BIM-based cable bridge routing autonomous optimization visual system

Abstract

This invention discloses a BIM-based autonomous optimization visualization system for cable tray routing, belonging to the field of engineering automation technology. It includes a heterogeneous professional semantic modeling module, a path identification module, a path generation module, a path conflict evolution module, a path reconstruction module, and an output module. The system generates high-quality candidate paths by constructing a cross-professional semantic graph, identifying cable thermally sensitive areas, constructing a multi-objective path evaluation function, and using a variable-weight genetic algorithm. It introduces a dynamic compression prediction mechanism during the construction phase to improve path robustness. Through versioned path reconstruction and 3D animation evolution functions, it achieves a visualized expression of the entire path design process. Finally, it outputs structured path data and obstacle avoidance strategies that can be used for construction guidance. This invention significantly improves the intelligence level, engineering feasibility, and system integration efficiency of cable tray route planning.

Description

Technical Field

[0001] This invention relates to the field of engineering automation technology, specifically to a BIM-based autonomous optimization visualization system for cable tray routing. Background Technology

[0002] In building electromechanical engineering, cable tray systems serve as crucial physical carriers for power and information transmission. Their layout design involves not only electromechanical disciplines but also requires comprehensive consideration of spatial coordination among multiple disciplines such as building structure, HVAC, and fire protection. However, existing cable tray routing design methods generally rely on designers manually drawing paths based on 2D drawings or BIM platforms. This lack of unified path evaluation standards and multi-objective optimization mechanisms leads to lengthy wiring paths, frequent turns, and recurring spatial conflicts, severely impacting construction efficiency and subsequent operation and maintenance.

[0003] Although there are BIM-based cable tray modeling tools that can provide some assistance with 3D layout, these tools are mostly limited to "geometric visualization" and do not yet have the ability to automatically optimize path solutions. Furthermore, they cannot perform path advantage analysis based on multiple dimensions such as spatial layout constraints, construction process priority, and material cost control, resulting in the path design process still heavily relying on human experience.

[0004] Furthermore, existing cable tray system designs lack dynamic adjustment and feedback mechanisms for path schemes. Path optimization is often a one-time calculation, unable to be quickly reconfigured according to changes in project stages (such as equipment relocation or new discoveries of professional collisions). At the same time, the visualization module is decoupled from the path optimization logic, resulting in visualization results that can only display static paths and do not support synchronous feedback of path changes and risk warnings.

[0005] Therefore, there is an urgent need for a cable tray routing system that can integrate multi-disciplinary BIM data, possess multi-dimensional path constraint analysis, autonomous optimization capabilities, and an interactive visualization feedback mechanism, so as to achieve a fundamental shift in cable route design from "graphical-driven" to "intelligent decision-driven". Summary of the Invention

[0006] The purpose of this invention is to provide a BIM-based autonomous optimization visualization system for cable tray routing to address the shortcomings of the prior art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a BIM-based cable tray routing autonomous optimization visualization system, comprising:

[0008] The semantic modeling module is used to extract the geometric features and spatial constraint information of different professional components from multiple BIM sub-models and construct a semantic constraint map.

[0009] The path identification module is used to identify heat-sensitive sections of the path based on cable load density, equipment energy consumption heat flow and system power requirements, generate a dynamic hot spot distribution map, and participate in the initial path screening as one of the path feasibility judgment factors.

[0010] The path generation module constructs a multi-objective evaluation function for paths based on path length, spatial accessibility, construction complexity, material cost, and hot spot avoidance factor. It then generates multiple sets of candidate path schemes that meet multiple constraints using a variable-weight genetic algorithm.

[0011] The path conflict evolution module is used to simulate the impact of dynamic spatial compression and equipment changes on paths during the building construction phase, predict the evolution of path conflicts, and perform secondary screening of the candidate path set based on the prediction results.

[0012] The path reconstruction module is used to embed path candidate schemes into the BIM model in real time, and supports path version switching, scheme comparison, conflict prompts and path comparison animation playback.

[0013] The output module is used to automatically generate a three-dimensional coordinate sequence, construction and installation steps, a list of cable tray quantities, and a construction obstacle avoidance plan for the final selected path, and output them to the construction platform in data format.

[0014] Preferably, the semantic modeling module specifically includes:

[0015] The structural, MEP, and HVAC sub-models from the building information model are partitioned, and the parametric family definition, constraint boundary, and logical purpose attributes of each component are extracted to construct a standardized component tag set.

[0016] Based on the spatial topological relationships and construction semantic dependencies between components, a graph neural network model is used to vectorize the component nodes and establish a cross-disciplinary semantic graph structure.

[0017] We introduce a relation weighting strategy based on attention mechanism to assign dynamic weights to node pairs of high-conflict-probability paths in the graph, and construct a multidimensional semantic relation graph for path generation constraint reasoning.

[0018] By inferring the relationship paths between structural nodes and device nodes in the semantic graph, potential wiring path corridors and obstacle avoidance strategies are automatically identified and used as prior inputs for the path optimization module.

[0019] Preferably, the path recognition module specifically includes:

[0020] Obtain the operating parameters of the equipment associated with the electrical system in the BIM model, and construct the cable energy flow density distribution model in a unit space of each area by combining the equipment category, nominal power and expected load curve;

[0021] A composite thermal sensing model based on thermal conduction and heat dissipation efficiency constraints is introduced to estimate the temperature rise of candidate sections of the cable tray path in multiple dimensions, and its thermal steady-state characteristics are evaluated by combining the material thermal capacity coefficient and cooling conditions.

[0022] The dynamic thermal field sequence of the path segment is generated by using time-varying load simulation, and the region with frequent thermal anomalies is extracted based on the convolutional spatiotemporal clustering algorithm to generate a three-dimensional hot spot distribution map.

[0023] The penalty factor matrix is ​​transformed into a path cost map and used as the feasibility weight input in the initial stage of path search, automatically avoiding high-risk thermally sensitive areas.

[0024] Preferably, the path generation module specifically includes:

[0025] Based on the spatial grid extracted from the BIM model, and combined with the cable laying specifications, a candidate layout channel map of cable trays is constructed, and each candidate path unit is assigned multiple evaluation dimensions such as path length, clearance, number of corners, hot spot overlap, and number of required supports.

[0026] A multi-objective path evaluation function is constructed, in which each objective dimension is fused by linear normalization and dimension weighting to form a fitness function, and the initial weights are dynamically set according to the priority strategy of the construction stage.

[0027] We designed a diverse genetic coding structure with an adaptive mutation intensity mechanism, using a joint encoding method of path segment sequences and direction vectors;

[0028] Through multi-generational evolution and inferior solution elimination strategies, multiple sets of global non-dominated path solutions are generated, and the optimal and suboptimal paths are selected as candidate solutions based on hot spot avoidance scores and cost constraints.

[0029] Preferably, the path conflict evolution module specifically includes:

[0030] Based on the building construction plan model, the installation sequence of components at each stage, temporary occupation information of the construction area and equipment location change log are extracted to generate a time-series spatial state change model.

[0031] A spatial compression prediction network is constructed. By combining the component encroachment probability and time span of the space where the path segment is located in different construction stages, a Bayesian dynamic network model is used to evaluate the path compression risk level.

[0032] Each path in the candidate path set is mapped to a time period, and the compression coefficient and dynamic conflict probability of the path segment in each construction stage are aggregated into a conflict evolution score.

[0033] A comprehensive evaluation matrix is ​​constructed based on conflict evolution score, path hotspot score, and cost score. Candidate paths are then sorted and screened in a second time to eliminate path schemes with prominent conflict trends.

[0034] Preferably, the path reconstruction module specifically includes:

[0035] Construct a path embedding interface based on the native BIM model, and automatically generate pluggable path component families through the 3D coordinate sequence of path segments and cable tray parameter templates;

[0036] The design path version control architecture adopts a scheme management mechanism based on difference vector coding, performs version alignment and incremental storage of path changes, and supports multi-dimensional parameter comparison between any two path schemes.

[0037] Build a conflict response engine to monitor geometric interference, insufficient clearance, and device occlusion events between the current embedded path and existing components in real time, and visualize the conflict through color coding or flashing prompts.

[0038] An animation of the path transition from the initial state to the final solution is generated based on the displacement sequence and timestamp information of the path components, which is used to show the path optimization process and the construction evolution trajectory.

[0039] Preferably, the output module specifically includes:

[0040] The three-dimensional component information of the final path scheme is analyzed, and a continuous path three-dimensional coordinate sequence is generated based on the path segment node coordinates and direction vectors, and the path structure is encoded in the form of a topological linked list.

[0041] Based on the installation procedure logic and on-site operation sequence, a parameterized construction step list is generated, including the installation starting point, inter-segment connection, support point setting and terminal access, with corresponding construction sequence numbers attached.

[0042] Based on the component distribution and clearance of the areas traversed by the path segment, high interference risk areas are extracted, and a path obstacle avoidance strategy table is constructed, which includes avoidance segment identification, angle adjustment suggestions, and minimum clearance prompts.

[0043] The three-dimensional coordinate sequence, construction steps, cable tray list, and obstacle avoidance strategy are packaged into an IFC extended format or a custom JSON structure and output synchronously to the construction platform.

[0044] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0045] 1. This invention constructs a BIM semantic fusion system for cable tray layout, integrating modules such as semantic modeling of heterogeneous components, dynamic hotspot identification, multi-objective path optimization, construction phase conflict evolution prediction, and 3D path visualization reconstruction. This significantly improves the intelligence, automation, and construction adaptability of path generation. Compared to traditional manual wiring or single geometric path generation tools, this invention can quickly generate high-quality path solutions under multi-dimensional constraints such as spatial accessibility, thermal safety, and installation convenience. Furthermore, it interfaces with the construction platform in a structured data manner, significantly shortening the design cycle and improving path execution stability.

[0046] 2. This invention achieves several breakthroughs in its technical approach: It is the first to introduce component semantic graphs based on graph neural networks into cable tray path planning, forming a reasonable spatial constraint system; it combines convolutional spatiotemporal clustering algorithms to achieve proactive identification and avoidance of cable thermally sensitive areas; it introduces a Bayesian dynamic model to predict spatial compression trends during the construction phase, effectively reducing the risk of path failure; and through component difference vector encoding, version management, and dynamic animation reconstruction, it achieves full-cycle visual management of the path from design to construction. The integration of these technologies significantly enhances the system's intelligent decision-making capabilities and its engineering implementation value. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0048] Figure 1 This is a mind map of the system modules of the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] For examples, please refer to Figure 1 As shown in this embodiment, the BIM-based cable tray routing autonomous optimization visualization system includes:

[0051] The semantic modeling module is used to extract the geometric features and spatial constraint information of different professional components from multiple BIM sub-models and construct a semantic constraint map.

[0052] The path identification module is used to identify heat-sensitive sections of the path based on cable load density, equipment energy consumption heat flow and system power requirements, generate a dynamic hot spot distribution map, and participate in the initial path screening as one of the path feasibility judgment factors.

[0053] The path generation module constructs a multi-objective evaluation function for paths based on path length, spatial accessibility, construction complexity, material cost, and hot spot avoidance factor. It then generates multiple sets of candidate path schemes that meet multiple constraints using a variable-weight genetic algorithm.

[0054] The path conflict evolution module is used to simulate the impact of dynamic spatial compression and equipment changes on paths during the building construction phase, predict the evolution of path conflicts, and perform secondary screening of the candidate path set based on the prediction results.

[0055] The path reconstruction module is used to embed path candidate schemes into the BIM model in real time, and supports path version switching, scheme comparison, conflict prompts and path comparison animation playback.

[0056] The output module is used to automatically generate a three-dimensional coordinate sequence, construction and installation steps, a list of cable tray quantities, and a construction obstacle avoidance plan for the final selected path, and output them to the construction platform in data format.

[0057] The semantic modeling module included in this invention aims to address the issues of semantic heterogeneity among multi-disciplinary components, fragmented geometric information, and unclear spatial relationships in BIM models. It constructs a unified semantic data structure to provide a highly readable and constrained data foundation for subsequent cable tray route planning. The specific implementation steps are as follows:

[0058] First, the data of multiple professional sub-models in the Building Information Model (BIM) such as structure, MEP, and HVAC is processed. The system automatically identifies the component types and their respective professional affiliations in different disciplines, and extracts the parametric family definitions (such as component type and family parameters), constraint boundary information (such as shape dimensions and spatial envelope), and logical purpose attributes (such as functional purpose, electrical rating, and cooling / heating medium attributes) of each component. After processing with unified mapping rules, a standardized set of component tags with unified structure and consistent semantics is formed for subsequent semantic modeling.

[0059] After standardizing the labels, the system constructs an initial cross-disciplinary component relationship diagram based on the spatial adjacency relationships (such as connection, nesting, and coplanarity) and construction semantic dependencies (such as installation sequence and shared trench layout) between components. Subsequently, a graph neural network (GNN) model is introduced to embed and encode the component nodes, generating vector representations with dimensional constraints and context awareness. This representation not only includes the attributes of the component itself but also implicitly contains its semantic role in the architectural space and its relationships with other components.

[0060] To enhance the identification capability of high-risk path segments (such as collision-prone and spatially compressed areas), the system further introduces a relationship weighting strategy based on an attention mechanism, dynamically assigning edge weights to path-related node pairs in the graph. The system learns and trains based on the historical conflict probabilities and path success rates between components in past projects, assigning higher attention weights to node connections with high conflict probabilities, thereby forming a multi-dimensional weighted semantic graph with semantic reasoning capabilities.

[0061] Finally, the system uses a path deduction algorithm between structural and equipment component nodes in the graph to identify potential wiring corridors with good passage potential in the building space, and analyzes their spatial characteristics and connection feasibility. Simultaneously, combining high-weight edges and conflict hotspot information from the attention graph, the system automatically generates a set of path avoidance strategies, including suggested detour paths and candidate passageways. This structured information serves as prior input to the path optimization module, providing semantic support and spatial guidance for subsequent path search and selection.

[0062] The path identification module aims to identify areas in the building space that pose a thermal risk to cable tray installation, based on electrical system operating parameters, and to dynamically avoid highly heat-sensitive areas during the path planning stage. This module generates a three-dimensional hotspot distribution map for initial path screening, based on the spatial distribution of cable thermal load and considering material properties and environmental conditions. The main steps include:

[0063] The system first extracts equipment parameters related to the electrical system from the BIM model, including equipment category (such as lighting, power, fire protection, etc.), nominal power, operating load level, and periodic start-stop characteristics. Combining cable layout rules with equipment power supply, a cable energy flow density distribution model is established within a unit space of the surrounding area of ​​each path segment, that is, the time-varying distribution of current load intensity within a given unit space, providing input basis for subsequent thermal impact analysis.

[0064] The specific implementation of this model is as follows:

[0065] The area where the path to be laid out in the BIM model is divided into regular cubic spatial cell grids (e.g., one cell per 0.5m³) to facilitate load distribution calculations in a three-dimensional coordinate system. Each spatial cell is numbered and bound with its center coordinates, volume information, and the components, equipment, or cable path segments it contains.

[0066] Based on the equipment load list in the system (including equipment nominal power, peak current, operating time, circuit number, etc.) and the cable tray wiring logic, a mapping relationship between cable segments and power supply equipment is established. The system automatically identifies the equipment served by each cable path and then obtains its expected load curve (obtained from equipment type or historical project experience database).

[0067] For each cable path segment li, calculate the electrical energy transmission volume (in W or A·m) per unit time based on the operating status of the equipment it serves: Where Ii is the cable load current and Li is the path segment length. The energy flow of all path segments is projected onto the spatial cell in which they are located, and the energy flow values ​​contributed by each path segment in the same cell are summed to form the total energy flow density of the cell.

[0068] For each spatial cell Sj, its unit volume energy flux density Dj can be defined as: Dj = ;in, Vj represents the total energy flow of all path segments traversing this cell, where Vj is the cell volume. This metric is used to measure cable concentration and heat accumulation potential.

[0069] For the cable energy flow density distribution model, the system introduces a composite thermal sensing model based on heat conduction theory and heat dissipation constraints. Combining the thermal capacity coefficient, thermal conductivity and local cooling capacity of the selected cable tray material in the path segment (such as ventilation conditions, shielding density, etc.), the steady-state temperature rise and temperature distribution curve of each path segment are estimated to form a path thermal stability assessment model.

[0070] To more realistically simulate the impact of equipment operation on cable load, the system simulates the thermal field response of the path segment over time based on the time-series load data during the project's operation, generating a dynamic thermal field sequence. Subsequently, a convolutional spatiotemporal clustering algorithm is used to automatically identify areas with frequent thermal anomalies, extract the path segments that form the critical temperature rise value, and output a three-dimensional structured hot spot distribution map.

[0071] Finally, the system transforms the temperature rise intensity, duration of action, and coverage area in the heat spot map into a penalty factor matrix in the path cost map, and uses it as one of the weight input variables in the path search phase. Highly thermally sensitive areas will be assigned higher path costs, thus being automatically identified as low-priority paths during path generation, achieving proactive avoidance of thermal risks.

[0072] The path generation module is responsible for constructing a multi-objective path evaluation mechanism based on semantic modeling and thermal risk analysis, taking into account spatial, engineering, and cost constraints, and generating high-quality candidate cable tray paths using an adaptive genetic algorithm. This module specifically includes the following steps:

[0073] The system first divides the building space into a three-dimensional spatial grid based on the structural geometry and clearance area data extracted from the BIM model. Then, combining this data with cable laying specifications (such as minimum clearance and maximum turning angles), it constructs a candidate cable tray access map that conforms to the laying standards. For each passable path unit in the map, the following evaluation dimensions are assigned: path length, clearance clearance, number of path turns, overlap with hotspot sections, and the number and type of support brackets required. This step forms the basic dataset for high-dimensional path evaluation.

[0074] To address the aforementioned multi-dimensional parameters, a unified multi-objective path evaluation function is constructed. Each evaluation factor, after linear normalization, is combined using a weighted fusion method to form the path fitness function. Different weighting strategies can be defined for different stages (such as scheme design, construction drawing design, and construction preparation). The system supports dynamic adjustment of weight distribution, achieving deep coupling between path planning and actual engineering needs.

[0075] In genetic algorithms, path schemes are expressed through a joint encoding of path segment sequences and direction vectors. Path segments are represented by three-dimensional grid numbers, and direction vectors indicate the path turning trend. The encoding structure is combined with an adaptive mutation intensity mechanism, meaning the system automatically adjusts gene mutation probabilities based on the current population diversity and evolutionary progress to enhance search diversity and prevent local optima traps.

[0076] Through multiple rounds of genetic evolution, including selection, crossover, mutation, and elimination of inferior solutions, the system generates a large number of path solutions. After completing fitness evaluation, a global non-dominated solution set (Pareto front) is constructed, which is the set of optimal paths that are not suppressed by other solutions on multiple objectives. The system then selects 1-3 sets of optimal and second-best paths from this set as the final candidate solution set based on hotspot avoidance index and construction cost constraints, and inputs them into the path reconstruction and comparison module.

[0077] The path conflict evolution module simulates path compression and potential collision risks caused by component installation progress, equipment location adjustments, or temporary construction occupation during building construction, and predicts and optimizes conflict trends for candidate path solutions. This module specifically includes the following steps:

[0078] The system first extracts the construction and installation sequence information of each component in the project, the temporary area occupancy plan for each construction stage, and the dynamic change log of equipment location based on the 4D-BIM construction plan model. The above information is uniformly mapped into a three-dimensional spatial model to generate a spatial state change model with time labels, that is, the spatial form and occupancy changes of components in different time periods, which is used to describe the dynamic evolution process of the construction environment.

[0079] For the areas traversed by candidate paths, the system constructs a spatial compression prediction network. Using the probability and duration of component encroachment in different construction stages as input variables, a Bayesian Dynamic Network model is introduced to probabilistically model the degree of spatial compression faced by the path segment and predict its risk level. The model supports effective simulation of construction environments with strong time dependence and complex spatial changes.

[0080] For each path in the candidate path set, the system projects its path segments onto the spatial state model of each construction stage, calculates its compression coefficient (space clearance reduction ratio) and dynamic conflict probability at different time periods, and aggregates them to form the path's Conflict Evolution Score. This score reflects the degree of spatial disturbance experienced by the path throughout the entire construction cycle and is an important indicator for evaluating the robustness of the path.

[0081] The system constructs a multi-objective comprehensive evaluation matrix based on conflict evolution scores, path hotspot scores, and construction cost scores. A weighted fusion method is then used to perform secondary sorting and elimination of candidate paths. Paths with prominent conflict trends are judged as low-priority or unqualified paths and removed from the final candidate set to ensure that the output paths have higher spatial adaptability and execution stability during the construction phase.

[0082] The path reconstruction module is used to embed visual representation, version management, and interactive feedback of candidate path solutions within the BIM model. It supports multi-version comparison of path solutions, conflict response, and dynamic evolution display, serving as a key visual interface for path planning results to design verification and construction delivery. Specifically, it includes the following steps:

[0083] The system is based on the native interface protocol of the BIM platform to construct a path embedding channel. For each candidate path scheme, its three-dimensional coordinate sequence (including the spatial position of path nodes) and cable tray parameter template (including cable tray type, width, height, connector type, etc.) are extracted. The system automatically generates a path component family that conforms to Revit, IFC and other standards. This component family supports insertion by segment, connection by node, and rendering by attribute, realizing the pluggable three-dimensional component expression of the path in the BIM model.

[0084] To support multi-scheme comparison and historical path tracking, the system is designed with a path version control architecture, employing a difference vector coding mechanism to archive and incrementally store each path change. Path difference data includes node changes, path length variations, and hotspot avoidance index changes. Through a multi-dimensional parameter comparison interface, the system can visually display and compare the differences between any two path versions in terms of spatial orientation, risk factors, and material usage.

[0085] After the path is embedded, the system activates the conflict response engine to detect in real time whether there are geometric interferences, insufficient clearance, or obstruction of critical equipment between the embedded path and existing components. After conflict identification, the system provides immediate visual prompts of the conflict through methods such as layer annotation, color coding, or path segment flashing, and marks the problematic path segment as "needs adjustment" for user decision-making reference or to initiate the path correction process.

[0086] The system records the coordinate change sequence and change timestamp information of all path segments in each version of the route plan. Based on this data, it generates a component animation transition sequence from the initial plan to the final plan, realizing a 3D dynamic demonstration of the route optimization process. This function allows users to visually track the path evolution logic, which helps project managers to review the results of route optimization and verify decisions.

[0087] The output module is used to perform structured coding, construction procedure conversion, and platform-compatible format output of the cable tray route scheme finally selected by the route generation module, achieving seamless data integration from route design to construction execution. This module specifically includes the following steps:

[0088] The system first parses the final path schemes determined in the path candidate set, extracting the 3D component information of its path segments, including the spatial coordinates of the start and end nodes, the path segment direction vectors, and data on turning and connecting points. Based on this, a continuous path 3D coordinate sequence is generated, and a topological linked list data structure is used to encode the logical order between path segments, constructing a path data model with spatial continuity and topological traceability.

[0089] Based on the standard cable tray construction process and on-site installation logic, the system automatically generates a parameterized construction step list. This list includes the starting point identifier for the installation path, the installation sequence of each cable tray segment, the setting of inter-segment connectors, the location and type of support points, and the terminal access method, etc., and assigns a construction sequence number to each construction action, forming a structured set of construction instructions that can be used for on-site operation guidance.

[0090] The system combines existing component distribution and clearance data along the path segment to identify critical sections with high collision probability or space compression risk, and constructs a path obstacle avoidance strategy table. This table includes information such as the location of sections to be avoided, recommended detour angles or path adjustment suggestions, and minimum clearance prompts within the area, to guide construction personnel in making reasonable installations in narrow areas.

[0091] The aforementioned 3D path coordinates, construction steps, cable tray material list, and obstacle avoidance strategies are uniformly packaged into a standardized data format, which can be the IFC extended format (Industry Foundation Classes), or exported as a custom structured JSON file according to the requirements of the actual platform. The system supports automatically pushing data packages to the BIM collaboration platform, intelligent construction terminal, or material prefabrication system, realizing digital connection and task-driven docking between design results and construction platform.

[0092] Example 2: To verify the technical advantages of the BIM-based cable tray routing autonomous optimization visualization system proposed in this invention in terms of path optimization efficiency, obstacle avoidance capability, and thermal stability, a BIM model of a mixed-use project was selected as the test scenario for system function verification and comparative experiments. The experiments included:

[0093] Implementation scenario:

[0094] Building area: approximately 42,000 square meters, 10 floors above ground and 2 floors below ground;

[0095] Includes BIM sub-models for multiple disciplines such as power supply, low voltage, fire protection, air conditioning, and water supply and drainage;

[0096] Experimental route: A typical cable tray route from the underground equipment floor to the floor power distribution room was selected. The route is about 60 meters long and involves 17 corners and crosses 8 types of component areas.

[0097] Comparison System:

[0098] This invention system (System A);

[0099] Traditional BIM visual modeling + manual wiring method (B system);

[0100] Commercial plug-in automatic routing (C system, without thermal assessment and conflict prediction capabilities).

[0101] Comparison of experimental data and results:

[0102]

[0103] Results Analysis and Technical Advantages:

[0104] The path quality is significantly better than existing methods: the path generated by system A is not only the shortest, but also has zero hot spot crossings, avoiding high temperature risks; while the paths of systems B and C cannot avoid some high-risk areas.

[0105] Accurate obstacle avoidance and dynamic compression prediction: The invention’s unique “conflict evolution module” successfully predicts the possible spatial compression area during later construction, ensuring sufficient clearance for the path; while traditional systems may fail during later construction.

[0106] Efficiency has been significantly improved: the time required for manual path layout has been reduced from approximately 90 minutes to a response time of seconds, and the system automatically outputs a component list and construction guidance documents.

[0107] 3D path reconstruction and evolution visualization: System A supports full path version animation playback and comparison, which helps with engineering collaboration and scheme review.

[0108] The above embodiments fully demonstrate the comprehensive technical advantages of the system of the present invention in terms of intelligent path planning, thermal sensitivity avoidance, construction adaptability and platform compatibility, providing a highly efficient, safe and visually interactive path design method for practical engineering applications.

[0109] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A BIM-based cable tray routing autonomous optimization visualization system, characterized by: include: The semantic modeling module is used to extract the geometric features and spatial constraint information of different professional components from multiple BIM sub-models and construct a semantic constraint map. The path identification module is used to identify heat-sensitive sections of the path based on cable load density, equipment energy consumption heat flow and system power requirements, generate a dynamic hot spot distribution map, and participate in the initial path screening as one of the path feasibility judgment factors. The path generation module constructs a multi-objective evaluation function for paths based on path length, spatial accessibility, construction complexity, material cost, and hot spot avoidance factor. It then generates multiple sets of candidate path schemes that meet multiple constraints using a variable-weight genetic algorithm. The path conflict evolution module is used to simulate the impact of dynamic spatial compression and equipment changes on paths during the building construction phase, predict the evolution of path conflicts, and perform secondary screening of the candidate path set based on the prediction results. The path reconstruction module is used to embed path candidate schemes into the BIM model in real time, and supports path version switching, scheme comparison, conflict prompts and path comparison animation playback. The output module is used to automatically generate a three-dimensional coordinate sequence, construction and installation steps, cable tray quantity list and construction obstacle avoidance plan for the final selected path, and output them to the construction platform in data format. The semantic modeling module specifically includes: The structural, MEP, and HVAC sub-models from the building information model are partitioned, and the parametric family definition, constraint boundary, and logical purpose attributes of each component are extracted to construct a standardized component tag set. Based on the spatial topological relationships and construction semantic dependencies between components, a graph neural network model is used to vectorize the component nodes and establish a cross-disciplinary semantic graph structure. We introduce a relation weighting strategy based on attention mechanism to assign dynamic weights to node pairs of high-conflict-probability paths in the graph, and construct a multidimensional semantic relation graph for path generation constraint reasoning. By inferring the relationship paths between structural nodes and device nodes in the semantic graph, potential wiring path corridors and obstacle avoidance strategies are automatically identified and used as prior inputs for the path optimization module.

2. The BIM-based cable tray routing autonomous optimization visualization system according to claim 1, characterized in that: The path recognition module specifically includes: Obtain the operating parameters of the equipment associated with the electrical system in the BIM model, and construct the cable energy flow density distribution model in a unit space of each area by combining the equipment category, nominal power and expected load curve; A composite thermal sensing model based on thermal conduction and heat dissipation efficiency constraints is introduced to estimate the temperature rise of candidate sections of the cable tray path in multiple dimensions, and its thermal steady-state characteristics are evaluated by combining the material thermal capacity coefficient and cooling conditions. The dynamic thermal field sequence of the path segment is generated by using time-varying load simulation, and the region with frequent thermal anomalies is extracted based on the convolutional spatiotemporal clustering algorithm to generate a three-dimensional hot spot distribution map. The penalty factor matrix is ​​transformed into a path cost map and used as the feasibility weight input in the initial stage of path search, automatically avoiding high-risk thermally sensitive areas.

3. The BIM-based cable tray routing autonomous optimization visualization system according to claim 1, characterized in that: The path generation module specifically includes: Based on the spatial grid extracted from the BIM model, and combined with the cable laying specifications, a candidate layout channel map of cable trays is constructed, and each candidate path unit is assigned multiple evaluation dimensions such as path length, clearance, number of corners, hot spot overlap, and number of required supports. A multi-objective path evaluation function is constructed, in which each objective dimension is fused by linear normalization and dimension weighting to form a fitness function, and the initial weights are dynamically set according to the priority strategy of the construction stage. We designed a diverse genetic coding structure with an adaptive mutation intensity mechanism, using a joint encoding method of path segment sequences and direction vectors; Through multi-generational evolution and inferior solution elimination strategies, multiple sets of global non-dominated path solutions are generated, and the optimal and suboptimal paths are selected as candidate solutions based on hot spot avoidance scores and cost constraints.

4. The BIM-based cable tray routing autonomous optimization visualization system according to claim 1, characterized in that: The path conflict evolution module specifically includes: Based on the building construction plan model, the installation sequence of components at each stage, temporary occupation information of the construction area and equipment location change log are extracted to generate a time-series spatial state change model. A spatial compression prediction network is constructed. By combining the component encroachment probability and time span of the space where the path segment is located in different construction stages, a Bayesian dynamic network model is used to evaluate the path compression risk level. Each path in the candidate path set is mapped to a time period, and the compression coefficient and dynamic conflict probability of the path segment in each construction stage are aggregated into a conflict evolution score. A comprehensive evaluation matrix is ​​constructed based on conflict evolution score, path hotspot score, and cost score. Candidate paths are then sorted and screened in a second time to eliminate path schemes with prominent conflict trends.

5. The BIM-based cable tray routing autonomous optimization visualization system according to claim 1, characterized in that: The path reconstruction module specifically includes: Construct a path embedding interface based on the native BIM model, and automatically generate pluggable path component families through the 3D coordinate sequence of path segments and cable tray parameter templates; The design path version control architecture adopts a scheme management mechanism based on difference vector coding, performs version alignment and incremental storage of path changes, and supports multi-dimensional parameter comparison between any two path schemes. Build a conflict response engine to monitor geometric interference, insufficient clearance, and device occlusion events between the current embedded path and existing components in real time, and visualize the conflict through color coding or flashing prompts. An animation of the path transition from the initial state to the final solution is generated based on the displacement sequence and timestamp information of the path components, which is used to show the path optimization process and the construction evolution trajectory.

6. The BIM-based cable tray routing autonomous optimization visualization system according to claim 1, characterized in that: The output module specifically includes: The three-dimensional component information of the final path scheme is analyzed, and a continuous path three-dimensional coordinate sequence is generated based on the path segment node coordinates and direction vectors, and the path structure is encoded in the form of a topological linked list. Based on the installation procedure logic and on-site operation sequence, a parameterized construction step list is generated, including the installation starting point, inter-segment connection, support point setting and terminal access, with corresponding construction sequence numbers attached. Based on the component distribution and clearance of the areas traversed by the path segment, high interference risk areas are extracted, and a path obstacle avoidance strategy table is constructed, which includes avoidance segment identification, angle adjustment suggestions, and minimum clearance prompts. The three-dimensional coordinate sequence, construction steps, cable tray list, and obstacle avoidance strategy are packaged into an IFC extended format or a custom JSON structure and output synchronously to the construction platform.

Images