perform constraint boundary compliance verification to confirm that the pipeline path corresponding to the individual in the population to be iterated is completely matched with the constraint boundary defined by the pipeline detection unit at the corresponding level. If the verification passes, proceed to the availability verification of the pipeline detection unit; if the verification fails, remove the corresponding invalid individual. Perform a pipeline detection unit availability check to confirm that there is a matching pipeline detection unit that can be called in real time for the path level corresponding to the individual in the population to be iterated. If the check passes, proceed to the iterative calculation stage; if the check fails, remove the corresponding invalid individual. Only valid individuals that have completed dual pre-verification and passed both verifications can enter the computation phase of this iteration, synchronously calling the pipeline detection unit at the corresponding level to complete real-time conflict verification.

3. The intelligent municipal pipeline routing method based on path optimization according to claim 2, characterized in that, The step of iteratively calculating valid individuals that have passed the double pre-verification includes: Set the constraint priority for iterative optimization, set the mandatory constraints of the standard as the highest priority unbreakable hard constraints, set the geological environment constraints and construction condition constraints as the second priority constraints, and set the economic constraints as the third priority constraints. The priority of each level of constraints is used as the core basis for fitness value calculation. The optimization objectives are weighted, with the lowest probability of pipeline space conflict as the core optimization objective and the lowest overall cost of pipeline engineering as the secondary optimization objective. The core optimization objective has a higher weight than the secondary optimization objective. For valid individuals that pass the double pre-validation, the fitness value is calculated according to the preset constraint priority and optimization target weight allocation. A preset number of individuals with the best fitness value are retained to enter the next iteration. At the same time, the conflict validation results of this iteration are input into deep reinforcement learning. When the preset maximum number of iterations is reached, or when the fitness value of the best individual does not improve after a preset number of generations, the iteration is terminated, and the pipeline layout path scheme corresponding to the individual with the best current fitness value is output.

4. The intelligent municipal pipeline routing method based on path optimization according to claim 3, characterized in that, The steps of inputting the iterative conflict verification results into deep reinforcement learning to adjust the search direction and constraint weights include: The adjustment trigger logic of deep reinforcement learning is set, with the conflict verification result of each iteration as the core triggering basis. When the number of pipeline conflict points detected in the next iteration exceeds the preset threshold, the search direction and constraint weight are immediately adjusted. Set rules for adjusting the search direction, lock the path region with concentrated conflicts based on the conflict verification results, narrow the search range of the next generation of the genetic algorithm, and strengthen the global search power of the conflict-free path region; Set rules for adjusting constraint weights, and based on the constraint type corresponding to the conflict point, simultaneously increase the weight ratio of the corresponding constraint priority to ensure that the adjusted constraint weights match the risk level of the on-site deployment. The adjusted search direction and constraint weights are then simultaneously incorporated into the dual pre-verification and fitness value calculation stages of the next iteration, forming a closed-loop adjustment chain for iterative optimization.

5. The intelligent municipal pipeline routing method based on path optimization according to claim 1, characterized in that, The steps of using hierarchical spatial indexing technology to divide the detection area, establish pipeline detection units, and delineate constraint boundaries include: The classification criteria are set at three levels: macro, meso, and micro. The spatial scope and detection accuracy corresponding to each level are clearly defined. The macro level corresponds to the entire planning red line scope, and the detection accuracy matches the requirements of the city's overall road network and planning control. The meso level corresponds to the construction scope of the block, and the detection accuracy matches the requirements of the topography and existing pipeline distribution within the block. The micro level corresponds to the pipeline laying point scope, and the detection accuracy matches the spatial layout requirements of single pipeline segments and fittings. Each level is matched with a unique pipeline detection unit. Each pipeline detection unit is assigned verification permissions and data access permissions that match the corresponding level. Macro-level pipeline detection units can only access global planning data, meso-level pipeline detection units can only access the existing pipeline network and geological data within the corresponding block, and micro-level pipeline detection units can only access the pipeline parameters and fitting size data of the corresponding point. Each pipeline detection unit can only perform pipeline spatial conflict verification within the corresponding spatial range within the scope of its corresponding level's authority. After verification, it outputs the spatial conflict verification result of the corresponding level for subsequent semantic topology analysis.

6. The intelligent municipal pipeline routing method based on path optimization according to claim 5, characterized in that, The steps involved in performing semantic topology analysis by combining pipeline functional attributes, ownership, and maintenance requirements, and defining constraint boundaries for pipeline layout at each level, include: Pre-matching corresponding semantic analysis dimensions for pipeline detection units at each level: macro level matching the overall pipeline planning direction and cross-block ownership division semantic dimensions; meso level matching the semantic dimensions of pipeline function adaptation and maintenance channel reservation within the block; micro level matching the semantic dimensions of single pipeline safety distance and pipe fitting connection requirements. Based on the spatial conflict verification results output by the pipeline detection unit at the corresponding level, and combined with the matching semantic analysis dimension, the semantic information of pipeline functional attributes, ownership, and daily maintenance requirements is extracted to complete the pipeline semantic topology analysis and identify non-geometric layout restrictions between pipelines. By combining the spatial conflict verification results with the non-geometric layout restrictions output by semantic topology analysis, the compliant range of pipelines allowed to be laid out and the prohibited range of pipelines within the corresponding level are defined, thus forming the pipeline layout constraint boundary of the corresponding level. Constraint boundaries are set for pipeline layout at each level, and output synchronously to the corresponding pipeline detection unit in the subsequent path optimization step according to the hierarchical correspondence, which is determined as the hard constraint benchmark for path optimization.

7. The intelligent municipal pipeline routing method based on path optimization according to claim 1, characterized in that, After the steps of real-time acquisition of pipeline laying parameters and comparison with the design model, triggering an early warning and generating a correction plan when the deviation exceeds the limit, the following steps are also included: The actual laying parameters of the laid pipelines and the real-time working conditions data of the construction site are collected in real time and updated to the three-dimensional digital base in a synchronous manner to complete the dynamic correction of the three-dimensional digital base and make the spatial data of the three-dimensional digital base consistent with the physical state of the construction site. Based on the corrected 3D digital base, the pipeline detection unit at the corresponding level is triggered to perform incremental spatial conflict verification. Combining the actual spatial location of the laid pipelines and the data of newly added obstacles on site, the spatial conflict verification results at the corresponding level are updated. Combined with the pipeline functional attributes, ownership, and maintenance requirements, supplementary semantic topology analysis is completed, and the layout constraint boundary of the unlaid pipeline segment in the corresponding level is dynamically adjusted. Based on the dynamically adjusted layout constraint boundary, a hybrid optimization method combining multi-objective genetic algorithm and deep reinforcement learning is called to perform path re-optimization only on the unlaid pipeline segments, lock the path scheme corresponding to the laid pipeline segments without modification, and generate a modified pipeline layout path scheme adapted to the actual working conditions of the construction site. The revised pipeline layout plan will be updated to the construction instructions, and the subsequent accurate layout and laying of pipelines will be completed based on the revised pipeline layout plan.

8. An intelligent municipal pipeline cabling system based on path optimization, characterized in that, include: The 3D base construction module performs preprocessing, including format unification and cleaning correction, on the collected urban basic geography, existing pipeline network, geological survey, planning red line and pipeline layout specifications data to obtain a standardized multi-source dataset. Based on the standardized multi-source dataset, a BIM parametric pipeline model and a GIS geospatial model are constructed respectively. The BIM parametric pipeline model and the GIS geospatial model are incorporated into the same coordinate system to complete spatial registration and fusion, so as to construct a 3D digital base with integrated indoor and outdoor rendering. The hierarchical conflict detection module, based on the three-dimensional digital base, uses hierarchical spatial indexing technology to divide the detection area into three progressively refined levels: macroscopic, mesoscopic, and microscopic. A corresponding pipeline detection unit is established for each level. Through each pipeline detection unit, the spatial conflict verification of pipelines within the corresponding level is completed. Semantic topology analysis is completed in combination with pipeline functional attributes, ownership, and maintenance requirements to delineate the constraint boundaries of pipeline layout at each level. The intelligent path optimization module constructs a multi-constraint optimization model based on the aforementioned constraint boundaries. It imports the constraints of regulatory requirements, geological environment, economic efficiency, and construction conditions into the pipeline detection units at each level to complete hierarchical weighting. Through a hybrid optimization method combining multi-objective genetic algorithms and deep reinforcement learning, the corresponding pipeline detection units at each level are synchronously invoked in each generation of the genetic algorithm to complete real-time conflict verification. The verification results are input into deep reinforcement learning to adjust the search direction and constraint weights, thereby generating the optimal pipeline layout path scheme. The construction layout and correction module generates construction instructions based on the pipeline layout path scheme. Using the three-dimensional digital base as a reference, it completes the accurate layout of the pipeline on site through RTK positioning and lidar point cloud matching. The pipeline laying parameters are collected in real time and compared with the design model. When the deviation exceeds the limit, an early warning is triggered and a correction plan is generated.

9. An electronic device, characterized in that, Includes a processor, wherein the processor runs a program for the intelligent municipal pipeline routing method based on path optimization as described in any one of claims 1-7.

10. A storage medium, characterized in that, The program stores the intelligent municipal pipeline routing method based on path optimization as described in any one of claims 1-7.

Images