Method and system for generating dynamic evacuation road network of buildings based on BIM model

Abstract

The present invention discloses a method and system for generating a dynamic evacuation road network for a building based on a BIM model. The method comprises: obtaining spatial layout data and multi-layer connection information from the building's BIM model to construct an initial road network framework; generating revised road network data based on the initial road network framework and access restriction information; calculating evacuation routes that meet standards based on the revised road network data and safety regulations to generate a preliminary evacuation road network; obtaining real-time data from sensors inside the building to determine the impact of dynamic changes on the preliminary evacuation road network and update information on impassable areas; and recalculating evacuation routes based on the updated impassable area information, adjusting the preliminary evacuation road network, and generating an optimized evacuation road network. Thus, the present invention's method for generating a dynamic evacuation road network for a building based on a BIM model improves the efficiency and accuracy of building emergency evacuations.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of emergency evacuation technology, and in particular to a method and system for generating a dynamic evacuation road network for a building based on a BIM model. Background Art

[0002] In the process of automatically generating a building's accessible road network, the building's geometric structure and spatial layout information must first be accurately extracted based on the BIM model and a 3D digital twin engine. However, buildings often contain complex structures such as stairs, elevators, corridors, and room dividers. The diversity and irregularity of these elements increase the complexity of road network generation. Especially in multi-story buildings, the connections between different floors (such as stairs, elevators, ramps, etc.) need to be accurately identified and modeled to ensure that the generated road network has actual traffic capacity.

[0003] Secondly, the road network generation process needs to consider the actual usage of the building. For example, some areas may be inaccessible due to equipment installation, temporary obstructions, or specific functional requirements. These dynamic factors may not be fully reflected in the BIM model, resulting in deviations from the generated road network. In addition, the design of a building's escape routes often needs to meet specific safety regulations, such as the shortest path and maximum evacuation time. These requirements need to be fully considered during road network generation.

[0004] Finally, once the road network is generated, effectively integrating it with emergency escape routes to generate evacuation routes for any room or location presents a technical challenge. This requires ensuring that the generated evacuation routes not only comply with safety regulations but also dynamically adjust to real-time changes within the building (such as fire and smoke spread) to ensure timely and effective evacuation. This process involves complex algorithm design and real-time data processing capabilities, requiring a high degree of flexibility and accuracy in technical implementation. Summary of the Invention

[0005] An embodiment of the present invention provides a method and system for generating a dynamic evacuation road network for a building based on a BIM model. The method extracts spatial layout and multi-layer connection information from the building's BIM model, combines access restrictions and real-time sensor data, and dynamically generates and optimizes the evacuation road network. Ultimately, it generates a final evacuation plan that meets safety regulations, thereby improving the efficiency and accuracy of emergency evacuation of the building.

[0006] To achieve the above-mentioned objectives, in a first aspect, the present invention provides a method for generating a dynamic evacuation road network for a building based on a BIM model, comprising: obtaining spatial layout data and multi-layer connection information from the building BIM model to construct an initial road network framework. Based on the initial road network framework and access restriction information, a revised road network data is generated. Based on the revised road network data and safety specification requirements, a standard evacuation route is calculated to generate a preliminary evacuation road network. Real-time data from sensors inside the building is obtained to determine the impact of dynamic changes on the preliminary evacuation road network and update the information on inaccessible areas. Based on the updated information on inaccessible areas, the evacuation route is recalculated, the preliminary evacuation road network is adjusted, and an optimized evacuation road network is generated. Through three-dimensional digital twin technology, the optimized evacuation road network is matched with the building BIM model to generate a visual three-dimensional road network diagram. Based on the visual three-dimensional road network diagram and real-time data stream, the evacuation path is dynamically updated to generate a final evacuation plan.

[0007] In a second aspect, the present invention provides a dynamic evacuation network generation system for buildings based on a BIM model, comprising: an initial network framework module, a network data generation module, a preliminary evacuation network generation module, an update module, an evacuation network generation module, a visual three-dimensional network map generation module, and an evacuation plan generation module. The initial network framework module is configured to acquire spatial layout data and multi-layer connectivity information from the building's BIM model to construct an initial network framework. The network data generation module is configured to generate revised network data based on the initial network framework and access restriction information. The preliminary evacuation network generation module is configured to calculate evacuation routes that meet standards based on the revised network data and safety regulations, thereby generating a preliminary evacuation network. The update module is configured to acquire real-time data from internal building sensors, determine the impact of dynamic changes on the preliminary evacuation network, and update information on impassable areas. The evacuation network generation module is configured to recalculate evacuation routes based on the updated impassable area information, adjust the preliminary evacuation network, and generate an optimized evacuation network. The generated three-dimensional visual road network map is used to match the optimized evacuation road network with the building BIM model using three-dimensional digital twin technology to generate a three-dimensional visual road network map. The generated evacuation plan module is used to dynamically update the evacuation path based on the three-dimensional visual road network map and real-time data stream to generate a final evacuation plan.

[0008] In a third aspect, the present invention provides an electronic device, comprising:

[0009] at least one processor; and

[0010] a memory communicatively coupled to the at least one processor;

[0011] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for generating a dynamic evacuation road network of a building based on a BIM model as described above.

[0012] In a fourth aspect, the present invention provides a computer-readable storage medium comprising a computer program and instructions. When the computer program or the instructions are run on a computer, the computer executes the method for generating a dynamic evacuation road network for a building based on a BIM model as described above.

[0013] Compared with the prior art, the method and system for generating a dynamic evacuation road network for a building based on a BIM model according to the present invention have the following beneficial effects:

[0014] 1. By accurately extracting geometric structure and spatial layout information from the building BIM model and combining it with multi-layer connection information, the present invention can construct an initial road network framework that is more consistent with the actual building structure, thereby improving the accuracy of the evacuation road network;

[0015] 2. The present invention takes into account the actual use of buildings, such as access restrictions such as equipment installation and temporary obstacles, as well as dynamic factors such as fire and smoke diffusion. It can update information on impassable areas in real time and recalculate evacuation routes, making the evacuation route network dynamically adaptable;

[0016] 3. In the process of generating the evacuation route network, the present invention combines the evacuation requirements in the safety regulations, such as the shortest path and the maximum evacuation time, to ensure that the generated evacuation routes meet the safety standards and improve the evacuation efficiency and safety;

[0017] 4. Through three-dimensional digital twin technology, the present invention matches the optimized evacuation road network with the building BIM model to generate a visual three-dimensional road network map, which facilitates users to intuitively understand the evacuation route and spatial layout, and improves the decision-making efficiency of emergency evacuation. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of a method for generating a dynamic evacuation road network for a building based on a BIM model in the first embodiment of the present invention;

[0019] Figure 2 This is a structural diagram of a building dynamic evacuation road network generation system based on a BIM model in Example 2 of the present invention;

[0020] Figure 3 It is a structural diagram of an electronic device in embodiment 3 of the present invention. DETAILED DESCRIPTION

[0021] The following is a further detailed description of the embodiments of the present invention in conjunction with the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely illustrative of the embodiments of the present invention and are not intended to limit the embodiments of the present invention. It should also be noted that, for ease of description, the accompanying drawings only illustrate portions of the embodiments of the present invention, rather than all structures.

[0022] To facilitate understanding, the main implementation concepts of the embodiments of the present invention are first briefly described.

[0023] There are multiple technical challenges in the process of generating a dynamic evacuation route network for a building based on a BIM model. First, the complexity and diversity of the building's internal structure increase the difficulty of generating the route network. In multi-story buildings, connecting elements such as stairs, elevators, and corridors need to be accurately identified and modeled to ensure that the generated route network has actual traffic capacity. Second, the actual use of the building, such as equipment installation areas and temporary obstacles, may make some areas inaccessible. These dynamic factors are difficult to fully reflect in the BIM model. In addition, the evacuation route design needs to comply with specific safety regulations, such as the shortest path and maximum evacuation time requirements. These factors need to be fully considered during the route network generation process. Finally, how to effectively combine the generated route network with the emergency escape route and dynamically adjust the evacuation route according to the real-time changes inside the building is a technical problem involving complex algorithm design and real-time data processing.

[0024] Consider a specific technology application scenario: a 30-story commercial complex housing a shopping mall, office space, and a hotel. While the building utilizes an advanced BIM system for management, the development and implementation of emergency evacuation plans still present challenges. The building's complex internal structure includes multiple atriums, skybridges connecting different areas, multiple elevator banks, and multiple emergency staircases. Furthermore, the layout of stores in the shopping mall frequently changes, while the office areas undergo irregular renovations and partition adjustments. In such a scenario, an evacuation network generated based on a static BIM model struggles to adapt to the building's dynamic changes. For example, a major renovation of a shopping mall floor temporarily blocked the existing evacuation route, and this information was not promptly reflected in the evacuation system. Furthermore, a large number of sensors, including smoke detectors, temperature sensors, and pedestrian flow monitoring devices, are installed throughout the building, collecting real-time data. However, effectively utilizing this data to dynamically update the evacuation network presents a key technical challenge.

[0025] Failure to effectively address this technical issue could lead to serious consequences. First, a static evacuation network cannot adapt to the dynamic changes of buildings, which may result in directing people into blocked or dangerous areas in an emergency. Second, failure to fully utilize real-time data may result in missed opportunities to adjust evacuation routes in a timely manner, increasing evacuation time and risks. In addition, if the evacuation network cannot keep pace with the actual conditions of the building, the results of the evacuation drill may deviate significantly from the actual situation, reducing the effectiveness of the emergency plan. From a technical perspective, this issue highlights the limitations of existing BIM systems in dynamic environments and the importance of real-time data processing and decision-making systems in complex building environments. Solving this problem will not only improve the safety performance of buildings, but also promote the development of BIM technology in a smarter and more dynamic direction.

[0026] In addressing the technical problem of generating a dynamic building evacuation network based on a BIM model, the present invention first considers how to accurately obtain the building's spatial layout and multi-layer connectivity information. Due to the complex internal structure of a building, including diverse connecting elements such as stairs, elevators, and corridors, extracting this information directly from the BIM model is challenging. To this end, the present invention proposes a method for extracting spatial layout data and multi-layer connectivity information from the building's BIM model to construct an initial network framework. This method effectively transforms complex building structures into basic data for network generation. However, the initial network framework alone is insufficient to address the actual usage of a building. For example, certain areas may be inaccessible due to equipment installation or temporary obstructions. To address this issue, the present invention proposes a step for generating revised network data based on the initial network framework and access restriction information. This step takes into account actual access restrictions, ensuring that the generated network is more realistic. Next, the present invention considers how to ensure that the generated evacuation routes comply with safety regulations. To this end, the present invention proposes a method for calculating standard evacuation routes based on the revised network data and safety regulations to generate a preliminary evacuation network. This step ensures that the generated network not only takes into account actual traffic conditions but also complies with relevant safety standards.

[0027] Given the dynamic changes in a building's internal environment, such as fires or other emergencies, a static evacuation network may not be able to respond promptly. Therefore, the present invention proposes a method for acquiring real-time data from sensors within the building, determining the impact of these dynamic changes on the preliminary evacuation network, and updating information on impassable areas. This innovative step enables the evacuation network to dynamically adjust based on real-time conditions. Based on the updated impassable area information, the present invention further proposes a method for recalculating evacuation routes, adjusting the preliminary evacuation network, and generating an optimized evacuation network. This ensures that the evacuation network can be continuously optimized to adapt to real-time changes within the building. To enhance the visualization and practicality of the evacuation network, the present invention incorporates 3D digital twin technology, matching the optimized evacuation network with the building's BIM model to generate a visualized 3D network diagram. This step not only improves the visualization of the network but also lays the foundation for subsequent dynamic updates. Finally, to enable real-time dynamic updates of the evacuation plan, the present invention proposes a method for dynamically updating evacuation routes and generating a final evacuation plan based on the visualized 3D network diagram and real-time data streams. This step enables the evacuation system to adjust evacuation strategies in real time based on the latest building status and environmental changes.

[0028] Example 1

[0029] Figure 1 This is a flow chart of a method for generating a dynamic evacuation road network for a building based on a BIM model in the first embodiment of the present invention. Figure 1 As shown, embodiment 1 provides a method for generating a dynamic evacuation road network for a building based on a BIM model, including: step S100, obtaining spatial layout data and multi-layer connection information from the building BIM model to construct an initial road network framework; step S200, generating revised road network data based on the initial road network framework and access restriction information; step S300, calculating evacuation routes that meet the standards based on the revised road network data and safety specifications, and generating a preliminary evacuation road network; step S400, obtaining real-time data from sensors inside the building, judging the impact of dynamic changes on the preliminary evacuation road network, and updating inaccessible area information; step S500, recalculating evacuation routes based on the updated inaccessible area information, adjusting the preliminary evacuation road network, and generating an optimized evacuation road network; step S600, matching the optimized evacuation road network with the building BIM model through three-dimensional digital twin technology to generate a visualized three-dimensional road network map; step S700, dynamically updating evacuation paths based on the visualized three-dimensional road network map and real-time data stream to generate a final evacuation plan.

[0030] BIM refers to a building information model, a digital representation of data including building geometry, spatial relationships, geographic information, and the properties and quantities of its components. This can be implemented using software such as Autodesk Revit and Bentley AECOsim Building Designer. A dynamic evacuation network is one that automatically adjusts to the real-time conditions within a building. This can be achieved using graph theory algorithms and real-time data processing techniques. Three-dimensional digital twin technology involves creating virtual replicas of physical entities, mapping the physical world to the digital world. This can be achieved using 3D engines such as Unity3D and Unreal Engine.

[0031] This invention combines BIM models, real-time sensor data, and 3D digital twin technology to automatically generate and optimize a building's dynamic evacuation network. This approach allows the evacuation network to dynamically adjust to real-time changes within the building, improving evacuation efficiency and safety.

[0032] The working principle of the present invention can be divided into the following key steps:

[0033] First, spatial layout data and multi-layer connectivity information are extracted from the building's BIM model. This step analyzes the BIM model's geometry to determine the locations and connectivity of key elements such as walls, partitions, stairs, elevators, and corridors. This information is used to construct an initial network framework, providing the foundation for subsequent network generation. Next, a revised network is generated based on the initial network framework and access restriction information. This step considers actual access restrictions, such as equipment installation areas and temporary obstacles, and uses geometric analysis and path planning tools to eliminate impassable areas, ensuring the generated network is consistent with actual conditions. Then, based on the revised network data and safety regulations, compliant evacuation routes are calculated to generate a preliminary evacuation network. This step utilizes the Dijkstra algorithm, combined with safety regulations' requirements for shortest paths and maximum evacuation time thresholds, to ensure the generated evacuation routes meet safety standards. Next, real-time data from internal building sensors is collected to determine the impact of dynamic changes on the preliminary evacuation network and update the impassable area information. This step dynamically marks affected areas as impassable by analyzing sensor data, such as fire or smoke spread, to achieve real-time network updates. Based on the updated information about impassable areas, evacuation routes are recalculated, the preliminary evacuation network is adjusted, and an optimized evacuation network is generated. This step uses the Dijkstra algorithm to recalculate evacuation routes and the K-means clustering algorithm to optimize node positions, ensuring real-time network optimization. Subsequently, using 3D digital twin technology, the optimized evacuation network is matched with the building BIM model to generate a visual 3D network diagram. This step uses tools such as Three.js to generate the 3D network diagram and uses the A-star algorithm to dynamically adjust evacuation routes, improving network visualization and practicality. Finally, based on the visual 3D network diagram and real-time data streams, evacuation routes are dynamically updated to generate the final evacuation plan. This step continuously analyzes real-time data, dynamically updates impassable areas, and recalculates evacuation routes using the A-star algorithm, ensuring the real-time and effectiveness of the evacuation plan. These steps work together to transform the static BIM model into a dynamic evacuation network, enabling continuous optimization of the evacuation plan based on real-time data. These technical features were chosen because they can effectively address complex internal building structures, dynamic changes in usage, and the need to comply with safety regulations, thereby significantly improving evacuation efficiency and safety.

[0034] As a preferred embodiment, the present invention can be applied to a large 30-story commercial complex. The building includes a shopping mall, office area, and hotel, and has a complex internal structure, including multiple atriums, sky bridges connecting different areas, multiple elevator groups, and multiple emergency escape stairs.

[0035] First, a BIM model of the building was created using Autodesk Revit software. Using Revit's API, the building's spatial layout data and multi-level connectivity information were extracted. For example, floor plans for each floor were obtained, including the location and dimensions of walls, doors, windows, and corridors; the location of stairways and elevators, and the floors they connected to, were extracted. Next, a Python-based path planning algorithm was used to construct an initial path network framework based on the extracted data. Given the frequent changes in store layouts within the mall and the irregular renovations and partition adjustments to office areas, the system regularly updates access restrictions. For example, if a major renovation on a particular floor temporarily blocks the existing evacuation route, this information is input into the system and used to revise the path network data. Based on this revised path network data and local building safety regulations, the Dijkstra algorithm was used to calculate compliant evacuation routes. For example, if the distance from any point to the nearest safe exit must not exceed 50 meters, the system generates a preliminary evacuation path network based on this information. Numerous sensors, including smoke detectors, temperature sensors, and pedestrian flow monitoring devices, are installed throughout the building. These sensors transmit data to the central processing system in real time via the MQTT protocol. The system uses machine learning algorithms (such as random forests) to analyze this data and determine whether any areas have become impassable. For example, if the temperature in a certain area suddenly rises and the smoke concentration increases, the system will mark that area as impassable. Based on this updated information about impassable areas, the system uses the A-star algorithm to recalculate evacuation routes and adjust the initial evacuation network. For example, if a fire is detected in an area on the third floor, the system will immediately replan evacuation routes for that floor and adjacent floors to avoid the danger zone. Next, a 3D digital twin model of the building is created using the Unity3D engine. The optimized evacuation network is mapped onto this 3D model, generating a visual 3D network diagram. This 3D model displays the building's internal conditions in real time, including impassable areas and recommended evacuation routes. Finally, the system continuously receives and analyzes real-time data streams to dynamically update evacuation routes. For example, if a sudden increase in traffic in a particular evacuation corridor is detected, the system will automatically adjust the evacuation plan, redirecting some traffic to other, less crowded corridors to avoid congestion.

[0036] In this way, the present invention can provide a dynamic, real-time, and visual evacuation system for the large commercial complex, greatly improving the evacuation efficiency and safety in emergency situations.

[0037] In this embodiment, the step S100 includes: step S101, parsing the geometric structure from the building BIM model to obtain the spatial layout data of the walls and partitions; step S102, determining the positional relationship of the walls and partitions based on the spatial layout data; step S103, extracting the multi-layer connection information of the stairs, elevators and corridors from the building BIM model; step S104, determining the connection point data based on the positional relationship and the multi-layer connection information; step S105, constructing a multi-layer road network structure through the connection point data; step S106, optimizing the multi-layer road network structure using a genetic algorithm to generate the initial road network framework.

[0038] Specifically, the geometry is parsed from the BIM model: This step can be accomplished using the BIM software's API or specialized geometric data extraction tools. For example, the Autodesk Revit API can be used to extract geometric information for structural elements such as walls, floors, and ceilings. This information typically includes coordinate points, length, width, and height. Spatial analysis algorithms are then used to identify and classify walls and partitions based on the parsed geometry. For example, a plane scanning algorithm can be used to detect planar structures perpendicular to the ground and classify them as walls or partitions. The output of this step can be a data structure containing the location, size, and orientation of walls and partitions. Positional relationships between walls and partitions are determined: Using the acquired spatial layout data, graph theory algorithms can be applied to establish the topological relationships between walls and partitions. For example, an adjacency matrix can be used to represent the connectivity between walls, or a quadtree structure can be used to quickly query walls and partitions within a specific area. Multi-layer connectivity information is extracted: This step requires particular attention to vertically connected elements such as stairs, elevators, and corridors. Attribute information in the BIM model can be used to identify these elements. For example, objects with the "Stairs," "Elevator," or "Corridor" attributes can be searched. For stairs, information such as their starting and ending floors, width, and inclination angle can be extracted. For elevators, information such as the range of floors they serve and their capacity can be extracted. Determining connection point data: Based on wall position relationships and multi-layer connectivity information, spatial interpolation algorithms can be used to generate connection points. For example, connection points can be set at corridor intersections, room entrances and exits, and stairway start and end points. These connection points will become nodes in the network. Constructing a multi-layer network structure: Using graph theory minimum spanning tree algorithms, such as Kruskal's algorithm or Prim's algorithm, a preliminary network structure is constructed based on the connection point data. This structure should include connection points on all floors and the connections between them. Using genetic algorithms to optimize the network structure: Genetic algorithms can be used to optimize network connectivity and path selection. For example, a fitness function can be defined to evaluate the efficiency of the network (such as average path length and congestion level). Crossover and mutation operations can then be used to generate and select the optimal network structure.

[0039] These steps are closely interconnected and interactive. For example, the positional relationships of walls and partitions directly influence the determination of connection points, while multi-layer connectivity information determines the network connections between different floors. Through this systematic approach, the present invention can effectively extract the necessary spatial information from the BIM model and transform this information into an initial network framework for evacuation route planning.

[0040] In particular, the extraction of multi-layer connectivity information and the application of genetic algorithms can effectively address vertical connectivity issues in complex building structures and optimize the road network structure. This not only improves the accuracy and practicality of the road network, but also lays a solid foundation for subsequent dynamic evacuation route planning.

[0041] For example, in one specific embodiment, consider a BIM model of a five-story office building. First, the Autodesk Revit API is used to extract the building's geometry, including the floor plan, wall location, and dimensions of each floor. Spatial analysis algorithms are then used to identify the room layout on each floor and determine the locations of corridors, office areas, and common areas. For multi-story connections, the system identifies two stairwells (located on the east and west sides of the building) and a central elevator shaft. The staircase's geometric information includes the height (18 cm) and width (30 cm) of each step, as well as the total width of the staircase (1.5 m). Elevator information includes its load capacity (1000 kg) and speed (2 m / s). When determining connection points, the system sets nodes at the entrances and exits of each room, the intersections of corridors, the starting and ending points of the stairs, and the elevator doors. For example, a connection point is set every 5 meters in the main corridor of each floor. The Kruskal algorithm is used to construct an initial multi-story network structure, connecting all nodes. Then, a genetic algorithm is applied for optimization, defining the fitness function as: f = 1 / (average path length + α * congestion coefficient), where α is a weight coefficient set to 0.5. Through multiple iterations (e.g., 100 generations), the connection structure is continuously adjusted, ultimately resulting in an optimized road network structure that balances path length and congestion. This optimized road network structure becomes the initial road network framework, providing the basis for subsequent dynamic evacuation route planning. It not only takes into account the actual spatial layout of buildings but also improves the efficiency of the road network through algorithmic optimization, thereby better supporting evacuation needs in emergency situations.

[0042] In this embodiment, the step S200 includes: step S201, extracting the spatial position relationship of walls and partitions based on the initial road network framework; step S202, identifying the access restriction information through geometric analysis, marking the equipment installation area and the temporary obstacle position; step S203, if there is an impassable area, excluding the impassable area through the path planning tool to generate a revised road network structure; step S204, extracting multi-layer connection information from the revised road network structure and determining the connection point data; step S205, constructing a revised multi-layer road network framework through a network modeling tool; step S206, using a path optimization algorithm to adjust the revised multi-layer road network framework to generate the revised road network data.

[0043] Specifically, first, the spatial positional relationships of walls and partitions are extracted based on the initial road network framework. This step can be achieved in various ways, such as using spatial analysis algorithms to identify and locate walls and partitions in the BIM model, or using image processing techniques to analyze building floor plans. The extracted spatial positional relationships can include parameters such as coordinates, length, width, and height of walls and partitions. Next, geometric analysis is used to identify access restriction information and mark the locations of equipment installation areas and temporary obstacles. Geometric analysis can use various algorithms, such as the convex hull algorithm and the minimum enclosing rectangle algorithm, to identify irregularly shaped obstacles. Access restriction information can include fixed obstacles (such as pillars and equipment) and temporary obstacles (such as construction areas and temporary isolation zones). These can be distinguished by assigning different weights or labels. Next, if impassable areas exist, these areas are excluded using a path planning tool to generate a revised road network structure. Path planning tools can use algorithms such as the A* algorithm, the Dijkstra algorithm, or the RRT (rapid randomized recursive tree) algorithm. This step ensures that the generated road network avoids all impassable areas, improving the practicality of the road network. Next, multi-layer connectivity information is extracted from the revised road network structure, and connection point data is determined. This multi-layer connectivity information includes the location and connectivity of vertical transportation facilities such as stairs, elevators, and ramps. Connection point data can include information such as the coordinates of each connection point, the floor it belongs to, and the connection type. This step solves the problem of connecting different floors in complex buildings and ensures the vertical connectivity of the generated road network. Next, a revised multi-layer road network framework is constructed using network modeling tools. Network modeling tools can use adjacency matrices or adjacency lists from graph theory to represent the road network structure, or specialized road network modeling software can be used. The constructed multi-layer road network framework should include information about nodes (representing traversable points) and edges (representing traversable paths). Finally, a path optimization algorithm is used to adjust the revised multi-layer road network framework to generate revised road network data. Path optimization algorithms can include genetic algorithms, ant colony algorithms, or simulated annealing algorithms. The optimization objectives can be minimizing total path length, maximizing travel efficiency, or balancing the load across paths.

[0044] Through this series of steps, the technical solution of the present invention can effectively solve the problem of the initial road network not matching the actual traffic conditions. This solution comprehensively considers static structural information and dynamic traffic restriction information. Through multi-step analysis and optimization, it generates road network data that not only conforms to the actual structure of the building but also takes into account real-time traffic conditions.

[0045] As a preferred embodiment, the technical solution of the present invention can be specifically implemented as follows:

[0046] First, a BIM model parsing tool is used to extract the geometric data of walls and partitions from the building information model. For example, a wall may be represented as a series of line segments or polygons, each containing information such as the coordinates of its start and end points, thickness, and so on. Partitions may be represented as thin walls or specific obstacle objects. Next, computational geometry algorithms are used for spatial analysis. For example, a planar scanline algorithm can be used to identify openings (such as doors) between adjacent walls, and polygon Boolean operations can be used to determine traversable areas. For equipment installation areas and temporary obstacles, a minimum enclosing rectangle algorithm can be used for rapid location and marking. After identifying impassable areas, a modified A* algorithm is applied for path planning. This algorithm can set obstacle weights to avoid impassable areas. For example, the weight of fixed obstacles can be set to infinity, while the weight of temporary obstacles can be set to a large but finite value (such as 1000). A depth-first search algorithm can be used to extract multi-level connectivity information. Starting from the entry point of each floor, all possible paths are searched until a connection point (such as a staircase or elevator) leading to another floor is found. Each connection point records its coordinates, floor location, and connection type. Graph data structures can be used in the network modeling phase. Each traversable point is a node in the graph, and traversable paths between nodes are edges. For multi-story buildings, a multi-layer graph structure can be used, with different layers connected through previously identified connection points. Finally, a genetic algorithm can be used for path optimization. The initial population can be a number of different road network structures, and a new road network structure is generated through crossover and mutation operations. The fitness function can comprehensively consider factors such as path length, congestion, and safety. For example, the fitness function can be set as: F = w1*L+w2*C+w3*S, where L is the total path length, C is the congestion index, S is the safety index, and w1, w2, and w3 are weight coefficients. After multiple generations of iteration, the road network structure with the highest fitness is finally selected as the optimization result.

[0047] Based on the above analysis, it can be seen that the technical solution of the present invention, by comprehensively considering static structure and dynamic factors and adopting a variety of advanced algorithms, effectively solves the technical problem of generating corrected road network data based on the initial road network framework and access restriction information, and provides more accurate and practical basic data for building evacuation route planning.

[0048] In this embodiment, the step S300 includes: step S301, based on the corrected road network data, combined with the evacuation requirements in the safety specifications, determining the shortest path and the maximum evacuation time threshold; step S302, using the Dijkstra algorithm, according to the corrected road network data and the maximum evacuation time threshold, calculating the evacuation route that meets the standards; step S303, based on the evacuation route, generating the preliminary evacuation road network, extracting key nodes and connection relationships; step S304, if there are areas in the preliminary evacuation road network that do not meet the standards, adjusting the Dijkstra algorithm parameters and recalculating the evacuation route; step S305, determining the final preliminary evacuation road network through iterative optimization.

[0049] Specifically, determining the shortest path and maximum evacuation time threshold based on the revised road network data and safety regulations is a key step. This step can be implemented in a variety of ways, for example: 1. Using the Floyd-Warshall algorithm to calculate the shortest path between all node pairs and setting the maximum evacuation time threshold according to the evacuation requirements in the safety regulations; 2. Using the A* algorithm combined with a heuristic function to quickly find the shortest path between key nodes, while taking into account the geometric characteristics of the building and the safety regulations; 3. Combining genetic algorithms and simulated annealing algorithms to optimize path selection and time threshold setting to adapt to complex building structures and evacuation requirements. Using the Dijkstra algorithm to calculate evacuation routes that meet the standards is the core step of the present invention. The implementation of the Dijkstra algorithm can consider the following aspects: 1. Using a priority queue (such as a Fibonacci heap) to optimize algorithm performance and improve the computational efficiency of large-scale road networks; 2. Introducing a dynamic adjustment mechanism for edge weights to update path weights based on real-time evacuation conditions and building status; 3. Combining parallel computing technologies, such as CUDA or OpenCL, to accelerate the path calculation process for large buildings. In the process of generating a preliminary evacuation network and extracting key nodes and connection relationships, the following implementation methods can be considered: 1. Using a minimum spanning tree algorithm from graph theory (such as the Kruskal algorithm or the Prim algorithm) to construct the preliminary evacuation network skeleton; 2. Using a clustering algorithm (such as K-means or DBSCAN) to identify and extract key nodes; 3. Using spatial indexing technology (such as R-tree or quadtree) to optimize the storage and query efficiency of nodes and connection relationships. If there are areas in the preliminary evacuation network that do not meet the standards, the present invention proposes a method for adjusting the Dijkstra algorithm parameters and recalculating the evacuation routes. This process can be achieved by: 1. Using a machine learning algorithm (such as a random forest or support vector machine) to predict areas that do not meet the standards and adjust the algorithm parameters accordingly; 2. Introducing a multi-objective optimization algorithm to dynamically adjust the weight parameters of the Dijkstra algorithm by considering multiple factors such as path length, evacuation time, and safety; 3. Using reinforcement learning technology to adaptively optimize algorithm performance by continuously trying and evaluating different parameter combinations. Determining the final preliminary evacuation network through iterative optimization is a key feature of the present invention. The following implementation methods can be considered for this process: 1. Use the Monte Carlo simulation method to run the optimization process multiple times and statistically analyze the distribution characteristics of the optimal solution; 2. Use genetic algorithms or particle swarm optimization algorithms to find the globally optimal evacuation road network solution through multi-generation evolution; 3. Combine expert systems and fuzzy logic, introduce human experience and judgment, and improve the practicality and reliability of the optimization results.

[0050] As a preferred embodiment, the technical solution of the present invention can be implemented according to the following steps:

[0051] 1. Data Preparation: Extract node and edge information from the corrected road network data and construct a graph structure G(V,E), where V is the node set and E is the edge set. Each edge e∈E contains information about its length l(e) and travel time t(e).

[0052] 2. Parameter setting: Set the maximum evacuation time threshold Tmax according to safety regulations. Initialize the weight parameter w of the Dijkstra algorithm, usually set to w = 0.7.

[0053] 3. Shortest Path Calculation: Use Dijkstra's algorithm to calculate the shortest path from each node to the nearest exit. For each edge e, define its weight as w*l(e)+(1-w)*t(e).

[0054] 4. Preliminary evacuation road network generation: Merge all shortest paths to form a preliminary evacuation road network N. Extract key nodes (such as intersections and turning points) and connection relationships in N.

[0055] 5. Compliance check: Check whether the total evacuation time of each path in N is less than or equal to Tmax. If there is a path that does not meet the requirements, go to step 6; otherwise, go to step 7.

[0056] 6. Parameter adjustment and recalculation: Adjust the weight parameter w, for example, w = w - 0.1. Return to step 3 and recalculate.

[0057] 7. Road network optimization: Use the K-means algorithm to cluster key nodes and merge nodes with close distances. Use the minimum spanning tree algorithm to optimize the connection relationship between nodes.

[0058] 8. Output results: Generate the final preliminary evacuation road network, including optimized node positions and connection relationships.

[0059] Through this implementation, the present invention can generate an evacuation road network that meets safety standards in a complex building environment. For example, in a multi-story office building, this method can effectively handle vertical connection structures such as stairs and elevators, while taking into account the special requirements of different floors (such as the safety doors of some floors are locked). Through iterative optimization, the evacuation road network finally generated can ensure that the evacuation time of all locations is within a safe range while ensuring the shortest path. Compared with the prior art, the technical solution of the present invention has significant advantages. Traditional evacuation route generation methods usually only consider the shortest path and ignore the requirements of evacuation time and safety regulations. The present invention, by introducing a maximum evacuation time threshold and an iterative optimization process, can ensure that the evacuation time meets safety standards while ensuring the shortest path. In addition, the method of the present invention can dynamically adapt to changes in the internal environment of the building, and quickly generate an updated evacuation road network through parameter adjustment and recalculation, which is of great significance in practical applications.

[0060] In this embodiment, the step S400 includes: step S401, obtaining real-time data from sensors inside the building, cleaning and standardizing the real-time data; step S402, calculating the range and speed of fire or smoke diffusion based on the standardized real-time data; step S403, judging whether the diffusion range exceeds the traffic capacity threshold according to a preset threshold rule; step S404, if it exceeds the traffic capacity threshold, marking the affected area as an inaccessible area; step S405, updating the traffic status of the preliminary evacuation road network based on the inaccessible area, and generating updated inaccessible area information.

[0061] Specifically, the technical solution proposed in this invention dynamically identifies dangerous areas within buildings through real-time data collection, processing, and analysis, and promptly updates the evacuation route network. This approach effectively addresses rapid changes in the building's internal environment, such as fire or smoke spread, ensuring the real-time effectiveness of evacuation routes. By setting capacity thresholds, the system can objectively determine whether an area is safe and accessible, avoiding the errors and delays that can result from human judgment. Furthermore, by continuously updating information on impassable areas, it provides accurate basic data for subsequent recalculation and optimization of evacuation routes, thereby improving the reliability and adaptability of the entire evacuation system. The technical solution of this invention includes several key features, each with multiple possible implementations: Real-time data from sensors within the building can be acquired using various types of sensors, such as temperature sensors, smoke sensors, and gas concentration sensors. These sensors can be connected to a central processing system via wired or wireless networks. Data cleaning and normalization can employ a variety of data processing algorithms, such as outlier detection, data interpolation, and data normalization. For example, the Z-score method can be used for data normalization, or the moving average method can be used to filter out noise. Calculating Fire or Smoke Dispersion: Computational fluid dynamics (CFD) models or simplified regional models can be used to simulate fire or smoke spread. For example, FDS (Fire Dynamics Simulator) software can be used for precise simulation, or CFAST (Consolidated Model of Fire and Smoke Transport) can be used for rapid estimation. Decision Thresholds: Threshold rules can be set based on multiple parameters, such as temperature, smoke density, and toxic gas concentration. Thresholds can be fixed or dynamically adjusted based on building characteristics. Marking Impassable Areas: Graph algorithms such as flood-fill can be used to mark continuous impassable areas, or gridding can be used to divide the building into small units and mark the affected units one by one. Updating Traffic Status: Graph algorithms such as depth-first search (DFS) or breadth-first search (BFS) can be used to update the connectivity of the road network, or dynamic programming algorithms can be used to recalculate the shortest path. These characteristics are closely related and interactive. Real-time data acquisition provides the foundation for subsequent data processing and analysis. Data cleaning and standardization ensure the accuracy of subsequent analysis. The calculation results of fire or smoke diffusion directly affect the threshold judgment, which in turn determines the marking of impassable areas. Finally, the information of impassable areas is used to update the access status of the evacuation road network, forming a complete dynamic update cycle.

[0062] The technical solution of the present invention addresses the impact of dynamic changes within buildings on the initial evacuation network by following the following process: First, a sensor network within the building acquires real-time environmental data. This data may include parameters such as temperature, smoke concentration, and toxic gas concentration. The frequency of data acquisition can be adjusted based on the building's characteristics and safety requirements, for example, collecting data every 5 or 10 seconds. Next, the collected data is cleaned and normalized. This step removes outliers, fills in missing data, and converts different types of data into a unified standard format. For example, median filtering can be used to remove outliers, linear interpolation can be used to fill in missing data, and then Min-Max normalization can be used to map all data to a range of 0-1. Based on this processed data, a pre-established model is used to calculate the spread and velocity of fire or smoke. This may involve complex physical models, such as those that consider the building's geometry, material properties, and ventilation conditions. The resulting calculation can be a three-dimensional diffusion field, indicating the level of danger at each spatial point. Based on these results, the system determines whether each area exceeds a preset capacity threshold. These thresholds may be based on safety standards. For example, when the smoke concentration exceeds 0.1m^-1, or the temperature exceeds 60℃, the area is considered unsuitable for passage. For areas that exceed the threshold, the system will mark them as impassable areas. This may involve updating the status flag of each spatial unit in the digital model of the building. The marking process needs to take into account the continuity of the space to ensure that there are no isolated impassable areas. Finally, the system re-evaluates the effectiveness of the preliminary evacuation road network based on the updated impassable area information. This may include deleting paths that pass through impassable areas, or adjusting paths to bypass these areas. The updated road network information will be used for subsequent evacuation planning and guidance.

[0063] Through this series of steps, the technical solution of the present invention can promptly reflect dynamic changes within the building, ensuring that the evacuation route network always remains up-to-date and secure. This dynamic update mechanism greatly improves the reliability and adaptability of the evacuation system, providing safer evacuation for building users.

[0064] As a preferred embodiment, the technical solution of the present invention can be implemented in a large commercial complex. The complex comprises multi-story shopping, office, and dining areas, with a total floor area of ​​approximately 100,000 square meters and a daily foot traffic of up to 50,000 people. A total of 5,000 sensors of various types are installed throughout the complex, including 2,000 temperature sensors, 2,000 smoke sensors, and 1,000 carbon monoxide sensors. These sensors are evenly distributed throughout the building, with an average of one sensor per 20 square meters. All sensors are connected to a central processing system via a wireless network, with data acquisition occurring every 5 seconds. Data cleaning and normalization utilizes a sliding window median filter algorithm to remove outliers, with a window size of 11 data points. Missing data is imputed using linear interpolation. All data are then normalized using the Z-score method. Fire or smoke spread calculations utilize a simplified regional model, dividing the entire building into 5,000 cubic units, each with a side length of 5 meters. The model takes into account the building's ventilation system and fire compartmentation, using the finite difference method to solve the diffusion equation and calculate temperature and smoke concentration changes within each cell. Capacity thresholds are set as follows: when a cell's temperature exceeds 50°C, its smoke concentration exceeds 0.08 m^-1, or its carbon monoxide concentration exceeds 100 ppm, the cell is marked as impassable. These thresholds are based on international safety standards and local fire regulations. Impassable areas are marked using a three-dimensional flood-fill algorithm, starting from the identified hazard source and spreading the marking outwards. The algorithm considers both vertical and horizontal connectivity to ensure continuity and integrity of the marking. Finally, the evacuation network is updated using the A* algorithm, which recalculates the shortest safe path from each possible starting point to the nearest exit. The algorithm's heuristic function takes into account distance and safety factors, prioritizing paths away from hazardous areas. This implementation method enables the technical solution of the present invention to rapidly respond to dynamic changes in complex, large buildings, promptly updating the evacuation network and providing safe and reliable evacuation guidance for large numbers of people. The system's response time from sensing danger to updating the evacuation route network is no more than 30 seconds, greatly improving evacuation efficiency in emergency situations.

[0065] In this embodiment, step S500 includes: step S501, recalculating the evacuation route using the Dijkstra algorithm based on the updated impassable area information; step S502, adjusting the preliminary evacuation road network based on the recalculated evacuation route to generate a new road network layout; step S503, extracting node location information from the new road network layout; step S504, optimizing the node location information through the K-means clustering algorithm and adjusting the road network spatial layout; step S505, generating an optimized evacuation road network based on the adjusted road network spatial layout.

[0066] Specifically, first, the Dijkstra algorithm is used to recalculate evacuation routes based on the updated information about impassable areas. The Dijkstra algorithm, a classic shortest path algorithm, is used in the present invention to calculate optimal evacuation routes. This algorithm can quickly find the shortest path from any starting point to a safe exit based on the latest information about impassable areas. For example, when an area becomes impassable due to a fire or other emergency, the Dijkstra algorithm can quickly replan routes to avoid these dangerous areas. Secondly, based on the recalculated evacuation routes, the preliminary evacuation network is adjusted to generate a new network layout. This step integrates the latest evacuation routes calculated by the Dijkstra algorithm into the existing network to form an updated network layout. This process may involve, for example, adding new paths, removing unsafe paths, and adjusting the weights of existing paths. Next, node location information is extracted from the new network layout. Nodes typically represent key locations in the network, such as intersections, corners, or safe exits. Accurate node location information is crucial for subsequent optimization. The K-means clustering algorithm is then used to optimize the node location information and adjust the spatial layout of the network. The K-means clustering algorithm is a commonly used data analysis method, which is innovatively applied to road network optimization in the present invention. The algorithm can group similar nodes and find the center point of each group, thereby optimizing the distribution of nodes. This step helps to simplify the road network structure, reduce redundant nodes, and improve evacuation efficiency. For example, in a large open space, K-means clustering can help determine the best intermediate gathering point, making the evacuation of personnel more orderly and efficient. Finally, based on the adjusted road network spatial layout, an optimized evacuation road network is generated. This final evacuation road network integrates the dynamically updated inaccessible area information, the shortest path calculated by the Dijkstra algorithm, and the node distribution optimized by K-means clustering to form a fully optimized evacuation plan.

[0067] For example, in a multi-story office building, if a fire breaks out on a certain floor, the technical solution of the present invention can rapidly update information about inaccessible areas and recalculate the shortest paths from each office to the emergency exit using the Dijkstra algorithm. Simultaneously, the K-means clustering algorithm can optimize the locations of meeting points on each floor, ensuring that evacuation avoids congestion in certain areas. This dynamic adjustment and optimization capability allows evacuation plans to adapt to changing circumstances, significantly improving evacuation efficiency and safety.

[0068] As a preferred embodiment, the technical solution of the present invention can be implemented as follows: First, the system receives updated information about impassable areas. This information may come from various sensors within the building, such as smoke detectors and heat sensors. For example, the system may receive information about thick smoke in a third-floor corridor and mark the area as impassable. Next, the system uses the Dijkstra algorithm to recalculate evacuation routes. The algorithm treats each traversable area of ​​the building as a node in a graph, with connections between nodes representing traversable paths. Each path is assigned a weight representing the time or difficulty required to traverse it. For example, a normal corridor might have a weight of 1, while a narrow or crowded passage might have a weight of 2 or 3. The weights of impassable areas are set to infinity to ensure that the algorithm does not select these paths. The Dijkstra algorithm begins at each starting point (such as an office) and calculates the shortest path to the nearest safe exit. The system then adjusts the initial evacuation route network based on the Dijkstra algorithm's calculation results, generating a new route network layout. This may involve removing routes that pass through impassable areas, adding new alternative routes, or adjusting the direction and traffic distribution of existing routes. Next, the system extracts node location information from the new road network layout. These nodes may include key locations such as corridor intersections, stairwells, and emergency exits. Each node is assigned a specific coordinate value, such as (x, y, z), where z represents the floor. The system then uses the K-means clustering algorithm to optimize the node location information. Assuming K=5 is selected, the algorithm divides all nodes into 5 clusters. The center point of each cluster is considered to be the best gathering point for the area. For example, in a large open office area, the algorithm may determine several gathering points near the main exit, near the stairs, etc. Finally, the system generates the final optimized evacuation road network based on the optimized node positions and path information. This road network not only contains the latest accessible paths, but also optimizes the distribution of personnel flow and gathering points, thereby improving the overall evacuation efficiency.

[0069] Based on the above analysis, it can be seen that in this way, the technical solution of the present invention can quickly respond to environmental changes in emergency situations, provide optimal evacuation routes, and improve overall evacuation efficiency by optimizing node distribution. This dynamic optimization capability significantly improves the adaptability and effectiveness of evacuation plans, providing building users with safer and more reliable evacuation guarantees. Compared with the existing technology, the technical solution of the present invention has significant advantages. Traditional evacuation road networks are usually static and cannot respond to sudden environmental changes in a timely manner. The present invention greatly improves the flexibility and adaptability of evacuation plans by updating information on impassable areas in real time and recalculating evacuation routes. In addition, traditional methods often only focus on calculating the shortest path, while ignoring the optimization of the overall road network layout. The present invention introduces the K-means clustering algorithm, which not only optimizes single paths, but also improves the spatial distribution of the entire road network, thereby improving evacuation efficiency at a macro level. This dynamic optimization and global consideration method makes the technical solution of the present invention show obvious superiority when dealing with complex and changeable emergency situations.

[0070] In this embodiment, step S600 includes: step S601, obtaining real-time data from sensors inside the building through the MQTT protocol and parsing it into a format recognizable by the three-dimensional engine; step S602, dynamically updating the impassable areas in the building BIM model based on the parsed data; step S603, matching the updated building BIM model with the optimized evacuation road network; step S604, generating a three-dimensional road network map through Three.js; step S605, using the A-star algorithm to adjust the evacuation routes in the three-dimensional road network map based on the real-time data; step S606, optimizing the spatial layout of the three-dimensional road network map through the random forest algorithm to generate a visual three-dimensional road network map.

[0071] Specifically, the MQTT protocol can be implemented in a variety of ways. For example, the Eclipse Paho MQTT client library can be used to implement MQTT communication. This library supports multiple programming languages, including Java, Python, and C++. The appropriate version can be selected based on the specific development environment. In actual applications, parameters such as the MQTT server address, port, and topic can be configured to ensure correct reception of sensor data. There are also many options for choosing a 3D engine. In addition to Three.js, powerful 3D engines such as Unity3D or Unreal Engine can also be used. These engines offer more powerful rendering capabilities and richer development tools, allowing for more complex 3D scenes and interactive effects. For data parsing, sensor data can be transmitted in JSON or XML formats, making it easy to parse the data into a format that the 3D engine can understand. For example, the JSON.parse() function can be used to parse JSON data and then map the parsed data to corresponding locations in the 3D scene. When implementing the A-star algorithm, for example, heuristic functions can be used to improve search efficiency. For example, Manhattan distance or Euclidean distance can be used as heuristic functions to more quickly find the optimal path. In addition, dynamic weight adjustment can be combined to dynamically adjust path weights based on real-time data to adapt to different evacuation situations.

[0072] When optimizing the spatial layout of a three-dimensional road network diagram, the random forest algorithm can consider using multiple features to train the model. For example, the spatial coordinates, connection relationships, pedestrian density, etc. of the nodes can be used as features, and the optimal spatial layout can be obtained through multiple iterative optimizations. There is a close correlation and interaction between these technical features. The real-time data obtained by the MQTT protocol directly affects the update of the BIM model, and the updated BIM model provides the basis for Three.js to generate a three-dimensional road network diagram. The A-star algorithm and the random forest algorithm perform path optimization and spatial layout adjustment based on these updated data. This multi-level technical integration enables the present invention to solve the complex evacuation road network generation problem more accurately and efficiently.

[0073] In practical applications, the technical solution of the present invention can significantly improve the efficiency and accuracy of evacuation network generation. For example, in the design of an evacuation plan for a multi-story office building, this can be achieved through the following steps: First, use the MQTT protocol to connect the building's temperature sensors, smoke sensors, and crowd density sensors. Set the MQTT server address to "mqtt.example.com", port 1883, and subscribe to the topic "building / sensors / #". Next, use Three.js to create a basic 3D scene and load a pre-prepared BIM model. The GLTFLoader in Three.js can be used to load BIM model files in GLTF format. Then, receive and parse the MQTT data in real time. For example, the received JSON formatted data may be as follows: {"temperature":35, "smoke":200, "density":0.8, "location":{"x":10, "y":5, "z":20}}. Based on this data, the impassable areas in the BIM model are dynamically updated. For example, when the smoke concentration exceeds 150, the corresponding area is marked as impassable. The optimal evacuation route was calculated using the A-star algorithm. The starting point was set to the current location (10, 5, 20) and the end point was the nearest safe exit. Euclidean distance was used as the heuristic function, and pedestrian density was considered as the path weight. Finally, the random forest algorithm was used to optimize the spatial layout of the 3D road network. The model was trained using node coordinates, connectivity, and pedestrian density as features, and the optimized spatial layout was obtained through training with 500 decision trees.

[0074] In this way, the present invention can generate and update a visual three-dimensional evacuation road network map in real time, providing intuitive and accurate evacuation guidance for building managers and users. Compared with the existing technology, the technical solution of the present invention has significant advantages. Traditional evacuation road network generation methods are usually based on static building models, which are difficult to adapt to real-time changes. The present invention, by introducing three-dimensional digital twin technology and real-time data analysis, can dynamically adjust the evacuation road network to better respond to emergencies. In addition, the A-star algorithm and random forest algorithm used in the present invention can generate optimal evacuation routes more quickly and accurately than traditional path planning methods, while taking into account a variety of complex factors, such as crowd density and obstacle distribution. This dynamic and intelligent evacuation road network generation method greatly improves evacuation efficiency and safety, and provides more reliable technical support for building emergency management.

[0075] In this embodiment, step S700 includes: step S701, obtaining real-time data from sensors inside the building and analyzing the spread of fire and smoke; step S702, determining the scope of the impassable area based on the analysis results; step S703, dynamically updating the impassable area in the visualized three-dimensional road network map through Three.js; step S704, using the A-star algorithm to recalculate the evacuation path based on the updated visualized three-dimensional road network map; step S705, updating the evacuation line inside the building based on the recalculated evacuation path; step S706, matching the updated evacuation line with the building BIM model through GIS tools to generate a dynamic road network map; step S707, generating a final evacuation plan based on the dynamic road network map.

[0076] Specifically, acquiring real-time data from sensors within buildings can be achieved in a variety of ways. For example, a distributed sensor network can be used, including temperature sensors, smoke sensors, and infrared sensors. These sensors transmit data in real time to a central processing system via wireless communication technologies such as ZigBee or Wi-Fi. Another approach is to leverage the Internet of Things (IoT) to connect various sensors to a cloud platform for real-time data collection and analysis. A variety of algorithms can be used to analyze fire and smoke spread. Computational fluid dynamics (CFD) models can be used to simulate the spread of fire and smoke, taking into account factors such as building geometry and ventilation conditions. Machine learning algorithms, such as support vector machines (SVMs) or deep learning networks, can also be used to predict fire and smoke spread trends, improving prediction accuracy by training on historical data. When determining the scope of impassable areas, multiple thresholds can be set to define the danger level. For example, when the temperature exceeds 60°C or the smoke concentration exceeds a certain threshold, the area is marked as impassable. These thresholds can be adjusted based on the characteristics and safety standards of each building. Three.js is a powerful JavaScript 3D library for dynamically updating and visualizing 3D road network maps. Three.js enables real-time rendering and interactive 3D scenes. When updating impassable areas, different colors or textures can be used to mark dangerous areas, such as red for high-risk areas and yellow for potentially dangerous areas. Additionally, animation effects such as flashing or gradients can be added to highlight recent changes. The A-star algorithm is an efficient path planning algorithm, particularly suitable for recalculating paths in dynamic environments. During implementation, algorithm performance can be optimized using different heuristic functions. For example, a heuristic function can be designed based on multiple factors such as distance, congestion, and danger level to find the safest and fastest evacuation path. When updating evacuation routes within a building, a segmented update strategy can be adopted. This strategy only updates the evacuation routes for the affected area, rather than recalculating the evacuation routes for the entire building. This significantly improves update efficiency. Furthermore, multiple alternative routes can be set to address the possibility that certain routes suddenly become unavailable. The use of GIS tools can improve the accuracy of matching evacuation routes with BIM models. For example, spatial analysis can be used to ensure that evacuation routes conform to the actual building structure, avoiding unreasonable paths such as those that pass through walls. GIS also provides a geographic coordinate system, facilitating accurate positioning and navigation within large or complex buildings. When generating dynamic road network maps, factors such as crowd density and movement speed can be considered. For example, cellular automata models can be used to simulate crowd movement, enabling more accurate prediction of congestion points and optimizing evacuation routes. The final evacuation plan can be generated using a multi-objective optimization algorithm, simultaneously considering multiple objectives such as evacuation time, route safety, and crowd distribution.

[0077] As a preferred embodiment, the technical solution of the present invention can be implemented in a large commercial complex. The complex comprises a multi-story shopping area, office space, and underground parking, with a total construction area of ​​approximately 100,000 square meters and a daily foot traffic of up to 50,000 people. First, approximately 1,000 sensors, including temperature, smoke, and infrared sensors, are installed throughout the building. These sensors transmit data to a central processing system in real time via a 5G network, with data updates occurring once per second. The central processing system uses deep learning algorithms to analyze the sensor data and predict fire and smoke spread. When a fire is detected, the system immediately initiates a dynamic evacuation plan generation process. Three.js is used to create a 3D visualization model of the entire building. In the event of a fire, the system updates the 3D model in real time, marking impassable areas in red and potentially dangerous areas in yellow. This update process has a latency of no more than 100 milliseconds, ensuring real-time information. The A-star algorithm is used to recalculate evacuation routes. This algorithm considers three key factors: distance, congestion, and danger level, with weights of 0.4, 0.3, and 0.3, respectively. The system recalculates the global evacuation route every 5 seconds to ensure real-time optimization of the route. The updated evacuation line is matched with the BIM model through GIS tools. The GIS system uses ESRI's ArcGIS platform, with an accuracy of centimeters. This ensures that the evacuation line is highly consistent with the actual building structure. The final dynamic road network map contains multiple levels of information: main evacuation routes, backup routes, danger zone markers, and estimated congestion points. The system also generates specific evacuation instructions for different areas, such as "Please use safety staircase No. 1 to evacuate the north area on the 3rd floor." Traditional evacuation systems usually rely on preset static routes and cannot adapt to dynamic changes in emergency situations. The present invention, through real-time data analysis and dynamic path planning, can adjust the evacuation strategy in a timely manner according to actual conditions. For example, some traditional systems may guide the crowd to an exit that has been blocked by fire, while the system of the present invention can identify this situation in real time and replan a safe route.

[0078] In addition, the visualization systems in the prior art can often only provide limited 2D plan views, which makes it difficult to intuitively display the three-dimensional structure of complex buildings. The 3D visualization model created by the present invention using Three.js technology can not only display the evacuation route more clearly, but also intuitively display the spatial distribution of dangerous areas, which helps managers make more accurate decisions. In terms of algorithms, many existing systems only consider the shortest path, while ignoring factors such as crowd density and degree of danger. The improved A-star algorithm adopted by the present invention comprehensively considers multiple factors and can generate safer and more efficient evacuation routes. This is especially important in buildings with high floors or complex structures, which can effectively avoid congestion in certain areas, thereby improving overall evacuation efficiency.

[0079] Finally, this invention uses GIS tools to precisely match evacuation routes to BIM models, overcoming the discrepancy between evacuation routes and actual building structures in traditional systems. This high-precision matching ensures the accuracy and enforceability of evacuation instructions, reducing potential confusion and misdirection in emergency situations.

[0080] In general, the present invention realizes a full-process dynamic evacuation system from data collection, analysis and processing to visual display, which significantly improves the building's response capability in emergency situations and the level of personnel safety.

[0081] Example 2

[0082] Figure 2 This is a structural diagram of a building dynamic evacuation road network generation system based on a BIM model in the second embodiment of the present invention. Figure 2 As shown, Embodiment 2 provides a system for generating a dynamic evacuation road network for a building based on a BIM model, comprising: an initial road network framework module, a road network data generation module, a preliminary evacuation road network generation module, an update module, an evacuation road network generation module, a visual three-dimensional road network map generation module, and an evacuation plan generation module. The initial road network framework module is configured to acquire spatial layout data and multi-layer connection information from the building BIM model to construct an initial road network framework. The road network data generation module is configured to generate revised road network data based on the initial road network framework and access restriction information. The preliminary evacuation road network generation module is configured to calculate evacuation routes that meet standards based on the revised road network data and safety regulations, thereby generating a preliminary evacuation road network. The update module is configured to acquire real-time data from sensors within the building, determine the impact of dynamic changes on the preliminary evacuation road network, and update information on impassable areas. The evacuation road network generation module is configured to recalculate evacuation routes based on the updated impassable area information, adjust the preliminary evacuation road network, and generate an optimized evacuation road network. The generated three-dimensional visual road network map is used to match the optimized evacuation road network with the building BIM model using three-dimensional digital twin technology to generate a three-dimensional visual road network map. The generated evacuation plan module is used to dynamically update the evacuation path based on the three-dimensional visual road network map and real-time data stream to generate a final evacuation plan.

[0083] In this embodiment, the module for constructing the initial road network framework includes: obtaining a spatial layout data unit, determining a position relationship unit, extracting a connection information unit, determining a connection point data unit, constructing a multi-layer road network structure unit, and generating an initial road network framework unit. The unit for obtaining spatial layout data is used to parse the geometric structure from the building BIM model and obtain the spatial layout data of walls and partitions. The unit for determining the position relationship is used to determine the position relationship of the walls and partitions based on the spatial layout data. The unit for extracting connection information is used to extract multi-layer connection information of stairs, elevators, and corridors from the building BIM model. The unit for determining the connection point data is used to determine the connection point data based on the position relationship and the multi-layer connection information. The unit for constructing a multi-layer road network structure is used to construct a multi-layer road network structure through the connection point data. The unit for generating the initial road network framework is used to optimize the multi-layer road network structure using a genetic algorithm and generate the initial road network framework.

[0084] In this embodiment, the road network data generation module includes: a spatial position relationship extraction unit, a marking unit, a road network structure generation unit, a connection point determination unit, a multi-layer road network framework construction unit, and a road network data generation unit. The spatial position relationship extraction unit is used to extract the spatial position relationship of walls and partitions based on the initial road network framework. The marking unit is used to identify access restriction information, mark equipment installation areas and temporary obstacle locations through geometric analysis. The road network structure generation unit is used to exclude impassable areas through path planning tools if there are impassable areas, and generate a revised road network structure. The connection point determination unit is used to extract multi-layer connection information from the revised road network structure and determine connection point data. The multi-layer road network framework construction unit is used to construct a revised multi-layer road network framework through a network modeling tool. The road network data generation unit is used to adjust the revised multi-layer road network framework using a path optimization algorithm to generate the revised road network data.

[0085] In this embodiment, the module for generating a preliminary evacuation road network includes: a determination unit, an evacuation route calculation unit, an extraction unit, a recalculation unit, and a preliminary evacuation road network determination unit. The determination unit is used to determine the shortest path and the maximum evacuation time threshold based on the revised road network data and the evacuation requirements in the safety specifications. The evacuation route calculation unit is used to use the Dijkstra algorithm to calculate an evacuation route that meets the standards based on the revised road network data and the maximum evacuation time threshold. The extraction unit is used to generate the preliminary evacuation road network based on the evacuation route and extract key nodes and connection relationships. The recalculation unit is used to adjust the Dijkstra algorithm parameters and recalculate the evacuation route if there are areas in the preliminary evacuation road network that do not meet the standards. The preliminary evacuation road network determination unit is used to determine the final preliminary evacuation road network through iterative optimization.

[0086] In this embodiment, the update module includes: a real-time data acquisition unit, a range speed calculation unit, a traffic capacity judgment unit, an inaccessible area marking unit, and an inaccessible area information generation unit. The real-time data acquisition unit is used to acquire real-time data from sensors inside the building, clean and standardize the real-time data. The range speed calculation unit is used to calculate the range and speed of fire or smoke diffusion based on the standardized real-time data. The traffic capacity judgment unit is used to judge whether the diffusion range exceeds the traffic capacity threshold according to a preset threshold rule. The inaccessible area marking unit is used to mark the affected area as an inaccessible area if it exceeds the traffic capacity threshold. The inaccessible area information generation unit is used to update the traffic status of the preliminary evacuation road network based on the inaccessible area and generate updated inaccessible area information.

[0087] In this embodiment, the evacuation road network generation module includes: an evacuation route recalculation unit, a road network layout generation unit, an information extraction unit, a layout adjustment unit, and an evacuation road network generation unit. The evacuation route recalculation unit is used to recalculate the evacuation route using the Dijkstra algorithm based on the updated inaccessible area information. The road network layout generation unit is used to adjust the preliminary evacuation road network based on the recalculated evacuation route to generate a new road network layout. The information extraction unit is used to extract node location information from the new road network layout. The layout adjustment unit is used to optimize the node location information through the K-means clustering algorithm and adjust the road network spatial layout. The evacuation road network generation unit is used to generate an optimized evacuation road network based on the adjusted road network spatial layout.

[0088] In this embodiment, the generation of a visual three-dimensional road network map includes: a parsing unit, a dynamic update unit, a matching unit, a three-dimensional road network map generation unit, an adjustment unit, and a visual three-dimensional road network map generation unit. The parsing unit is used to obtain real-time data from sensors inside the building through the MQTT protocol and parse it into a format recognizable by the three-dimensional engine. The dynamic update unit is used to dynamically update the impassable areas in the building BIM model based on the parsed data. The matching unit is used to match the updated building BIM model with the optimized evacuation road network. The generation of a three-dimensional road network map unit is used to generate a three-dimensional road network map through Three.js. The adjustment unit is used to use the A-star algorithm to adjust the evacuation routes in the three-dimensional road network map according to the real-time data. The generation of a visual three-dimensional road network map unit is used to optimize the spatial layout of the three-dimensional road network map through the random forest algorithm to generate a visual three-dimensional road network map.

[0089] In this embodiment, the evacuation plan generation module includes: an analysis unit, a range determination unit, a dynamic update area unit, a recalculation evacuation path unit, an evacuation line update unit, a dynamic road network map generation unit, and an evacuation plan generation unit. The analysis unit is used to obtain real-time data from sensors inside the building and analyze the spread of fire and smoke. The range determination unit is used to determine the scope of the impassable area based on the analysis results. The dynamic update area unit is used to dynamically update the impassable area in the visual three-dimensional road network map through Three.js. The recalculation evacuation path unit is used to use the A-star algorithm to recalculate the evacuation path based on the updated visual three-dimensional road network map. The evacuation line update unit is used to update the evacuation line inside the building based on the recalculated evacuation path. The dynamic road network map generation unit is used to match the updated evacuation line with the building BIM model through GIS tools to generate a dynamic road network map. The evacuation plan generation unit is used to generate a final evacuation plan based on the dynamic road network map.

[0090] The various variations and specific examples of the method for generating a dynamic evacuation road network for a building based on a BIM model provided in Example 1 are also applicable to the system for generating a dynamic evacuation road network for a building based on a BIM model provided in this embodiment. Through the above detailed description of a method for generating a dynamic evacuation road network for a building based on a BIM model, those skilled in the art can clearly understand the implementation method of the system for generating a dynamic evacuation road network for a building based on a BIM model in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here.

[0091] Example 3

[0092] Figure 3 This is a schematic diagram of the structure of an electronic device in the third embodiment of the present invention. Figure 3As shown, the third embodiment further provides an electronic device 300 , which may include: a processor 301 and a memory 302 .

[0093] Memory 302 is used to store programs. Memory 302 may include volatile memory (volatile memory), such as random-access memory (RAM), such as static random-access memory (SRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), etc. Memory may also include non-volatile memory (non-volatile memory), such as flash memory. Memory 302 is used to store computer programs (such as applications and functional modules that implement the above-mentioned methods), computer instructions, etc. The above-mentioned computer programs and computer instructions may be partitioned and stored in one or more memories 302. Furthermore, the above-mentioned computer programs, computer instructions, data, etc. may be called by processor 301.

[0094] The aforementioned computer programs, computer instructions, etc. may be stored in partitions in one or more memories 302 , and the aforementioned computer programs, computer instructions, etc. may be called by the processor 301 .

[0095] The processor 301 is configured to execute the computer program stored in the memory 302 to implement the various steps in the method involved in the above embodiment.

[0096] For details, please refer to the relevant description in the previous method embodiment.

[0097] The processor 301 and the memory 302 may be independent structures or integrated structures. When the processor 301 and the memory 302 are independent structures, the memory 302 and the processor 301 may be coupled via a bus 303 .

[0098] The electronic device of this embodiment can execute the technical solution in the above method. Its specific implementation process and technical principles are the same and will not be repeated here.

[0099] Example 4

[0100] Embodiment 4 also provides a computer-readable storage medium, including a computer program and instructions. When the computer program or instructions are run on a computer, the computer executes the method for generating a dynamic evacuation road network of a building based on a BIM model according to any embodiment of the present invention.

[0101] Computer-readable storage media include: USB flash drives, mobile hard drives, ROM, RAM, magnetic disks, or optical disks, and other media that can store program codes.

[0102] This embodiment also provides a computer program product, which includes: a computer program, which is stored in a readable storage medium. At least one processor of an electronic device can read the computer program from the readable storage medium, and at least one processor executes the computer program so that the electronic device executes the solution provided by any of the above embodiments.

[0103] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in the present disclosure can be achieved. This is not limited herein.

[0104] Note that the above are only preferred embodiments of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments, and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments and may include many other equivalent embodiments without departing from the concept of the present invention. The scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for generating a dynamic evacuation road network for a building based on a BIM model, characterized in that: include: Obtain spatial layout data and multi-layer connectivity information from the building BIM model to construct the initial road network framework; Generating revised road network data based on the initial road network framework and traffic restriction information; Calculating evacuation routes that meet safety standards based on the revised road network data and safety regulations to generate a preliminary evacuation road network; Obtaining real-time data from sensors inside the building, determining the impact of dynamic changes on the preliminary evacuation network, and updating information on impassable areas; Recalculating evacuation routes based on the updated information of the impassable area, adjusting the preliminary evacuation road network, and generating an optimized evacuation road network; By using three-dimensional digital twin technology, the optimized evacuation road network is matched with the building BIM model to generate a visual three-dimensional road network map; Dynamically update the evacuation path based on the visualized three-dimensional road network map and real-time data stream to generate a final evacuation plan; The acquiring of real-time data from sensors inside the building, determining the impact of dynamic changes on the preliminary evacuation road network, and updating information on impassable areas includes: Acquire real-time data from sensors inside the building, and clean and standardize the real-time data; Calculating the range and speed of fire or smoke spread based on the standardized real-time data; According to a preset threshold rule, determining whether the diffusion range exceeds a traffic capacity threshold; If the capacity threshold is exceeded, the affected area is marked as an impassable area; updating the traffic status of the preliminary evacuation road network according to the impassable area and generating updated impassable area information; The recalculating the evacuation routes based on the updated impassable area information, adjusting the preliminary evacuation road network, and generating an optimized evacuation road network includes: Recalculating the evacuation route using the Dijkstra algorithm according to the updated impassable area information; Adjusting the preliminary evacuation road network according to the recalculated evacuation routes to generate a new road network layout; Extracting node location information from the new road network layout; Optimizing the node location information through the K-means clustering algorithm and adjusting the road network spatial layout; Based on the adjusted road network spatial layout, an optimized evacuation road network is generated.

2. The method for generating a dynamic evacuation road network for a building based on a BIM model according to claim 1, wherein: The process of acquiring spatial layout data and multi-layer connection information from the building BIM model and constructing an initial road network framework includes: Analyze the geometric structure from the building BIM model to obtain the spatial layout data of walls and partitions; Determining the positional relationship between the walls and partitions according to the spatial layout data; Extract multi-level connectivity information of stairs, elevators, and corridors from the building BIM model; determining connection point data according to the positional relationship and the multi-layer connection information; Constructing a multi-layer road network structure through the connection point data; A genetic algorithm is used to optimize the multi-layer road network structure and generate the initial road network framework.

3. The method for generating a dynamic evacuation road network for a building based on a BIM model according to claim 1, wherein: Generating the revised road network data according to the initial road network framework and traffic restriction information includes: Extracting the spatial position relationship between walls and partitions according to the initial road network framework; Through geometric analysis, identify access restriction information and mark equipment installation areas and temporary obstacle locations; If there are impassable areas, the impassable areas are excluded through the path planning tool to generate a revised road network structure; Extracting multi-layer connection information from the revised road network structure and determining connection point data; Use network modeling tools to build a revised multi-layer road network framework; A path optimization algorithm is used to adjust the revised multi-layer road network framework to generate the revised road network data.

4. The method for generating a dynamic evacuation road network for a building based on a BIM model according to claim 1, wherein: Calculating evacuation routes that meet safety standards based on the revised road network data and safety regulations to generate a preliminary evacuation road network includes: Based on the revised road network data and in combination with the evacuation requirements in the safety regulations, determining the shortest path and the maximum evacuation time threshold; Using the Dijkstra algorithm, according to the corrected road network data and the maximum evacuation time threshold, calculate an evacuation route that meets the safety standards; According to the evacuation routes, the preliminary evacuation road network is generated, and key nodes and connection relationships are extracted; If there are areas in the preliminary evacuation road network that do not meet safety standards, adjust the Dijkstra algorithm parameters and recalculate the evacuation route; The final preliminary evacuation road network is determined through iterative optimization.

5. The method for generating a dynamic evacuation road network for a building based on a BIM model according to claim 1, wherein: The method of matching the optimized evacuation road network with the building BIM model through the three-dimensional digital twin technology to generate a visual three-dimensional road network diagram includes: Obtain real-time data from sensors inside the building through the MQTT protocol and parse it into a format that can be recognized by the 3D engine; Dynamically update the impassable areas in the building BIM model based on the parsed data; Match the updated building BIM model with the optimized evacuation road network; Generate a 3D road network map using Three.js; Adopting the A-star algorithm to adjust the evacuation routes in the three-dimensional road network map according to the real-time data; The spatial layout of the three-dimensional road network map is optimized by a random forest algorithm to generate a visual three-dimensional road network map.

6. A building dynamic evacuation road network generation system based on BIM model, characterized in that: include: Constructing the initial road network framework module, which is used to obtain spatial layout data and multi-layer connection information from the building BIM model to construct the initial road network framework; A road network data generation module is used to generate revised road network data based on the initial road network framework and traffic restriction information; A preliminary evacuation road network generation module is used to calculate evacuation routes that meet safety standards based on the revised road network data and safety regulations, and generate a preliminary evacuation road network; An update module is used to obtain real-time data from sensors inside the building, determine the impact of dynamic changes on the preliminary evacuation road network, and update information on impassable areas; An evacuation road network generation module is used to recalculate evacuation routes based on the updated impassable area information, adjust the preliminary evacuation road network, and generate an optimized evacuation road network; Generating a visual three-dimensional road network diagram for matching the optimized evacuation road network with the building BIM model through three-dimensional digital twin technology to generate a visual three-dimensional road network diagram; An evacuation plan generation module is used to dynamically update the evacuation path based on the visual three-dimensional road network map and real-time data stream to generate a final evacuation plan; The acquiring of real-time data from sensors inside the building, determining the impact of dynamic changes on the preliminary evacuation road network, and updating information on impassable areas includes: Acquire real-time data from sensors inside the building, and clean and standardize the real-time data; Calculating the range and speed of fire or smoke spread based on the standardized real-time data; According to a preset threshold rule, determining whether the diffusion range exceeds a traffic capacity threshold; If the capacity threshold is exceeded, the affected area is marked as an impassable area; updating the traffic status of the preliminary evacuation road network according to the impassable area and generating updated impassable area information; The recalculating the evacuation routes based on the updated impassable area information, adjusting the preliminary evacuation road network, and generating an optimized evacuation road network includes: Recalculating the evacuation route using the Dijkstra algorithm according to the updated impassable area information; Adjusting the preliminary evacuation road network according to the recalculated evacuation routes to generate a new road network layout; Extracting node location information from the new road network layout; Optimizing the node location information through the K-means clustering algorithm and adjusting the road network spatial layout; Based on the adjusted road network spatial layout, an optimized evacuation road network is generated.

7. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for generating a dynamic evacuation road network of a building based on a BIM model as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that The method comprises a computer program and instructions. When the computer program or the instructions are run on a computer, the computer is caused to execute the method for generating a dynamic evacuation road network of a building based on a BIM model according to any one of claims 1 to 5.

Images