Park security risk dynamic visualization method and system based on BIM-GIS semantic fusion

Abstract

The invention provides a park security risk dynamic visualization method and system based on BIM-GIS semantic fusion, and belongs to the technical field of digital twinning and emergency management. The method aims at solving the problems that in the prior art, calculation of a building information model and a geographic information system model is separated, and visual occlusion exists in staticizing and three-dimensional visualization of a risk evolution model. The method comprises the following steps: constructing an attribute voxel grid fusing a building information model and a geographic information system model as a unified space grid; and based on the dynamic permeability model, calculating a flux of the risk medium propagating and diffusing between adjacent units of the attribute voxel grid. Wherein the dynamic permeability model can dynamically determine the permeability according to material attributes of adjacent units and real-time environmental parameters, and particularly, when a boundary material is a fragile material and the real-time temperature exceeds a heat-resistant threshold value of the boundary material, the permeability is adjusted from a first preset value representing a blocking state to a second preset value representing a connected state. The invention further provides a corresponding system. According to the method, the simulation accuracy is improved through the dynamic permeability model, indoor and outdoor seamless calculation is realized through a unified grid, and the emergency decision-making efficiency and the plan reliability can be improved in combination with adaptive visualization and dynamic path planning.

Description

Technical Field

[0001] This application relates to the fields of digital twin and emergency management technology, and in particular to a method and system for dynamic evolution and adaptive visualization of security risks. Background Technology

[0002] In security management of smart parks and large buildings, Building Information Modeling (BIM) is typically used to describe the detailed internal structure of a building, and combined with Geographic Information Systems (GIS) to present the macroscopic outdoor environment. To simulate emergencies such as fires and toxic gas leaks, existing technologies usually generate a computational mesh based on the fused 3D model and use methods based on fluid dynamics and other models to calculate the diffusion process of hazardous media within the building.

[0003] However, existing technologies have the following shortcomings in application: First, the data fusion of Building Information Modeling (BIM) and Geographic Information Systems (GIS) is mostly limited to the overlay at the visualization level. The two are separate in terms of underlying data structure and computational topology. This makes it impossible to continuously and realistically calculate the cross-spatial physical diffusion process of risk media from indoors to outdoors or from outdoors to indoors during risk simulation, resulting in data fragmentation and computational discontinuity. Second, most existing disaster simulations are based on static, fixed building physical structures, while in real disasters, the physical environment is dynamically changing. For example, high temperatures may cause glass curtain walls to shatter, creating new air convection channels, and explosions may cause walls to collapse, blocking existing channels. Existing technologies cannot predict and update these structural changes in real time based on physical laws, leading to static risk evolution models and distorted simulation results. Furthermore, in 3D emergency command scenarios, non-critical structures such as building exterior walls and floors often obscure core risk elements such as internal fire sources and trapped personnel, making it difficult for command personnel to intuitively "see through" the building interior, resulting in a heavy information cognitive burden and low decision-making efficiency. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for dynamic evolution and adaptive visualization of safety risks, in order to solve the problems existing in the prior art, such as the separation of calculation between building information model and geographic information system model, the static nature of risk evolution model, and visual occlusion in 3D visualization.

[0005] To achieve the above objectives, this application provides a method for dynamic evolution of security risks, comprising the following steps: Step 1: Constructing a unified spatial grid, wherein the unified spatial grid is an attribute voxel grid integrating a building information model and a geographic information system model; Step 2: Calculating the flux of the risk medium propagating and diffusing between adjacent units of the attribute voxel grid based on a dynamic permeability model, wherein the dynamic permeability model is used to dynamically determine the permeability between adjacent units based on the material properties of the adjacent units and real-time environmental parameters; the calculation of the dynamic permeability model includes: obtaining the material properties of the boundary corresponding to the adjacent unit and real-time environmental physical quantities; when the boundary material is a fragile material, and the real-time temperature in the real-time environmental physical quantities exceeds a preset heat resistance threshold of the fragile material, adjusting the permeability between the adjacent units from a first preset value representing the blocking state to a second preset value representing the connecting state, wherein the second preset value is greater than the first preset value.

[0006] Optionally, the attribute information of the attribute voxel mesh includes the material type and physical state of each mesh cell.

[0007] Optionally, the real-time environmental physical quantities are derived from IoT sensor data or simulation results based on the risk evolution process.

[0008] Optionally, the real-time environmental physical quantity also includes pressure; and when the pressure borne by the boundary of the adjacent unit exceeds the preset structural yield strength, the permeability between the adjacent units is increased.

[0009] Optionally, the method further includes: performing adaptive perspective visualization rendering based on the propagation and diffusion results of the risk medium.

[0010] Optionally, the adaptive perspective visualization rendering includes: calculating the visual importance score of each grid cell, wherein the visual importance score is directly proportional to the risk level of the grid cell and inversely proportional to the distance of the grid cell from the viewing point; and adjusting the rendering transparency of the corresponding components in the 3D model according to the visual importance score.

[0011] Optionally, the adaptive perspective visualization rendering further includes: when the importance score of the occluding component located between the viewing point and the occluded high-importance target is lower than that of the high-importance target, further reducing the rendering transparency of the occluding component.

[0012] Optionally, the method further includes: performing dynamic emergency path planning based on the dynamically evolving risk distribution.

[0013] Optionally, the dynamic emergency path planning includes: associating the passage cost of each node in the path planning network with the real-time risk level of the node's location; and setting the passage cost of any node to infinity when the risk level of any node exceeds a preset lethal threshold.

[0014] This application also provides a dynamic evolution system for safety risks, comprising: a data fusion module for constructing an attribute voxel grid that integrates a building information model and a geographic information system model; a risk evolution simulation engine for calculating the flux of a risk medium propagating and diffusing between adjacent units of the attribute voxel grid based on a dynamic permeability model; wherein the risk evolution simulation engine is configured to: acquire the material properties of the boundary corresponding to the adjacent unit and real-time environmental physical quantities; when the boundary material is a fragile material and the real-time temperature in the real-time environmental physical quantity exceeds a preset heat resistance threshold of the fragile material, adjust the permeability between the adjacent units from a first preset value representing the blocking state to a second preset value representing the connected state, wherein the second preset value is greater than the first preset value; an adaptive visualization module for performing adaptive perspective visualization rendering based on the propagation and diffusion results of the risk medium; and a dynamic path planning module for performing dynamic emergency path planning based on the dynamically evolving risk distribution.

[0015] Compared with existing technologies, the technical solution provided in this application has the following beneficial effects: 1. By introducing a dynamic permeability model, it is possible to predictively simulate structural damage to physical boundaries caused by high temperature or high pressure during disasters, such as the "broken window effect," based on real-time environmental parameters and the physical properties of materials. This solves the shortcomings of existing technologies that rely on static models and cannot reflect dynamic environmental changes, making risk evolution prediction closer to physical reality and improving simulation accuracy. 2. By constructing a unified attribute voxel grid, the data structure of indoor and outdoor spaces is unified at the physical calculation level, realizing seamless and continuous diffusion simulation of risk media in indoor and outdoor spaces, overcoming the problems of data fragmentation and discontinuous calculation in existing technologies. 3. Through adaptive perspective visualization rendering, the transparency of obstructing components can be automatically reduced according to the risk distribution and observation viewpoint, enabling "perspective" observation of core risk points inside the building, significantly reducing the information cognitive burden of command personnel and improving emergency decision-making efficiency. 4. By conducting real-time path planning based on dynamically evolving risk fields, safe evacuation paths that better match the actual situation on site can be generated, avoiding the risk of static plans failing during sudden disaster changes, and improving the reliability of emergency plans and the success rate of personnel evacuation. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram illustrating the dynamic evolution of risk media provided in the embodiments of this application; Figure 2 A flowchart illustrating a method for dynamic evolution and adaptive visualization of security risks provided in this application embodiment; Figure 3 This is a functional module architecture diagram of a security risk dynamic evolution and adaptive visualization system provided in an embodiment of this application.

[0018] The main reference numerals in the attached figures are explained as follows: 100 - Indoor; 101 - Outdoor; 200 - Physical boundary; 300 - Risk flux; 400 - Breakpoint; S10 - Step to construct a unified spatial grid; S20 - Step to calculate risk evolution based on a dynamic permeability model; S30 - Step to perform adaptive perspective visualization rendering; S40 - Step to perform dynamic emergency path planning; 610 - Data fusion module; 620 - Risk evolution simulation engine; 630 - Adaptive visualization module; 640 - Dynamic path planning module. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0020] Example 1 This embodiment provides a complete workflow for a method of dynamic evolution and adaptive visualization of safety risks applied in fire scenarios. (Refer to...) Figure 2 The document illustrates a flowchart of a dynamic evolution and adaptive visualization method for security risks provided in an embodiment of this application. This method can be executed by a computer system and specifically includes the following steps.

[0021] Step S10: Construct a unified spatial grid. This step aims to address the disconnect between traditional Building Information Modeling (BIM) and Geographic Information System (GIS) models at the data and computational levels, providing a unified computational foundation for subsequent integrated indoor and outdoor risk evolution calculations. In a specific application scenario, such as the comprehensive emergency management of a high-tech park, the system first loads BIM data describing a laboratory building within the park, along with GIS data covering the entire park. The BIM data can adopt standard formats such as industrial basic data types, containing precise 3D geometric, material, and semantic information of building components such as beams, slabs, columns, walls, doors, and windows. The GIS data can include vector or raster data describing the park's terrain, roads, green spaces, and other building outlines.

[0022] Understandably, since Building Information Modeling (BIM) typically uses a local coordinate system while Geographic Information Systems (GIS) use a geographic coordinate system, the primary task of data fusion is coordinate system alignment and transformation. The system reads preset coordinate transformation parameters (e.g., the latitude, longitude, elevation, and rotation angle of the building model's origin in the geographic coordinate system) to transform all geometric entities in the BIM to a unified geographic coordinate system, thereby achieving precise spatial registration with the GIS model.

[0023] After coordinate alignment, the system discretizes the entire fused 3D space to generate a unified attribute voxel mesh. A voxel, as the smallest discrete unit in 3D space, can be analogous to a pixel in a 2D image. In this embodiment, the voxel size can be set to 0.5m × 0.5m × 0.5m, which achieves a good balance between computational efficiency and simulation accuracy. It should be noted that for large-scale outdoor open spaces, non-uniform mesh structures such as octrees can also be used to finely subdivide key areas (such as the area around buildings) while coarsely subdividing open areas to optimize storage and computational overhead.

[0024] Subsequently, the system assigns attribute information to each voxel unit in the three-dimensional discretized spatial mesh. This process involves traversing each voxel, determining the component of the original model (Building Information Model or Geographic Information System model) where its center point is located, and inheriting the attributes of that component. As an optional implementation, the attribute information includes at least material type and physical state. For example, if the center point of a voxel falls within a glass curtain wall in the Building Information Model, the material type attribute of that voxel is marked as "glass," and its initial physical state is marked as "intact." Similarly, other voxels may be marked with material types such as "concrete," "steel," "air," and "soil," and physical states such as "open," "closed," "intact," and "damaged." This attribute information is stored in a database or data structure associated with the voxel mesh, providing crucial input parameters for subsequent physical simulation calculations.

[0025] Step S20 involves calculating the risk evolution based on a dynamic permeability model. This step is the core of this application's scheme, used to simulate the dynamic propagation process of risky media (such as smoke and heat generated by a fire) within a unified spatial grid. When a safety event occurs, for example, if an IoT smoke sensor and a temperature sensor in a chemical laboratory on the third floor of the experimental building simultaneously issue alarm signals, the system marks the voxel cell corresponding to the sensor's location as the initial risk source. Specifically, the risk concentration (such as smoke concentration) and temperature value of that voxel cell can be instantly increased as the initial conditions for the entire evolution calculation.

[0026] Subsequently, the risk evolution simulation engine begins iterative calculations at each time step (e.g., 0.1 seconds). The core of the calculation lies in solving the flux of the risk medium propagating and diffusing between adjacent voxel units. The magnitude of this flux mainly depends on the concentration gradient, temperature gradient, and permeability at the interface between adjacent voxels. In traditional methods, permeability is usually static; for example, walls are always considered as blocks, and doors are always considered as closed. This application's embodiment employs a dynamic permeability model, the key being that the permeability between adjacent units changes dynamically based on the unit's material properties and real-time environmental parameters.

[0027] Reference Figure 1This is a schematic diagram illustrating the dynamic evolution of the risk medium provided in this application embodiment. A core computational logic of this model lies in simulating the dynamic failure process of physical boundaries. Specifically, in each computational iteration, the system acquires the material properties of the boundaries corresponding to adjacent voxel units, as well as real-time environmental physical quantities (e.g., temperature) obtained through IoT sensors or the previous round of simulation calculations. When the system detects that the material property of the physical boundary 200 between two adjacent voxel units (one located indoors 100, and the other outdoors 101) is a fragile material (e.g., ordinary glass), and the real-time temperature of the environment where the boundary is located exceeds the preset heat resistance threshold of the material (e.g., the heat resistance threshold T_break for glass can be set to 800 degrees Celsius), the system triggers a state change. Figure 1 As shown in (a), when the temperature is below the threshold, the physical boundary 200 remains intact, and its permeability κ is set to a first preset value characterizing the blockage state (e.g., a minimum of 0.0), at which point almost no risk flux 300 is generated. Accordingly, as Figure 1 As shown in (b), when the temperature exceeds the threshold, the system determines that the glass window has thermally cracked due to high temperature. It then updates the physical state attribute of the boundary to "broken" and adjusts the permeability κ between the window and the adjacent voxel from a first preset value (0.0) to a second preset value (e.g., a maximum value of 1.0) that represents the connectivity state. Since the second preset value is greater than the first preset value, this abrupt change simulates the failure of the physical boundary, forming a new break point 400 between the indoor 100 and the outdoor 101. This allows the originally blocked risk medium (such as high-temperature flue gas) to rapidly diffuse from the indoor to the outdoor with a large risk flux 300, which is the so-called "broken window effect".

[0028] In this way, risk evolution is no longer limited to static building structures, but can simulate the physical boundary failure process caused by sudden environmental changes during a disaster with high fidelity, thereby achieving accurate prediction of the cross-space, continuous, and dynamic evolution of safety risks. As an optional implementation method, the sources of real-time environmental physical quantities can be diverse, including real-time data uploaded by IoT sensors deployed on-site (such as temperature, pressure, and gas concentration sensors), or simulation calculation results based on the risk evolution process itself, such as the temperature rise of surrounding voxels calculated by a fire source model.

[0029] Step S30: Perform adaptive perspective visualization rendering. This step aims to address the visual occlusion problem in 3D visualization, helping emergency command personnel quickly and intuitively grasp core risk information within the building. On the large screen in the remote emergency command center, the system adaptively renders the 3D model with transparency based on the real-time risk distribution results obtained from risk evolution calculations (e.g., smoke concentration and temperature field of each voxel) and the user's current viewing point.

[0030] In one specific implementation, this step first calculates the visual importance score of each cell in the spatial grid or its corresponding 3D model component. This score is a comprehensive evaluation index used to quantify the attention value of an object in the current emergency scenario. As a preferred implementation, the calculation of the visual importance score can follow these principles: the score is directly proportional to the risk level of the cell, i.e., areas with higher smoke concentration and higher temperature have higher importance scores; the score is inversely proportional to the distance of the cell from the observation point, i.e., objects closer to the observer are generally considered more important; in addition, the system can directly mark special targets such as alarm sources, trapped personnel, and critical equipment as having the highest importance. One possible calculation formula is: ,in It is the visual importance score of voxel V. It is its risk level. It is its distance from the observation point C. It is a Boolean value indicating whether it is a critical target. These are adjustable weighting coefficients.

[0031] Based on the calculated visual importance score, the system dynamically adjusts the rendering transparency of corresponding components in the 3D model. For example, the transparency α of a component can be correlated with its importance score. A reverse mapping is performed: components with higher scores are less transparent (α value close to 1.0), while components with lower scores are more transparent (α value close to 0.0). In this way, high-risk areas will be prominently displayed as solids, while ordinary walls without risk will become translucent.

[0032] Furthermore, to achieve an "X-ray" perspective of key internal risk points, this step includes a crucial line-of-sight occlusion judgment logic. The system emits a virtual ray from the observation point towards each high-importance target in the scene (such as an internal fire source). If the ray path crosses one or more other components, the system compares the visual importance scores of these occluding components with those of the occluded high-importance target. When a low-importance occluding component (such as an exterior wall) lies in the line-of-sight path between the observation point and a high-importance target it occludes, the system will force a further reduction in the rendering transparency of the occluding component, for example, by multiplying its current transparency value by a decay factor (such as 0.2), making it nearly completely transparent or only displayed as a wireframe outline. In this way, the commander's line of sight can "penetrate" the building's exterior walls, allowing a clear view of the location of fire points inside the building, the extent of smoke spread, and icons of trapped personnel. This significantly reduces the cognitive burden and improves decision-making efficiency.

[0033] Step S40: Perform dynamic emergency route planning. This step aims to provide real-time, safe evacuation routes for people within the building by utilizing the dynamically evolving risk distribution results. The system maintains a route planning network corresponding to the building's spatial topology, which consists of a series of nodes (such as rooms, corridor corners, and stairwells) and edges connecting the nodes (such as corridors, doors, and staircases).

[0034] In traditional static path planning, the travel cost of an edge typically only considers its physical length. However, in this embodiment, the travel cost of a path is dynamically changing. Specifically, the system associates the travel cost of each node or edge in the path planning network with the real-time risk level of that location. For example, the travel cost of an edge can be defined as: ,in It is the physical length of the edge. It is the average risk level (e.g., average flue gas concentration) of the voxel area traversed by the edge. This is a risk weighting coefficient. Therefore, the higher the risk level, the greater the cost of passage in that area.

[0035] More importantly, when the dynamic evolution calculation determines that the risk level (such as flue gas concentration or temperature) of any node exceeds a preset lethal threshold (e.g., carbon monoxide concentration exceeding 12800 ppm or temperature exceeding 150 degrees Celsius), the system sets the passage cost of that node or all edges passing through that node to a maximum value (i.e., equivalent to infinity). This achieves dynamic "blocking" of the dangerous area at the level of path planning algorithms such as Dijkstra's algorithm or A* algorithm, thereby ensuring that any subsequently calculated path will automatically avoid this area.

[0036] In the fire scenario of this embodiment, when the risk evolution simulation engine calculates that the risk level of the third-floor corridor exceeds the safety threshold due to smoke spread, the dynamic path planning module will immediately update the passage cost of that corridor in the path network to infinity. At this time, if the system needs to replan the evacuation route for people in the building, it will calculate based on the updated cost network and automatically generate a new route that avoids the fire corridor on the third floor and guides people to evacuate via an unaffected safety staircase on the other side. This route information can be pushed to relevant personnel through mobile applications, indoor broadcasts, or smart signs, thereby achieving dynamic and reliable emergency evacuation guidance.

[0037] Example 2 This embodiment is a further extension of the dynamic permeability model in Embodiment 1, aiming to demonstrate that the model can respond not only to temperature changes, but also to other real-time environmental physical quantities such as pressure, thus making it suitable for simulating more types of disaster scenarios such as explosions and high-pressure gas leaks.

[0038] As an optional implementation, the dynamic permeability model incorporates pressure-based decision-making logic. In this case, real-time environmental physical quantities include not only temperature but also pressure. For components such as walls, doors, and windows in a building, the permeability κ at the corresponding boundary in the attribute voxel mesh can be defined as a function related to the pressure difference ΔP across the boundary.

[0039] Scenario 1: A dust explosion occurs in a workshop within a chemical plant. The explosion instantly generates a powerful shock wave, creating a rapidly propagating high-pressure zone. The risk evolution simulation engine calculates the propagation process of this shock wave within a unified spatial grid. When the shock wave reaches a non-load-bearing wall, the system acquires the instantaneous pressure values ​​of voxels on both sides of the wall and calculates the pressure ΔP borne by the wall. The system pre-stores the structural yield strength of the wall material. When ΔP is detected to be much greater than Upon detection, the system determined that the wall structure had been destroyed by the shock wave. Accordingly, the system updated the physical state of all voxel units corresponding to the wall to "broken," and instantly increased its permeability κ from a value close to 0 to 1.0. This simulated the physical process of the wall being destroyed, creating a large gap, allowing any subsequent toxic or harmful gases or fire to rapidly diffuse into the adjacent space through this newly formed gap, making the simulation results more consistent with the realities of an explosion disaster.

[0040] Scenario 2: Simulates a leak in a liquid hazardous chemical storage tank within a sealed warehouse. The leaked liquid forms a pool on the ground and continues to accumulate. The system, based on the real-time calculated liquid depth h, uses hydrostatic formulas... (Where ρ is the liquid density and g is the acceleration due to gravity) Calculate the static pressure exerted by the accumulated liquid on the bottom of the warehouse sealing door. The sealing strip at the bottom of this door is designed with a certain pressure resistance limit. In the dynamic permeability model, the permeability κ of this door can be defined as a piecewise or continuous function of pressure P. For example, when the pressure P is less than the sealing strip's withstand limit... When the pressure P exceeds a certain threshold, the permeability κ is a minimum value representing normal gap leakage (e.g., 1e-6); when the pressure P exceeds a certain threshold... At that time, the permeability κ does not abruptly become 1, but increases accordingly with the increase of pressure, in order to simulate the gradual process of the sealing strip deforming due to high pressure, the sealing performance gradually failing, and the leakage continuously increasing.

[0041] By using pressure as another key input variable in the dynamic permeability model, the solution provided in this application can extend the simulation capability from thermal disasters such as fires to force-induced disasters such as explosions and high-pressure leaks, greatly enhancing the applicability and simulation fidelity of the system.

[0042] Example 3 This embodiment aims to illustrate in detail the various implementations of the adaptive perspective visualization module, demonstrating that it can provide customized and more targeted visualization strategies based on different user roles and emergency task requirements, rather than being limited to the general "X-ray" perspective mode described in Embodiment 1.

[0043] In the same fire incident, the system can provide differentiated visualizations for users with different roles.

[0044] Mode 1: Augmented Reality-Assisted Decision-Making Mode for On-Site Firefighters. Firefighters entering the fire scene can wear augmented reality glasses, which run the client program of this application system. In this mode, in addition to performing conventional perspective processing on walls obscuring fire points, the adaptive visualization module will adjust the calculation logic of visual importance scores according to the specific task requirements of the firefighters. For example, additional weighting terms are added to the visual importance calculation formula, such as... .in, Used to determine whether voxel V corresponds to fire protection facilities (such as fire hydrants, fire extinguishers, sprinkler heads), This is used to determine whether it belongs to a ventilation system duct. This is done by assigning weights to these two items. and By setting a high positive value, the system can highlight the location of these critical facilities in a conspicuous manner, such as by brightening or flashing, within the firefighters' field of vision, even if they are hidden behind walls or in ceilings. Simultaneously, highlighting ventilation ducts helps firefighters quickly determine the possible paths of smoke diffusion, thereby enabling them to develop more effective smoke extraction and attack strategies.

[0045] Mode Two: Simplified Route Navigation Mode for People Being Evacuated from Buildings. For the general public in dangerous buildings, the primary need is to quickly find the safest escape route. Complex 3D building models and risk field information may actually cause confusion. Therefore, when a user selects the "trapped person" role through the mobile application, the system switches to this mode. In this mode, the adaptive visualization module no longer performs complex 3D perspective rendering but instead performs dimensionality reduction. The system projects and simplifies the 3D model of the user's floor and the optimal evacuation route calculated by the dynamic route planning module into a 2D planar evacuation map. On this map, a bright green dynamic arrow clearly indicates the direction of travel, while a red semi-transparent area covers and marks currently known danger zones. This simplified and information-focused view allows trapped people to effortlessly understand escape instructions even under stress, thereby maximizing evacuation efficiency and success rate.

[0046] Mode 3: Structural Engineer Mode for Post-Disaster Assessment. After a disaster, the damage to buildings needs to be assessed. Structural engineers can use this mode at this time. The adaptive visualization module reads log data recorded by the risk evolution simulation engine throughout the simulation process, detailing every change in the physical state of each voxel unit. In this mode, the visualization module iterates through all components and uses special color coding (e.g., highlighting in red) in the 3D model for all components whose physical state changed during the disaster (e.g., windows changing from "intact" to "broken," or steel beams whose strength decreased due to high temperatures). Engineers can generate a structural damage assessment map of the building with a single click, visually viewing the location and extent of all damaged components, providing precise data support for subsequent reinforcement, repair, or demolition decisions.

[0047] The implementation of the above-mentioned multiple modes demonstrates that the visualization solution provided in this application has high flexibility and scalability. It can provide users with different roles with information that best matches their cognitive habits and task needs by adjusting the visualization strategy, thus realizing the transformation from "information presentation" to "knowledge push".

[0048] Example 4 This embodiment describes a specific, deployable distributed system architecture for the security risk dynamic evolution and adaptive visualization system provided in this application. (Refer to...) Figure 3 This is a functional module architecture diagram of a security risk dynamic evolution and adaptive visualization system provided in this application embodiment. To balance the massive computational load in large-scale scenarios and support access from various types of terminal devices, the system can adopt a typical client-server separation deployment architecture.

[0049] like Figure 3 As shown, the entire system can be logically divided into a data layer, an engine layer, and an application layer.

[0050] The server-side architecture, housing the engine and data layers, can consist of a cluster of high-performance servers or cloud platform instances, responsible for all data storage and intensive computing tasks. It primarily includes the following core functional modules: 1. Data Fusion Module 610: This module reads raw data from the Building Information Modeling (BIM) database and Geographic Information System (GIS) database in the data layer, performs coordinate transformation, data cleaning, and model fusion, ultimately generating an attribute voxel mesh covering the entire scene. This process is typically performed as a preprocessing step, and the generated unified spatial mesh model is stored in a dedicated spatial database for use by other modules. 2. Risk Evolution Simulation Engine 620: As the core computing unit of the system, it runs continuously as a background service. This engine loads the attribute voxel mesh model and subscribes to sensor data in real-time from the IoT platform in the data layer. Internally, it implements an evolutionary algorithm based on a dynamic permeability model, updating the distribution of the entire risk field (temperature, pressure, smoke concentration, etc.) at each time step and writing the results to shared memory or a cache in real-time. 3. Dynamic Path Planning Module 640: This module also runs as a background service, subscribing to the risk field data output by the Risk Evolution Simulation Engine 620. Once a significant change in the risk field is detected (e.g., a new area becomes dangerous), it automatically triggers an update to the path cost network and can recalculate the optimal global or local evacuation path based on preset strategies or front-end requests. 4. Adaptive Visualization Module 630: In a distributed architecture, this module's functionality is further subdivided into server-side visualization computing services and client-side rendering execution. The server-side visualization computing service is responsible for performing computationally intensive tasks.

[0051] The client layer carries the application layer and can be an application installed on an emergency command center PC workstation, a field commander's tablet, or a firefighter's augmented reality helmet. The client is mainly responsible for graphics rendering and user interaction. It caches the 3D geometric models of buildings and scenes locally but does not need to perform complex physical simulations and occlusion detection.

[0052] The specific workflow under this distributed architecture is as follows: 1. Interaction begins on the client side. When a user (such as a commander) moves or rotates their virtual camera in the 3D scene, the client sends its latest viewpoint information (including camera position coordinates and orientation vector) to the server-side visualization computing service via the network. 2. Upon receiving the request, the visualization computing service immediately requests the latest full-scene risk field data from the risk evolution simulation engine. 3. The risk evolution simulation engine reads the data from its internal cache and returns it to the visualization computing service. 4. The visualization computing service performs all the complex calculations on the server side, including combining the client's viewpoint information and risk field data to calculate the visual importance score for each component in the scene. 5. Subsequently, the visualization computing service performs line-of-sight occlusion judgment to identify all occluded components that need to be forcibly made transparent. Finally, instead of transmitting large 3D models or voxel data, it generates a set of lightweight rendering instructions. 6. The visualization computing service returns this set of rendering instructions to the client. 7. After receiving the instruction, the client only needs to iterate through the instruction list, find the corresponding component in the local rendering engine, and update its material properties such as transparency to the values ​​specified in the instruction. This process involves very little computation and can be completed very quickly.

[0053] This architecture, which separates computation from rendering, greatly reduces the performance requirements of client hardware, enabling complex dynamic simulations and adaptive visualizations to run smoothly on a variety of terminal devices, and significantly enhancing the system's scalability, availability, and cross-platform capabilities.

[0054] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for dynamic visualization of park safety risks based on BIM-GIS semantic fusion, characterized in that, Includes the following steps: Step 1: Construct a unified spatial grid, which is an attribute voxel grid that integrates building information model and geographic information system model; Step 2: Based on the dynamic permeability model, calculate the flux of the risk medium propagating and diffusing between adjacent cells of the attribute voxel grid, wherein the dynamic permeability model is used to dynamically determine the permeability between adjacent cells according to the material properties of the adjacent cells and real-time environmental parameters. The calculation of the dynamic permeability model includes: obtaining the material properties of the boundary corresponding to the adjacent unit and the real-time environmental physical quantity; when the material of the boundary is a fragile material and the real-time temperature in the real-time environmental physical quantity exceeds the preset heat resistance threshold of the fragile material, the permeability between the adjacent units is adjusted from a first preset value representing the blocking state to a second preset value representing the connecting state, wherein the second preset value is greater than the first preset value.

2. The method according to claim 1, characterized in that, The attribute information of the attribute voxel mesh includes the material type and physical state of each mesh cell.

3. The method according to claim 1, characterized in that, The real-time environmental physical quantities are derived from IoT sensor data or simulation results based on the risk evolution process.

4. The method according to claim 1, characterized in that, The real-time environmental physical quantities also include pressure; and when the pressure borne by the boundary of the adjacent unit exceeds the preset structural yield strength, the permeability between the adjacent units is increased.

5. The method according to claim 1, characterized in that, The method further includes: Based on the propagation and diffusion results of the aforementioned risk medium, adaptive perspective visualization rendering is performed.

6. The method according to claim 5, characterized in that, The adaptive perspective visualization rendering includes: Calculate the visual importance score for each grid cell, which is directly proportional to the risk level of the grid cell and inversely proportional to the distance of the grid cell from the observation viewpoint; Adjust the rendering transparency of the corresponding components in the 3D model based on the visual importance score.

7. The method according to claim 6, characterized in that, The adaptive perspective visualization rendering also includes: When the importance score of an occluding component located between the viewpoint and the occluded high-importance target is lower than that of the high-importance target, the rendering transparency of the occluding component is further reduced.

8. The method according to claim 1, characterized in that, The method further includes: Dynamic emergency response path planning is carried out based on the dynamically evolving risk distribution.

9. The method according to claim 8, characterized in that, The dynamic emergency path planning includes: The passage cost of each node in the path planning network is correlated with the real-time risk level of the node's location; When the risk level of any node exceeds the preset lethal threshold, the passage cost of that node is set to infinity.

10. A dynamic visualization system for park safety risks based on BIM-GIS semantic fusion, characterized in that, include: The data fusion module is used to construct an attribute voxel grid that integrates building information model and geographic information system model; A risk evolution simulation engine is used to calculate the flux of risk media propagating and diffusing between adjacent cells of the attribute voxel grid based on a dynamic permeability model. The risk evolution simulation engine is configured to: acquire the material properties and real-time environmental physical quantities of the boundary corresponding to the adjacent unit; when the boundary material is a fragile material and the real-time temperature in the real-time environmental physical quantity exceeds the preset heat resistance threshold of the fragile material, adjust the permeability between the adjacent units from a first preset value representing the blocking state to a second preset value representing the connected state, wherein the second preset value is greater than the first preset value. An adaptive visualization module is used to perform adaptive perspective visualization rendering based on the propagation and diffusion results of the risk medium. The dynamic path planning module is used for dynamic emergency path planning based on the dynamically evolving risk distribution.

Images