A power transformation engineering electromechanical model rapid collaboration method based on BIM technology

Abstract

The application discloses a power transformation engineering electromechanical model rapid cooperation method based on BIM technology, and belongs to the technical field of BIM technology and electric power engineering design. Standardized integrated processing is performed on acquired BIM electromechanical model data to generate a standardized electromechanical model; hierarchical collision detection rules are established based on the standardized electromechanical model, collision detection is performed, and the detection result is structured to generate structured collision data; a collision relationship graph is constructed based on the structured collision data, is decomposed according to graph connectivity to obtain a plurality of collision sub-networks, and complexity indexes of the collision sub-networks are calculated; for each collision sub-network, collision components are quantitatively scored, a comprehensive priority index is calculated, and a component adjustment priority ranking is generated according to the comprehensive priority index; a plurality of candidate adjustment schemes are generated according to the component adjustment priority ranking, a multi-objective optimization model is constructed and solved, and a Pareto optimal solution set is obtained.

Description

Technical Field

[0001] This application relates to the fields of BIM technology and power engineering design technology, and more specifically, to a rapid collaborative method for electromechanical models of substation engineering based on BIM technology. Background Technology

[0002] With the continuous expansion of power grid construction, the electromechanical systems of substation projects are becoming increasingly complex, encompassing multiple disciplines such as electrical equipment, cable trays, pipelines, HVAC, fire protection, and structural supports. These systems are highly coupled in spatial relationships. During the design and construction phases, BIM technology is typically used for 3D modeling and comprehensive layout to improve design accuracy and construction controllability. However, due to independent professional divisions, inconsistent model standards, and limited spatial resources, numerous spatial collisions and clearance conflicts easily arise between electromechanical components.

[0003] In existing technologies, collision detection primarily focuses on collision identification itself. While it can output a collision list, it lacks systematic analysis of the relationships between collisions, the scope of their impact, and the priority order for handling them. During the adjustment phase, it typically relies on manual experience to modify each collision item by item, which is inefficient and prone to triggering new chain collisions. Although some automated optimization methods introduce algorithms for solving problems, they often focus on single collisions and lack a comprehensive consideration of the overall structural complexity of the subsystem and the global coordination relationship, making it difficult to balance multiple objectives such as cost, space occupation, and construction feasibility.

[0004] Furthermore, during multiple rounds of adjustments, model updates and re-detections often employ a globally repetitive calculation method, resulting in high computational overhead and low collaborative efficiency. The integration of schemes between different collision regions also lacks an effective conflict coordination mechanism, affecting the consistency of the overall model and the feasibility of engineering implementation.

[0005] In summary, how to construct a rapid collaborative method for electromechanical models of substation engineering in a BIM environment that can characterize collision-related structures, achieve zoned collaborative optimization, and take into account global consistency has become an urgent technical problem to be solved. Summary of the Invention

[0006] To overcome a series of shortcomings in existing technologies, the purpose of this application is to provide a rapid collaborative method for electromechanical models of substation engineering based on BIM technology, which includes the following steps:

[0007] Step S1: Standardize and integrate the acquired BIM MEP model data to generate a standardized MEP model;

[0008] Step S2: Establish hierarchical collision detection rules based on the standardized electromechanical model, perform collision detection, and perform structured processing on the detection results to generate structured collision data;

[0009] Step S3: Construct a collision relationship graph based on the structured collision data, decompose it according to the graph connectivity to obtain several collision sub-networks, and calculate the complexity index of each collision sub-network.

[0010] Step S4: For each collision sub-network, the collision components are quantitatively scored, a comprehensive priority index is calculated, and the component priority adjustment sorting is generated based on the comprehensive priority index.

[0011] Step S5: Generate multiple candidate adjustment schemes according to the priority of the component adjustment, construct a multi-objective optimization model and solve it to obtain the Pareto optimal solution set.

[0012] In some embodiments, the method for constructing a collision relationship map is as follows:

[0013] Extract information about the model components involved in the collision from structured collision log data;

[0014] A node set is constructed using model components that participate in the collision as graph nodes, while model components that do not participate in the collision are not included in the node set;

[0015] Based on the collision relationships between model component pairs recorded in the structured collision log, establish graph edges for model component pairs with collision relationships;

[0016] For each collision event, the weights of the graph edges are calculated, including: normalizing the collision severity level value according to the preset maximum level value to map it to a preset value range; normalizing the collision volume or spacing violation amount according to the maximum value in the same type of collision data to map it to a preset value range; and weighting and summing the normalized severity level value and the normalized collision amount value according to the preset weight coefficient to obtain the edge weight of the corresponding graph edge, which is used to characterize the tightness of the collision relationship.

[0017] When there are multiple collision events between the same pair of model components, the edge weights corresponding to each collision event are accumulated, and the accumulated result is used as the final edge weight between the pair of model components.

[0018] After all nodes and edges are established, a collision relationship graph is generated.

[0019] Topological feature analysis is performed on the collision relationship graph to obtain topological feature parameters such as total number of nodes, total number of edges, maximum node degree, average node degree, and graph density;

[0020] The collision relationship graph is stored in the form of an adjacency matrix and an adjacency list, respectively. The adjacency matrix is ​​used for fast lookup of node connection relationships, and the adjacency list is used for traversal of node adjacency relationships.

[0021] In some embodiments, the method for calculating the complexity index of each collision subnetwork is as follows:

[0022] Obtain statistical data for each collision subnetwork, including the total number of nodes, actual number of edges, severity level of each collision event within the subnetwork, number of design disciplines involved, total number of nodes in the entire graph, and total number of disciplines in the project.

[0023] Based on the statistical data, the following indicators are calculated: node size index, connection density index, collision severity weighted index, and professional crossover index. Among them, the node size index is the ratio of the total number of nodes in the collision subnetwork to the total number of nodes in the whole graph; the connection density index is the ratio of the actual number of edges in the collision subnetwork to the number of edges in the whole graph with the corresponding number of nodes; the collision severity weighted index is the weighted average obtained after assigning preset weights to the severity levels of each collision event in the subnetwork; and the professional crossover index is the ratio of the number of design disciplines involved in the collision subnetwork to the total number of disciplines in the project.

[0024] The node size index, connection density index, collision severity index, and professional crossover index are linearly weighted and summed according to preset weight coefficients to obtain the comprehensive complexity index value.

[0025] In some embodiments, the collision components in each collision sub-network are quantitatively evaluated, and a comprehensive priority index is calculated, including the following steps:

[0026] Obtain basic information about each collision component in the collision sub-network. The basic information includes the system to which the component belongs, spatial constraints, pipeline connection complexity, movable range, engineering volume and schedule impact of adjustment, degree centrality of the component in the collision relationship graph, and regulatory compliance of the adjustment direction.

[0027] Based on the aforementioned basic information, an importance score, adjustment difficulty score, cost impact score, construction period impact score, collision correlation score, and regulatory compliance score are calculated for each collision component. The collision correlation score is normalized based on the degree centrality of the component in the collision relationship graph, and the regulatory compliance score is a binary score.

[0028] The six scoring dimensions are linearly weighted according to preset weighting coefficients to calculate the comprehensive priority index of each collision component.

[0029] In some embodiments, the method for generating multiple candidate adjustment schemes based on the component adjustment priority is as follows:

[0030] Based on the comprehensive priority index of each collision component in the collision sub-network, the components are sorted from low to high, and the components to be adjusted are determined in turn, with priority given to adjusting the components with lower comprehensive priority indices, so as to ensure that the positions of the components with higher comprehensive priority indices remain unchanged.

[0031] For each component to be adjusted, based on the used and available space model of its spatial region, a set of feasible adjustment directions that meet the minimum clearance requirement is enumerated. The set of directions includes positive and negative directions along the X, Y, and Z axes and oblique merging directions.

[0032] For each adjustment direction, a trial calculation is performed according to a preset step size to calculate the minimum feasible adjustment amount of the component and form a set of adjustment actions;

[0033] The adjustment actions of the different components are arranged and combined to generate at least ten significantly different candidate adjustment schemes, each of which includes the adjustment actions of the corresponding component.

[0034] The candidate adjustment schemes are recorded as a list of triplets containing the adjustment component code, adjustment direction vector, and adjustment amount value, and are distinguished by the difference in the selection of the main adjustment component or the difference in the adjustment direction.

[0035] In some embodiments, step S5 includes the following steps:

[0036] For all collision components in the collision subnetwork, sort them from low to high according to the comprehensive priority index, and determine the components to be adjusted in turn, so as to prioritize the adjustment of components with lower comprehensive priority index and ensure that the position of components with higher comprehensive priority remains unchanged.

[0037] For each component to be adjusted, based on the used and available space of its spatial region, the adjustment directions that meet the minimum clearance requirements are enumerated, including positive and negative X, Y, and Z axes and diagonal combination directions; the feasible minimum adjustment amount in each direction is calculated with a preset step size to form a set of adjustment actions.

[0038] Arrange and combine the adjustment actions of different components to be adjusted to generate multiple candidate adjustment schemes. Each scheme is recorded as a triplet list of the component code, adjustment direction vector and adjustment amount value.

[0039] A multi-objective optimization model is constructed based on the candidate adjustment schemes. The optimization objectives include minimizing adjustment costs, minimizing space occupation, and maximizing construction convenience. The constraints include design specification constraints and spatial function constraints.

[0040] An improved genetic algorithm is used to solve the problem. The improvements include real-number encoded chromosomes, an elite-preserving selection operator, and a directional mutation mechanism based on the spatial proximity of components.

[0041] The improved genetic algorithm generates a Pareto optimal solution set during the iterative process, with each solution satisfying the optimization objective and constraints.

[0042] In some embodiments, the rapid collaborative method for electromechanical models of substation engineering further includes the following steps:

[0043] Step S6: Select the solution with the best overall performance from the Pareto optimal solution set as the optimal solution for the corresponding collision sub-network, and perform secondary collision detection.

[0044] Step S7: When a new collision is generated by the secondary collision detection, update the collision relationship map and re-optimize the local region where the new collision is generated. Repeat the priority evaluation and optimization solution steps until no new collision is generated.

[0045] Step S8: When no new collision is generated by the secondary collision detection, check whether there is a conflict between the optimal solutions of each collision sub-network; if there is a conflict, coordinate and regenerate the adjustment scheme based on the comprehensive priority index of the conflicting components; if there is no conflict, merge the optimal solutions of each collision sub-network to form the final solution.

[0046] Step S9: Generate adjustment scripts for updating the BIM MEP model based on the final solution and execute them in batches to complete the collaborative update of the MEP model.

[0047] In some embodiments, step S6 includes the following steps:

[0048] Using all non-dominated solutions in the Pareto optimal solution set as the decision candidate set, the optimization objective function value and constraint violation amount of each non-dominated solution are extracted to construct a multi-attribute decision matrix;

[0049] Based on the decision matrix, the weights of each evaluation index are determined using the entropy weight method, the positive ideal solution and the negative ideal solution are determined according to the TOPSIS method, and the relative closeness of each non-dominated solution is calculated.

[0050] Based on the relative proximity ranking, the non-dominated solution with the highest proximity is selected as the optimal solution for the corresponding collision subnetwork.

[0051] The optimal solution is implemented by identifying the set of components that have been adjusted, and by extending the bounding box of each adjusted component outward by a preset safety margin to form an influence domain, and then determining the union of these domains as the secondary collision detection area.

[0052] Collision detection is performed only on components and their adjacent components within the secondary collision detection area. The determination is completed by combining bounding box pre-screening with precise geometric intersection calculation, and the detection results are output.

[0053] In some embodiments, step S7 includes the following steps:

[0054] When a new collision is generated by secondary collision detection, the components involved in the new collision are identified, and the components and their corresponding collision relationships are updated in the collision relationship graph to form an updated graph structure.

[0055] Using the components involved in the new collision as seed nodes, a breadth-first search is performed in the collision relationship graph starting from the seed nodes, traversing neighbor nodes whose distance does not exceed a preset hop count threshold, and determining the subgraph formed by the neighbor nodes and their connecting edges as a local re-optimization region.

[0056] While keeping the component adjustment scheme outside the local re-optimization region unchanged, only the components within the local re-optimization region are re-executed with the steps of quantitative scoring, priority ranking, and candidate scheme generation, and the local optimization solution is completed.

[0057] After local re-optimization is completed, incremental collision detection is performed with the influence domain of the new collision component as the detection range to verify whether the new collision has been eliminated.

[0058] If a new collision still exists at the same location after a preset number of local re-optimizations, the hop count threshold is increased to expand the local re-optimization area, and the process of determining the local re-optimization area, performing local re-optimization, and performing incremental collision detection is executed again until the collision is eliminated or the preset termination condition is met.

[0059] In some embodiments, step S8 includes the following steps:

[0060] After confirming that no new collisions are generated by the secondary collision detection, the optimal solution of each collision sub-network is obtained, and the adjusted spatial position information and bounding box data of the adjusted components are extracted to construct a global spatial index structure.

[0061] Based on the global spatial index structure, spatial relationship retrieval is performed on the adjusted components between different collision sub-networks to determine whether there is direct spatial overlap, insufficient spacing, or channel blockage and functional space encroachment caused by adjustment vectors. The identified conflicts are classified and recorded according to conflict type and severity to generate a conflict list.

[0062] When the conflict list is not empty, each conflict record in the conflict list is coordinated and processed. The comprehensive priority index of the components of the two conflicting parties is compared, the priority component is determined and its adjustment scheme remains unchanged. For non-priority components, after excluding restricted directions, a set of feasible adjustment directions is regenerated, and candidate scheme generation and multi-objective optimization are re-executed. When the difference in the comprehensive priority index of the two conflicting parties does not exceed a preset threshold, the one with the smaller adjustment displacement is selected as the priority component. When there is a component with no feasible adjustment scheme, it is marked as awaiting coordination review and output to the design coordination problem list. After the manual decision-making results are entered, conflict coordination continues. After the conflict coordination is completed, cross-collision sub-network conflict detection is performed again until the conflict list is empty.

[0063] When the conflict list is empty, the optimal solutions of each collision subnetwork are merged to form an overall candidate adjustment scheme;

[0064] The overall candidate adjustment scheme is sequentially subjected to global channel connectivity verification, global elevation coordination verification, and global support and hanger installability verification. When all verifications meet the preset constraints, the overall candidate adjustment scheme is determined as the final solution and the model is updated.

[0065] Compared with the prior art, this application has the following beneficial effects:

[0066] This application constructs the BIM electromechanical collision problem as a weighted collision relationship graph and decomposes it into collision subnetworks. Through a multi-objective optimization and incremental re-optimization mechanism driven by complexity assessment and component priority quantification, it achieves rapid and automated adjustment of the partitioned collaboration, adaptive iteration and global consistency of the electromechanical model of substation engineering. Attached Figure Description

[0067] Figure 1 This is a schematic diagram of a rapid collaborative method for electromechanical models of substation engineering based on BIM technology, as disclosed in an embodiment of this application. Detailed Implementation

[0068] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be described in more detail below with reference to the accompanying drawings. In the drawings, the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The described embodiments are some embodiments of this invention, but not all embodiments.

[0069] 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.

[0070] The embodiments and directional terms described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0071] like Figure 1 As shown, a rapid collaborative method for electromechanical models of substation engineering based on BIM technology includes the following steps:

[0072] Step S1: Standardize and integrate the acquired BIM MEP model data to generate a standardized MEP model;

[0073] Step S2: Establish hierarchical collision detection rules based on the standardized electromechanical model and perform collision detection. Then, perform structured processing on the detection results to generate structured collision data.

[0074] Step S3: Construct a collision relationship graph based on the structured collision data, and decompose the collision relationship graph according to graph connectivity to obtain several collision sub-networks, while calculating the complexity index of each collision sub-network.

[0075] Step S4: Quantitatively evaluate the collision components in each collision sub-network, calculate the comprehensive priority index, and generate a component adjustment priority ranking based on the comprehensive priority index.

[0076] Step S5: Generate multiple candidate adjustment schemes according to the priority of the component adjustment, construct a multi-objective optimization model and solve it to obtain the Pareto optimal solution set;

[0077] Step S6: Select the solution with the best overall performance from the Pareto optimal solution set as the optimal solution for the corresponding collision sub-network, and perform secondary collision detection.

[0078] Step S7: When a new collision is generated by the secondary collision detection, update the collision relationship map and re-optimize the local region where the new collision is generated. Repeat the priority evaluation and optimization solution steps until no new collision is generated.

[0079] Step S8: When no new collision is generated by the secondary collision detection, check whether there is a conflict between the optimal solutions of each collision sub-network; if there is a conflict, coordinate and regenerate the adjustment scheme based on the comprehensive priority index of the conflicting components; if there is no conflict, merge the optimal solutions of each collision sub-network to form the final solution.

[0080] Step S9: Generate adjustment scripts for updating the BIM MEP model based on the final solution and execute them in batches to complete the collaborative update of the MEP model.

[0081] The aforementioned method for rapid collaborative processing of substation electromechanical models based on BIM technology, as described in this application, achieves unified expression and collaborative foundation construction of multi-disciplinary model data through standardized integration processing of BIM electromechanical model data. Based on this, it establishes hierarchical collision detection rules and structures the collision results. By constructing a collision relationship graph and decomposing it into several collision sub-networks according to graph connectivity, it achieves localized expression and complex quantitative analysis of collision problems. Furthermore, it quantitatively evaluates colliding components and calculates a comprehensive priority index to form a component adjustment priority ranking to guide the subsequent collaborative optimization process. Based on the priority ranking, it generates multiple candidate adjustment schemes and obtains the Pareto index through a multi-objective optimization model. The optimal solution set is obtained by selecting the solution with the best overall performance as the optimal solution for the collision subnetwork, and the adjustment effect is verified by secondary collision detection. When a new collision occurs, the collision relationship graph is updated and the local area is re-optimized to achieve iterative resolution of the collision problem. When no new collision occurs, the conflict relationship between the optimal solutions of each collision subnetwork is detected and coordinated, and the solutions are merged to form the final solution in the absence of conflict. Finally, a model adjustment script is generated based on the final solution and executed in batches to achieve automated collaborative updating of the electromechanical model, thereby improving the efficiency of collision handling of electromechanical models in substation engineering, reducing manual coordination costs, and improving the accuracy and consistency of model collaborative design.

[0082] In some embodiments, the method for standardizing and integrating the electromechanical model data is as follows:

[0083] The electromechanical model data is processed by coordinate system one. A global unified coordinate system is established with the positive X-axis of the substation building grid as the reference direction. The local coordinate system of each electromechanical model is mapped to the global unified coordinate system through an affine transformation matrix. At the same time, the translation vector and rotation angle parameters during the mapping process are recorded.

[0084] The elevation benchmark of the electromechanical model data is unified. The absolute elevation corresponding to the ±0.000 elevation is used as the benchmark to convert the relative elevation of each electromechanical model into an absolute elevation.

[0085] The component coding rules of the electromechanical model data are uniformly processed. The components are uniquely coded and assigned values ​​according to the coding system of "professional code-system number-floor number-component type-serial number", and the correspondence between component codes and original design drawing numbers is established.

[0086] The standardized integration processing method for electromechanical model data described in this application achieves standardized integration of multi-disciplinary model data through unified coordinate system, unified elevation benchmark, and unified component coding rules. Specifically, a globally unified coordinate system is established with the positive X-axis of the substation building grid. Affine transformation is used to map each local coordinate to the global coordinate and record translation and rotation parameters. The absolute elevation corresponding to ±0.000 is used as the benchmark to uniformly convert the relative elevation of each model to the absolute elevation. Unique coding is performed according to the rule of "professional code-system number-floor number-component type-serial number", and a correspondence between the code and the original drawing number is established. This achieves consistent model spatial benchmark, unified elevation, and traceable component management, providing a standardized data foundation for subsequent collaborative analysis and optimization.

[0087] In some embodiments, the method for establishing hierarchical collision detection rules is as follows:

[0088] Rules are set for the spatial overlap relationship between model components with three-dimensional geometric entities in the BIM MEP model, hard collision detection rules are established, and a threshold for the intersection volume of three-dimensional geometric entities is set; when the intersection volume between two model components is greater than the threshold, it is determined to be a hard collision.

[0089] Rules are set for the net space required for the installation, operation and maintenance of the model components, and soft collision detection rules are established; protection spacing parameters are set for different types of model components according to electrical engineering design specifications, and when the actual distance between model components is less than the corresponding protection spacing parameter, it is judged as a soft collision.

[0090] Rules are set for the work space required for construction procedures and cross-operations, and spacing collision detection rules are established; when the arrangement of model components results in insufficient construction work space or passage space, it is judged as spacing collision.

[0091] A hierarchical execution mechanism is established for the hard collision detection rules, soft collision detection rules, and spacing collision detection rules, setting the hard collision detection rules to be executed first, while the soft collision detection rules and spacing collision detection rules are executed in parallel.

[0092] The collision results obtained from different rules are classified and labeled to form a hierarchical collision detection rule system.

[0093] The hierarchical collision detection rule establishment method described in this application achieves hierarchical identification and classification management of collision problems by setting rules based on the spatial relationships and operational requirements between model components. Specifically, a hard collision detection rule is established by setting a threshold for the intersection volume of three-dimensional geometric entities; when the intersection volume of components exceeds the threshold, it is determined to be a hard collision. A soft collision detection rule is established by setting a protection spacing parameter based on the net space requirements for installation, operation, and maintenance; when the spacing between components is less than the protection spacing, it is determined to be a soft collision. A spacing collision detection rule is established based on the construction sequence and the space requirements for cross-operations; when the working or passage space is insufficient, it is determined to be a spacing collision. On this basis, a hierarchical execution mechanism is established, prioritizing the execution of hard collision detection rules and executing soft collision and spacing collision detection rules in parallel, and classifying and labeling the detection results, thereby forming a hierarchical collision detection rule system and improving the accuracy and efficiency of collision identification.

[0094] In some embodiments, the method for generating structured collision data is as follows:

[0095] Collision detection is performed on the electromechanical model to obtain collision events;

[0096] The collision event is determined according to the preset collision determination rules, and the corresponding collision volume or spacing violation is calculated.

[0097] The collision events are classified into severity levels based on their impact on engineering safety, functional implementation, and construction feasibility.

[0098] For each collision event, a structured collision record is established, which includes the collision number, collision type, collision severity level, a list of involved component codes, the three-dimensional bounding box coordinates of the collision location, the collision volume or spacing violation, the relevant design specialty code, the first detection timestamp, and the associated design specification clause number.

[0099] The structured collision records are stored in a database, and a retrieval index is established based on collision type, severity level, involved profession, and spatial region to form structured collision data.

[0100] The structured collision data generation method described in this application achieves unified recording and standardized management of collision information by performing collision detection on an electromechanical model and acquiring collision events. Based on this, it determines the type of collision events according to preset collision judgment rules and calculates the collision volume or spacing violation. Simultaneously, it classifies the severity level according to the impact of collision events on engineering safety, functional implementation, and construction feasibility. A structured collision record is established for each collision event, including the collision number, collision type, severity level, a list of involved component codes, the three-dimensional bounding box coordinates of the collision location, the collision volume or spacing violation, the relevant professional code, the first detection timestamp, and the associated design specification clause number. The structured collision record is stored in a database, and a retrieval index is established based on collision type, severity level, involved professional field, and spatial region, thereby forming searchable and analyzable structured collision data and improving the efficiency of collision information management and collaborative processing.

[0101] In some embodiments, the method for constructing a collision relationship map is as follows:

[0102] Extract information about the model components involved in the collision from structured collision log data;

[0103] A node set is constructed using model components that participate in the collision as graph nodes, while model components that do not participate in the collision are not included in the node set;

[0104] Based on the collision relationships between model component pairs recorded in the structured collision log, establish graph edges for model component pairs with collision relationships;

[0105] For each collision event, the weights of the graph edges are calculated, including: normalizing the collision severity level value according to the preset maximum level value to map it to a preset value range; normalizing the collision volume or spacing violation amount according to the maximum value in the same type of collision data to map it to a preset value range; and weighting and summing the normalized severity level value and the normalized collision amount value according to the preset weight coefficient to obtain the edge weight of the corresponding graph edge, which is used to characterize the tightness of the collision relationship.

[0106] When there are multiple collision events between the same pair of model components, the edge weights corresponding to each collision event are accumulated, and the accumulated result is used as the final edge weight between the pair of model components.

[0107] After all nodes and edges are established, a collision relationship graph is generated.

[0108] Topological feature analysis is performed on the collision relationship graph to obtain topological feature parameters such as total number of nodes, total number of edges, maximum node degree, average node degree, and graph density;

[0109] The collision relationship graph is stored in the form of an adjacency matrix and an adjacency list, respectively. The adjacency matrix is ​​used for fast lookup of node connection relationships, and the adjacency list is used for traversal of node adjacency relationships.

[0110] The collision relationship graph construction method described in this application extracts information about the model components involved in the collision from structured collision records, and uses the model components involved in the collision as graph nodes and the collision relationships between components as graph edges to achieve a graph representation of collision relationships. Based on this, weights are calculated for the graph edges corresponding to each collision event. The collision severity level and the collision volume or spacing violation are normalized separately, and the edge weights are obtained by weighted summation according to preset weight coefficients. When multiple collision events exist for the same model component, the edge weights are accumulated to characterize the tightness of the collision relationship. After establishing the nodes and edges, a collision relationship graph is generated, and topological feature analysis is performed on the graph to obtain characteristic parameters such as the number of nodes, the number of edges, and the graph density. Simultaneously, the collision relationship graph is stored in both adjacency matrix and adjacency list formats, thereby achieving efficient querying and traversal analysis of collision relationships and improving the organization and collaborative processing capabilities of collision problems.

[0111] In some embodiments, the collision relationship graph is decomposed according to graph connectivity to obtain several collision sub-networks, including the following steps:

[0112] Data integrity is verified for each node in the collision relationship graph. The component code corresponding to the node is checked to see if it can be uniquely matched in the standardized electromechanical model. Nodes that cannot be uniquely matched are marked as abnormal and an abnormal log is generated. The abnormal nodes are excluded from subsequent decomposition processing.

[0113] Set the node that passes the integrity check as an unaccessed node;

[0114] Select a node from the unvisited nodes as the starting node, and use a depth-first search to traverse all nodes connected to it, marking visited nodes as visited during the traversal; take the set of nodes covered by a complete traversal and its corresponding set of edges as a collision subnetwork, and repeat the above traversal process until all valid nodes have been visited, thus obtaining several collision subnetworks.

[0115] Sort each collision subnetwork according to the number of nodes;

[0116] Determine whether the number of nodes in each collision subnetwork exceeds a preset upper limit threshold. For collision subnetworks that exceed the preset upper limit threshold, use a splitting method based on edge betweenness centrality to subdivide them. By iteratively calculating the edge betweenness of each edge and deleting the edge with the highest edge betweenness, the superscale subnetwork is divided into several subnetworks with the number of nodes not exceeding the preset upper limit threshold.

[0117] The collision graph decomposition method described in this application ensures the validity of the decomposed data by verifying the data integrity of the graph nodes and excluding abnormal nodes that cannot uniquely match the component codes. Based on this, valid nodes are set to an unvisited state. A starting node is selected from the unvisited nodes, and a depth-first search is used to traverse connected nodes and mark their access status. The set of nodes and edges covered by one traversal is considered as a collision subnetwork, and this traversal is repeated until all valid nodes are visited, resulting in several collision subnetworks. Subsequently, each collision subnetwork is sorted according to the number of nodes, and subnetworks with more than a preset upper limit threshold are further subdivided using a splitting method based on edge betweenness centrality. By iteratively deleting the edges with the highest betweenness, the oversized subnetwork is divided into several subnetworks that meet the size requirements, thereby achieving modular partitioning of the collision problem and improving the efficiency of subsequent collaborative optimization.

[0118] In some embodiments, the method for calculating the complexity index of each collision subnetwork is as follows:

[0119] Obtain statistical data for each collision subnetwork, including the total number of nodes, actual number of edges, severity level of each collision event within the subnetwork, number of design disciplines involved, total number of nodes in the entire graph, and total number of disciplines in the project.

[0120] Based on the statistical data, the following indicators are calculated: node size index, connection density index, collision severity weighted index, and professional crossover index. Among them, the node size index is the ratio of the total number of nodes in the collision subnetwork to the total number of nodes in the whole graph; the connection density index is the ratio of the actual number of edges in the collision subnetwork to the number of edges in the whole graph with the corresponding number of nodes; the collision severity weighted index is the weighted average obtained after assigning preset weights to the severity levels of each collision event in the subnetwork; and the professional crossover index is the ratio of the number of design disciplines involved in the collision subnetwork to the total number of disciplines in the project.

[0121] The node size index, connection density index, collision severity index, and professional crossover index are linearly weighted and summed according to preset weight coefficients to obtain the comprehensive complexity index value.

[0122] The method for calculating the complexity index of the collision subnetwork described in this application achieves a quantitative representation of the scale and coupling degree of the collision problem by obtaining statistical data of each collision subnetwork. The statistical data includes the total number of nodes in the subnetwork, the actual number of edges, the severity level of the collision event, the number of design disciplines involved, the total number of nodes in the entire graph, and the total number of disciplines in the project. Based on this, the node scale index, the connection density index, the collision severity weighted index, and the discipline crossover index are calculated respectively, and then linearly weighted and summed according to preset weight coefficients to obtain the comprehensive complexity index value. This achieves a unified measurement of the complexity of the collision subnetwork and provides a quantitative basis for subsequent collaborative optimization and priority decision-making.

[0123] In some embodiments, the collision components in each collision sub-network are quantitatively evaluated, and a comprehensive priority index is calculated, including the following steps:

[0124] Obtain basic information about each collision component in the collision sub-network. The basic information includes the system to which the component belongs, spatial constraints, pipeline connection complexity, movable range, engineering volume and schedule impact of adjustment, degree centrality of the component in the collision relationship graph, and regulatory compliance of the adjustment direction.

[0125] Based on the aforementioned basic information, an importance score, adjustment difficulty score, cost impact score, construction period impact score, collision correlation score, and regulatory compliance score are calculated for each collision component. The collision correlation score is normalized based on the degree centrality of the component in the collision relationship graph, and the regulatory compliance score is a binary score.

[0126] The six scoring dimensions are linearly weighted according to preset weighting coefficients to calculate the comprehensive priority index of each collision component.

[0127] The quantitative evaluation method for collision components described in this application achieves quantitative analysis of component adjustment priorities by acquiring basic information about collision components in each collision sub-network. This basic information includes the component's system, spatial constraints, pipeline connection complexity, movable range, engineering workload and schedule impact involved in the adjustment, degree centrality of the component in the collision relationship graph, and regulatory compliance of the adjustment direction. Based on this, the method calculates the component's importance score, adjustment difficulty score, cost impact score, schedule impact score, collision correlation score, and regulatory compliance score. The collision correlation score is obtained by degree centrality normalization, and the regulatory compliance score uses a binary evaluation. A comprehensive priority index is obtained by linearly weighting and summing each score dimension according to preset weight coefficients, thereby quantitatively determining the order of collision component adjustments and improving the rationality and efficiency of collaborative optimization decisions.

[0128] In some embodiments, the method for generating multiple candidate adjustment schemes based on the component adjustment priority is as follows:

[0129] Based on the comprehensive priority index of each collision component in the collision sub-network, the components are sorted from low to high, and the components to be adjusted are determined in turn, with priority given to adjusting the components with lower comprehensive priority indices, so as to ensure that the positions of the components with higher comprehensive priority indices remain unchanged.

[0130] For each component to be adjusted, based on the used and available space model of its spatial region, a set of feasible adjustment directions that meet the minimum clearance requirement is enumerated. The set of directions includes positive and negative directions along the X, Y, and Z axes and oblique merging directions.

[0131] For each adjustment direction, a trial calculation is performed according to a preset step size, with a horizontal step size of not less than 10mm and a vertical step size of not less than 5mm, in order to calculate the minimum feasible adjustment amount of the component and form a set of adjustment actions.

[0132] The adjustment actions of the different components are arranged and combined to generate at least ten significantly different candidate adjustment schemes, each of which includes the adjustment actions of the corresponding component.

[0133] The candidate adjustment schemes are recorded as a list of triplets containing adjustment component codes, adjustment direction vectors, and adjustment magnitude values, and are distinguished by differences in the selection of major adjustment components or differences in adjustment directions, in order to avoid redundant schemes and ensure the diversity of Pareto front exploration.

[0134] The candidate adjustment scheme generation method described in this application determines the adjustment order of components based on the comprehensive priority index of the collision components, thereby prioritizing the adjustment of low-priority components while maintaining the stable position of high-priority components. Based on this, for each component to be adjusted, feasible adjustment directions that meet the minimum clearance requirement are enumerated using the space used and available space models. Trial calculations are performed according to a preset step size to determine the minimum feasible adjustment amount, forming a set of component adjustment actions. Furthermore, the adjustment actions of different components are arranged and combined to generate multiple sets of significantly different candidate adjustment schemes, which are recorded in a triplet list containing the component code, adjustment direction vector, and adjustment amount value. These schemes are distinguished by differences in the adjusted components or adjustment directions, thereby avoiding redundant schemes and ensuring the diversity of subsequent multi-objective optimization searches.

[0135] In some embodiments, step S5 includes the following steps:

[0136] For all collision components in the collision subnetwork, sort them from low to high according to the comprehensive priority index, and determine the components to be adjusted in turn, so as to prioritize the adjustment of components with lower comprehensive priority index and ensure that the position of components with higher comprehensive priority remains unchanged.

[0137] For each component to be adjusted, based on the used and available space of its spatial region, the adjustment directions that meet the minimum clearance requirements are enumerated, including positive and negative X, Y, and Z axes and diagonal combination directions; the feasible minimum adjustment amount in each direction is calculated with a preset step size to form a set of adjustment actions.

[0138] Arrange and combine the adjustment actions of different components to be adjusted to generate multiple candidate adjustment schemes. Each scheme is recorded as a triplet list of the component code, adjustment direction vector and adjustment amount value.

[0139] A multi-objective optimization model is constructed based on the candidate adjustment schemes. The optimization objectives include minimizing adjustment costs, minimizing space occupation, and maximizing construction convenience. The constraints include design specification constraints and spatial function constraints, which are embedded in the model in the form of penalty functions.

[0140] An improved genetic algorithm is used to solve the problem. The improvements include real-number encoded chromosomes, an elite-preserving selection operator, and a directional mutation mechanism based on the spatial proximity of components.

[0141] The improved genetic algorithm generates a Pareto optimal solution set during the iterative process, with each solution satisfying the optimization objective and constraints.

[0142] The iteration termination condition is that the improvement of the Pareto front hypervolume index is less than 0.1% for 20 consecutive generations, or the total number of iterations reaches the preset upper limit. When either condition is met, the iteration stops and the Pareto optimal solution set is output.

[0143] In step S5 of this application, components in the collision sub-network are sorted from low to high according to their comprehensive priority index. Low-priority components are adjusted first while keeping the positions of high-priority components unchanged. For the component to be adjusted, the adjustment direction that satisfies the minimum clearance is enumerated by combining the used and available space. An adjustment action set is formed by trial calculation according to a preset step size. Multiple candidate adjustment schemes are generated by permutation and combination, and represented by a triplet list of component code, adjustment direction vector, and adjustment amount value. On this basis, a multi-objective optimization model is constructed with the objectives of minimizing adjustment cost, minimizing space occupation, and maximizing construction convenience, and with design specifications and spatial functional constraints as conditions. An improved genetic algorithm with real number encoding, elite retention, and directed mutation mechanism is used to solve the problem. Pareto optimal solution set is generated iteratively. The calculation is terminated and the result is output when the Pareto front hypervolume improves by less than 0.1% for 20 consecutive generations or reaches the iteration limit.

[0144] In some embodiments, step S6 includes the following steps:

[0145] Using all non-dominated solutions in the Pareto optimal solution set as the decision candidate set, the optimization objective function value and constraint violation amount of each non-dominated solution are extracted to construct a multi-attribute decision matrix;

[0146] Based on the decision matrix, the weights of each evaluation index are determined using the entropy weight method, the positive ideal solution and the negative ideal solution are determined according to the TOPSIS method, and the relative closeness of each non-dominated solution is calculated.

[0147] Based on the relative proximity ranking, the non-dominated solution with the highest proximity is selected as the optimal solution for the corresponding collision subnetwork; when the proximity difference does not exceed the preset threshold, the solution with fewer adjustment components is selected, and the evaluation data is archived.

[0148] The optimal solution is executed to identify the set of components that have been adjusted; the influence domain is formed by extending the bounding box of each adjusted component outward by a preset safety margin, and the union of these domains is determined as the secondary collision detection area;

[0149] Collision detection is performed only on components and their adjacent components within the secondary collision detection area. The determination is completed by combining bounding box pre-screening with precise geometric intersection calculation, and the detection results are output.

[0150] In step S6 of this application, the non-dominated solutions in the Pareto optimal solution set are constructed into a multi-attribute decision matrix. The objective function value and constraint violation amount of each scheme are extracted. The entropy weight method is used to determine the weight of the evaluation index and the TOPSIS method is combined to calculate the relative closeness of each scheme. The optimal solution is selected according to the closeness ranking. When the closeness difference does not exceed the preset threshold, the scheme with fewer adjusted components is selected first and the data is archived. After the optimal solution is executed, the set of adjusted components is identified. The influence domain is formed by expanding the safety margin outward with the adjusted bounding box and the secondary collision detection area is determined. Collision detection is performed only on the components and their adjacent components in this area. The judgment is completed by pre-screening of the bounding box and accurate geometric intersection calculation, thereby improving the verification efficiency and reducing the computational cost.

[0151] In some embodiments, step S7 includes the following steps:

[0152] When a new collision is generated by secondary collision detection, the components involved in the new collision are identified, and the components and their corresponding collision relationships are updated in the collision relationship graph to form an updated graph structure.

[0153] Using the components involved in the new collision as seed nodes, a breadth-first search is performed in the collision relationship graph starting from the seed nodes, traversing neighbor nodes whose distance does not exceed a preset hop count threshold, and determining the subgraph formed by the neighbor nodes and their connecting edges as a local re-optimization region.

[0154] While keeping the component adjustment scheme outside the local re-optimization region unchanged, only the components within the local re-optimization region are re-executed with the steps of quantitative scoring, priority ranking, and candidate scheme generation, and the local optimization solution is completed.

[0155] After local re-optimization is completed, incremental collision detection is performed with the influence domain of the new collision component as the detection range to verify whether the new collision has been eliminated.

[0156] If a new collision still exists at the same location after a preset number of local re-optimizations, the hop count threshold is increased to expand the local re-optimization area, and the process of determining the local re-optimization area, performing local re-optimization, and performing incremental collision detection is executed again until the collision is eliminated or the preset termination condition is met.

[0157] In step S7 of this application, when a new collision is generated during secondary collision detection, relevant components are identified and the collision relationship graph is updated. Using the new collision component as a seed node, a breadth-first search is performed in the graph to traverse neighbor nodes whose distance does not exceed a preset hop count threshold and to determine the local re-optimization region. While keeping the component adjustment scheme outside the region unchanged, only the components within the local region are re-quantified, prioritized, and candidate schemes are generated to complete the local optimization solution. Subsequently, incremental collision detection is performed with the influence domain of the new collision component as the range to verify the collision elimination. When collisions still exist after multiple consecutive local re-optimizations, the hop count threshold is increased to expand the re-optimization range and the above process is repeated until the collision is eliminated or the termination condition is met.

[0158] In some embodiments, step S8 includes the following steps:

[0159] After confirming that no new collisions are generated by the secondary collision detection, the optimal solution of each collision sub-network is obtained, and the adjusted spatial position information and bounding box data of the adjusted components are extracted to construct a global spatial index structure.

[0160] Based on the global spatial index structure, spatial relationship retrieval is performed on the adjusted components between different collision sub-networks to determine whether there is direct spatial overlap, insufficient spacing, or channel blockage and functional space encroachment caused by adjustment vectors. The identified conflicts are classified and recorded according to conflict type and severity to generate a conflict list.

[0161] When the conflict list is not empty, each conflict record in the conflict list is coordinated and processed. The comprehensive priority index of the components of the two conflicting parties is compared, the priority component is determined and its adjustment scheme remains unchanged. For non-priority components, after excluding restricted directions, a set of feasible adjustment directions is regenerated, and candidate scheme generation and multi-objective optimization are re-executed. When the difference in the comprehensive priority index of the two conflicting parties does not exceed a preset threshold, the one with the smaller adjustment displacement is selected as the priority component. When there is a component with no feasible adjustment scheme, it is marked as awaiting coordination review and output to the design coordination problem list. After the manual decision-making results are entered, conflict coordination continues. After the conflict coordination is completed, cross-collision sub-network conflict detection is performed again until the conflict list is empty.

[0162] When the conflict list is empty, the optimal solutions of each collision subnetwork are merged to form an overall candidate adjustment scheme;

[0163] The overall candidate adjustment scheme is sequentially subjected to global channel connectivity verification, global elevation coordination verification, and global support and hanger installability verification. When all verifications meet the preset constraints, the overall candidate adjustment scheme is determined as the final solution and the model is updated.

[0164] In step S8 of this application, after no new collisions are generated during secondary collision detection, the optimal solutions for each collision sub-network are obtained. The spatial position and bounding box data of the adjusted components are extracted and a global spatial index is established to achieve cross-sub-network spatial relationship retrieval. Based on this, conflicts such as spatial overlap, insufficient spacing, and encroachment on channel or functional space between different sub-networks are identified and a conflict list is generated. When a conflict exists, the priority component is determined according to the component comprehensive priority index and its adjustment scheme remains unchanged. For non-priority components, feasible adjustment directions are regenerated and optimization solutions are performed. When there is no feasible solution, manual coordination is performed until the conflict list is empty. Subsequently, the optimal solutions of each sub-network are merged to form an overall adjustment scheme, and global channel connectivity, elevation coordination, and support and hanger installability verification are performed in sequence. When all constraints are met, it is determined as the final solution and the model is updated.

[0165] In some embodiments, the method for generating the adjusted script is as follows:

[0166] Based on the application programming interface (API) specification of the BIM electromechanical model operating environment, determine the scripting language and interface calling rules that are compatible with it, and convert the final solution into the corresponding interface calling instructions;

[0167] In accordance with the principle of atomicity of operations, each component's single adjustment operation is encapsulated into an indivisible atomic operation unit;

[0168] Construct a spatial dependency model among the components within the collision sub-network, and determine the execution order of each atomic operation unit based on the spatial dependency model, so that the atomic operation units corresponding to the components that are depended by other components are prioritized.

[0169] Arrange the atomic operation units according to the determined execution order to form an ordered execution sequence of component adjustment operations within the collision subnetwork, and generate a corresponding script file based on the ordered execution sequence.

[0170] The above-mentioned method for generating adjustment scripts in this application determines the script language and interface calling rules based on the application programming interface specifications of the BIM MEP model operating environment, and converts the final solution into corresponding interface calling instructions. On this basis, according to the principle of atomicity of operation, the adjustment operations of each component are encapsulated into atomic operation units, and a spatial dependency model between components is constructed to determine the execution order, so that the dependent components are executed first. Then, the atomic operation units are arranged according to the execution order to generate an ordered execution sequence, and a script file is formed accordingly, thereby realizing the automation and ordered execution of MEP model adjustment operations.

[0171] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A rapid collaborative method for electromechanical models of substation engineering based on BIM technology, characterized in that, Includes the following steps: Step S1: Standardize and integrate the acquired BIM MEP model data to generate a standardized MEP model; Step S2: Establish hierarchical collision detection rules based on the standardized electromechanical model, perform collision detection, and perform structured processing on the detection results to generate structured collision data; Step S3: Construct a collision relationship graph based on the structured collision data, decompose it according to the graph connectivity to obtain several collision sub-networks, and calculate the complexity index of each collision sub-network. Step S4: For each collision sub-network, the collision components are quantitatively scored, a comprehensive priority index is calculated, and the component priority adjustment sorting is generated based on the comprehensive priority index. Step S5: Generate multiple candidate adjustment schemes according to the priority of the component adjustment, construct a multi-objective optimization model and solve it to obtain the Pareto optimal solution set.

2. The rapid collaborative method for electromechanical models of substation engineering according to claim 1, characterized in that, The method for constructing a collision relationship graph is as follows: Extract information about the model components involved in the collision from structured collision log data; A node set is constructed using model components that participate in the collision as graph nodes, while model components that do not participate in the collision are not included in the node set; Based on the collision relationships between model component pairs recorded in the structured collision log, establish graph edges for model component pairs with collision relationships; For each collision event, the weights of the graph edges are calculated, including: normalizing the collision severity level value according to the preset maximum level value to map it to a preset value range; normalizing the collision volume or spacing violation amount according to the maximum value in the same type of collision data to map it to a preset value range; and weighting and summing the normalized severity level value and the normalized collision amount value according to the preset weight coefficient to obtain the edge weight of the corresponding graph edge, which is used to characterize the tightness of the collision relationship. When there are multiple collision events between the same pair of model components, the edge weights corresponding to each collision event are accumulated, and the accumulated result is used as the final edge weight between the pair of model components. After all nodes and edges are established, a collision relationship graph is generated. Perform topological feature analysis on the collision relationship graph to obtain topological feature parameters such as total number of nodes, total number of edges, maximum node degree, average node degree, and graph density; The collision relationship graph is stored in the form of an adjacency matrix and an adjacency list, respectively. The adjacency matrix is ​​used for fast lookup of node connection relationships, and the adjacency list is used for traversal of node adjacency relationships.

3. The rapid collaborative method for electromechanical models of substation engineering according to claim 1, characterized in that, The method for calculating the complexity index of each collision subnetwork is as follows: Obtain statistical data for each collision subnetwork, including the total number of nodes, actual number of edges, severity level of each collision event within the subnetwork, number of design disciplines involved, total number of nodes in the entire graph, and total number of disciplines in the project. Based on the statistical data, the following indicators are calculated: node size index, connection density index, collision severity weighted index, and professional crossover index. Among them, the node size index is the ratio of the total number of nodes in the collision subnetwork to the total number of nodes in the whole graph; the connection density index is the ratio of the actual number of edges in the collision subnetwork to the number of edges in the whole graph with the corresponding number of nodes; the collision severity weighted index is the weighted average obtained after assigning preset weights to the severity levels of each collision event in the subnetwork; and the professional crossover index is the ratio of the number of design disciplines involved in the collision subnetwork to the total number of disciplines in the project. The node size index, connection density index, collision severity index, and professional crossover index are linearly weighted and summed according to preset weighting coefficients to obtain the comprehensive complexity index value.

4. The rapid collaborative method for electromechanical models of substation engineering according to claim 1, characterized in that, The collision components in each collision sub-network are quantitatively evaluated, and a comprehensive priority index is calculated, including the following steps: Obtain basic information about each collision component in the collision sub-network. The basic information includes the system to which the component belongs, spatial constraints, pipeline connection complexity, movable range, engineering volume and schedule impact of adjustment, degree centrality of the component in the collision relationship graph, and regulatory compliance of the adjustment direction. Based on the aforementioned basic information, an importance score, adjustment difficulty score, cost impact score, construction period impact score, collision correlation score, and regulatory compliance score are calculated for each collision component. The collision correlation score is normalized based on the degree centrality of the component in the collision relationship graph, and the regulatory compliance score is a binary score. The six scoring dimensions are linearly weighted according to preset weighting coefficients to calculate the comprehensive priority index of each collision component.

5. The rapid collaborative method for electromechanical models of substation engineering according to claim 1, characterized in that, The method for generating multiple candidate adjustment schemes based on the component adjustment priority is as follows: Based on the comprehensive priority index of each collision component in the collision sub-network, the components are sorted from low to high, and the components to be adjusted are determined in turn, with priority given to adjusting the components with lower comprehensive priority indices, so as to ensure that the positions of the components with higher comprehensive priority indices remain unchanged. For each component to be adjusted, based on the used and available space model of its spatial region, a set of feasible adjustment directions that meet the minimum clearance requirement is enumerated. The set of directions includes positive and negative directions along the X, Y, and Z axes and oblique merging directions. For each adjustment direction, a trial calculation is performed according to a preset step size to calculate the minimum feasible adjustment amount of the component and form a set of adjustment actions; The adjustment actions of the different components are arranged and combined to generate at least ten significantly different candidate adjustment schemes, each of which includes the adjustment actions of the corresponding component. The candidate adjustment schemes are recorded as a list of triplets containing the adjustment component code, adjustment direction vector, and adjustment amount value, and are distinguished by the difference in the selection of the main adjustment component or the difference in the adjustment direction.

6. The rapid collaborative method for electromechanical models of substation engineering according to claim 1, characterized in that, Step S5 includes the following steps: For all collision components in the collision subnetwork, sort them from low to high according to the comprehensive priority index, and determine the components to be adjusted in turn, so as to prioritize the adjustment of components with lower comprehensive priority index and ensure that the position of components with higher comprehensive priority remains unchanged. For each component to be adjusted, based on the used and available space of its spatial region, the adjustment directions that meet the minimum clearance requirements are enumerated, including positive and negative X, Y, and Z axes and diagonal combination directions; the feasible minimum adjustment amount in each direction is calculated with a preset step size to form a set of adjustment actions.

7. Arrange and combine the adjustment actions of different components to be adjusted to generate multiple candidate adjustment schemes. Each scheme is recorded as a triplet list of the component code, adjustment direction vector and adjustment amount value. A multi-objective optimization model is constructed based on the candidate adjustment schemes, and the optimization objectives include: The adjustment aims to minimize costs, space requirements, and ease of construction; constraints include design code constraints and spatial function constraints. An improved genetic algorithm is used to solve the problem. The improvements include real-number encoded chromosomes, an elite-preserving selection operator, and a directional mutation mechanism based on the spatial proximity of components. The improved genetic algorithm generates a Pareto optimal solution set during the iterative process, with each solution satisfying the optimization objective and constraints.

8. The rapid collaborative method for electromechanical models of substation engineering according to any one of claims 1-6, characterized in that, It also includes the following steps: Step S6: Select the solution with the best overall performance from the Pareto optimal solution set as the optimal solution for the corresponding collision sub-network, and perform secondary collision detection. Step S7: When a new collision is generated by the secondary collision detection, update the collision relationship map and re-optimize the local region where the new collision is generated. Repeat the priority evaluation and optimization solution steps until no new collision is generated. Step S8: When no new collision is generated by the secondary collision detection, check whether there is a conflict between the optimal solutions of each collision sub-network. If a conflict exists, coordination will be carried out based on the comprehensive priority index of the conflicting components, and an adjustment plan will be regenerated. If there is no conflict, the optimal solutions of each collision subnetwork are merged to form the final solution; Step S9: Generate adjustment scripts for updating the BIM MEP model based on the final solution and execute them in batches to complete the collaborative update of the MEP model.

9. The rapid collaborative method for electromechanical models of substation engineering according to claim 7, characterized in that, Step S6 includes the following steps: Using all non-dominated solutions in the Pareto optimal solution set as the decision candidate set, the optimization objective function value and constraint violation amount of each non-dominated solution are extracted to construct a multi-attribute decision matrix; Based on the decision matrix, the weights of each evaluation index are determined using the entropy weight method, the positive ideal solution and the negative ideal solution are determined according to the TOPSIS method, and the relative closeness of each non-dominated solution is calculated. Based on the relative proximity ranking, the non-dominated solution with the highest proximity is selected as the optimal solution for the corresponding collision subnetwork. The optimal solution is implemented by identifying the set of components that have been adjusted, and by extending the bounding box of each adjusted component outward by a preset safety margin to form an influence domain, and then determining the union of these domains as the secondary collision detection area. Collision detection is performed only on components and their adjacent components within the secondary collision detection area. The determination is completed by combining bounding box pre-screening with precise geometric intersection calculation, and the detection results are output.

10. The rapid collaborative method for electromechanical models of substation engineering according to claim 1, characterized in that, Step S7 includes the following steps: When a new collision is generated by secondary collision detection, the components involved in the new collision are identified, and the components and their corresponding collision relationships are updated in the collision relationship graph to form an updated graph structure. Using the components involved in the new collision as seed nodes, a breadth-first search is performed in the collision relationship graph starting from the seed nodes, traversing neighbor nodes whose distance does not exceed a preset hop count threshold, and determining the subgraph formed by the neighbor nodes and their connecting edges as a local re-optimization region. While keeping the component adjustment scheme outside the local re-optimization region unchanged, only the components within the local re-optimization region are re-executed with the steps of quantitative scoring, priority ranking, and candidate scheme generation, and the local optimization solution is completed. After local re-optimization is completed, incremental collision detection is performed with the influence domain of the new collision component as the detection range to verify whether the new collision has been eliminated. If a new collision still exists at the same location after a preset number of local re-optimizations, the hop count threshold is increased to expand the local re-optimization area, and the process of determining the local re-optimization area, performing local re-optimization, and performing incremental collision detection is executed again until the collision is eliminated or the preset termination condition is met.

11. The rapid collaborative method for electromechanical models of substation engineering according to claim 1, characterized in that, Step S8 includes the following steps: After confirming that no new collisions are generated by the secondary collision detection, the optimal solution of each collision sub-network is obtained, and the adjusted spatial position information and bounding box data of the adjusted components are extracted to construct a global spatial index structure. Based on the global spatial index structure, spatial relationship retrieval is performed on the adjusted components between different collision sub-networks to determine whether there is direct spatial overlap, insufficient spacing, or channel blockage and functional space encroachment caused by adjustment vectors. The identified conflicts are classified and recorded according to conflict type and severity to generate a conflict list. When the conflict list is not empty, each conflict record in the conflict list is coordinated and processed. The comprehensive priority index of the components of the two conflicting parties is compared, the priority component is determined and its adjustment scheme remains unchanged. For non-priority components, after excluding restricted directions, a set of feasible adjustment directions is regenerated, and candidate scheme generation and multi-objective optimization are re-executed. When the difference in the comprehensive priority index of the two conflicting parties does not exceed a preset threshold, the one with the smaller adjustment displacement is selected as the priority component. When there is a component with no feasible adjustment scheme, it is marked as awaiting coordination review and output to the design coordination problem list. After the manual decision-making results are entered, conflict coordination continues. After the conflict coordination is completed, cross-collision sub-network conflict detection is performed again until the conflict list is empty. When the conflict list is empty, the optimal solutions of each collision subnetwork are merged to form an overall candidate adjustment scheme; The overall candidate adjustment scheme is sequentially subjected to global channel connectivity verification, global elevation coordination verification, and global support and hanger installability verification. When all verifications meet the preset constraints, the overall candidate adjustment scheme is determined as the final solution and the model is updated.

Images