A traditional village fire scene cascade failure road vulnerability assessment method

Abstract

The application discloses a kind of traditional village fire scene cascade failure road vulnerability assessment method, it is related to building group fire prevention and control and personnel evacuation technical field.Firstly, the spatial vector data of traditional village building group, single building and road is acquired, and building group fire spread network and road network are constructed;Subsequently, the road vulnerability assessment system is constructed, and the accessibility index is introduced to dynamically represent the road traffic efficiency, and the vulnerability index is defined to quantify the influence of disaster on road system;Finally, a multi-layer network cascade failure simulation framework is proposed, which can quantify the dynamic cascade failure process and generate multi-scenario vulnerability heat map to locate high-risk areas.The application fully considers the influence of traditional village building fire on road network, establishes a fire-road double-layer coupled network model, realizes dynamic quantification, and provides a scientific decision-making tool for traditional village building fire and road evacuation risk, which has important application value for improving the resilience of traditional village.

Description

Technical Field

[0001] This invention relates to the technical fields of fire prevention and control in building complexes and personnel evacuation in fire scenarios, specifically to a method for assessing the vulnerability of roads in traditional village fire scenarios with cascading failures. Background Technology

[0002] Because traditional villages mostly use wooden structures with compact layouts and outdated fire safety planning, fires can spread rapidly and easily cause serious consequences. Building-borne fires not only directly lead to economic losses but also trigger a cascading failure of adjacent road networks, specifically by blocking evacuation routes and delaying fire and rescue responses, thus exacerbating the severity of the disaster.

[0003] In the field of traditional village fire research, previous studies have largely focused on the characteristics of building fires themselves, such as the dynamic mechanisms of fire spread, fire simulation based on Pyrosim software, fire risk assessment, and fire safety planning. Regarding research on traditional village road networks, existing studies mainly concentrate on the layout and morphological analysis of traditional village roads and basic transportation network planning. Although some studies have begun to pay attention to the impact of disasters on traditional village infrastructure, they focus primarily on earthquake disasters, analyzing the damage patterns of earthquakes to traditional village roads, such as road cracking, collapse, and landslide blockage. However, existing research has not considered the impact of building fires on the road network itself and its evacuation functions. Therefore, it cannot fully reveal the multifaceted impact of building fires on traditional villages.

[0004] Existing research largely focuses on the characteristics of building fires themselves (fire spread mechanisms in building complexes, fire simulation, and fire risk assessment) or single systems such as road network layout and traffic planning, lacking a systematic analysis of the dynamic coupling relationship between fire spread and road network functional failure. Single-layer network models can only analyze the vulnerability of a single system and cannot capture the mutual influence between building fires and road networks. Furthermore, single-layer network assessment models may underestimate the actual vulnerability of infrastructure under disasters because they ignore the cascading failure effects caused by node failures in multi-layer networks. Although multi-layer network models have been applied in urban roads, railways, aviation, and power-transportation scenarios, research on the specific scenario of traditional villages remains lacking, especially regarding the multi-layer coupling mechanism between building fires and road networks. While vulnerability assessment schemes based on complex networks and road functionality have been developed for urban transportation, these schemes are not all applicable to traditional villages. Because the wooden buildings and dense layout of traditional villages exacerbate the speed of fire spread, existing single-layer network assessment techniques cannot model the coupling relationship between building fire spread networks and road networks, resulting in an inability to predict the dynamic impact of fires on evacuation routes and rescue paths. These methods also failed to reveal the cascading failure mechanism of fire on roads, and therefore failed to propose intervention strategies to prevent cascading failure.

[0005] In the field of disaster analysis, complex network theory has been widely applied to assess infrastructure vulnerability under various disaster scenarios. However, these studies primarily focus on individual systems and single-layer networks. Scholars utilize complex network theory to construct urban infrastructure network models, identifying key nodes and links through topological characteristics (including node degree, betweenness centrality, shortest path, clustering coefficient, and network efficiency) or system performance indicators (travel cost, accessibility, or traffic flow, etc.) to assess network vulnerability under disasters such as earthquakes, rainstorms, floods, and hurricanes. However, single-layer network models often fail to capture the complex dependencies between various interconnected systems, making it difficult to accurately assess the robustness or vulnerability of infrastructure, thus exhibiting certain limitations. In contrast, research on multi-layer interdependent networks shows that removing only a small number of nodes can lead to the collapse of the multi-layer network structure, a significant difference from the scenario of numerous node failures in single-layer complex networks. Therefore, multi-layer network models are introduced to analyze the complex relationships between different infrastructure networks. In the field of transportation infrastructure, research has begun to explore the application of multi-layer network models in scenarios such as railway systems-urban rail transit, railway-urban road networks, railway-aviation networks, and power-transportation networks. These studies show that multilayer network research models can not only more accurately describe the interactions and dependencies between different systems, but also better assess the robustness and vulnerability of multilayer network cascading failures. Currently, the coupling relationship and cascading failure relationship between fire spread networks and road networks in traditional village buildings have become problems that urgently need to be solved by those skilled in the art. Summary of the Invention

[0006] To address the lack of research on the coupling relationship between traditional village fires and road networks in existing technologies, this invention proposes a vulnerability assessment method for cascading failures of roads in traditional village fire scenarios. This method effectively reveals the dynamic damage mechanism and cascading failure process of road networks under traditional village fire scenarios. Based on this, vulnerable nodes in the two-layer network of traditional villages are identified, and targeted prevention and control measures are taken to improve the overall disaster resilience of villages.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] The present invention provides a method for assessing the vulnerability of roads in a cascading failure scenario during a traditional village fire, comprising the following steps:

[0009] S1. Prepare building cluster parameter data, define a two-layer coupled network under the traditional village fire scenario, and construct the building fire spread network G. B and road network G R Finally, a two-layer network G={G B G R};

[0010] S1.1 Preparation of building group parameter data: including the collection of building group parameters, fire parameters and road parameters;

[0011] This invention uses a combination of UAV oblique photography technology and field survey to collect building parameters, fire parameters, and road parameters of the target building complex.

[0012] The building parameters include: the location, area, scale, and geographical distribution characteristics of each individual building;

[0013] The fire parameters include: dimensions of each individual building, fire resistance rating, building layout, building opening ratio, quantity of combustibles, and other fire parameter information;

[0014] The road parameters include: the connection relationship between each individual building and the road network, the length and width of each road segment, and the geographical layout of village roads;

[0015] This invention is selected from the ancient city street area. The architectural style is a typical log cabin wooden structure. The buildings in the building group are closely connected and arranged in a continuous area. There are a total of 195 buildings in the target area.

[0016] S1.2 Construct a two-layer coupled network model for traditional village fire scenarios;

[0017] The two-layer coupled network model includes: interlayer coupling matrix of building fire spread network. Interlayer coupling matrix of road network ;

[0018] A two-layer coupled network model is constructed, including a building fire spread network layer and a road network layer. The model can realistically display the impact of building fires on the road network; denoted as G={G B G R}, where G B For building sprawl networks, G R For road network;

[0019] In building fire spread calculations, to measure the spread of fire in a building... Impact on road function, introducing the interlayer coupling matrix of building fire spread network. Interlayer coupling matrix of road network ;in element Represents building nodes To its adjacent road network nodes actual distance To calculate the shortest evacuation time; inter-layer coupling matrix ,element , This indicates the vertical distance from the building to the adjacent road, describing the relationship between the fire source and the heated road section;

[0020] S1.3 Constructing a building fire spread network G based on a directed graph B The fire spread network is constructed by using individual buildings as nodes and the fire spread paths between buildings as directed edges.

[0021] Constructing a building fire spread network based on directed graph theory Define individual buildings within a building complex as nodes; that is, a single building is a directed graph. One of the nodes, the set of nodes is The village has M building nodes and a total of 195 individual buildings, i.e., M=195. The fire spread paths between buildings are defined as directed edges, resulting in a set of edges. The number of edges is N;

[0022] At the same time, determine the initial fire-starting building node. Attributes include building material (wood structure, brick-concrete structure), building area, building height, building ignition heat release rate, and fire risk parameters; determine edge attributes, including distance between building nodes, wind direction, and building heat radiation threshold.

[0023] S1.3.1 Determine the status of building nodes ;

[0024] Define node t at any time. The state is The expression is as follows:

[0025] ;

[0026] In the formula, For nodes The time of ignition, set This is the initial fire-starting building node. ,set up When it is not the initial ignition point and has not been ignited. Use inf to represent it; For nodes The rate of heat release in a fire, For nodes The temperature is a function of time t;

[0027] S1.3.2 Simulation of fire spread paths in buildings; including two parts: simulation of individual building areas and simulation of fire spread processes between buildings;

[0028] The method for simulating individual building areas is as follows: using the FDS tool from NIST (National Institute of Standards and Technology) in the United States, the fire heat release rate curve of the individual building is obtained. ;

[0029] The methods for simulating the spread of fires between buildings include the following:

[0030] Setting initial conditions: Simulation environment conditions (wind speed, temperature), building complex simulation conditions (number of building nodes, simulation step size, and total number of simulation steps), node initial conditions (initial fire node). The remaining nodes This invention only considers the fire spread caused by thermal radiation and fire plumes, and sets... Time Node Ignition conditions are kW / m 2 The specific formula is as follows:

[0031] ;

[0032] ;

[0033] ;

[0034] ;

[0035] In the formula, For timber, a power consumption of 12.5 kW / m² is generally acceptable. 2 ; For nodes The heat flux received by the wooden exterior walls or openings; The wall emissivity is set to 0.8; Let be the Stefan-Boltzmann law constant, taken as 5.67 × 10⁻⁸ W / (m²). 2 ·K); This represents the number of fire ignition points; For the first The fire ignition point relative to the building node The radiation angle coefficient of the wooden exterior wall; For nodes The thermal radiation power is calculated using the ignition source calculation method of the smoke control system standard (NFPA 92B); The wall-to-wall convective heat transfer coefficient can be calculated using correlations from large-space natural convection heat transfer experiments. In this invention, it is taken as 7.6 W / (m²). 2 ·K); The wall temperature of node j in the building that did not catch fire; for The equivalent superposition temperature of the fire plumes at each ignition point. The thermal radiation absorption coefficient, Radiation fraction, Let be the heat release rate of the i-th node at time t. Let be the temperature difference between the fire plume at the i-th ignition node and the environment;

[0036] Update all times Node status: Heat release rate (HRR) and temperature interpolation curves are determined through field simulation. Interpolation is used to determine all ignited nodes (that satisfy the following conditions). exist Moment and The fire ignition point was calculated using the thermal radiation calculation formula. For each undisturbed building node numbered j... Heat flux generated at any time Determine whether node j of the unburned building has been ignited. If node j of the unburned building has been ignited, then... The status of node j of the building that did not catch fire has been updated to ;

[0037] S1.3.3 Construct an adjacency matrix of ignition relationships between buildings by simulating the fire spread path in buildings. and spread matrix ;

[0038] The ignition relationship between buildings is determined by simulating the fire spread path in buildings, and an adjacency matrix is ​​constructed. Matrix element a( , )=1 indicates the initial fire-starting building node. Ignitionable building nodes Conversely, a( , =0; the adjacency matrix of the fire spread throughout the entire village is . ;

[0039] Based on the simulation of building fire spread paths, using the spread matrix Describe the global ignition relationship of the building complex under the initial fire;

[0040] The fire spread vector is denoted as p(i,j) in the formula. 1 indicates that there is only one node starting from the initial burning building. Point to building node The propagation path (direct or indirect); considering time weights in this invention, then p(i,j) = ( For nodes Direct or indirect ignition nodes (time);

[0041] S1.3.5 Adjacency Matrix Constructed Based on Building Node States and spread matrix By iterating through and calculating all fire scenarios in the village, the adjacency matrix of fire spread between buildings is obtained. and spread matrix ;

[0042] The fire scenario considers only one initial ignition point. In this invention, there are a total of M ignition scenarios for the entire village, where M is the number of building nodes (M=195). Based on the obtained fire spread adjacency matrix, a graph traversal algorithm is used to determine the ignition scenario. After calculating all fire scenarios, the fire spread adjacency matrix for all fire scenarios in the entire village is: Calculated using the following formula:

[0043]

[0044] In the formula, This indicates a union operation on the elements at corresponding positions in the matrix, where M is the number of building nodes and i represents the initial fire point building node number.

[0045] Obtaining the adjacency matrix of fire spread in the building complex In the case of an initial fire source building node Starting with a depth-first recursive traversal algorithm, we traverse the fire spread adjacency matrix (if a(i,j)=1, then include it in the set). ), resulting in a set of nodes The initial fire-starting building node is At that time, the set of all buildings ignited corresponds to The scenario is either the largest connected subgraph of the initial burning building node or the fire spread loss scenario when the burning building is i; after traversal, the initial building node is... Spreading Matrix The values ​​of the elements in the i-th row are as follows:

[0046]

[0047] In the above formula, Indicates the initial fire building node as The ignition state, Indicates the initial fire building node The time of the fire, Represents building nodes Initial fire building node Ignition time; i.e., spread matrix The value of the i-th row is The column number corresponding to the element indicates the node. After the fire starts, the node numbers that can be ignited are represented by the value in column j. The row number corresponding to the element indicates the node that can be ignited. Scenes of the fire spreading.

[0048] After calculating all fire scenarios in the entire village, the adjacency matrix of fire spread between buildings is obtained. and spread matrix This describes the global ignition relationship of a cluster of buildings under a specified initial fire building.

[0049] S1.4. Constructing a road network based on a directed graph: By defining a two-layer coupled network under a traditional village fire scenario and constructing a building fire spread network, a two-layer network G={G B G R};

[0050] The construction of road network G R The method is as follows: using road intersections as nodes and road segments between road intersections as edges, a road network G is constructed. R Based on road network G R Determine the road link status;

[0051] The road link status includes normal status and failure status;

[0052] In road network G R In the middle, road intersections are nodes, using Indicated by m, the number of road intersections; the road segments between intersections are called edges, denoted by m. Let n represent the set of edges, where n is the number of edges; Represents the adjacency matrix of the road network. Road nodes To the node The connection status is as follows:

[0053]

[0054] In the formula, This refers to the actual distance of the road segment;

[0055] The road link status is determined based on the road network diagram, and the road link status is defined; the road link status includes normal status and failure status.

[0056] In this invention, the normal functioning of the road network depends primarily on the intensity of thermal radiation from adjacent burning buildings. Considering the functional requirements of roads in traditional village fire scenarios, the road link status is defined as normal and completely failed states.

[0057] Normal state: If no building adjacent to the road link is on fire, the road link is in a normal state and the connection is maintained. ;

[0058] Failure status: The rule for complete failure of a road link is defined as the vertical distance from the burning building to the centerline of the adjacent road. The heat flux received at the location is In a fire, the human body undergoes heat convection with the outside environment. When the heat flux incident on the human skin exceeds 2.5 kW / m², the heat transfer can be significant. 2 This can cause skin burns and trigger a series of serious health problems. This invention defines the complete failure state of a road link using the NFPA 92B point source thermal radiation calculation formula as follows:

[0059]

[0060] In the formula, It is a building that caught fire. At any moment The heat flux radiated to adjacent roads; The radiation fraction is taken as 1 / 3; The fire started at the building node. exist The rate of heat release at any given moment; Perpendicular distance from the burning building to the centerline of the adjacent road ;

[0061] The final two-layer network G={G} constructed in this invention B G R} and inter-layer links and ;

[0062] S2. Construct a road vulnerability assessment system;

[0063] The method for constructing a road vulnerability assessment system is as follows: Define and calculate the road accessibility index A and the road vulnerability index. The overall vulnerability index of the village was obtained. ;

[0064] The accessibility index A dynamically represents road traffic efficiency.

[0065] The road vulnerability index To characterize building nodes The extent to which the fire affects road accessibility;

[0066] The road network accessibility index and road vulnerability index are calculated. The formula for calculating the road accessibility index A is as follows:

[0067]

[0068] In the formula, s is the starting point of the evacuation route. e is the evacuation endpoint M represents the number of building nodes; Defined as a road node To the finish line The cost of the shortest path travel distance is calculated using the Dijkstra algorithm for the shortest path. It is the accessibility index for a single node, and is inversely proportional to distance cost, meaning that the lower the distance cost, the better the accessibility; The road accessibility index reflects the average evacuation capacity of a village; in the event of a fire, a highly accessible road network can ensure that fire rescue can reach the scene quickly, while also enabling the rapid evacuation of villagers.

[0069] The road vulnerability index is defined as the situation where road network accessibility decreases due to disturbances caused by increased village evacuation lengths. Since vulnerability is a relative value, the relative rate of decline indicates the vulnerability of roads under the influence of multi-layer network cascading failures. 'i' represents the initial fire point building node number, and fire event i (building node...) Road vulnerability in fire scenarios Used to identify the most vulnerable fire scenarios; Road Vulnerability Index The specific formula is as follows:

[0070]

[0071] In the formula, For building nodes The accessibility value of the entire road network in a fire scenario. This represents the road network accessibility value under normal conditions.

[0072] Based on the system expected loss calculation method, the overall vulnerability index of the entire village is... The calculation is performed using the formula shown below:

[0073]

[0074] In the formula, Let M be the probability of fire scenario i occurring, and M be the number of building nodes; this invention assumes that all buildings in the village have the same probability of ignition, i.e., ... ;

[0075] S3. Construct a two-layer network cascaded failure simulation framework for traditional village building fire scenarios, perform a full fire scenario traversal of traditional villages, and obtain the overall vulnerability index set;

[0076] The method for constructing a two-layer network cascading failure simulation framework under the fire scenario of traditional village buildings is as follows: setting the initial multi-layer network state, the initial ignition node time, cascading failure simulation iteration, and cascading failure termination judgment.

[0077] This invention provides a multi-layered network cascading failure simulation framework for traditional village building fire scenarios, based on a thermal radiation flux threshold for dynamic simulation of cascading failures. This framework includes a thermal radiation flux threshold to dynamically simulate building ignition and road failure.

[0078] Building ignition conditions: Considering only fire spread caused by heat radiation and fire plume. Time Node Ignition conditions are kW / m 2 .

[0079] Road failure conditions: If the road section adjacent to the building where the fire occurred... The thermal radiation flux received at the center location kW / m 2 Then the road section Invalid.

[0080] S3.1, Set the initial multi-layer network state, set the time t=0; including: setting the building sprawl network. For the non-fired state and road network To establish a normal connection state, establish an inter-layer coupling matrix. ;

[0081] Establish a building sprawl network The building was not on fire at this time; there was no fire in the building, and the fire had no spread path. Status nodes of each building node , To simulate the total time step, For time step, the spreading matrix =0.

[0082] Establish a road network Road matrix , Road nodes Actual distance when connected, set of failed road links ;

[0083] Establish interlayer coupling matrix ; matrix elements , Represents building nodes To the nearest road node Euclidean distance; interlayer coupling matrix Matrix elements , Indicates the distance from the building to the adjacent road section The vertical distance;

[0084]

[0085] This state constitutes the baseline scenario for cascading failure simulation, and the initial reachability index is calculated based on Dijkstra's algorithm. Save as the initial reachability exponent vector ; This represents the road network accessibility value under normal conditions.

[0086] S3.2, Set the initial multi-layer network state; set the initial fire ignition node time t=1, and update the building node state. and spread matrix ;

[0087] Set the initial building on fire Update building node status and spread matrix ;

[0088] S3.3, Set the initial state of the multi-layer network; Set the cascade failure simulation iteration time t>1, perform road failure detection and building spread detection, and update the reachability index matrix at each time step;

[0089] Update the state of the burning building node at each time t. ;

[0090] The method for road failure detection is: reading the inter-layer coupling matrix. Determine the initial fire location of the building. If there are adjacent roads, calculate the thermal radiation flux of the associated roads. ,like kW / m 2 If a road marker is invalid, add one. To the set of failed links Update the road network adjacency matrix .

[0091] The method for detecting building fire spread is as follows: Building spread is detected according to the building fire spread network construction method described in step S1; if adjacent buildings... Thermal radiation flux received satisfy kW / m 2 If the building is marked as ignited, add a spread link. to ,renew Combustion node state is ,Record Ignition time p(i,j)= That is, update the spread matrix ;

[0092] Update the reachability index matrix at each time step , by formula and Calculate the accessibility index;

[0093] S3.4 Cascade Failure Termination Judgment: By updating the reachability index matrix at each time point, cascade failure termination judgment is performed. The entire fire scenario of traditional villages is traversed to obtain the final overall vulnerability index set.

[0094] The method for determining cascading failure termination is as follows: [Determine...] and Check if adjacent buildings can be ignited. If a new building node is ignited, proceed to step S3.3; otherwise, the cascading failure terminates, and the output is... And calculate the vulnerability index in this scenario. Save to vulnerability index set To conduct a full traversal of fire scenarios in traditional villages;

[0095] The method for traversing the entire fire scene of a traditional village is as follows: traverse all building nodes in the traditional village. Repeat steps S3.1–S3.4; calculate the overall vulnerability index of the entire village. The overall vulnerability index is obtained, and the dynamic cascading failure process ends.

[0096] S4. Based on the overall vulnerability index set, generate multi-scenario vulnerability heatmaps, output multi-scenario vulnerability heatmaps and locate high-risk areas to complete the method;

[0097] Output a set of vulnerability indices for the cascading failures of various fire scenarios in traditional villages. After obtaining the results, the vulnerability index corresponding to each fire scenario is visualized on the village plan, and a multi-scenario vulnerability heat map is drawn. Based on the heat map, high-risk areas can be located intuitively.

[0098] The beneficial effects of this invention are:

[0099] (1) This invention, by establishing a two-layer coupling model of “building fire spread network-road network” and an inter-layer dependency matrix, reveals for the first time the complete cascading failure dynamic evolution mechanism of “building fire → road failure → evacuation / rescue obstruction → further building spread obstruction” in a traditional village fire scenario. This overcomes the fundamental defect of existing technologies (single-layer network models or focusing only on a single system) in failing to capture cross-system interactions, and provides a theoretical tool for understanding complex disaster chains.

[0100] (2) This invention achieves precise physical quantification of the fire spread process and its impact on road traffic capacity by introducing dynamic failure judgment rules based on physical mechanisms (thermal radiation flux), building ignition rules, and road failure rules. Through iterative simulation framework, the time-varying dynamic characterization of cascading failure processes is realized by dynamically updating the building node status, road link status, and spread / failure relationships. This can accurately simulate key dynamic processes such as the gradual blockage of evacuation routes due to fires in adjacent buildings and the delay of rescue routes due to multiple fires.

[0101] (3) This invention quantifies the impact of disasters by defining and dynamically calculating accessibility and vulnerability indices; it directly and dynamically characterizes the traffic efficiency of the road network (especially evacuation and rescue efficiency), with values ​​changing in real time as fire spreads and roads fail; it accurately quantifies the relative loss of road system functions under specific fire scenarios; and it comprehensively assesses the overall vulnerability level of the entire village road network under fire threats. This indicator system provides an objective and comparable measurement standard for disaster impact.

[0102] (4) This invention generates multi-scenario vulnerability heatmaps through full-scene traversal simulation and vulnerability result spatial visualization, achieving precise spatial positioning of high-risk areas (such as areas where a fire in a specific building would cause severe loss of road function or cascading spread). This provides clear objectives and scientific basis for fire protection planning (such as adding fire hydrants and firebreaks), emergency resource deployment (such as material storage points and temporary shelter locations), and targeted risk modifications (such as fireproofing key vulnerable buildings and widening bottleneck evacuation routes). Attached Figure Description

[0103] Figure 1 This is a flowchart of the evaluation method of the present invention.

[0104] Figure 2 The following is a case study of a traditional village building complex in Yunnan Province according to the present invention; wherein, part (a) is a UAV survey map and road layout map of the wooden building complex, and part (b) is a typical wooden building structure diagram.

[0105] Figure 3 This invention provides a simulation process for the spread of fires in building complexes.

[0106] Figure 4 The diagrams are of the building fire spread network and road network of the present invention; wherein, part (a) is the building fire spread network diagram and part (b) is the road network diagram.

[0107] Figure 5 This is a diagram of the two-layer network topology of Tongle Village's buildings and roads according to the present invention.

[0108] Figure 6 This is a flowchart of the cascading failure simulation framework of the present invention.

[0109] Figure 7 This is the vulnerability index distribution of the present invention.

[0110] Figure 8 This is a multi-scenario vulnerability heatmap for the present invention. Detailed Implementation

[0111] The present invention will be further described in detail below with reference to specific embodiments. The preferred embodiments described below are only examples, and those skilled in the art can conceive of other obvious variations.

[0112] like Figure 1 As shown, a traditional method for assessing the vulnerability of roads in a cascading failure scenario during a village fire includes the following steps:

[0113] S1. Preparation of building cluster parameter data, definition of a two-layer coupled network under the traditional village fire scenario, and construction of building fire spread network and road network, ultimately constructing a two-layer network G={G B G R};

[0114] S1.1 Preparation of building group parameter data: including the collection of building group parameters, fire parameters and road parameters;

[0115] This invention is based on a combination of UAV oblique photography technology and field survey to collect building parameters, fire parameters and road parameters of the target building complex, with 195 individual buildings;

[0116] The number of individual buildings in the complex is 195;

[0117] The building parameters include: the location, area, scale, and geographical distribution characteristics of each individual building;

[0118] The fire parameters include: dimensions of each individual building, fire resistance rating, building layout, building opening ratio, quantity of combustibles, and other fire parameter information;

[0119] The road parameters include: the connection relationship between each individual building and the road network, the length and width of each road segment, and the geographical layout of village roads;

[0120] This example is selected from a certain ancient city district. The architectural style is a typical log cabin-style wooden structure. The buildings within the complex are closely connected and arranged in a continuous cluster. There are a total of 195 buildings in the target area. Figure 2 As shown, (a) is a UAV survey map of the wooden building complex and a road layout map, and (b) is a structural diagram of a typical wooden building.

[0121] S1.2 Construct a two-layer coupled network model for traditional village fire scenarios;

[0122] The two-layer coupled network model includes: interlayer coupling matrix of building fire spread network. Interlayer coupling matrix of road network ;

[0123] A two-layer coupled network model is constructed, including a building fire spread network layer and a road network layer. The model can realistically display the impact of building fires on the road network; denoted as G={G B G R}, where G B For building sprawl networks, G R For road network.

[0124] In building fire spread calculations, to measure the spread of fire in a building... Impact on road function, introducing the interlayer coupling matrix of building fire spread network. Interlayer coupling matrix of road network ;in element Represents building nodes To its adjacent road network nodes actual distance To calculate the shortest evacuation time; inter-layer coupling matrix ,element , This indicates the vertical distance from the building to the adjacent road, describing the relationship between the fire source and the heated road section;

[0125] S1.3 Constructing a building fire spread network G based on a directed graph B The fire spread network is constructed by using individual buildings as nodes and the fire spread paths between buildings as directed edges.

[0126] Constructing a building fire spread network includes: defining building node states, simulating the heat release rate curve of a single building fire, simulating the fire spread process between buildings; and determining the adjacency matrix of ignition relationships between buildings based on the simulation results. Generate a spread matrix describing the global ignition relationship and time. ;

[0127] Constructing a building fire spread network based on directed graph theory Define individual buildings within a building complex as nodes; that is, a single building is a directed graph. One of the nodes, the set of nodes is The village has M building nodes and a total of 195 individual buildings, i.e., M=195. The fire spread paths between buildings are defined as directed edges, resulting in a set of edges. The number of edges is N;

[0128] At the same time, determine the initial fire-starting building node. Attributes include building material (wood structure, brick-concrete structure), building area, building height, building ignition heat release rate, and fire risk parameters; determine edge attributes, including distance between building nodes, wind direction, and building heat radiation threshold.

[0129] S1.3.1 Determine the status of building nodes ;

[0130] Define node t at any time. The state is The expression is as follows:

[0131] ;

[0132] In the formula, For nodes The time of ignition, set This is the initial fire-starting building node. ,set up When it is not the initial ignition point and has not been ignited. Use inf to represent it; For nodes The rate of heat release in a fire, For nodes The temperatures are all functions of time t, where t is the time of day.

[0133] S1.3.2 Simulation of fire spread paths in buildings; including two parts: simulation of individual building areas and simulation of fire spread processes between buildings;

[0134] The method for simulating individual building areas is as follows: using the FDS tool from NIST (National Institute of Standards and Technology) in the United States, the fire heat release rate curve of the individual building is obtained. ;

[0135] The method for simulating the fire spread process between buildings is as follows: Figure 3 The fire spread simulation process for the building complex shown includes the following:

[0136] Setting initial conditions: Simulation environment conditions (wind speed, temperature), building complex simulation conditions (number of building nodes, simulation step size, and total number of simulation steps), node initial conditions (initial fire node). The remaining nodes This invention only considers the fire spread caused by thermal radiation and fire plumes, and sets... Time Node Ignition conditions are kW / m 2 The specific formula is as follows:

[0137] ;

[0138] ;

[0139] ;

[0140] ;

[0141] In the formula, For timber, a power consumption of 12.5 kW / m² is generally acceptable. 2 ; For nodes The heat flux received by the wooden exterior walls or openings; The wall emissivity is set to 0.8; Let be the Stefan-Boltzmann law constant, taken as 5.67 × 10⁻⁸ W / (m²). 2 ·K); This represents the number of fire ignition points; For the first The fire ignition point relative to the building node The radiation angle coefficient of the wooden exterior wall; For nodes The thermal radiation power is calculated using the ignition source calculation method of the smoke control system standard (NFPA 92B); The wall-to-wall convective heat transfer coefficient can be calculated using correlations from large-space natural convection heat transfer experiments. In this invention, it is taken as 7.6 W / (m²). 2 ·K); The wall temperature of node j in the building that did not catch fire; for The equivalent superposition temperature of the fire plumes at each ignition point. The thermal radiation absorption coefficient, Radiation fraction, Let be the heat release rate of the i-th node at time t. Let be the temperature difference between the fire plume at the i-th ignition node and the environment;

[0142] Update all times Node status: Heat release rate (HRR) and temperature interpolation curves are determined through field simulation. Interpolation is used to determine all ignited nodes (that satisfy the following conditions). exist Moment and The fire ignition point was calculated using the thermal radiation calculation formula. For each undisturbed building node numbered j... Heat flux generated at any time Determine whether node j of the unburned building has been ignited. If node j of the unburned building has been ignited, then... The status of node j of the building that did not catch fire has been updated to ;

[0143] S1.3.3 Construct an adjacency matrix of ignition relationships between buildings by simulating the fire spread path in buildings. and spread matrix ;

[0144] The ignition relationship between buildings is determined by simulating the fire spread path in buildings, and an adjacency matrix is ​​constructed. Matrix element a( , )=1 indicates the initial fire-starting building node. Ignitionable building nodes Conversely, a( , =0; the adjacency matrix of the fire spread throughout the entire village is . ;

[0145] Based on the simulation of building fire spread paths, using the spread matrix Describe the global ignition relationship of the building complex under the initial fire;

[0146] The fire spread vector is denoted as p(i,j) in the formula. 1 indicates that there is only one node starting from the initial burning building. Point to building node The propagation path (direct or indirect); considering time weights in this invention, then p(i,j) = ( For nodes Direct or indirect ignition nodes (time);

[0147] S1.3.5 Adjacency Matrix Constructed Based on Building Node States and spread matrix By iterating through and calculating all fire scenarios in the village, the adjacency matrix of fire spread between buildings is obtained. and spread matrix ;

[0148] The fire scenario considers only one initial ignition point. In this invention, there are a total of M ignition scenarios for the entire village, where M is the number of building nodes (M=195 in this embodiment). Based on the obtained fire spread adjacency matrix, a graph traversal algorithm is used to determine the fire scenario. After calculating all fire scenarios, the fire spread adjacency matrix for all fire scenarios in the entire village is: Calculated using the following formula:

[0149] ;

[0150] In the formula, This indicates a union operation on the elements at corresponding positions in the matrix, where M is the number of building nodes and i represents the initial fire point building node number.

[0151] Obtaining the adjacency matrix of fire spread in the building complex In the case of an initial fire source building node Starting with a depth-first recursive traversal algorithm, we traverse the fire spread adjacency matrix (if a(i,j)=1, then include it in the set). ), resulting in a set of nodes The initial fire-starting building node is At that time, the set of all buildings ignited corresponds to The scenario is either the largest connected subgraph of the initial burning building node or the fire spread loss scenario when the burning building is i; after traversal, the initial building node is... Spreading Matrix The values ​​of the elements in the i-th row are as follows:

[0152] ;

[0153] In the above formula, Indicates the initial fire building node as The ignition state, Indicates the initial fire building node The time of the fire, Represents building nodes Initial fire building node Ignition time; i.e., spread matrix The value of the i-th row is The column number corresponding to the element indicates the node. After the fire starts, the node numbers that can be ignited are represented by the value in column j. The row number corresponding to the element indicates the node that can be ignited. Scenes of the fire spreading.

[0154] After calculating all fire scenarios in the entire village, the adjacency matrix of fire spread between buildings is obtained. and spread matrix This describes the global ignition relationship of a cluster of buildings under a specified initial fire-starting building. The building spread network results for 195 fire scenarios in the village in this embodiment are as follows: Figure 4 Part a shows the building sprawl network diagram.

[0155] S1.4. Constructing a road network based on a directed graph: By defining a two-layer coupled network under a traditional village fire scenario and constructing a building fire spread network, a two-layer network G={G B G R};

[0156] The construction of road network G R The method is as follows: using road intersections as nodes and road segments between road intersections as edges, a road network G is constructed. R Based on road network G R Determine the road link status;

[0157] The road link status includes normal status and failure status;

[0158] In road network G R In the middle, road intersections are nodes, using Indicated by m, the number of road intersections; the road segments between intersections are called edges, denoted by m. Let n represent the set of edges, where n is the number of edges; Represents the adjacency matrix of the road network. Road nodes To the node The connection status is as follows:

[0159]

[0160] In the formula, This refers to the actual distance of the road segment;

[0161] The road network constructed in this embodiment of the village Figure 4 Part b is a road network diagram;

[0162] The road link status is determined based on the road network diagram, and the road link status is defined; the road link status includes normal status and failure status.

[0163] In this invention, the normal functioning of the road network depends primarily on the intensity of thermal radiation from adjacent burning buildings. Considering the functional requirements of roads in traditional village fire scenarios, the road link status is defined as normal and completely failed states.

[0164] Normal state: If no building adjacent to the road link is on fire, the road link is in a normal state and the connection is maintained. ;

[0165] Failure status: The rule for complete failure of a road link is defined as the vertical distance from the burning building to the centerline of the adjacent road. The heat flux received at the location is In a fire, the human body undergoes heat convection with the outside environment. When the heat flux incident on the human skin exceeds 2.5 kW / m², the heat transfer can be significant. 2 This can cause skin burns and trigger a series of serious health problems. This invention defines the complete failure state of a road link using the NFPA 92B point source thermal radiation calculation formula as follows:

[0166] ;

[0167] In the formula, It is a building that caught fire. At any moment The heat flux radiated to adjacent roads; The radiation fraction is taken as 1 / 3; The fire started at the building node. exist The rate of heat release at any given moment; The vertical distance from the building on fire to the centerline of the adjacent road ;

[0168] The final two-layer network G={G} constructed in this invention B G R} and inter-layer links and The result is as follows Figure 5 As shown;

[0169] S2. Construct a road vulnerability assessment system;

[0170] The method for constructing a road vulnerability assessment system is as follows: Define and calculate the road accessibility index A and the road vulnerability index. The overall vulnerability index of the village was obtained. ;

[0171] The accessibility index A dynamically represents road traffic efficiency.

[0172] The road vulnerability index To characterize building nodes The extent to which the fire affects road accessibility;

[0173] The road network accessibility index and road vulnerability index are calculated. The formula for calculating the road accessibility index A is as follows:

[0174] ;

[0175] In the formula, s is the starting point of the evacuation route. e is the evacuation endpoint M represents the number of building nodes; Defined as a road node To the finish line The cost of the shortest path travel distance is calculated using the Dijkstra algorithm for the shortest path. It is the accessibility index for a single node, and is inversely proportional to distance cost, meaning that the lower the distance cost, the better the accessibility; The road accessibility index reflects the average evacuation capacity of a village; in the event of a fire, a highly accessible road network can ensure that fire rescue can reach the scene quickly, while also enabling the rapid evacuation of villagers.

[0176] The road vulnerability index is defined as the situation where road network accessibility decreases due to disturbances caused by increased village evacuation lengths. Since vulnerability is a relative value, the relative rate of decline indicates the vulnerability of roads under the influence of multi-layer network cascading failures. 'i' represents the initial fire point building node number, and fire event i (building node...) Road vulnerability in fire scenarios Used to identify the most vulnerable fire scenarios; road vulnerability. The specific formula is as follows:

[0177] ;

[0178] In the formula, For building nodes The accessibility value of the entire road network in a fire scenario. This represents the road network accessibility value under normal conditions.

[0179] Based on the system expected loss calculation method, the overall vulnerability index of the entire village is... The calculation is performed using the formula shown below:

[0180] ;

[0181] In the formula, Let M be the probability of fire scenario i occurring, and M be the number of building nodes; this invention assumes that all buildings in the village have the same probability of ignition, i.e., ... ;

[0182] S3. Construct a two-layer network cascaded failure simulation framework for traditional village building fire scenarios, perform a full fire scenario traversal of traditional villages, and obtain the overall vulnerability index set;

[0183] The method for constructing a two-layer network cascading failure simulation framework under the fire scenario of traditional village buildings is as follows: setting the initial multi-layer network state, the initial ignition node time, cascading failure simulation iteration, and cascading failure termination judgment.

[0184] This invention provides a multi-layered network cascading failure simulation framework for traditional village building fire scenarios, based on a thermal radiation flux threshold for dynamic simulation of cascading failures. This framework includes a thermal radiation flux threshold to dynamically simulate building ignition and road failure.

[0185] Building ignition conditions: Considering only fire spread caused by heat radiation and fire plume. Time Node Ignition conditions are kW / m 2 .

[0186] Road failure conditions: If the road section adjacent to the building where the fire occurred... The thermal radiation flux received at the center location kW / m 2 Then the road section Invalid.

[0187] S3.1, Set the initial multi-layer network state, set the time t=0; including: setting the building sprawl network. For the non-fired state and road network To establish a normal connection state, establish an inter-layer coupling matrix. ;

[0188] Establish a building sprawl network The building was not on fire at this time; there was no fire in the building, and the fire had no spread path. Status nodes of each building node , To simulate the total time step, For time step, the spreading matrix =0;

[0189] Establish a road network Normal connection status; road matrix , Road nodes Actual distance when connected, set of failed road links ;

[0190] Establish interlayer coupling matrix ; matrix elements , Represents building nodes To the nearest road node Euclidean distance; interlayer coupling matrix Matrix elements , Indicates the distance from the building to the adjacent road section The vertical distance;

[0191]

[0192] This state constitutes the baseline scenario for cascading failure simulation, and the initial reachability index is calculated based on Dijkstra's algorithm. Save as the initial reachability exponent vector ; This represents the road network accessibility value under normal conditions.

[0193] S3.2, Set the initial multi-layer network state; set the initial fire ignition node time t=1, and update the building node state. and spread matrix ;

[0194] Set the initial building on fire Update building node status and spread matrix ;

[0195] S3.3, Set the initial state of the multi-layer network; Set the cascade failure simulation iteration time t>1, perform road failure detection and building spread detection, and update the reachability index matrix at each time step;

[0196] Update the state of the burning building node at each time t. ;

[0197] The method for road failure detection is: reading the inter-layer coupling matrix. Determine the initial fire location of the building. If there are adjacent roads, calculate the thermal radiation flux of the associated roads. ,like kW / m 2 Then add To the set of failed links Update the road network adjacency matrix ;

[0198] The method for detecting building fire spread is as follows: Building spread is detected according to the building fire spread network construction method described in step S1; if adjacent buildings... Thermal radiation flux received satisfy kW / m 2 Then add a propagation link. to ,renew Combustion state ,Record Ignition time p(i,j)= That is, update the spread matrix ;

[0199] Update the reachability index matrix at each time step , by formula and Calculate the accessibility index;

[0200] S3.4 Cascade Failure Termination Judgment: By updating the reachability index matrix at each time point, cascade failure termination judgment is performed. The entire fire scenario of traditional villages is traversed to obtain the final overall vulnerability index set.

[0201] The method for determining cascading failure termination is as follows: [Determine...] and Check if adjacent buildings can be ignited. If a new building node is ignited, proceed to step S3.3; otherwise, the cascading failure terminates, and the output is... And calculate the vulnerability index in this scenario. Save to vulnerability index set To conduct a full traversal of fire scenarios in traditional villages;

[0202] The method for traversing the entire fire scene of a traditional village is as follows: traverse all building nodes in the traditional village. Repeat steps S3.1–S3.4; calculate the overall vulnerability index of the entire village. The dynamic cascading failure process has ended;

[0203] This embodiment proposes a multi-layer network cascade failure simulation framework for traditional village building fire scenarios;

[0204] According to the cascading failure simulation framework (such as...) Figure 6 As shown): Initialize the multi-layer network state (time t=0); Set the initial fire node (t=1); Cascade failure simulation iteration (t>1); Cascade failure termination judgment; Full scene traversal; Calculate the road vulnerability under the influence of cascade failure in various fire scenarios in the village of this invention case, and obtain the vulnerability index set;

[0205] S4. Based on the overall vulnerability index set, generate multi-scenario vulnerability heatmaps, output multi-scenario vulnerability heatmaps and locate high-risk areas to complete the method;

[0206] After outputting the multi-scenario vulnerability index results, they are visualized on a village plan to form a multi-scenario vulnerability heat map. High-risk areas can then be intuitively located based on the heat map. For example... Figure 7 The figure shows the vulnerability index distribution of this invention. The thermal vulnerability of roads under the influence of cascading failures in various fire scenarios in the case village is shown in the figure. Figure 8 The figure shows a multi-scenario vulnerability heatmap of the present invention. Descriptive statistical results of road vulnerability under cascading failures in various fire scenarios are shown in Table 1.

[0207] Table 1: Road Vulnerability under Cascaded Failures in Various Fire Scenarios

[0208]

[0209] From the perspective of the village as a whole, the frequency statistics of the road network vulnerability index VI under various fire scenarios are as follows: Figure 8 The frequency distribution shows that road network vulnerability exhibits a significantly left-skewed distribution (mean VI = 2.86%), and the overall vulnerability value of villages is... The road vulnerability value was 0.0286, with 35.9% of fire scenarios showing a vulnerability of 0 and 48.72% of fire scenarios having a vulnerability value between 0 and 5%, indicating that the overall vulnerability of the village is low. This means that although different fire scenarios may have some impact on the overall evacuation function of the village, the village road network can still maintain good robustness and ensure overall evacuation needs.

Claims

1. A traditional village fire scene cascade failure road vulnerability assessment method, characterized in that, Includes the following steps: S1. Prepare building group parameter data, define a two-layer network under the traditional village fire scenario, and construct a building fire spread network and a road network, and finally construct a two-layer network. S2. Construct a road vulnerability assessment system; The way to build the road vulnerability evaluation system is: defining and calculating the road accessibility index A and the road vulnerability index , obtaining the overall vulnerability index of the village ; The accessibility index A dynamically represents road traffic efficiency. said road vulnerability index to characterize architectural nodes a degree of impact of a fire on road accessibility; S3. Construct a two-layer network cascaded failure simulation framework for traditional village building fire scenarios, perform a full fire scenario traversal of traditional villages, and obtain the overall vulnerability index set; The method for constructing a two-layer network cascading failure simulation framework under the fire scenario of traditional village buildings is as follows: setting the initial multi-layer network state, the initial ignition node time, dynamic simulation of cascading failure, and judgment of cascading failure termination. S4. Based on the overall vulnerability index set, generate multi-scenario vulnerability heat maps, output multi-scenario vulnerability heat maps and locate high-risk areas, and complete the assessment of the vulnerability of roads in the cascading failure scenario of traditional village fires.

2. The method according to claim 1, wherein, The preparation of building cluster parameter data, the definition of a two-layer network under a traditional village fire scenario, and the construction of a building fire spread network and a road network are all part of the final construction of the two-layer network. The specific steps for constructing the two-layer network are as follows: S1.1 Preparation of building group parameter data: including the collection of building group parameters, fire parameters and road parameters; The building parameters include: the location, area, scale, and geographical distribution characteristics of each individual building; The fire parameters include: dimensions of each individual building, fire resistance rating, building layout, building opening ratio, and quantity of combustibles; The road parameters include: the connection relationship between each individual building and the road network, the length and width of each road segment, and the geographical layout of village roads; S1.2 Construct a two-layer network model for traditional village fire scenarios; The two-layer network model includes: the inter-layer coupling matrix of the building fire spread network and the inter-layer coupling matrix of the road network; S1.3 Constructing a building fire spread network based on a directed graph; using individual buildings as nodes in the building fire spread network and the fire spread paths between buildings as directed edges to construct the building fire spread network. The construction of the building fire spread network includes: defining the building node status, simulating the heat release rate curve of a single building fire and simulating the fire spread process between buildings; determining the adjacency matrix of the ignition relationship between buildings based on the simulation results, and generating a spread matrix describing the global ignition relationship and time. S1.4 Construct a road network based on a directed graph, and combine it with the building fire spread network constructed in step S1.3 to finally construct a two-layer network; The method for constructing the road network is as follows: using road intersections as nodes and road segments between road intersections as edges, a road network is constructed, and the road link status is determined based on the road network; The road link status includes normal status and failure status.

3. The method according to claim 2, wherein, In the inter-layer coupling matrix of the building fire spread network and the inter-layer coupling matrix of the road network, the elements in the inter-layer coupling matrix of the building fire spread network are... Represents building nodes To its adjacent road network nodes actual distance ; The road network layer coupling matrix is represents the vertical distance of the building to the adjacent road.

4. The method according to claim 2, wherein, The definition of building node states, in simulating the heat release rate curve of a single building fire and the fire spread between buildings, is expressed as follows: ； In the formula, For building node status, The time it takes for the building node to be ignited. The fire heat release rate of the building node. Let t represent the temperature of the building node, and t represent time.

5. The method according to claim 2, wherein, The definition of building node status, simulation of the heat release rate curve of a single building fire, and simulation of the fire spread between buildings are described. The simulation of the fire spread between buildings includes two parts: simulation of the single building area and simulation of the fire spread between buildings. The monolithic building area simulation is done by using the FDS tool from NIST, USA to obtain the monolithic building fire heat release rate curve ; The method for simulating the fire spread process between buildings is as follows: simulating preset environmental conditions, building group simulation conditions, node initial conditions, and ignition conditions. The ignition condition is a preset heat flux threshold .

6. The method according to claim 2, wherein, The road link status includes a normal state and a failed state, and the expression for determining the failed state is: ； In the formula, The fire occurred at a specific point in time. The heat flux radiated to adjacent roads; The radiation fraction is taken as 1 / 3; The fire started at the building's structural node. The rate of heat release at any given moment; The vertical distance from the building on fire to the centerline of the adjacent road.

7. The method according to claim 1, wherein, The method for constructing the road vulnerability assessment system is as follows: Define and calculate the road accessibility index A and the road vulnerability index. The overall vulnerability index of the village was obtained. In this process, the overall vulnerability index of the village is obtained. The method is as follows: calculate the road network accessibility index and the road vulnerability index, where the formula for calculating the road accessibility index A is: ； In the formula, s is the starting point of the evacuation route. e is the evacuation endpoint M represents the preset number of building nodes. This represents the shortest path travel distance from the road node to the destination. Let A be the accessibility index for a single node, and let A be the road accessibility index. Road vulnerability index The specific calculation formula is as follows: ； In the formula, is the reachability value of the entire road network under the building node fire scenario, is the road network reachability value under normal circumstances; Based on the system expected loss calculation method, the overall vulnerability index of the entire village The overall vulnerability index is calculated The calculation formula is shown as follows: ； In the formula, is the probability of the fire scenario occurring, M is the preset number of building node quantities, and i represents the initial ignition point building node number.

8. The method according to claim 1, wherein, The construction of a two-layer network cascaded failure simulation framework under the traditional village building fire scenario involves traversing the entire fire scenario of traditional villages to obtain a set of overall vulnerability indices. The specific steps for obtaining the set of overall vulnerability indices are as follows: S3.1, Set the initial multi-layer network state, set the time t=0; including: setting the building sprawl network. For the non-fired state and road network To establish a normal connection state, establish an inter-layer coupling matrix. ; S3.2, set the initial state of the multi-layer network; set the initial fire node time t = 1, update the building node state and the spread matrix ; S3.3, Set the initial state of the multi-layer network; Set the cascade failure simulation iteration time t>1, perform road failure detection and building spread detection, and update the reachability index matrix at each time step; S3.4 Cascade Failure Termination Judgment: By updating the reachability index matrix at each time point, cascade failure termination judgment is performed. The entire fire scenario of traditional villages is traversed to obtain the final overall vulnerability index set.

9. The method according to claim 8, wherein, In the road failure detection and building sprawl detection, the road failure detection method is as follows: the road is subjected to thermal radiation flux kW / m 2 Mark roads as invalid and update the road network adjacency matrix; the building sprawl detection method is: the thermal radiation flux received by adjacent buildings. kW / m 2 The building has been marked as ignited; update the node status and spread matrix.

10. The method according to claim 1, wherein, The process involves generating multi-scenario vulnerability heatmaps based on an overall vulnerability index set, outputting these heatmaps, and locating high-risk areas. In the generated multi-scenario vulnerability heatmaps, the vulnerability indices corresponding to each fire scenario are visualized on a village plan, forming a multi-scenario vulnerability heatmap. High-risk areas can be intuitively located based on these multi-scenario vulnerability heatmaps.

Images