An air-ground collaborative emergency response method and system for subway sudden failure

Abstract

The application discloses a kind of air-ground collaborative emergency response method and system for subway sudden failure, belong to urban traffic emergency management technical field, method includes: the multi-layer composite traffic network of target city is built;Based on the topological characteristics and passenger flow function attributes of multi-layer composite traffic network, high-risk fragile node is identified;With the goal of preferentially covering high-risk fragile node, generate emergency bus stop layout scheme;Determine the deployment point of unmanned aerial vehicle, and update the traffic risk value according to the observation coverage state of unmanned aerial vehicle to road network edge;Based on the updated traffic impedance, generate the optimal emergency path from emergency bus stop to evacuation target;According to multi-layer composite traffic network, high-risk fragile node, emergency bus stop layout scheme, unmanned aerial vehicle deployment point and optimal emergency path, generate air-ground collaborative emergency response decision scheme.The application effectively improves the emergency response efficiency and system resilience under the scene of urban rail transit sudden interruption.

Description

Technical Field

[0001] This invention relates to the field of urban traffic emergency management technology, and in particular to an air-ground coordinated emergency response method and system for sudden subway failures. Background Technology

[0002] The networked operation of urban rail transit has led to increasingly complex transfer relationships between lines, with passenger flow exhibiting high density and strong fluctuations, significantly enhancing the coupling and overall integrity of system operation. When key stations, important sections, or large transfer hubs experience disruptions due to equipment failure, extreme weather, or unforeseen events, local disturbances can rapidly spread through the network topology, triggering a chain reaction across multiple modes of transportation, including pedestrian route reselection, reduced transfer capacity, and a surge in pressure on public transport connections. In severe cases, this can even lead to widespread functional paralysis of the city's integrated transportation system.

[0003] Currently, most studies focus on single-layer network analysis of rail transit or public transport networks, lacking unified modeling of cross-layer collaborative relationships between rail transit, public transport, and road networks. Furthermore, existing emergency resource allocation methods are mostly based on static networks, rarely considering the impact of real-time information acquisition on emergency decision-making during disaster evolution, making it difficult to meet the rapid response needs in complex scenarios. In addition, research on UAVs in urban transportation mainly focuses on traffic inspection, road monitoring, and logistics delivery, with relatively insufficient research on their deep integration with rail transit emergency evacuation and public transport coordinated scheduling. Summary of the Invention

[0004] The purpose of this invention is to provide an air-ground collaborative emergency response method and system for sudden subway failures. By integrating bus connections and UAV dynamic perception, it can identify high-risk nodes, optimize the allocation of emergency resources, and dynamically plan evacuation routes, thereby improving the emergency response efficiency and resilience of the integrated transportation system.

[0005] To achieve the above objectives, in a first aspect, the present invention provides an air-ground coordinated emergency response method for sudden subway failures, comprising the following steps: S1. Construct a multi-layered composite transportation network for the target city. The multi-layered composite transportation network includes a road network layer, a public transport network layer, a subway network layer, and cross-layer coupling connections between each network layer. S2. Based on the topological characteristics and passenger flow functional attributes of the multi-layer composite transportation network, identify high-risk and vulnerable nodes in the multi-layer composite transportation network. S3. With the goal of prioritizing coverage of high-risk and vulnerable nodes, generate an emergency bus stop layout plan; S4. Determine the deployment points of UAVs in the multi-layered complex transportation network, and update the traffic risk value of the covered roadside based on the observation coverage status of the UAVs over the road network. S5. Based on the updated traffic impedance, generate the optimal emergency route from the emergency bus stop to the evacuation target in the multi-layered complex traffic network. S6. Based on the multi-layered complex transportation network, high-risk and vulnerable nodes, emergency bus stop layout plan, drone deployment points and optimal emergency routes, generate an air-ground coordinated emergency response decision plan.

[0006] Preferably, the construction of the multi-layered composite transportation network in S1 specifically includes: S11, Construct road networks separately Public transport network and subway network ,in For a set of nodes, Let be the set of edges. For weight set; S12. For nodes belonging to different network layers and If the spatial distance between the two Not exceeding the preset interlayer coupling radius Then, a cross-layer coupling edge is established between the two. The determination method is as follows: ; in, Indicates belonging to the network layer nodes Belonging to the network layer nodes Spatial distance between them , , , These represent the road network layer, the public transport network layer, and the subway network layer, respectively. Represents network layer With network layer The preset interlayer coupling radius between them Indicates at node With nodes Establish cross-layer coupling edges between them. This indicates that no cross-layer coupling edges are established; S13, Construct a multi-layered, complex transportation network The node set is The edge set is ,in This represents the set of cross-layer coupling edges between the public transport network layer and the road network layer. This represents the set of cross-layer coupling edges between the subway network layer and the road network layer. This is the set of cross-layer coupling edges between the public transport network layer and the subway network layer.

[0007] Preferably, high-risk vulnerable nodes are identified in S2 and evaluated using a comprehensive vulnerability index, specifically including: Calculate each node in a multi-layered complex transportation network Comprehensive vulnerability index: ; in, For nodes The comprehensive vulnerability index value, For the normalized degree centrality, For normalized betweenness centrality, To achieve normalized proximity centrality, The normalized passenger flow intensity This is the normalized transfer load. , , , , These are the weighting coefficients for the corresponding indicators; All nodes in the multi-layered complex transportation network are sorted in descending order based on the comprehensive vulnerability index value, and the top N nodes are selected as high-risk vulnerable nodes, where N is a positive integer.

[0008] Preferably, the emergency bus stop layout plan generated in S3 adopts a risk-priority maximum coverage site selection model, specifically including: S31. Set up a set of candidate emergency bus stops and demand node set ,in Indicates the first One candidate emergency bus stop The total number of candidate emergency bus stops. Indicates the first Each demand node , The total number of demand nodes, with the weight of each demand node being the comprehensive vulnerability index value. ; S32. Define coverage determination rules: If candidate sites With demand nodes Spatial distance between Not exceeding the preset service radius Then the demand node is considered Overwritten: ; in, To cover the identifier variable, Indicates demand node Candidate sites cover, This indicates that it was not covered. Indicates candidate sites With demand nodes Spatial distance between them This is the preset service radius.

[0009] Preferably, the risk-priority maximum coverage location model further includes: S33, in the During the round-robin site selection process, each candidate site that has not yet been selected is calculated. Marginal revenue: ; in, For the first Round of candidate sites marginal revenue, Represents the set of demand nodes Before removal The set of demand nodes that have been covered by the wheel The remaining set of demand nodes For demand nodes The comprehensive vulnerability index value, Indicates candidate sites Availability, This indicates that the candidate site is available. Indicates unavailable; S34. Select the candidate site with the highest marginal benefit as the result of this round of site selection: ; in, For the first Emergency bus stops selected in rotation Indicates taking the objective function The variable that achieves the maximum value , Represents the set of candidate sites Before removal The selected emergency bus stops The remaining set of candidate sites; S35. Iterate through S33 to S34 until a preset number of emergency bus stops are selected, obtaining the final set of emergency bus stops. ,in This represents the total number of pre-set emergency bus stop locations. S36. Evaluate emergency bus stop layout schemes using expected weighted coverage: ; in, For the expected weighted coverage, To be gathered at the final emergency bus stop The set of required nodes to be covered For demand nodes The comprehensive vulnerability index value.

[0010] Preferably, determining the drone deployment point in S4 specifically includes: S41. Set up a set of candidate observation points for the UAV. ,in Indicates the first Candidate observation points, This represents the total number of candidate observation points. S42. Determine each candidate observation point. The set of roadside areas that can be covered within the observation radius: ; in, Candidate observation points Coverable roadside collection, It refers to the side of a road in a road network. , roadside The index numbers of the two endpoint nodes, It is the set of road network edges in a multi-layered complex transportation network. Indicates candidate observation points and the side of the road midpoint Spatial distance between them The preset observation radius of the drone; S43. Calculate candidate observation points Information gain: ; in, Candidate observation points Information gain Represents the set of roads that can be covered. Remove the set of covered road edges The remaining roadside groups were then gathered. Indicates the side of the road Information value; S44. In each round, select the candidate observation point with the largest information gain: ; in, For the first The selected drone deployment sites Indicates taking the objective function The variable that achieves the maximum value , Represents the set of candidate observation points Before removal Round the selected drone deployment site set The set of remaining candidate observation points; Iteratively execute steps S42 to S44 until a preset number of drone deployment points are selected, resulting in a set of drone deployment points. ,in This represents the preset total number of drone deployment points.

[0011] Preferably, in S4, the traffic risk value is updated based on the drone's coverage of the road network edges, specifically including: S45, For the set of drone deployment points Any roadside covered by any drone Update the roadside Traffic risk value: ; in, For the updated roadside The passage risk value, roadside The original passage risk value, For drone information update intensity coefficient, It is a binary indicator variable. Indicates the side of the road Covered by at least one drone from the set of drone deployment points. This indicates that the area was not covered by any drones. S46. Based on the updated traffic risk value, calculate the traffic impedance along the corresponding roadside: ; in, For the updated roadside The conduction impedance, roadside The original passing impedance, This represents the risk penalty coefficient.

[0012] Preferably, S5 specifically includes: S51. Obtain information from emergency bus stops to evacuation destinations. The candidate emergency paths constitute the candidate path set. ,in Indicates the first Candidate emergency paths, The total number of candidate emergency routes; S52. Calculate each candidate emergency path Total path cost: ; in, Candidate emergency paths Total path cost, Candidate emergency paths The set of roadside included. For the updated roadside The conduction impedance; S53. Select the candidate emergency path with the minimum total cost from the candidate path set as the optimal emergency path: ; in, This is the optimal emergency path. Indicates taking the objective function Variable that achieves minimum value And obtain the set of roadsides included in the optimal emergency path. .

[0013] Preferably, the method further includes evaluating the adequacy of unmanned aerial vehicle (UAV) services, specifically including: Determine the set of drone deployment points Covered roadside collection : ; in, The set of roadside locations covered by the set of drone deployment points. Quantifiers indicating existence This indicates the existence of a set of drone deployment points. drone deployment sites , Indicates drone deployment points and the side of the road midpoint Spatial distance between them The preset observation radius of the drone; Calculate road information coverage Emergency path observation rate Corridor Concentration Non-corridor drone ratio : ; ; ; ; in, Road information coverage rate, representing the proportion of roadsides covered by drones to the total number of road network edges; Emergency path observation rate represents the proportion of the number of optimal emergency path edges covered by drones to the total number of optimal emergency path edges. The corridor concentration represents the proportion of the number of optimal emergency path edges covered by drones to the total number of road edges covered by drones. The percentage of non-corridor drones represents the proportion of drone coverage that was not used to observe optimal emergency routes; Indicates the number of elements in the set. This is the set of road edges included in the optimal emergency path. It is the set of road network edges in a multi-layered complex transportation network. This represents the intersection of the set of roadsides covered by drones and the set of roadsides for the optimal emergency path; When both conditions are met , and At that time, the drone service was deemed sufficient.

[0014] Secondly, the present invention also provides an air-ground coordinated emergency response system for sudden subway failures, comprising: The data acquisition module is used to acquire road network data, public transportation network data, subway network data, and passenger flow data for the target city. The composite network construction module is used to construct a multi-layer composite transportation network based on the data obtained by the data acquisition module. The multi-layer composite transportation network includes a road network layer, a public transport network layer, a subway network layer, and cross-layer coupling connections between each network layer. The vulnerability identification module is used to identify high-risk vulnerable nodes in multi-layered complex transportation networks based on the topological characteristics and passenger flow functional attributes of the network. The emergency bus layout module is used to generate emergency bus stop layout schemes with the goal of prioritizing coverage of high-risk and vulnerable nodes. The UAV deployment and path optimization module is used to determine the deployment points of UAVs in a multi-layered complex transportation network, update the traffic risk value of the covered roadside based on the observation coverage status of the UAVs on the road network, and generate the optimal emergency path from the emergency bus stop to the evacuation target based on the updated traffic impedance. The decision generation module is used to generate air-ground coordinated emergency response decision schemes based on multi-layered complex transportation networks, high-risk and vulnerable nodes, emergency bus stop layout schemes, drone deployment points, and optimal emergency routes.

[0015] Therefore, the present invention adopts the above-mentioned air-ground coordinated emergency response method and system for sudden subway failures, and achieves the following technical effects: (1) This invention constructs a three-layer composite transportation network of subway-bus-road and adopts the Comprehensive Vulnerability Index (CVI-DS) to integrate topology and passenger flow characteristics, thereby achieving accurate identification of high-risk vulnerable nodes in the composite transportation network. This overcomes the limitations of traditional methods that rely solely on a single centrality index and provides a reliable quantitative basis for emergency resource allocation.

[0016] (2) The present invention adopts the risk-priority maximum coverage location model (R-MEXCLP) for emergency bus station layout, with high vulnerable nodes as the priority coverage target. Compared with the traditional maximum coverage model, the expected weighted coverage rate is increased by about 126.6%, and the coverage rate of the top 30 high vulnerable nodes reaches 0.3000, realizing the priority protection of core risk areas under the condition of limited bus resources.

[0017] (3) This invention constructs an integrated emergency response mechanism of "aerial perception - ground connection - cross-level evacuation" through the synergistic linkage of the UAV Information Gain Model (UAV-RIG) and the BIR-KSP dynamic path optimization algorithm. The UAV dynamically observes the road network and updates the traffic risk value, while the path optimization algorithm integrates real-time traffic information to generate the optimal evacuation path. Experimental results show that the core traffic corridor guarantee capacity reaches 98.62%, the emergency path observation rate reaches 33.08%, and the proportion of non-corridor UAVs reaches 97.24%, meeting the service adequacy judgment conditions. Moreover, the system performance fluctuation is less than 0.05 under different parameter disturbance conditions, demonstrating excellent stability and robustness.

[0018] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating an air-ground coordinated emergency response method for sudden subway failures according to the present invention. Figure 2 This is a schematic diagram of the spatial distribution of high-risk vulnerable nodes identified using the Comprehensive Vulnerability Index (CVI-DS) in an embodiment of the present invention; Figure 3 This is a spatial distribution diagram of the emergency bus stop layout scheme generated by the risk-first maximum coverage location model (R-MEXCLP) in this embodiment of the invention; Figure 4 This is a schematic diagram of the layout of UAV deployment points and service coverage in an embodiment of the present invention; Figure 5 This is a schematic diagram of the optimal emergency path generated by the dynamic path optimization algorithm (BIR-KSP) in an embodiment of the present invention. Detailed Implementation

[0020] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Example 1 This embodiment takes the public transportation system of a city's main urban area as the research object and obtains the following data: The road network data comes from open-source map platforms and Shapefile vector files, and includes road nodes, road edges, road levels, and connectivity relationships, which are used to construct the city's basic road network and emergency access space.

[0023] The public transport network data comes from the city's publicly available public transport route data, which includes the locations of bus stops and their connections to the routes, and is used to construct the ground-level public transport layer.

[0024] The subway network data comes from the city's publicly available subway line data, which includes the location of subway stations and their connections to the tracks, and is used to construct the rail transit layer.

[0025] The passenger flow data for the automated fare collection system was selected from the city's actual subway card swipe records in May 2021, including station entry and exit volumes and passenger flow matrices at origin and destination points, to construct passenger flow load indicators. This data contains 679 passenger flow statistics files, covering 8 lines, 160 stations, and 330 directed intervals.

[0026] In this embodiment, the transfer threshold between buses and subways is set to 500 meters. The composite transportation network constructed in this embodiment contains 5638 nodes.

[0027] This embodiment is implemented sequentially according to S1 to S6 below. Figure 1 The overall flow of the method in this embodiment is illustrated. The following sections provide a detailed description of the specific implementation of each step.

[0028] S1. Construct a multi-layered composite transportation network for the target city. The multi-layered composite transportation network includes a road network layer, a public transport network layer, a subway network layer, and cross-layer coupling connections between each network layer. The construction of a multi-layered, complex transportation network specifically includes: S11, Construct road networks separately Public transport network and subway network ,in For a set of nodes, Let be the set of edges. For weight set; S12. Establish cross-layer coupling connections between different network layers based on the principle of spatial proximity. For nodes belonging to different network layers... and If the spatial distance between the two Not exceeding the preset interlayer coupling radius Then, a cross-layer coupling edge is established between the two. The determination method is as follows: ; in, Indicates belonging to the network layer nodes Belonging to the network layer nodes Spatial distance between them , , , These represent the road network layer, the public transport network layer, and the subway network layer, respectively. Represents network layer With network layer The preset inter-layer coupling radius is used; in this embodiment, the inter-layer coupling radius is 500m, the same as the transfer threshold. Indicates at node With nodes Establish cross-layer coupling edges between them. This indicates that no cross-layer coupling edges are established; S13, Construct a multi-layered, complex transportation network The node set is The edge set is ,in This represents the set of cross-layer coupling edges between the public transport network layer and the road network layer. This represents the set of cross-layer coupling edges between the subway network layer and the road network layer. This is the set of cross-layer coupling edges between the public transport network layer and the subway network layer.

[0029] This composite network can simultaneously represent road traffic, bus operation, subway service, and multi-modal transfer relationships, serving as the foundation for subsequent vulnerability identification, emergency station site selection, and route optimization.

[0030] S2. Based on the topological characteristics and passenger flow functional attributes of the multi-layer composite transportation network, identify high-risk and vulnerable nodes in the multi-layer composite transportation network. High-risk vulnerable nodes are identified and evaluated using the Comprehensive Vulnerability Index (CVI-DS). This index comprehensively considers the node's topological importance, passenger flow function, and transfer load, avoiding a one-sided evaluation caused by relying solely on a single centrality indicator. Specifically, it includes: Calculate each node in a multi-layered complex transportation network Comprehensive vulnerability index: ; in, For nodes The comprehensive vulnerability index value, The normalized degree centrality, degree centrality This indicates the number of neighboring nodes of a node. For normalized betweenness centrality, betweenness centrality This indicates the frequency of a node appearing in all shortest paths. For normalized proximity centrality, proximity centrality This represents the reciprocal of the average shortest path length from a node to all other nodes. All three types of centrality are calculated based on the edge weight structure of a multi-layered complex transportation network. The normalized passenger flow intensity This is the normalized transfer load. , , , , These are the weighting coefficients for the corresponding indicators; in this embodiment, the weighting coefficients are set to... , , , , Betweenness centrality The highest weight is given to highlight the control role of nodes in the network propagation process.

[0031] In this embodiment, all nodes in the complex transportation network are sorted in descending order based on the comprehensive vulnerability index value, and the top 20 nodes are selected as high-risk vulnerable nodes.

[0032] Figure 2 This is a schematic diagram illustrating the spatial distribution of high-risk, vulnerable nodes. Figure 2 It can be seen that high-risk and vulnerable nodes are mainly concentrated in the central urban area and the intersection of main transportation corridors, showing obvious characteristics of central agglomeration and cross-layer diffusion.

[0033] S3. With the goal of prioritizing coverage of high-risk and vulnerable nodes, generate an emergency bus stop layout plan; This embodiment uses the risk-priority maximum coverage location model (R-MEXCLP) to generate an emergency bus stop layout scheme. The goal of this model is not simply to cover the most nodes, but to prioritize covering critical demand nodes with higher overall vulnerability indicators. Specifically, it includes: S31. Set up a set of candidate emergency bus stops and demand node set ,in Indicates the first One candidate emergency bus stop The total number of candidate emergency bus stops. Indicates the first Each demand node , The total number of demand nodes, with the weight of each demand node being the comprehensive vulnerability index value. Candidate emergency bus stops are selected from existing bus stations and road nodes with stopping conditions in the city, while the required nodes are the high-risk and vulnerable nodes identified in S2. In this embodiment, the service radius... The distance is set to 1500m. Road nodes with suitable stopping conditions refer to road sections with a width of not less than 3.5 meters, no non-motorized vehicle lane separation, and no construction encroachment on the surrounding area. Candidate stops are selected based on the following criteria: node degree not less than 2, and distance from the nearest bus stop not less than 100m and not more than 500m.

[0034] S32. Define coverage determination rules: If candidate sites With demand nodes Spatial distance between Not exceeding the preset service radius Then the demand node is considered Overwritten: ; in, To cover the identifier variable, Indicates demand node Candidate sites cover, This indicates that it was not covered. Indicates candidate sites With demand nodes Spatial distance between them This is the preset service radius.

[0035] S33, in the During the round-robin site selection process, each candidate site that has not yet been selected is calculated. Marginal revenue: ; in, For the first Round of candidate sites marginal revenue, Represents the set of demand nodes Before removal The set of demand nodes that have been covered by the wheel The remaining set of demand nodes For demand nodes The comprehensive vulnerability index value, Indicates candidate sites Availability, This indicates that the candidate site is available. Indicates unavailable; S34. Select the candidate site with the highest marginal benefit as the result of this round of site selection: ; in, For the first Emergency bus stops selected in rotation Indicates taking the objective function The variable that achieves the maximum value , Represents the set of candidate sites Before removal The selected emergency bus stops The remaining set of candidate sites; S35. Iterate through S33 to S34 until a preset number of emergency bus stops are selected, obtaining the final set of emergency bus stops. ,in The total number of emergency bus stop locations is preset; in this embodiment, the number of locations is set to 20, which matches the number of high-risk and vulnerable nodes.

[0036] S36. Evaluate emergency bus stop layout schemes using expected weighted coverage: ; in, For the expected weighted coverage, To be gathered at the final emergency bus stop The set of required nodes to be covered For demand nodes The comprehensive vulnerability index value.

[0037] Calculations show that the expected weighted coverage rate in this embodiment is 0.2604, and the coverage rate of the Top 30 highly vulnerable nodes is 0.3000, indicating that limited public transportation resources can more effectively cover core risk areas.

[0038] Figure 3 This is a spatial distribution diagram of the emergency bus stop layout plan. Figure 3 It can be seen that emergency bus stops are mainly concentrated in the central urban area and near core transportation corridors, which is highly consistent with the spatial distribution of high-risk and vulnerable nodes.

[0039] S4. Determine the deployment points of UAVs in the multi-layered complex transportation network, and update the traffic risk value of the covered roadside based on the observation coverage status of the UAVs over the road network. This embodiment uses a UAV-RIG (Unmanned Aerial Vehicle Information Gain Model) to determine UAV deployment points. The UAV observation points not only serve the final emergency path but also supplement road information in non-corridor areas, thus avoiding the cyclical verification problem caused by deploying UAVs entirely along the path. Specifically, this includes: S41. Set up a set of candidate observation points for the UAV. Observation radius The m depth was set to 2000m. Candidate observation points for the drones were selected from existing drone take-off and landing sites and open areas with safe take-off and landing conditions within the city. Indicates the first Candidate observation points, This represents the total number of candidate observation points. S42. Determine each candidate observation point. The set of roadside areas that can be covered within the observation radius: ; in, Candidate observation points Coverable roadside collection, It refers to the side of a road in a road network. , roadside The index numbers of the two endpoint nodes, It is the set of road network edges in a multi-layered complex transportation network. Indicates candidate observation points and the side of the road midpoint Spatial distance between them The preset observation radius of the drone; S43. Calculate candidate observation points Information gain: ; in, Candidate observation points Information gain Represents the set of roads that can be covered. Remove the set of covered road edges The remaining roadside groups were then gathered. Indicates the side of the road The information value is calculated based on a comprehensive assessment of roadside traffic flow and connectivity importance: ,in This represents the average daily traffic flow along the roadside. For the betweenness centrality of road edges in the network, , This is the weighting coefficient. In this embodiment... =0.6、 =0.4.

[0040] S44. In each round, select the candidate observation point with the largest information gain: ; in, For the first The selected drone deployment sites Indicates taking the objective function The variable that achieves the maximum value , Represents the set of candidate observation points Before removal Round the selected drone deployment site set The set of remaining candidate observation points; Iteratively execute steps S42 to S44 until a preset number of drone deployment points are selected, resulting in a set of drone deployment points. ,in This represents the preset total number of drone deployment points. In this embodiment, a total of 18 drone deployment points were selected.

[0041] The traffic risk value is updated based on the drone's coverage of the road network edges, specifically including: S45, For the set of drone deployment points Any roadside covered by any drone Update the roadside Traffic risk value: ; in, For the updated roadside The passage risk value, roadside The original traffic risk value is determined comprehensively based on the road grade and design speed: For expressways... Main road corresponding Secondary trunk roads correspond to Branch road corresponding =0.2, The value ranges from 0 to 1, with higher values ​​indicating greater risk of passage. For drone information update intensity coefficient, It is a binary indicator variable. Indicates the side of the road Covered by at least one drone from the set of drone deployment points. This indicates that the area was not covered by any drones. S46. Based on the updated traffic risk value, calculate the traffic impedance along the corresponding roadside: ; in, For the updated roadside The conduction impedance, roadside The original passing impedance, This is a risk penalty coefficient used to adjust the severity of the penalty imposed on traffic risk on traffic resistance. Its value range is set according to the safety tolerance of the road traffic environment and emergency response requirements. Under normal emergency scenarios, A value of 1.0 indicates that the risk value is proportionally converted into an increase in the passage impedance; in scenarios with higher safety requirements, The size can be increased appropriately; in this embodiment, it is taken as... .

[0042] Figure 4 A schematic diagram showing the layout of drone deployment points and service coverage. (Source: [Insert Source Here]) Figure 4 It can be seen that the drone deployment points are mainly distributed in the central urban area and near key transportation corridors, achieving key coverage of major rail transit transfer areas and high-risk passenger flow gathering areas.

[0043] S5. Based on the updated traffic impedance, generate the optimal emergency route from the emergency bus stop to the evacuation target in the multi-layered complex traffic network. This embodiment employs the Biometric Path Optimization (BIR-KSP) algorithm to generate the optimal emergency path. This algorithm dynamically updates roadside risks by incorporating UAV observation information, thereby altering the edge weight structure during the path search process. This ensures that the emergency path not only considers the traditional shortest path cost but also integrates the benefits of reduced information uncertainty. Specifically, it includes: S51. Obtain information from emergency bus stops to evacuation destinations. The candidate emergency paths constitute the candidate path set. ,in Indicates the first Candidate emergency paths, This represents the total number of candidate emergency paths; in this embodiment... Take 5.

[0044] S52. Calculate each candidate emergency path Total path cost: ; in, Candidate emergency paths Total path cost, Candidate emergency paths The set of roadside included. For the updated roadside The conduction impedance; S53. Select the candidate emergency path with the minimum total cost from the candidate path set as the optimal emergency path: ; in, This is the optimal emergency path. Indicates taking the objective function Variable that achieves minimum value And obtain the set of roadsides included in the optimal emergency path. .

[0045] Figure 5 This is a schematic diagram of the optimal emergency route. Figure 5 It can be seen that the optimized emergency routes mainly cover the city's core transportation corridors and high-risk transfer areas, realizing coordinated linkage between the subway network, bus network and drone aerial response layer.

[0046] S6. Based on the multi-layered complex transportation network, high-risk and vulnerable nodes, emergency bus stop layout plan, drone deployment points and optimal emergency routes, generate an air-ground coordinated emergency response decision plan.

[0047] This embodiment also constructs an air-ground coordinated emergency evacuation simulation platform based on the Python programming language and the Three.js graphics library. The platform consists of a data layer, a simulation calculation layer, and a front-end visualization layer: the data layer receives urban passenger flow data and spatial topology data; the simulation calculation layer performs logical calculations for passenger flow dynamic evolution, UAV collaborative reconnaissance, and emergency bus connections; the front-end visualization layer performs 3D dynamic rendering of the complex transportation network, passenger flow status, bus connection routes, and UAV flight trajectories. The platform supports setting parameters such as interruption stations, number of UAVs, bus capacity, and scheduling schemes, and can display the passenger evacuation process and resource coordination status under different schemes in real time.

[0048] This embodiment uses the simulation platform to dynamically simulate and verify the above decision-making scheme. To verify the rationality of the UAV deployment, this embodiment sets evaluation indicators such as road information coverage, emergency path observation rate, corridor concentration, and the ratio of UAVs to non-corridor routes. Specifically, these include: Determine the set of drone deployment points Covered roadside collection : ; in, The set of roadside locations covered by the set of drone deployment points. Quantifiers indicating existence This indicates the existence of a set of drone deployment points. drone deployment sites , Indicates drone deployment points and the side of the road midpoint Spatial distance between them The preset observation radius of the drone; Calculate road information coverage Emergency path observation rate Corridor Concentration Non-corridor drone ratio : ; ; ; ; in, Road information coverage rate, representing the proportion of roadsides covered by drones to the total number of road network edges; Emergency path observation rate represents the proportion of the number of optimal emergency path edges covered by drones to the total number of optimal emergency path edges. The corridor concentration represents the proportion of the number of optimal emergency path edges covered by drones to the total number of road edges covered by drones. The percentage of non-corridor drones represents the proportion of drone coverage that was not used to observe optimal emergency routes; Indicates the number of elements in the set. This is the set of road edges included in the optimal emergency path. It is the set of road network edges in a multi-layered complex transportation network. This represents the intersection of the set of roadsides covered by drones and the set of roadsides for the optimal emergency path; The criteria for determining the adequacy of drone services in this embodiment are as follows: ; in, This indicates that drone services are adequate. This indicates that drone services are inadequate.

[0049] The verification results of this embodiment are as follows: Core transportation corridor support capacity: 98.62%; Road information coverage 19.89%; Emergency path observation rate 33.08%; Non-core area coverage: 61.11%; Coverage along roadsides: 43,756; Emergency routes covered: 1206; Total number of roads: 3646; Based on the above data, the following calculations were performed: ; ; ; The above indicators meet the criteria for determining the adequacy of drone services: Emergency path observation rate ; Corridor Concentration ; Non-corridor drone ratio ; therefore This demonstrates that the drone deployment scheme in this embodiment effectively supports emergency routes and has the ability to supplement road information in non-route areas.

[0050] Further, the differences between different methods were evaluated using relative lift rates: ; in, This invention represents the method of the present invention. Representation of the benchmark method, Indicates the indicator to be compared. This refers to the relative increase rate.

[0051] Calculations show that the expected coverage rate in this embodiment reaches 0.2604, which is about 126.6% higher than that of the traditional CVI-only baseline model, indicating that the method of the present invention has a significant performance improvement in emergency resource allocation.

[0052] This embodiment also provides an air-ground coordinated emergency response system for sudden subway failures, used to execute the above-described methods. The system includes the following modules: Data acquisition module: This module is used to acquire road network data, public transport network data, subway network data, and passenger flow data from the automated fare collection system for the target city. The data acquisition module then outputs the acquired data to the composite network construction module.

[0053] Composite network building blocks: Connected to the data acquisition module, this module constructs a multi-layered composite transportation network based on the acquired data. Specifically, it first constructs road subnetworks, bus subnetworks, and subway subnetworks separately. Then, based on the principle of spatial proximity, it establishes cross-layer coupling connections between the subnetworks, and finally merges them to obtain the multi-layered composite transportation network. This module outputs the completed network to the vulnerability identification module.

[0054] Vulnerability identification module: Connected to the composite network construction module, this module receives data from a multi-layered composite transportation network. Based on the network's topological characteristics and passenger flow functional attributes, it evaluates all nodes using a comprehensive vulnerability index, identifying the nodes with the highest comprehensive vulnerability index values ​​as high-risk vulnerable nodes. This module then outputs the high-risk vulnerable nodes and their index values ​​to the emergency bus deployment module and the drone deployment and route optimization module.

[0055] Emergency bus layout module: Connected to the vulnerability identification module, this module receives high-risk vulnerable nodes and their comprehensive vulnerability index values. With the goal of prioritizing coverage of high-risk vulnerable nodes, it generates an emergency bus stop layout scheme using a risk-first maximum coverage site selection model. Specifically, high-risk vulnerable nodes are designated as demand nodes, and their comprehensive vulnerability index values ​​are used as the weights of these demand nodes. A greedy iterative approach is then used to select the candidate sites with the highest marginal benefit. This module outputs the generated layout scheme to the decision generation module and the UAV deployment and path optimization module.

[0056] Drone deployment and path optimization module: Connected to the composite network construction module, vulnerability identification module, and emergency bus layout module, this module receives multi-layered composite transportation networks, high-risk vulnerable nodes, and emergency bus stop layout schemes. It determines UAV deployment points within the multi-layered composite transportation network, updates the traffic risk values ​​of covered road edges based on UAV observations of road network edges, and generates the optimal emergency path from the emergency bus stop to the evacuation target based on the updated traffic impedance. This module outputs the UAV deployment points and the optimal emergency path to the decision generation module.

[0057] Decision generation module: Connecting to modules for composite network construction, vulnerability identification, emergency bus deployment, and UAV deployment and path optimization, this module receives data on multi-layered composite transportation networks, high-risk vulnerable nodes, emergency bus stop layout plans, UAV deployment points, and optimal emergency routes. It then generates an air-ground coordinated emergency response decision plan. The decision plan includes a list of high-risk vulnerable nodes and corresponding safeguards, an emergency bus deployment table, UAV flight plans, and an optimal emergency route map with estimated travel time. This module pushes the decision plan to on-site commanders, bus drivers, and UAV pilots via mobile terminals to guide on-site response efforts.

[0058] Simulation verification module: Connected to the decision generation module, this module receives air-ground coordinated emergency response decision plans. Based on the Python programming language and the Three.js graphics library, it constructs an air-ground coordinated emergency evacuation simulation platform to perform 3D visual dynamic simulation verification of the decision plans. This module includes a data layer, a simulation calculation layer, and a front-end visualization layer, capable of displaying in real-time the dynamic evolution of passenger flow after subway disruption, UAV-assisted reconnaissance, and emergency bus connection processes.

[0059] The modules are connected sequentially to form a complete system encompassing data acquisition, network construction, vulnerability identification, public transportation layout, UAV deployment and path optimization, and decision generation and verification. Each module is implemented using conventional computer program modules, runs on general-purpose computing devices, and data transfer between modules is accomplished through standard data interfaces.

[0060] Taking a scenario where a core transportation hub, Station A, experiences a service disruption due to equipment failure, causing subway operations to be interrupted at the station and in adjacent sections, the emergency command center activates the system described in this invention: Step 1: The system calls upon the city's existing road network data, public transport network data, and subway network data, and combines them with real-time passenger flow data to quickly construct a multi-layered, complex transportation network.

[0061] Step 2: Based on real-time passenger flow data and network topology, the system conducts a comprehensive vulnerability assessment of all subway stations within the affected area and identifies the stations with the highest comprehensive vulnerability index values ​​as key protection targets.

[0062] Step 3: Based on the spatial distribution of high-risk and vulnerable nodes, the system adopts a risk-priority maximum coverage location selection model. Under the constraint of a preset service radius, it selects the optimal emergency bus stop from the surrounding bus stations and sends the dispatch instructions to the corresponding buses through mobile terminals.

[0063] Step 4: The system dispatches multiple drones from the nearest drone take-off and landing point, plans observation routes according to the principle of maximizing information gain, conducts real-time reconnaissance of the road network in the affected area, and transmits road traffic status information back to the command center in real time to update the traffic impedance of each roadside.

[0064] Step 5: The system integrates real-time traffic information transmitted back by drones and uses a dynamic path optimization algorithm to calculate the optimal path from each emergency bus stop to the surrounding normally operating subway station.

[0065] Step Six: The system integrates the emergency bus shuttle plan, the drone reconnaissance plan, and the optimal evacuation route into a complete air-ground coordinated emergency response decision plan, which is then pushed to on-site commanders, bus drivers, and drone pilots via mobile terminals to guide on-site handling work.

[0066] Therefore, this invention employs the aforementioned air-ground collaborative emergency response method and system for sudden subway failures. By constructing a three-layer composite network of subway, bus, and road and integrating UAV dynamic perception, it achieves accurate identification and priority coverage of high-risk and vulnerable nodes. Experimental results show that the expected weighted coverage rate of emergency buses reaches 0.2604, an improvement of 126.6% compared to traditional methods; the core transportation corridor guarantee capability under UAV collaboration reaches 98.62%; and the emergency path observation rate reaches 33.08%, meeting the service adequacy judgment criteria. The system performance fluctuation is less than 0.05 under different parameter disturbances, verifying that this invention has good stability and robustness, and can effectively support emergency decision-making and evacuation command in scenarios of sudden urban rail transit disruptions.

[0067] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for air-ground coordinated emergency response to sudden subway failures, characterized in that, Includes the following steps: S1. Construct a multi-layered composite transportation network for the target city. The multi-layered composite transportation network includes a road network layer, a public transport network layer, a subway network layer, and cross-layer coupling connections between each network layer. S2. Based on the topological characteristics and passenger flow functional attributes of the multi-layer composite transportation network, identify high-risk and vulnerable nodes in the multi-layer composite transportation network. S3. With the goal of prioritizing coverage of high-risk and vulnerable nodes, generate an emergency bus stop layout plan; S4. Determine the deployment points of UAVs in the multi-layered complex transportation network, and update the traffic risk value of the covered roadside based on the observation coverage status of the UAVs over the road network. S5. Based on the updated traffic impedance, generate the optimal emergency route from the emergency bus stop to the evacuation target in the multi-layered complex traffic network. S6. Based on the multi-layered complex transportation network, high-risk and vulnerable nodes, emergency bus stop layout plan, drone deployment points and optimal emergency routes, generate an air-ground coordinated emergency response decision plan.

2. The method according to claim 1, characterized in that, The construction of the multi-layered complex transportation network in S1 specifically includes: S11, Construct road networks separately Public transport network and subway network ,in For a set of nodes, Let be the set of edges. For weight set; S12. For nodes belonging to different network layers and If the spatial distance between the two Not exceeding the preset interlayer coupling radius Then, a cross-layer coupling edge is established between the two. The determination method is as follows: ； in, Indicates belonging to the network layer nodes Belonging to the network layer nodes Spatial distance between them , , , These represent the road network layer, the public transport network layer, and the subway network layer, respectively. Represents network layer With network layer The preset interlayer coupling radius between them Indicates at node With nodes Establish cross-layer coupling edges between them. This indicates that no cross-layer coupling edges are established; S13, Construct a multi-layered, complex transportation network The node set is The edge set is ,in This represents the set of cross-layer coupling edges between the public transport network layer and the road network layer. This represents the set of cross-layer coupling edges between the subway network layer and the road network layer. This is the set of cross-layer coupling edges between the public transport network layer and the subway network layer.

3. The method according to claim 1, characterized in that, S2 identifies high-risk vulnerable nodes and evaluates them using a comprehensive vulnerability index, specifically including: Calculate each node in a multi-layered complex transportation network Comprehensive vulnerability index: ； in, For nodes The comprehensive vulnerability index value, For the normalized degree centrality, For normalized betweenness centrality, To achieve normalized proximity centrality, The normalized passenger flow intensity This is the normalized transfer load. , , , , These are the weighting coefficients for the corresponding indicators; All nodes in the multi-layered complex transportation network are sorted in descending order based on the comprehensive vulnerability index value, and the top N nodes are selected as high-risk vulnerable nodes, where N is a positive integer.

4. The method according to claim 1, characterized in that, S3 generates an emergency bus stop layout plan, employing a risk-priority maximum coverage site selection model, specifically including: S31. Set up a set of candidate emergency bus stops and demand node set ,in Indicates the first One candidate emergency bus stop The total number of candidate emergency bus stops. Indicates the first Each demand node , The total number of demand nodes, with the weight of each demand node being the comprehensive vulnerability index value. ; S32. Define coverage determination rules: If candidate sites With demand nodes Spatial distance between Not exceeding the preset service radius Then the demand node is considered Overwritten: ； in, To cover the identifier variable, Indicates demand node Candidate sites cover, This indicates that it was not covered. Indicates candidate sites With demand nodes Spatial distance between them This is the preset service radius.

5. The method according to claim 4, characterized in that, The risk-priority maximum coverage location selection model further includes: S33, in the During the round-robin site selection process, each candidate site that has not yet been selected is calculated. Marginal revenue: ； in, For the first Round of candidate sites marginal revenue, Represents the set of demand nodes Before removal The set of demand nodes that have been covered by the wheel The remaining set of demand nodes For demand nodes The comprehensive vulnerability index value, Indicates candidate sites Availability, This indicates that the candidate site is available. Indicates unavailable; S34. Select the candidate site with the highest marginal benefit as the result of this round of site selection: ； in, For the first Emergency bus stops selected in rotation Indicates taking the objective function The variable that achieves the maximum value , Represents the set of candidate sites Before removal The selected emergency bus stops The remaining set of candidate sites; S35. Iterate through S33 to S34 until a preset number of emergency bus stops are selected, obtaining the final set of emergency bus stops. ,in This represents the total number of pre-set emergency bus stop locations. S36. Evaluate emergency bus stop layout schemes using expected weighted coverage: ； in, For the expected weighted coverage, To be gathered at the final emergency bus stop The set of required nodes to be covered For demand nodes The comprehensive vulnerability index value.

6. The method according to claim 1, characterized in that, Determining drone deployment points in S4 specifically includes: S41. Set up a set of candidate observation points for the UAV. ,in Indicates the first Candidate observation points, This represents the total number of candidate observation points. S42. Determine each candidate observation point. The set of roadside areas that can be covered within the observation radius: ； in, Candidate observation points Coverable roadside collection, It refers to the side of a road in a road network. , roadside The index numbers of the two endpoint nodes, It is the set of road network edges in a multi-layered complex transportation network. Indicates candidate observation points and the side of the road midpoint Spatial distance between them The preset observation radius of the drone; S43. Calculate candidate observation points Information gain: ； in, Candidate observation points Information gain Represents the set of roads that can be covered. Remove the set of covered road edges The remaining roadside groups were then gathered. Indicates the side of the road Information value; S44. In each round, select the candidate observation point with the largest information gain: ； in, For the first The selected drone deployment sites Indicates taking the objective function The variable that achieves the maximum value , Represents the set of candidate observation points Before removal Round the selected drone deployment site set The set of remaining candidate observation points; Iteratively execute steps S42 to S44 until a preset number of drone deployment points are selected, resulting in a set of drone deployment points. ,in This represents the preset total number of drone deployment points.

7. The method according to claim 6, characterized in that, In S4, the traffic risk value is updated based on the drone's coverage of the road network edges, specifically including: S45, For the set of drone deployment points Any roadside covered by any drone Update the roadside Traffic risk value: ； in, For the updated roadside The passage risk value, roadside The original passage risk value, For drone information update intensity coefficient, It is a binary indicator variable. Indicates the side of the road Covered by at least one drone from the set of drone deployment points. This indicates that the area was not covered by any drones. S46. Based on the updated traffic risk value, calculate the traffic impedance along the corresponding roadside: ； in, For the updated roadside The conduction impedance, roadside The original passing impedance, This represents the risk penalty coefficient.

8. The method according to claim 1, characterized in that, S5 specifically includes: S51. Obtain information from emergency bus stops to evacuation destinations. The candidate emergency paths constitute the candidate path set. ,in Indicates the first Candidate emergency paths, The total number of candidate emergency routes; S52. Calculate each candidate emergency path Total path cost: ； in, Candidate emergency paths Total path cost, Candidate emergency paths The set of roadside included. For the updated roadside The conduction impedance; S53. Select the candidate emergency path with the minimum total cost from the candidate path set as the optimal emergency path: ； in, This is the optimal emergency path. Indicates taking the objective function Variable that achieves minimum value And obtain the set of roadsides included in the optimal emergency path. .

9. The method according to claim 8, characterized in that, The method also includes evaluating the adequacy of drone services, specifically including: Determine the set of drone deployment points Covered roadside collection : ； in, The set of roadside locations covered by the set of drone deployment points. Quantifiers indicating existence This indicates the existence of a set of drone deployment points. drone deployment sites , Indicates drone deployment points and the side of the road midpoint Spatial distance between them The preset observation radius of the drone; Calculate road information coverage Emergency path observation rate Corridor Concentration Non-corridor drone ratio : ； ； ； ； in, Road information coverage rate, representing the proportion of roadsides covered by drones to the total number of road network edges; Emergency path observation rate represents the proportion of the number of optimal emergency path edges covered by drones to the total number of optimal emergency path edges. The corridor concentration represents the proportion of the number of optimal emergency path edges covered by drones to the total number of road edges covered by drones. The percentage of non-corridor drones represents the proportion of drone coverage that was not used to observe optimal emergency routes; Indicates the number of elements in the set. This is the set of road edges included in the optimal emergency path. It is the set of road network edges in a multi-layered complex transportation network. This represents the intersection of the set of roadsides covered by drones and the set of roadsides for the optimal emergency path; When both conditions are met , and At that time, the drone service was deemed sufficient.

10. An air-ground coordinated emergency response system for sudden subway failures, used to execute the method according to any one of claims 1-9, characterized in that, include: The data acquisition module is used to acquire road network data, public transportation network data, subway network data, and passenger flow data for the target city. The composite network construction module is used to construct a multi-layer composite transportation network based on the data obtained by the data acquisition module. The multi-layer composite transportation network includes a road network layer, a public transport network layer, a subway network layer, and cross-layer coupling connections between each network layer. The vulnerability identification module is used to identify high-risk vulnerable nodes in multi-layered complex transportation networks based on the topological characteristics and passenger flow functional attributes of the network. The emergency bus layout module is used to generate emergency bus stop layout schemes with the goal of prioritizing coverage of high-risk and vulnerable nodes. The UAV deployment and path optimization module is used to determine the deployment points of UAVs in a multi-layered complex transportation network, update the traffic risk value of the covered roadside based on the observation coverage status of the UAVs on the road network, and generate the optimal emergency path from the emergency bus stop to the evacuation target based on the updated traffic impedance. The decision generation module is used to generate air-ground coordinated emergency response decision schemes based on multi-layered complex transportation networks, high-risk and vulnerable nodes, emergency bus stop layout schemes, drone deployment points, and optimal emergency routes.

Images