A three-dimensional visualization-based urban road accident emergency evacuation simulation method and system

Abstract

The application discloses a kind of city road accident emergency evacuation simulation method and system based on three-dimensional visualization, comprising: according to the traffic state dataset of multiple-source information of real-time acquisition construction;According to its construction contains real-time traffic simulation environment of fine model;Simulate the traffic capacity and conflict situation after accident in environment, obtain the influence result of control scheme;According to this, dynamically adjust evacuation path and control scheme, generate optimization strategy dataset;Strategy data is analyzed in combination with risk assessment model, and risk degree spatial distribution data is calculated;It is stereoscopic in color gradient and heat map in three-dimensional platform, and generates risk level distribution diagram;According to this, automatic matching and priority allocation of rescue resources are carried out, and final rescue deployment scheme is determined.The application realizes the intelligent management of whole process from data acquisition to rescue deployment, significantly improves traffic accident emergency response efficiency and rescue resource utilization rate, and provides technical support for highway safety management.

Description

Technical Field

[0001] This invention belongs to the field of highway safety management technology, and in particular relates to a simulation method and system for emergency evacuation of urban road accidents based on three-dimensional visualization. Background Technology

[0002] Emergency evacuation simulation for highway traffic accidents is a key technological area for improving emergency response capabilities, ensuring the safety of life and property, and rapidly restoring traffic order. Especially in scenarios involving serious accidents such as hazardous material leaks and fires, scientific evacuation plans and rescue deployments are of great significance in controlling the impact of accidents and reducing secondary risks.

[0003] Currently, simulation methods in this field generally suffer from the following limitations: In terms of data fusion, most systems struggle to dynamically integrate real-time collected multi-source information such as traffic density, accident locations, and weather conditions into the simulation process, leading to discrepancies between the generated evacuation strategies and the actual on-site conditions. Regarding traffic dynamics modeling, for complex road network nodes such as interchanges and ramp intersections, existing methods often fail to accurately simulate the real processes of vehicle merging, diversion, and conflict interactions, making it difficult to assess the actual impact of traffic signal control and regulation measures on traffic capacity. In terms of hazard assessment, due to the interplay of multiple factors such as accident type, hazardous material diffusion, and secondary accident risks, the degree of danger exhibits a complex spatial distribution. Traditional simulation systems lack the ability to fuse and intuitively present multi-dimensional hazard information, making it difficult for command personnel to quickly identify high-risk areas and rationally determine evacuation and rescue priorities, thus affecting the timeliness and accuracy of overall emergency decision-making.

[0004] Therefore, how to achieve real-time fusion and closed-loop feedback of multi-source information in simulation, realistically reproduce the dynamic behavior of complex traffic nodes, and intuitively present the spatial distribution characteristics of danger has become an urgent need to improve the simulation effectiveness of emergency evacuation for highway accidents. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a three-dimensional visualization-based urban road accident emergency evacuation simulation method, comprising the following steps: An initial dataset is constructed based on real-time collected multi-source information of highway sections, and then integrated through a 3D visualization platform to obtain a dynamically updated traffic status dataset. Based on the dynamically updated traffic state dataset, a real-time traffic simulation environment, including a refined model of interchanges and ramp intersections, is constructed in a 3D visualization platform. The changes in traffic capacity and traffic flow conflicts at the intersection after an accident are simulated in the real-time traffic simulation environment to obtain the impact of different signal control and management schemes on traffic efficiency. Based on the traffic efficiency impact results, the evacuation routes and traffic control schemes are dynamically adjusted through the evacuation route planning algorithm to generate an optimized evacuation organization strategy dataset. A hazard assessment model was constructed to analyze the evacuation organization strategy dataset, calculate the hazard level of each area, and obtain the spatial distribution data of the hazard level. The spatial distribution data of the hazard level is presented in a three-dimensional visualization platform in the form of color gradient and heat map to generate a visualized hazard level distribution map. Based on the visualized hazard level distribution map, the regional hazard level and rescue resources are automatically matched and prioritized to determine the final rescue force deployment plan.

[0006] Optionally, an initial dataset is constructed based on real-time collected multi-source information of highway sections, and then integrated through a 3D visualization platform to obtain a dynamically updated traffic status dataset, including: Feature extraction was performed on the initial dataset to separate traffic flow density values, accident coordinates, and meteorological condition parameters. Calculate the correlation coefficient between traffic flow density and various meteorological condition parameters, and determine the significance of the correlation based on preset thresholds to generate feature correlation labels; Based on the aforementioned feature association markers, meteorological parameters with significant correlations are selected and superimposed on the traffic density data along with the accident coordinates to form an intermediate dataset with correlation features. The intermediate dataset was trained using the random forest algorithm to construct a traffic congestion classification model; By inputting real-time intermediate data into the classification model, road segment status prediction results, including congestion warning or normal passage labels, are obtained; The prediction results are rendered in a 3D visualization platform, and the congested areas and accident locations are dynamically updated and displayed to form a visualized traffic status dataset.

[0007] Optionally, based on the dynamically updated traffic state dataset, a real-time traffic simulation environment, including a refined model of interchanges and ramp intersections, is constructed in a 3D visualization platform, including: Based on traffic status data, a real-time traffic scenario framework with a multi-layered structure is constructed within the platform; By employing complex node modeling methods, layered geometric modeling of interchange areas is carried out, the inter-layer connection relationships are processed, and a refined interchange geometry is formed. Based on the refined interchange geometry, geometric modeling is performed on the ramp intersection area to analyze the lane transition logic and determine the path distribution; By combining real-time data streams, the detailed parameters of lane transition sections in the traffic model are adjusted to form a simulation environment that adapts to multi-layered structures. Dynamic traffic data is loaded into the simulation environment to generate real-time traffic scene images, and local model optimization is triggered when the traffic flow exceeds a preset threshold.

[0008] Optionally, the simulation environment simulates changes in intersection capacity and traffic flow conflicts after an accident, obtaining the impact of different signal control and management schemes on traffic efficiency, including: Obtain traffic flow data in each direction at the intersection before and after the accident, and calculate a dynamic change sequence of traffic capacity; Analyze the dynamic change sequence to detect the overlapping area between the merging and diverging directions of traffic flow and determine the conflict distribution; The number of conflict points per unit of time is counted to form a conflict intensity index; Clustering algorithms are used to group the conflict intensity indices under different traffic light timing schemes to form timing scheme categories; Based on the preset conflict intensity threshold, each category is marked as a low-conflict or high-conflict timing scheme; The low-conflict timing scheme was combined with various traffic control schemes for simulation operation to obtain the corresponding traffic efficiency values. Establish a correlation model between low-conflict timing schemes and traffic efficiency values ​​to determine the optimal combination of signal timing and control schemes.

[0009] Optionally, based on the impact results, the evacuation routes and traffic control schemes are dynamically adjusted using an evacuation route planning algorithm to generate an optimized evacuation organization strategy dataset, including: Obtain impact data including real-time traffic flow and driver / passenger distribution; Call the evacuation route planning algorithm to calculate the initial evacuation route and the corresponding traffic control plan; The initial evacuation routes are assessed for congestion risk. If the risk exceeds a preset threshold, the route allocation ratio is dynamically adjusted. Update the traffic light control parameters in the traffic control plan based on the adjusted route; Based on the updated plan and routes, the shortest evacuation time is recalculated, and the route connectivity is verified. The optimized evacuation routes and traffic control plans are integrated to form an evacuation organization strategy dataset.

[0010] Optionally, a hazard assessment model is constructed to analyze the evacuation organization strategy dataset, calculate the hazard level of each area, and obtain hazard spatial distribution data, including: Based on the accident type information, determine the preliminary scope of the hazard impact; Based on the regional labeling data, the hazard level of each region is initially labeled to obtain initial values; A risk assessment model is constructed to perform weighted calculations on the initial values. If the initial value of a certain area exceeds a preset threshold, the correlation with its surrounding areas is analyzed to obtain adjusted risk level data. The adjusted data is divided into spatial grids to obtain the hazard level distribution of each grid cell. Analyze the difference in hazard level between each grid cell and its surrounding cells, and smooth out cells whose differences exceed the threshold. By combining evacuation strategy optimization information, high-risk areas are prioritized to determine the final spatial distribution of risk.

[0011] Optionally, the spatial distribution data of the hazard level is presented in a three-dimensional visualization platform in the form of color gradients and heat maps to generate a visualized hazard level distribution map, including: Interpolation processing is performed on the spatial distribution data of hazard level to generate continuous three-dimensional grid data; Set a threshold for hazard level classification, assign corresponding hazard levels based on the hazard values ​​of grid points, and form a tiered data structure. According to the color gradient mapping rules, color values ​​are assigned to the hierarchical volume data to obtain three-dimensional voxel data with color attributes; The three-dimensional voxel data is rendered using a heatmap to generate a three-dimensional thermal distribution image. The 3D thermal distribution image is loaded into a 3D visualization platform and a legend is overlaid to generate a hazard level distribution map.

[0012] Optionally, based on the visualized hazard level distribution map, automatic matching and priority allocation of regional hazard levels and rescue resources are performed to determine the final rescue force deployment plan, including: Extract the geographical location information of high-risk areas from the aforementioned hazard level distribution map to identify key target areas; Based on the hazard level data of key target areas, the areas are stratified according to preset thresholds, and the priority levels of each area are divided. Based on the data on available relief resources, a preliminary resource allocation list is generated according to priority levels. By using the automatic matching function, the list is compared with the actual needs of high-risk areas to adjust and form a detailed allocation of resources; Based on the allocation details and geographical information, if a certain area is insufficient in resources, resources will be transferred from neighboring areas to determine the final allocation plan; The final allocation scheme is validated for rationality to obtain the optimized deployment result, and execution instructions are generated and sent to the scheduling system.

[0013] On the other hand, this invention proposes a three-dimensional visualization-based urban road accident emergency evacuation simulation system for implementing the method, comprising: The data acquisition and processing module is used to collect and integrate multi-source information of road segments in real time to generate dynamic traffic status datasets. The three-dimensional simulation environment construction module is connected to the data acquisition and processing module and is used to construct a real-time traffic simulation environment, including a refined model of complex nodes, in the three-dimensional visualization platform based on the traffic state dataset. The capacity and impact analysis module is connected to the three-dimensional simulation environment construction module and is used to simulate changes in capacity and traffic flow conflict after an accident in the simulation environment, and analyze the impact results of different control schemes. The evacuation strategy optimization module is connected to the traffic capacity and impact analysis module, and is used to dynamically adjust and generate an optimized evacuation organization strategy based on the impact results. The hazard assessment and spatialization module is connected to the evacuation strategy optimization module and is used to analyze the evacuation strategy in conjunction with the assessment model, calculate and output hazard spatial distribution data. The stereoscopic visualization module, connected to the hazard assessment and spatialization module, is used to render and present the spatial distribution data of the hazard in a stereoscopic manner on a three-dimensional platform in the form of color gradients and heat maps. The rescue plan generation module is connected to the three-dimensional visualization presentation module. It is used to link the deployment function based on the visualization results, automatically match resources and priorities, and generate the final rescue force deployment plan.

[0014] Compared with the prior art, the present invention has the following advantages and technical effects: This invention constructs a dynamic traffic state dataset by real-time collection of highway traffic density, accident location, and meteorological conditions data. A real-time traffic scene model is then generated on a 3D visualization platform, allowing for detailed simulation of complex node areas. This analysis of post-accident capacity changes and traffic conflict situations optimizes signal timing and control schemes. Combining evacuation route planning algorithms and hazard assessment models, the invention dynamically adjusts evacuation strategies, calculates hazard level distribution, and visually presents hazardous areas using color gradients and heatmaps. Ultimately, it coordinates the deployment of rescue forces and prioritizes resource allocation to high-risk areas. This invention achieves intelligent management of the entire process from data collection to rescue deployment, significantly improving the efficiency of traffic accident emergency response and the utilization rate of rescue resources, providing technical support for highway safety management. Attached Figure Description

[0015] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic diagram of the method flow according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating how the present invention automatically matches high-risk areas with priority allocation of rescue resources to determine the final deployment plan for rescue forces. Figure 3 This is a schematic diagram of the system structure according to an embodiment of the present invention. Detailed Implementation

[0016] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0017] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0018] Example 1 This embodiment provides a simulation method for emergency evacuation of urban road accidents based on three-dimensional visualization, including the following steps: An initial dataset is constructed based on real-time collected multi-source information of highway sections, and then integrated through a 3D visualization platform to obtain a dynamically updated traffic status dataset. Based on the dynamically updated traffic state dataset, a real-time traffic simulation environment, including a refined model of interchanges and ramp intersections, is constructed in a 3D visualization platform. The changes in traffic capacity and traffic flow conflicts at the intersection after an accident are simulated in the real-time traffic simulation environment to obtain the impact of different signal control and management schemes on traffic efficiency. Based on the traffic efficiency impact results, the evacuation routes and traffic control schemes are dynamically adjusted through the evacuation route planning algorithm to generate an optimized evacuation organization strategy dataset. A hazard assessment model was constructed to analyze the evacuation organization strategy dataset, calculate the hazard level of each area, and obtain the spatial distribution data of the hazard level. The spatial distribution data of the hazard level is presented in a three-dimensional visualization platform in the form of color gradient and heat map to generate a visualized hazard level distribution map. Based on the visualized hazard level distribution map, the regional hazard level and rescue resources are automatically matched and prioritized to determine the final rescue force deployment plan.

[0019] As a specific implementation method, such as Figure 1 As shown, the specific steps include: Step S101: Real-time data collection of traffic flow density, accident location coordinates, and meteorological conditions related to the highway section is performed using a multi-source sensor network to construct an initial dataset. This dataset is then transmitted to a 3D visualization platform for preliminary integration, resulting in a dynamically updated traffic status dataset. A multi-source sensor network collects real-time data on traffic density, accident location coordinates, and meteorological conditions along highway sections to form an initial dataset. Based on this initial dataset, traffic density values, accident coordinates, and meteorological condition parameters are extracted to obtain a separate feature set. The correlation coefficient between traffic density and meteorological condition parameters is calculated using this feature set. The absolute value of the correlation coefficient is checked against a preset threshold. If the threshold is exceeded, a significant correlation between meteorological conditions and traffic density is determined; otherwise, it is marked as having no significant correlation, resulting in a feature association label. Meteorological condition parameters significantly correlated with traffic density are selected based on the feature association labels and superimposed onto the traffic density values ​​along with the accident coordinates to generate an intermediate traffic state dataset with associated features. A random forest algorithm is used to train this intermediate traffic state dataset with associated features to obtain a traffic congestion classification model, resulting in a trained model. The trained model is then fed the real-time updated intermediate traffic state dataset with associated features to determine the current road segment's congestion level. If the congestion level is high, a congestion warning label is output; otherwise, a normal passage label is output, yielding a road segment status prediction result. The road segment condition prediction results are transmitted to a 3D visualization platform for rendering, dynamically updating the congestion area display and accident coordinate labels in the traffic condition dataset, and generating a visualized traffic condition dataset.

[0020] In one possible implementation, in a highway traffic condition monitoring scenario, real-time data acquisition via a multi-source sensor network serves as the starting point for the entire system. Sensors are distributed throughout the road segment, collecting data on traffic density, accident location coordinates, and weather conditions to form an initial dataset. Assuming a certain road segment has a peak-hour traffic density of 80 vehicles per kilometer, an accident location at 116.5°E, 39.9°N, and weather conditions indicating heavy rain with visibility as low as 200 meters, this data is integrated into the initial dataset, laying the foundation for subsequent analysis.

[0021] Specifically, after feature extraction from the initial dataset, traffic density values, accident coordinates, and meteorological condition parameters are separated to form a set of distinct features. The traffic density remains at 80 vehicles per kilometer, the accident coordinates remain unchanged, and the meteorological condition parameters are refined to include rainfall of 10 millimeters per hour and visibility of 200 meters. This separation facilitates subsequent targeted analysis and improves the accuracy of data processing.

[0022] In one possible implementation, calculating the correlation coefficient between traffic flow density and meteorological condition parameters is a crucial step. Assume that statistical analysis shows a correlation coefficient of -0.75 between traffic flow density and rainfall, and -0.68 between traffic flow density and visibility. If a preset threshold of 0.5 is set, then the absolute values ​​of both exceed the threshold, indicating a significant correlation between meteorological conditions and traffic flow density, and this is marked as a significant correlation feature. This step helps to identify the core factors affecting traffic conditions, providing a basis for subsequent modeling.

[0023] Based on feature association labels, rainfall and visibility were selected as parameters significantly correlated with traffic density. These, combined with accident coordinates, were overlaid onto the traffic density values ​​to form an intermediate traffic state dataset with associated features. Assuming the above road segment, the overlaid dataset shows a traffic density of 80 vehicles per kilometer, rainfall of 10 millimeters per hour, visibility of 200 meters, and the accident location at 116.5°E and 39.9°N. This provides more comprehensive feature support for subsequent model training.

[0024] Specifically, a random forest algorithm is used to train an intermediate dataset to generate a traffic congestion classification model. During training, the model learns the impact of traffic density, weather conditions, and accident locations on congestion levels, ultimately outputting three categories: high, medium, and low. This model effectively captures the complex relationships between features, improving prediction accuracy.

[0025] In one possible implementation, an intermediate dataset updated in real time is input into the trained model to determine the current level of congestion on a road segment. If the model outputs a high-level congestion warning label, a congestion warning label is generated; if it outputs a medium or low-level congestion label, a normal traffic flow label is generated. For example, if at a certain moment the traffic density rises to 100 vehicles per kilometer and the rainfall increases to 15 millimeters per hour, the model predicts high congestion and outputs a warning label. This helps to promptly alert relevant departments to take traffic management measures.

[0026] The prediction results are transmitted to a 3D visualization platform for rendering, dynamically updating the display of congested areas and accident coordinates in the traffic status dataset. Assuming the platform highlights congested road sections in red and marks accident locations with yellow icons, users can intuitively understand road conditions. This visualization method greatly improves information transmission efficiency, facilitates rapid decision-making, reduces the risk of traffic accidents, and provides drivers with real-time traffic information, optimizing their travel experience.

[0027] Step S102: Based on the dynamically updated traffic state dataset, a real-time traffic scene model is constructed in the 3D visualization platform. Complex node modeling techniques are used to perform refined geometric modeling of interchanges and ramp intersections, determining a simulation environment that includes multi-layered structures and lane transition sections. By collecting real-time traffic status data and combining it with dynamically updated data streams, an initial traffic information database is constructed to obtain a complete dataset covering interchanges and ramp intersections, resulting in a preliminary traffic status description. For this preliminary description, 3D visualization technology is used to generate a framework structure for the real-time scene on the platform, loading the basic elements of the traffic model and determining a multi-layered spatial layout. Based on the spatial layout, complex node modeling techniques are employed to perform layered geometric modeling of the interchange area, handling the connection relationships of the multi-layered structure to obtain a refined interchange geometry. Starting from this refined geometry, detailed geometric modeling is implemented for ramp intersections, analyzing the transition logic of lane changes and determining the path distribution in the intersection area. Using the path distribution data and considering the dynamic update requirements of the real-time scene, the detailed features of the lane change segments in the traffic model are adjusted to obtain simulation environment parameters suitable for the multi-layered structure. Based on these simulation environment parameters, dynamically updated traffic status data streams are loaded to generate a real-time traffic scene simulation, judging the flow changes in each area of ​​the scene. If the flow exceeds a preset threshold, local model optimization is triggered, resulting in the final simulation output.

[0028] In one possible implementation, when building the initial traffic information database, sensor devices deployed at key nodes of the highway can collect real-time vehicle traffic data at interchanges and ramp intersections. Assuming an interchange area covers traffic entry points in four main directions, and hourly vehicle traffic data are collected at 1200, 1500, 800, and 1000 vehicles respectively, this data will be entered into the database as basic information for subsequent analysis. This approach ensures the comprehensiveness and real-time nature of the data, providing a reliable basis for subsequent scenario construction.

[0029] For the application of 3D visualization technology, when generating a real-time scene framework structure on the platform, the interchange area can be divided into three parts: the upper-level main road, the lower-level auxiliary roads, and the connecting ramps. The geometric features of each part are scaled proportionally according to the actual road width and height. For example, the main road width is set to 12 meters, and the ramp slope is controlled at 5%, thus forming a spatial layout that closely resembles real road conditions. This layered design helps to clearly display the structural features of complex road sections, facilitating subsequent modeling and dynamic adjustments.

[0030] When performing layered geometric modeling of interchange areas, a layer-by-layer separation approach can be used to handle the connection relationships of multi-level structures. Taking a three-level interchange as an example, the upper level is the main highway, the middle level is the turning ramps, and the lower level is the ground-level auxiliary road. During modeling, it is necessary to ensure that the vertical spacing between each level is at least 4.5 meters to meet actual engineering standards. Through this refined modeling, the three-dimensional shape of the interchange can be accurately reproduced, providing a high-fidelity foundation for the simulation environment.

[0031] For detailed geometric modeling of ramp intersection areas, when analyzing lane transition logic, attention can be paid to the length of the transition section where the lane width gradually decreases from 3.5 meters to 3 meters, typically set at 50 meters, to ensure smooth lane changes for vehicles within this area. Path distribution is allocated based on actual traffic flow, for example, 60% of the main road's straight lanes and 40% of the turning ramp lanes, thereby optimizing vehicle flow logic in the simulation scenario.

[0032] When adjusting the detailed features of lane transition sections in the traffic model, simulation environment parameters can be dynamically adjusted based on real-time traffic flow data. For example, if the traffic flow in a certain ramp area surges to 2000 vehicles per hour, exceeding the preset threshold of 1500 vehicles, the system will automatically optimize the lane curvature radius of that area from 30 meters to 35 meters to reduce the risk of collisions when vehicles change lanes. This dynamic adjustment significantly improves the adaptability of the simulation environment.

[0033] When generating real-time traffic scene simulations, loading dynamically updated data streams allows for a clear visualization of traffic flow changes in different areas. For example, if the traffic flow at an interchange entrance suddenly increases from 1000 vehicles per hour to 1800 vehicles per hour, the system will trigger local area model optimization, reallocating lane resources to ensure the simulation reflects real-world road conditions. This real-time nature helps in the timely identification of potential problem areas. Through these various implementation methods, it can be seen that every step from data acquisition to simulation output is closely aligned with the highway traffic scenario, with each stage supporting the others to ensure the integrity and consistency from basic data to the final image. This method not only improves the accuracy of the traffic model but also provides reliable support for subsequent analysis.

[0034] Step S103: Based on the traffic state dataset in the simulation environment, simulate the dynamic changes in intersection capacity after an accident, analyze the merging and diverging conflicts of traffic flow, and obtain the impact results of signal timing and traffic control schemes on traffic efficiency. Traffic state data before and after the accident is loaded into the traffic simulation environment dataset. The changes in traffic flow in each direction at the intersection are calculated using this data to obtain a dynamic capacity change sequence. Based on this dynamic capacity change sequence, overlapping areas between merging and diverging traffic directions are detected to determine the conflict distribution. The number of conflict points per unit time is counted based on this conflict distribution to obtain a conflict intensity index. A clustering algorithm is used to group the conflict intensity index under different signal timing schemes to obtain timing scheme categories. If the conflict intensity index under the same timing scheme category is below a preset threshold, the signal timing scheme is marked as a low-conflict timing scheme; otherwise, it is marked as a high-conflict timing scheme. Simulations are conducted using a combination of low-conflict timing schemes and traffic control schemes to obtain corresponding traffic efficiency values. A linear regression model is used to establish a correlation between low-conflict timing schemes and traffic efficiency values ​​to determine the optimal combination of signal timing and traffic control schemes.

[0035] In one possible implementation, when loading traffic state data before and after the accident in a traffic simulation environment, vehicle traffic records in each direction at the intersection can be collected using sensors and cameras. For example, if the number of vehicles passing through the intersection per minute is recorded as 200 in the east-to-west direction, 180 in the west-to-east direction, and 150 in the north-south direction within 5 minutes before and after the accident, time series data can be constructed for subsequent analysis.

[0036] It should be noted that data collection needs to cover all key road sections within the accident's impact area to ensure data integrity and accurately reflect fluctuations in traffic conditions.

[0037] The calculation of the dynamic change sequence of traffic capacity can be based on the collected traffic flow data to analyze the changing trend of vehicle traffic volume in each direction before and after the accident. For example, if the traffic flow in the east-to-west direction drops to 120 vehicles per minute after the accident, while the traffic flow in the north-south direction increases to 180 vehicles per minute due to detours, the percentage decrease or increase in traffic capacity can be obtained by comparison, providing a basis for subsequent conflict detection.

[0038] When detecting areas where traffic flows overlap with merging and diverging directions, attention can be paid to the intersection points of vehicle trajectories within the intersection. Assuming that vehicle trajectories frequently intersect in the merging areas from east to west and north to south, forming multiple conflict points, spatial distribution analysis can be used to determine the dense areas of conflict, laying the foundation for subsequent indicator calculations.

[0039] When counting the number of conflict points per unit of time, the number of conflict points within an intersection can be set to 10 per minute. By accumulating the data, a conflict intensity index can be obtained, reflecting the potential risk level of traffic operations. This method helps to identify high-risk periods and areas.

[0040] When using clustering algorithms to group the conflict intensity index of traffic light timing schemes, the timing schemes for different time periods within a day can be divided into three categories: morning peak, off-peak, and evening peak. Assuming the conflict intensity index for the morning peak is 8, for the off-peak it is 3, and for the evening peak it is 6, the conflict characteristics under different categories can be analyzed through grouping.

[0041] When labeling low-conflict and high-conflict timing schemes, assuming a preset threshold of 5, the timing schemes during off-peak hours are labeled as low-conflict schemes, while the morning and evening peak hours are labeled as high-conflict schemes. This classification helps to select more suitable signal control strategies.

[0042] The simulation of the combination of low-conflict timing scheme and traffic control scheme can be assumed to be a control method that extends the green light time and restricts left turns during off-peak hours. The simulation results show that the traffic efficiency value increases to 2,500 vehicles per hour, reflecting the applicability of the combined scheme.

[0043] When establishing the correlation between low-conflict timing schemes and traffic efficiency values ​​using a linear regression model, the relationship between different green light durations and vehicle throughput can be analyzed. Assuming that an increase of 10 seconds in green light time leads to a 5% increase in traffic efficiency, the optimal timing scheme is determined to be 40 seconds of green light and 20 seconds of red light. This, combined with traffic restriction measures, forms the optimal combination scheme. This approach effectively improves intersection capacity, reduces the probability of conflicts, and provides a scientific basis for traffic management.

[0044] Step S104: By influencing the result data, a pre-established evacuation route planning algorithm is invoked to dynamically adjust the evacuation route planning and traffic control scheme, generating an optimized evacuation organization strategy dataset. The impact data obtained includes real-time traffic flow and driver / passenger distribution. Using this data, a pre-established evacuation route planning algorithm is used to calculate initial evacuation routes and traffic control plans. The initial evacuation routes are then assessed for congestion risk using the algorithm. If the risk assessment indicates that the congestion risk exceeds a preset threshold, the evacuation route allocation ratio is dynamically adjusted to obtain adjusted evacuation routes. Based on the adjusted evacuation routes, the traffic light timing control parameters in the traffic control plan are updated to obtain an updated traffic control plan. Using the updated traffic control plan and the adjusted evacuation routes, Dijkstra's algorithm is executed to recalculate the shortest evacuation time, resulting in an optimized evacuation time. The connectivity of the evacuation routes is verified using the optimized evacuation time and the impact data, resulting in optimized evacuation routes and traffic control plans. Finally, the optimized evacuation routes and traffic control plans are integrated to form an evacuation organization strategy dataset.

[0045] In one possible implementation, within the field of traffic simulation and evacuation route planning, real-time traffic flow and driver / passenger distribution can be obtained by deploying cameras and sensors at intersections to collect vehicle traffic data and driver / passenger density information in real time. Assuming an intersection experiences a peak traffic flow of 2000 vehicles per hour and a driver / passenger density of 5 people per square meter, this data will be input into the system as the basis for influencing the results and providing a foundation for subsequent route planning. This data collection method ensures the timeliness of information, laying the groundwork for dynamically adjusting the plan.

[0046] When invoking a pre-established evacuation route planning algorithm, a preliminary main route from the accident site to the safe area can be generated based on historical traffic data and current real-time data. Assuming the route length is 5 kilometers and the estimated travel time is 20 minutes, this can be combined with traffic control measures to restrict vehicle access to certain sections, forming a preliminary control strategy. This approach enables rapid response to emergencies and ensures the feasibility of evacuation routes.

[0047] For congestion risk assessment, traffic flow data at key nodes along a route can be analyzed to determine the presence of potential congestion. For example, if a route has a vehicle density of 100 vehicles per kilometer, exceeding a preset threshold of 80 vehicles per kilometer, the system will determine that the route has a high risk of congestion. Based on this result, the allocation ratio of evacuation routes is dynamically adjusted, for example, diverting 50% of the traffic flow to alternative routes to alleviate pressure on the main routes. This assessment and adjustment mechanism can effectively reduce the likelihood of congestion occurring.

[0048] When updating traffic control plans, traffic light timing parameters can be optimized based on the adjusted evacuation routes. For example, if the original green light duration was 30 seconds, it can be extended to 45 seconds to prioritize the passage of vehicles along the main evacuation routes. Such parameter adjustments can improve traffic efficiency on key road sections, allowing more time for evacuation.

[0049] When executing the shortest path algorithm, evacuation time can be recalculated based on the updated route and control measures. Assuming the optimized route length is reduced to 4.5 kilometers, the evacuation time is reduced to 18 minutes. This optimization can significantly improve evacuation efficiency and ensure the safety of people and vehicles.

[0050] When verifying the connectivity of evacuation routes, real-time data can be used to check for temporary roadblocks or abnormal traffic flow along the routes. If a section of road is temporarily closed due to an accident, the system will automatically replan the route to ensure connectivity. This verification mechanism improves the reliability of the plan.

[0051] The final integrated evacuation organization strategy dataset can include multiple optimized routes and corresponding control schemes, allowing relevant departments to flexibly utilize them in different scenarios. Assuming the dataset contains three main routes and two backup schemes, it can handle varying degrees of traffic pressure. This comprehensive strategy set can provide diverse response methods for emergencies, improving the adaptability of overall evacuation management.

[0052] Step S105: Based on the optimized evacuation organization strategy dataset, and combined with the hazard assessment model, comprehensively analyze the accident types and hazard area labeling information, calculate the hazard level distribution of different areas, and obtain the hazard spatial distribution data: The process involves acquiring optimized evacuation organization strategy data and the output of the hazard assessment model, classifying accident type information, and determining the preliminary hazard impact range for each type of accident. Based on the classified impact ranges and combined with regional labeling data, a preliminary hazard level label is applied to each region, yielding an initial hazard level value for each region. Using these initial hazard level values, a pre-established assessment model is employed to perform a weighted calculation of the hazard level for each region. If the initial hazard level value of a region exceeds a preset threshold, a correlation analysis is performed on its surrounding regions to obtain adjusted regional hazard level data. For the adjusted regional hazard level data, combined with spatial distribution information, all regions are divided into grids to obtain the hazard level distribution of each grid unit, determining the gridded hazard distribution result. Based on the gridded hazard distribution result, the hazard level difference between each grid unit and its surrounding units is analyzed. If the hazard level difference between a grid unit and its surrounding units exceeds a preset threshold, it is smoothed to obtain smoothed hazard distribution data. Using the smoothed hazard distribution data and optimized evacuation organization strategy information, high-risk areas are prioritized to determine the final spatial hazard distribution result.

[0053] In one possible implementation, the system first acquires the optimized evacuation organization strategy dataset and the output of the hazard assessment model. These results typically include the specific type of accident, such as a chemical leak or a fire. Classifying the accident type information allows for categorization into point-source and area-source accidents, thereby quickly determining the initial hazard impact range for each type of accident.

[0054] Specifically, for a chemical leak accident, the initial impact area may be defined as a circular area with a radius of 500 meters centered on the accident point.

[0055] In one embodiment, based on the classified impact range and combined with regional labeling data, such as building density and population distribution, a preliminary hazard level label is made for each region.

[0056] High-density residential areas were initially categorized as Level 8, while remote vacant areas were categorized as Level 2, thus obtaining the initial hazard level for each area. This initial categorization helps to quickly identify high-risk areas and avoid delays in emergency response.

[0057] It should be noted that the initial hazard level is used to perform a weighted calculation using a pre-established assessment model that comprehensively considers factors such as wind direction and terrain. If the initial value of a certain area exceeds the threshold, such as level 7, a correlation analysis of the surrounding areas is triggered.

[0058] When the hazard level of the central area is level 9, the hazard levels of its adjacent areas are automatically adjusted upwards by 1-2 levels due to the diffusion effect, resulting in adjusted regional hazard level data. This correlation analysis can more accurately reflect the propagation characteristics of the hazard and improve the accuracy of the assessment.

[0059] Based on the adjusted regional hazard level data, a gridded division is performed in conjunction with spatial distribution information. Typically, the entire affected area is divided into 100m x 100m grid units, and the hazard level distribution of each grid is obtained. Assuming the hazard level of the grid at the accident center is level 9 and the edge grids are level 4, the gridded hazard distribution is determined. This gridded approach facilitates refined management and supports subsequent precise evacuation.

[0060] In one possible implementation, the differences between each grid cell and its surrounding cells are analyzed based on the gridded hazard distribution results. If a grid cell has a hazard level of 8, while the surrounding cells average only 3, and the difference exceeds a threshold such as 4, then smoothing is performed to adjust it to around 6, resulting in smoothed hazard distribution data. Smoothing can eliminate assessment noise caused by isolated high-risk points, ensure distribution continuity, and effectively reduce the risk of misjudgment.

[0061] Specifically, by combining smoothed hazard distribution data with optimized evacuation organization strategies, such as path connectivity and traffic allocation, high-risk areas are prioritized.

[0062] Grids with a hazard level of 8 or above have the highest priority, and evacuation resources are allocated first to determine the spatial distribution of hazard levels. This prioritization significantly shortens the evacuation time for high-risk groups, improves overall emergency response efficiency, and forms a closed-loop optimization with the dynamic adjustment of evacuation routes in the early stages, ensuring the consistency and practicality of the strategy.

[0063] Step S106: Based on the spatial distribution data of hazard levels, color gradient and heatmap technologies are used in a 3D visualization platform to intuitively present the three-dimensional distribution of hazardous areas and generate a visualized hazard level distribution map. Acquire spatial distribution data of hazard levels. Obtain continuous 3D hazard grid data through interpolation. Determine hazard level classification thresholds for the 3D hazard grid data. If the hazard level of a grid point is higher than the highest threshold, assign the highest hazard level; otherwise, assign the corresponding hazard level, resulting in graded hazard volume data. Assign color values ​​based on the graded hazard volume data using color gradient mapping rules, obtaining 3D voxel data with color attributes. Use heatmap technology to render the color-attributed 3D voxel data, obtaining a 3D thermal distribution image. Load the 3D thermal distribution image into a 3D visualization platform and overlay color gradient legends to generate a hazard level distribution map.

[0064] Based on the obtained spatial distribution data of hazard levels, continuous three-dimensional hazard grid data can be obtained through interpolation. This interpolation method typically employs inverse distance weighting or kriging methods to extend discrete regional hazard level points into a continuous spatial field, facilitating subsequent three-dimensional analysis.

[0065] Understandably, thresholds are usually pre-set based on accident type and historical data. For example, the degree of danger is divided into four levels: 0-3 is low risk, 3-5 is medium risk, 5-7 is high risk, and 7 and above is extremely risky.

[0066] In one embodiment, for an ammonia leak scenario, the highest threshold is set to 9.0. If a grid point has a calculated hazard level of 8.2, it is directly assigned the highest hazard level; if it is 5.8, it falls into the high-risk level. This classification ensures the hierarchical nature of the hazard assessment, facilitating the rapid identification of key areas. Color values ​​are assigned based on the hazard level data using a color gradient mapping rule, which intuitively expresses the risk intensity.

[0067] Preferably, a gradient spectrum from blue to red is used, with low-risk areas assigned blue, medium-risk areas green, high-risk areas yellow, and critically dangerous areas red.

[0068] In one possible implementation, a voxel with a hazard level of 4.5 is mapped to green, and one with a hazard level of 8.7 is mapped to dark red, thus generating 3D voxel data with color attributes. This mapping not only improves the readability of the visualization but also helps decision-makers quickly perceive changes in risk gradients. Using heatmap technology to perform volume rendering on the color-attributed 3D voxel data yields a 3D thermal distribution image.

[0069] Specifically, volume rendering uses ray casting or texture slicing to accumulate and project the colors of internal voxels, forming a semi-transparent thermal cloud map.

[0070] In ammonia diffusion simulations, the rendered image clearly shows high-risk areas as red clusters concentrated near the leak source, gradually transitioning to yellow and green towards the outside. This rendering method preserves internal structural details, avoiding the limitations of only displaying the surface, and helps to comprehensively understand the three-dimensional distribution of hazards in space. By loading a three-dimensional thermal distribution image and overlaying a color gradient legend into a 3D visualization platform, a hazard level distribution map is finally generated.

[0071] Understandably, the legend is placed on one side of the image, indicating the correspondence between colors and levels, to facilitate interpretation.

[0072] In one embodiment, a 3D model of the plant area is overlaid on the platform. With the heat map coinciding with the buildings, managers can visually identify which pipelines pose the highest risk, allowing them to prioritize adjustments to evacuation routes. This visualization significantly improves the targeting and efficiency of emergency response, supporting more precise implementation of optimized evacuation strategies.

[0073] Step S107: By visualizing the hazard level distribution map and linking it with the rescue force deployment module, the system automatically matches areas with high hazard levels with the priority of rescue resource allocation to determine the final rescue force deployment plan. like Figure 2 As shown, by constructing a hazard level distribution map, the geographical location information of high-risk areas is obtained to determine key monitoring target areas. Based on the hazard level data of the target areas, a preset threshold is used for stratification to obtain the priority level classification results for each area. For the priority level classification results, the available inventory data of rescue forces is obtained. If the priority level of a certain area is higher than the preset standard, more rescue forces are allocated to it, determining a preliminary resource allocation list. Through the automatic matching function of the deployment module, the preliminary resource allocation list is compared with the actual needs of high-risk areas to obtain an adjusted force allocation detail. Based on the force allocation detail and combined with the geographical information data of the area matching, if the rescue forces in a certain area are insufficient, resources are transferred from neighboring areas to determine the final allocation plan. The data of the final allocation plan is obtained, and a logistic regression model is used to verify the rationality of the allocation plan, resulting in an optimized rescue force deployment result. Based on the optimized rescue force deployment result, specific execution instructions are generated and transmitted to the rescue force dispatch system to complete the deployment task.

[0074] In one possible implementation, when constructing the hazard level distribution map, a geographic information system (GIS) can be used to combine hazard data with the latitude and longitude information of specific areas to form an intuitive visual distribution map. For identifying high-risk areas, a hazard level threshold of 80 can be set. For example, if the hazard level of an area is 85, it can be marked as a key monitoring target area. This method can quickly locate geographical locations requiring urgent attention, providing a basis for subsequent resource allocation.

[0075] When performing stratified processing of hazard level data, hazard levels and prohibition levels can be divided into three levels based on preset thresholds: high priority, medium priority, and low priority. For example, if the high priority threshold is a hazard level greater than 75, and a certain area has a hazard level of 80, it is classified as a high priority area. This stratified processing helps to clarify the priority order of resource allocation and ensures that high-risk areas receive more attention.

[0076] Matching rescue force inventory data with priority levels allows for querying the current number of available rescue teams and their equipment status through the system database. For example, if a high-priority area requires 10 rescue teams, but the inventory shows only 6, more resources should be prioritized for allocation to that area, creating a preliminary resource allocation list. This method ensures resources are directed to where they are most needed.

[0077] The automatic matching function in the deployment module can adjust the allocation details by comparing the initial list with the actual needs of the region. For example, if a region actually needs 8 rescue vehicles, but the initial list only allocates 5, the system will automatically add 3 more, resulting in the adjusted allocation. This automated matching improves allocation efficiency.

[0078] When allocating resources in neighboring areas using geographic information data, the availability of rescue forces in adjacent areas can be analyzed for deployment. For example, if the target area lacks two rescue teams, while neighboring areas have three surplus teams, two teams can be deployed to the target area. This method effectively addresses localized resource shortages.

[0079] When using logistic regression models to validate the rationality of resource allocation schemes, historical data analysis can be used to correlate resource allocation with rescue effectiveness, thereby optimizing the final deployment results. For example, if a scheme indicates over-allocation in a certain area, the model might suggest reducing resource input. This validation improves the overall rationality of the deployment.

[0080] When generating execution instructions and transmitting them to the dispatch system, the optimized deployment results can be transformed into specific task orders, including information such as the rescue team number and target area location, and directly distributed to each unit. This method ensures that the instructions are clear and executable, guaranteeing the smooth implementation of the mission.

[0081] On the other hand, such as Figure 3 As shown, this embodiment also proposes a three-dimensional visualization-based urban road accident emergency evacuation simulation system for implementing the method, including: The data acquisition and processing module is used to collect and integrate multi-source information of road segments in real time to generate dynamic traffic status datasets. The three-dimensional simulation environment construction module is connected to the data acquisition and processing module and is used to construct a real-time traffic simulation environment, including a refined model of complex nodes, in the three-dimensional visualization platform based on the traffic state dataset. The capacity and impact analysis module is connected to the three-dimensional simulation environment construction module and is used to simulate changes in capacity and traffic flow conflict after an accident in the simulation environment, and analyze the impact results of different control schemes. The evacuation strategy optimization module is connected to the traffic capacity and impact analysis module, and is used to dynamically adjust and generate an optimized evacuation organization strategy based on the impact results. The hazard assessment and spatialization module is connected to the evacuation strategy optimization module and is used to analyze the evacuation strategy in conjunction with the assessment model, calculate and output hazard spatial distribution data. The stereoscopic visualization module, connected to the hazard assessment and spatialization module, is used to render and present the spatial distribution data of the hazard in a stereoscopic manner on a three-dimensional platform in the form of color gradients and heat maps. The rescue plan generation module is connected to the three-dimensional visualization presentation module. It is used to link the deployment function based on the visualization results, automatically match resources and priorities, and generate the final rescue force deployment plan.

[0082] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A simulation method for emergency evacuation of urban road accidents based on three-dimensional visualization, characterized in that, include: An initial dataset is constructed based on real-time collected multi-source information of highway sections, and then integrated through a 3D visualization platform to obtain a dynamically updated traffic status dataset. Based on the dynamically updated traffic state dataset, a real-time traffic simulation environment, including a refined model of interchanges and ramp intersections, is constructed in a 3D visualization platform. The changes in traffic capacity and traffic flow conflicts at the intersection after an accident are simulated in the real-time traffic simulation environment to obtain the impact of different signal control and management schemes on traffic efficiency. Based on the traffic efficiency impact results, the evacuation routes and traffic control schemes are dynamically adjusted through the evacuation route planning algorithm to generate an optimized evacuation organization strategy dataset. A hazard assessment model was constructed to analyze the evacuation organization strategy dataset, calculate the hazard level of each area, and obtain the spatial distribution data of the hazard level. The spatial distribution data of the hazard level is presented in a three-dimensional visualization platform in the form of color gradient and heat map to generate a visualized hazard level distribution map. Based on the visualized hazard level distribution map, the regional hazard level and rescue resources are automatically matched and prioritized to determine the final rescue force deployment plan.

2. The method according to claim 1, characterized in that, An initial dataset is constructed based on real-time multi-source information collected from highway sections. This dataset is then integrated using a 3D visualization platform to obtain a dynamically updated traffic status dataset, including: Feature extraction was performed on the initial dataset to separate traffic flow density values, accident coordinates, and meteorological condition parameters. Calculate the correlation coefficient between traffic flow density and various meteorological condition parameters, and determine the significance of the correlation based on preset thresholds to generate feature correlation labels; Based on the aforementioned feature association markers, meteorological parameters with significant correlations are selected and superimposed on the traffic density data along with the accident coordinates to form an intermediate dataset with correlation features. The intermediate dataset was trained using the random forest algorithm to construct a traffic congestion classification model; By inputting real-time intermediate data into the classification model, road segment status prediction results, including congestion warning or normal passage labels, are obtained; The prediction results are rendered in a 3D visualization platform, and the congested areas and accident locations are dynamically updated and displayed to form a visualized traffic status dataset.

3. The method according to claim 1, characterized in that, Based on the dynamically updated traffic state dataset, a real-time traffic simulation environment, including refined models of interchanges and ramp intersections, is constructed in a 3D visualization platform, comprising: Based on traffic status data, a real-time traffic scenario framework with a multi-layered structure is constructed within the platform; By employing complex node modeling methods, layered geometric modeling of interchange areas is carried out, the inter-layer connection relationships are processed, and a refined interchange geometry is formed. Based on the refined interchange geometry, geometric modeling is performed on the ramp intersection area to analyze the lane transition logic and determine the path distribution; By combining real-time data streams, the detailed parameters of lane transition sections in the traffic model are adjusted to form a simulation environment that adapts to multi-layered structures. Dynamic traffic data is loaded into the simulation environment to generate real-time traffic scene images, and local model optimization is triggered when the traffic flow exceeds a preset threshold.

4. The method according to claim 1, characterized in that, The simulation environment simulates changes in intersection capacity and traffic flow conflicts after an accident, obtaining the impact of different signal control and management schemes on traffic efficiency, including: Obtain traffic flow data in each direction at the intersection before and after the accident, and calculate a dynamic change sequence of traffic capacity. Analyze the dynamic change sequence to detect the overlapping area between the merging and diverging directions of traffic flow and determine the conflict distribution; The number of intersections and conflict points per unit of time is counted to form a conflict intensity index. Clustering algorithms are used to group the conflict intensity indices under different traffic light timing schemes to form timing scheme categories; Based on the preset conflict intensity threshold, each category is marked as a low-conflict or high-conflict timing scheme; The low-conflict timing scheme was combined with various traffic control schemes for simulation operation to obtain the corresponding traffic efficiency values. Establish a correlation model between low-conflict timing schemes and traffic efficiency values ​​to determine the optimal combination of signal timing and control schemes.

5. The method according to claim 1, characterized in that, Based on the impact results, evacuation routes and traffic control plans are dynamically adjusted using evacuation route planning algorithms to generate an optimized evacuation organization strategy dataset, including: Obtain impact data including real-time traffic flow and driver / passenger distribution; Call the evacuation route planning algorithm to calculate the initial evacuation route and the corresponding traffic control plan; The initial evacuation routes are assessed for congestion risk. If the risk exceeds a preset threshold, the route allocation ratio is dynamically adjusted. Update the traffic light control parameters in the traffic control plan based on the adjusted route; Based on the updated plan and routes, the shortest evacuation time is recalculated, and the route connectivity is verified. The optimized evacuation routes and traffic control plans are integrated to form an evacuation organization strategy dataset.

6. The method according to claim 1, characterized in that, A hazard assessment model was constructed to analyze the evacuation organization strategy dataset, calculate the hazard level of each area, and obtain hazard spatial distribution data, including: Based on the accident type information, determine the preliminary scope of the hazard impact; Based on the regional labeling data, the hazard level of each region is initially labeled to obtain initial values; A risk assessment model is constructed to perform weighted calculations on the initial values. If the initial value of a certain area exceeds a preset threshold, the correlation with its surrounding areas is analyzed to obtain adjusted risk level data. The adjusted data is divided into spatial grids to obtain the hazard level distribution of each grid cell. Analyze the difference in hazard level between each grid cell and its surrounding cells, and smooth out cells whose differences exceed the threshold. By combining evacuation strategy optimization information, high-risk areas are prioritized to determine the final spatial distribution of risk.

7. The method according to claim 1, characterized in that, The spatial distribution data of the hazard level is presented in a three-dimensional visualization platform in the form of color gradients and heat maps to generate a visualized hazard level distribution map, including: Interpolation processing is performed on the spatial distribution data of hazard level to generate continuous three-dimensional grid data; Set a threshold for hazard level classification, assign corresponding hazard levels based on the hazard values ​​of grid points, and form a tiered data structure. According to the color gradient mapping rules, color values ​​are assigned to the hierarchical volume data to obtain three-dimensional voxel data with color attributes; The three-dimensional voxel data is rendered using a heatmap to generate a three-dimensional thermal distribution image. The 3D thermal distribution image is loaded into a 3D visualization platform and a legend is overlaid to generate a hazard level distribution map.

8. The method according to claim 1, characterized in that, Based on the visualized hazard level distribution map, automatic matching and priority allocation of regional hazard levels and rescue resources are performed to determine the final rescue force deployment plan, including: Extract the geographical location information of high-risk areas from the aforementioned hazard level distribution map to identify key target areas; Based on the hazard level data of key target areas, the areas are stratified according to preset thresholds, and the priority levels of each area are divided. Based on the data on available relief resources, a preliminary resource allocation list is generated according to priority levels. By using the automatic matching function, the list is compared with the actual needs of high-risk areas to adjust and form a detailed allocation of resources; Based on the allocation details and geographical information, if a certain area is insufficient in resources, resources will be transferred from neighboring areas to determine the final allocation plan; The final allocation scheme is validated for rationality to obtain the optimized deployment result, and execution instructions are generated and sent to the scheduling system.

9. A simulation system for emergency evacuation of urban road accidents based on three-dimensional visualization, characterized in that, For implementing the method according to any one of claims 1-8, comprising: The data acquisition and processing module is used to collect and integrate multi-source information of road segments in real time to generate dynamic traffic status datasets. The three-dimensional simulation environment construction module is connected to the data acquisition and processing module and is used to construct a real-time traffic simulation environment, including a refined model of complex nodes, in the three-dimensional visualization platform based on the traffic state dataset. The capacity and impact analysis module is connected to the three-dimensional simulation environment construction module and is used to simulate changes in capacity and traffic flow conflict after an accident in the simulation environment, and analyze the impact results of different control schemes. The evacuation strategy optimization module is connected to the traffic capacity and impact analysis module, and is used to dynamically adjust and generate an optimized evacuation organization strategy based on the impact results. The hazard assessment and spatialization module is connected to the evacuation strategy optimization module and is used to analyze the evacuation strategy in conjunction with the assessment model, calculate and output hazard spatial distribution data. The stereoscopic visualization module, connected to the hazard assessment and spatialization module, is used to render and present the spatial distribution data of the hazard in a stereoscopic manner on a three-dimensional platform in the form of color gradients and heat maps. The rescue plan generation module is connected to the three-dimensional visualization presentation module. It is used to link the deployment function based on the visualization results, automatically match resources and priorities, and generate the final rescue force deployment plan.

10. An electronic device comprising a memory, a processor, and a computing program stored in the memory and executable on the processor, characterized in that, When the processor executes the computing program, it implements the method of any one of claims 1-8.

Images