Subway station fire safety evacuation path planning method based on multi-objective optimization

Abstract

The closed space and high personnel density of a subway station enable personnel evacuation in a fire scene to face a severe challenge. The invention provides a subway station fire safety evacuation path planning method based on multi-objective optimization. Establishing a subway station evacuation path network structure model, and describing the network structure model from the equivalent length, the congestion degree, the risk degree and the availability of an evacuation channel; according to the fire source occurrence point and the influence on the surroundings, an evacuation path starting point and terminal point determination principle is formulated; starting from the angles of evacuation direction priority, the risk degree of an evacuation channel and the like, taking evacuation path selection cost interval control and availability as constraint conditions, and constructing a personnel safety evacuation path planning multi-target optimization model; solving the model through an improved NSGA-II algorithm, and carrying out loop removal and subset combination processing on paths in the path set; and screening out a plurality of optimal evacuation paths considering both efficiency and safety through comprehensive evaluation.

Description

Technical Field

[0001] The present invention relates to the field of subway station fire safety evacuation path planning, and in particular to a subway station fire safety evacuation path planning method based on multi-objective optimization. Background Art

[0002] As a vital component of modern urban public transportation, the efficiency and safety of subways directly impact the stability of urban transportation systems. However, the unique environment of subway stations makes evacuation during emergencies such as fires complex and challenging. Firstly, subway stations are often located underground, with enclosed spaces and complex structures, making evacuation susceptible to various factors such as fire, smoke spread, and congestion. Secondly, the high passenger volume during peak hours exacerbates safety risks during evacuation.

[0003] Currently, most evacuation route planning methods prioritize the shortest path and select routes solely based on geometric distance or time cost. However, this single-objective approach to route planning has significant limitations in practical applications. Specifically, it fails to fully utilize all available evacuation routes within subway stations, nor does it comprehensively consider the safety of routes under fire scenarios, the dynamic distribution of occupants, route priorities, and the changing nature of multiple influencing factors. Furthermore, most methods only plan a single optimal evacuation route, ignoring the importance of providing multiple alternative routes. Due to the dynamic nature of fire scenarios, evacuees may find it difficult to evacuate smoothly along the planned route due to congestion or route failure, resulting in confusion, delays, and even more serious safety hazards. Summary of the Invention

[0004] In light of this, this paper proposes a method for planning safe evacuation routes for subway station fires based on multi-objective optimization. This method, based on the actual layout of subway stations, comprehensively considers the directional priority, equivalent length, congestion level, risk level, and availability of evacuation routes. Combining a multi-objective optimization algorithm with a strategy for generating and selecting multiple optimal routes, it provides an efficient and safe solution for subway evacuations in fire scenarios.

[0005] According to one aspect of this application, a method for planning a fire safety evacuation path for a subway station based on multi-objective optimization is proposed:

[0006] S1. Based on the actual layout and spatial structure of subway stations, as well as the video surveillance cameras and other devices installed in subway stations, a subway station evacuation path network structure model is established. The network structure model is described in terms of the equivalent length, congestion, risk level, and availability of evacuation channels.

[0007] Specifically, step S1 is as follows:

[0008] S101. Based on the actual layout and spatial structure of the subway station, extract key nodes (such as exits, intersections) and channel information, and establish an evacuation path network node diagram.

[0009] S102. Utilize video surveillance cameras and other devices installed in subway stations to collect dynamic data such as passenger flow, channel status, and fire source impact range in real time.

[0010] S103. Calculate the equivalent length, congestion, risk level, and availability of the evacuation passage based on the collected data.

[0011] S104. Assign the calculated channel attributes to the edge weights of the network model to form a network structure model that can dynamically reflect the evacuation status, providing an input basis for subsequent path optimization.

[0012] S2. Based on the fire source location and its impact on the surrounding area, combined with public safety management regulations, formulate principles for determining the starting and ending points of evacuation routes.

[0013] Specifically, step S2 is as follows:

[0014] S201. When the fire source impact range does not endanger the safety of the train, the nodes near the fire source and the train nodes are used as evacuation starting points.

[0015] S202. If the fire source affects a range that endangers the safety of the train, only the nodes near the fire source will be used as evacuation starting points.

[0016] S203. Prioritize the ground exits of subway stations and emergency evacuation passage exits as evacuation destinations.

[0017] S204. If the fire smoke spread time is earlier than the time when personnel arrive at a certain node or path, the affected node and path are removed from the evacuation network, and the starting point and end point of the safe evacuation are re-determined.

[0018] S3. Starting from the perspectives of evacuation direction priority and evacuation channel risk level, and taking evacuation path selection cost interval control and availability as constraints, a multi-objective subway station fire safety evacuation path optimization model is constructed with section evacuation direction penalty value, section evacuation time, section congestion and section risk level as core indicators.

[0019] Specifically, step S3 is as follows:

[0020] S301. Most subway stations are located underground, and most exits lead to safe areas on the ground. Therefore, upward evacuation is preferred, which can quickly guide people away from underground enclosed spaces and avoid being stranded under limited ventilation conditions. Set the objective function to minimize the penalty value F for the evacuation direction of the path P .

[0021] S302: Setting the objective function to minimize the path congestion F G By minimizing the congestion on the route, the potential risk of casualties can be significantly reduced, ensuring that all people on site can evacuate to a safe area quickly and in an orderly manner.

[0022] S303, setting the objective function to minimize the path evacuation time F T By minimizing the evacuation time, it can ensure that in an emergency, the escape time of personnel is shortened to the greatest extent possible, thus improving overall efficiency.

[0023] S304. When planning safe evacuation routes in a subway station fire scenario, attention should also be paid to potential threats such as the distance between the planned route and the fire spread path, toxic gases, high temperatures, and smoke, which may cause harm to personnel. Therefore, a comprehensive road section risk level model must be established to ensure that the planned route can effectively avoid potential harm factors and provide a safe evacuation plan for personnel. Set the objective function to minimize the path risk level F Q .

[0024] S305. In order to ensure that the model obtains a safe and feasible evacuation path during the optimization process, it is necessary to set constraints related to the cost interval control and availability of the evacuation path selection.

[0025] S4. Taking minimizing the penalty value of the evacuation direction of the road section, the path evacuation time, the path congestion and the path risk level as the optimization goal, the model is solved by the improved NSGA-II algorithm to generate a Pareto solution set that meets the multi-objective optimization and provide multiple alternative evacuation path sets.

[0026] S5. De-looping and subset merging are performed on the paths in the set of multiple alternative evacuation paths to reduce unnecessary path output.

[0027] S6. The processed evacuation path set is input into a hierarchical analysis model based on the fusion entropy weight method. The optimal evacuation paths are screened by comprehensively evaluating each path's evacuation direction penalty, evacuation time, congestion, and risk level. This screening process avoids generating a single path and ensures the inclusion of multiple optimal paths, providing diverse evacuation solutions that balance efficiency and safety, thereby improving evacuation flexibility and adaptability.

[0028] Specifically, step S6 is as follows:

[0029] S601: Take the processed evacuation path set as input and perform dimensionless processing on each indicator.

[0030] S602: Calculate the weight of each goal, and solve the priority problem between goals by combining the subjective weight obtained by the AHP method and the objective weight obtained by the entropy weight method, so as to make the decision more reasonable.

[0031] S603. Evaluate each candidate path according to the following comprehensive scoring formula.

[0032] S604: Sort the paths according to the comprehensive scores, and give priority to paths with higher comprehensive scores.

[0033] Beneficial Effects of the Invention: This invention proposes a multi-objective optimization-based method for planning fire-safe evacuation routes in subway stations. This method comprehensively considers the equivalent length, congestion, risk level, and availability of evacuation corridors. By constructing a dynamic evacuation route network model, it achieves more scientific and efficient route planning. Compared with traditional evacuation methods that prioritize the shortest path, this invention considers the priority of evacuation directions and the risk level of evacuation corridors. Using evacuation route selection cost interval control and availability as constraints, it constructs a multi-objective subway station fire-safe evacuation route optimization model with core indicators including the penalty value of the evacuation direction, the evacuation time, the congestion level, and the risk level. This model can provide multiple alternative evacuation routes that balance safety and efficiency based on dynamic information such as fire source location, smoke spread, and passenger flow distribution, thus avoiding the congestion and safety hazards caused by single-route planning. Furthermore, through an improved NSGA-II algorithm and a multi-objective comprehensive evaluation model, this invention achieves optimal route selection, effectively improving evacuation efficiency and personnel safety, making it suitable for fire emergency evacuation in complex subway station scenarios.

[0034] Those skilled in the art will become more aware of the above and other objects, advantages and features of the present invention based on the following detailed description of the present invention in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a flow chart of a method for planning a fire safety evacuation route in a subway station based on multi-objective optimization according to an embodiment of the present application.

[0036] Figure 2 This is a diagram of the subway station evacuation path network structure model according to one embodiment of the present application.

[0037] Figure 3 A flowchart of building a multi-objective model for an embodiment of the present application.

[0038] Figure 4 This is a diagram of a hierarchical analysis model based on the fusion entropy weight method according to an embodiment of the present application. DETAILED DESCRIPTION

[0039] The implementation process mainly includes the following 6 steps.

[0040] S1. Based on the actual layout and spatial structure of subway stations, as well as the video surveillance cameras and other devices installed in subway stations, a subway station evacuation path network structure model is established. The network structure model is described in terms of the equivalent length, congestion, risk level, and availability of evacuation channels.

[0041] Specifically, step S1 is as follows:

[0042] S101. According to the actual layout and spatial structure of the subway station, a graph model of nodes and edges is established, where node v i represents the key position of the evacuation path, e ij Edge represents node v i To node v j The attributes of each edge include equivalent length, congestion, risk level and channel availability.

[0043] S102. Utilize video surveillance cameras and other devices installed in subway stations, combined with target detection algorithms, to collect dynamic data such as passenger flow, channel status, and fire source impact range in real time.

[0044] S103. Monitor the flow of people and the impact of fire sources on the paths in real time through devices such as video surveillance cameras, and dynamically calculate attributes such as the congestion and risk level of each path after a fire occurs.

[0045] S104: If the congestion or risk level of a certain path exceeds a preset evacuation path selection cost range, the path is determined to be an unusable path and is eliminated in the subsequent multi-objective optimization process.

[0046] S2. Based on the fire source location and its impact on the surrounding area, combined with public safety management regulations, formulate principles for determining the starting and ending points of evacuation routes.

[0047] Specifically, step S2 is as follows:

[0048] S201. If the location of the fire source and its impact do not endanger the use of trains as evacuation tools, nodes and train nodes within a certain range of the fire source location are used as starting points for safe evacuation.

[0049] S202: If the location of the fire source and its impact endanger the safe operation of the train and the train cannot be used as an evacuation tool, the nodes within a certain range of the fire source location will be used as the starting point for safe evacuation.

[0050] S203. In consideration of the structural characteristics of the subway station, the terminus is preferentially set as the ground exit and emergency evacuation passage exit of the subway station.

[0051] S204. When a fire occurs, smoke spread is considered. If the time it takes for the fire smoke to spread to a certain node or path is less than the time it takes for station personnel to move from the planned evacuation starting point to that node, then the nodes and paths affected by the smoke spread need to be deleted from the safe evacuation path network structure, and new safe evacuation starting points and end points are determined based on the adjusted network structure.

[0052] S3. Starting from the perspectives of evacuation direction priority and evacuation channel risk level, and taking evacuation path selection cost interval control and availability as constraints, a multi-objective subway station fire safety evacuation path optimization model is constructed with section evacuation direction penalty value, section evacuation time, section congestion and section risk level as core indicators.

[0053] Specifically, step S3 is as follows:

[0054] S301. Most subway stations are located underground, and most exits lead to safe areas on the ground. Therefore, upward evacuation is preferred, which can quickly guide people away from underground enclosed spaces and avoid being stranded under limited ventilation conditions. Set the objective function to minimize the penalty value F for the evacuation direction of the path P , and its calculation formula is as follows.

[0055] F P =∑ ij P ij

[0056]

[0057] Among them, P ij Represents node v i To node v j Path between ij The penalty value of the direction up; o ij is the direction weight, positive value is the preferred direction, negative value is the non-preferred direction; c ij Select the variable for the path, select 1 and not select 0. ij c ij When ≤0, the path selection does not conform to the priority direction, and the penalty value P ij Increase by 1; when o ij c ij When >0, the path selection complies with the priority direction and the penalty value remains unchanged.

[0058] S302: Setting the objective function to minimize the path congestion F G The calculation formula is as follows. By minimizing the path congestion, the potential risk of casualties can be greatly reduced, ensuring that all people on site can evacuate to a safe area quickly and orderly. In order to more reasonably calculate the path congestion, the panic coefficient f is introduced. ijThe panic index is an important parameter that measures the impact of a crowd's psychological state on evacuation behavior during an emergency, directly influencing changes in congestion. In fire evacuation scenarios, as panic intensifies, individuals' behavior within a confined space becomes more disordered, often leading to significantly increased congestion in densely populated areas. Under high panic indexes, people may be more inclined to choose routes that appear intuitively safe or fast, leading to concentrated traffic on certain sections of road and further exacerbating localized congestion.

[0059]

[0060]

[0061] Among them, C ij Indicates that at node v i To node v j Path between ij The channel capacity on n ij The number of people to be evacuated; ij represents the panic coefficient; k, A and B are all expressed as constants; r ij represents the sum of the radii between people; d ij Indicates the center point distance between people; h ij and t ij They respectively represent the tangent and hairline directions of the force between people.

[0062] S303, setting the objective function to minimize the path evacuation time F T , and its calculation formula is as follows. By minimizing the path evacuation time, it can ensure that the escape time of personnel is minimized to the greatest extent in an emergency, thereby improving the overall efficiency.

[0063]

[0064] Among them, v ij Indicates that at node v i To node v j Path between ij The speed of people on the stairs is generally set by default at 1.20m / s for flat evacuation passages, 0.86m / s for upward movement in stairs, and 0.65m / s for downward movement in stairs. ij For path e ij The equivalent length is calculated as follows:

[0065]

[0066] K ij =k ij -k1

[0067] L ij =L const(i,j)

[0068] Z ij =sinθ×L ij

[0069] Among them, K ij is the effective width of the path, k ij Refers to the path e ij The inherent width of the path e ij The width affected by the upper obstacle; L ij is the actual length of the path, L const Refers to node v i To node v j The actual distance between each pair of nodes is usually a constant; Z ij is the difficulty of the path, θ is the inclination angle of the path; α1, α2, and α3 are the influence coefficients of the effective width of the path, the actual length of the path, and the difficulty of the path on the safe evacuation of personnel, respectively, and α1+α2+α3=1.

[0070] S304. When planning safe evacuation routes in a subway station fire scenario, attention should also be paid to potential threats such as the distance between the planned route and the fire spread path, toxic gases, high temperatures, and smoke, which may cause harm to personnel. Therefore, a comprehensive road section risk level model must be established to ensure that the planned route can effectively avoid potential harm factors and provide a safe evacuation plan for personnel. Set the objective function to minimize the path risk level F Q , and its calculation formula is as follows.

[0071] F Q =Q ij F G

[0072] Among them, n ij The number of people to be evacuated; Q ij is the risk coefficient, and the specific calculation formula is as follows:

[0073]

[0074]

[0075]

[0076] Among them, q ij For the entire planned path E=(e 12 -e 23 -…-e ij ) and flame spread path E f The average distance, n is the number of road sections, d k For the planned road section eij The vertical distance to the nearest flame spreading section. α and β quantify the potential damage to the human body caused by toxic gases and temperature. N represents the product of the toxic gas concentration and the maximum exposure time an individual can safely tolerate at that concentration, providing an important basis for assessing the threat of toxic gases in fires. ρ is the simulated toxic gas concentration in the environment. M is the product of the ambient temperature and the maximum exposure time an individual can safely tolerate at that temperature. T is the simulated ambient temperature, which facilitates temperature analysis and ensures safe evacuation of personnel.

[0077] S305. To ensure that the model obtains a safe and feasible evacuation path during the optimization process, it is necessary to set constraints related to the cost range control and availability of the evacuation path selection:

[0078]

[0079] Where A is a set of connectable node pairs, that is, a set of edges that allow connections, and i,j∈A means that nodes i and j can be connected. fire The time it takes for a fire to start and become a direct threat to people. This means that during each optimization or iteration, the total evacuation time should be less than 0.8 times the time it takes for a fire to start and become a direct threat to people. The congestion on the path should not exceed 0.75, and the risk level of the path should not exceed 0.5.

[0080] S4. Taking minimizing the penalty value of the evacuation direction of the road section, the path evacuation time, the path congestion and the path risk level as the optimization goal, the model is solved by the improved NSGA-II algorithm to generate a Pareto solution set that meets the multi-objective optimization and provide multiple alternative evacuation path sets.

[0081] Specifically, step S4 is as follows:

[0082] S401. Randomly generate an initial population, where each individual corresponds to a feasible evacuation path plan.

[0083] S402: For each individual, calculate the fitness values of the four optimization objectives respectively.

[0084] S403. Perform non-dominated sorting on each individual in the population, and divide the population into multiple levels according to the relationship between the advantages and disadvantages of the optimization objectives.

[0085] S404. Calculate the crowding distance of each individual in the population. The larger the crowding distance, the more likely the individual is to be retained to maintain population diversity.

[0086] S405. Use the tournament selection method to select superior individuals from the current population as parents. Perform a path crossover on the selected parent individuals to generate a child path solution, ensuring that the crossover path still meets the constraints. Randomly change some edges in the child path to generate a new path, ensuring that the mutation operation does not cause the path to become invalid.

[0087] S406: Merge the parent and offspring populations to form a new large population. Perform non-dominated sorting and crowding calculation on the merged population. Based on the non-dominated rank and crowding distance, select the top N individuals from the merged population as the next generation population.

[0088] S407. When the iteration reaches the maximum number of generations or the convergence condition, the optimization is terminated and the Pareto solution set of the current population is output, that is, the non-dominated solutions of each objective in the multi-objective optimization.

[0089] S408. Extract multiple representative evacuation paths from the Pareto solution set and provide them for subsequent comprehensive evaluation.

[0090] S5. De-looping and subset merging are performed on the paths in the set of multiple alternative evacuation paths to reduce unnecessary path output.

[0091] Specifically, step S5 is as follows:

[0092] S501. Deloop the paths, converting the path set into a directed graph. Use an adjacency list to store the nodes and their edge relationships. Detect loops using topological sorting. Initialize the in-degree table and remove nodes with an in-degree of 0. If sorting cannot be completed, a loop exists in the graph. Locate the edges or nodes that cause the loop and repair the loop. Delete any extra edges or nodes that cause loops, ensuring that all paths between nodes no longer form loops. Use the DFS method to regenerate all loop-free paths from the repaired graph, recording the complete node sequence for each path as output.

[0093] S502: Subset and merge the paths. If most of the nodes of two paths are the same, they can be considered redundant paths. Compare the similarity of the two paths. If the repetition rate is higher than 0.9, merge them into one. The similarity calculation formula is as follows:

[0094]

[0095] S6. The processed evacuation path set is input into a hierarchical analysis model based on the fusion entropy weight method. The optimal evacuation paths are screened by comprehensively evaluating each path's evacuation direction penalty, evacuation time, congestion, and risk level. This screening process avoids generating a single path and ensures the inclusion of multiple optimal paths, providing diverse evacuation solutions that balance efficiency and safety, thereby improving evacuation flexibility and adaptability.

[0096] Specifically, step S6 is as follows:

[0097] S601: Taking the processed evacuation path set as input, adopting the range normalization method to perform dimensionless processing on each indicator.

[0098] S602. Calculate the weight of each goal. By combining the subjective weight obtained by the AHP method and the objective weight obtained by the entropy weight method, the priority problem between goals is solved to make the decision more reasonable. The combined weight formula is as follows:

[0099]

[0100] Among them, the subjective weights are α1, α2, α3, α4, the objective weights are β1, β2, β3, β4, and the combined weights are w1, w2, w3, w4.

[0101] S603. Evaluate each candidate path according to the following comprehensive scoring formula:

[0102]

[0103] Among them, W is the weight matrix and K is the decision matrix.

[0104] S604: Sort the paths according to the comprehensive scores, give priority to paths with higher comprehensive scores, and guide station personnel to evacuate safely through sound and light information.

[0105] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0106] The above is only the most effective implementation scheme of the present invention. It should be pointed out that for ordinary technicians in this technical field, appropriate improvements and modifications can be made without departing from the working principle of the present invention. These improvements and modifications should also be considered as the scope of protection of the present invention.

Claims

1. A method for planning fire safety evacuation paths in subway stations based on multi-objective optimization, characterized in that The method comprises the following steps: (1) Based on the actual layout and spatial structure of subway stations, as well as the video surveillance cameras and other devices arranged in subway stations, a subway station evacuation path network structure model is established. The network structure model is described from the perspective of the equivalent length, congestion, risk level, and availability of the evacuation channel. (2) Based on the location of the fire source and its impact on the surrounding area, and in combination with public safety management regulations, formulate principles for determining the starting and ending points of evacuation routes. (3) Starting from the perspectives of evacuation direction priority and evacuation channel risk level, and taking evacuation path selection cost interval control and availability as constraints, a multi-objective subway station fire safety evacuation path optimization model is constructed with section evacuation direction penalty value, section evacuation time, section congestion and section risk level as core indicators. (4) Taking minimizing the penalty value of the evacuation direction of the road section, the path evacuation time, the path congestion and the path risk level as the optimization goal, the model is solved by the improved NSGA-II algorithm to generate a Pareto solution set that meets the multi-objective optimization and provide multiple alternative evacuation path sets. (5) De-looping and subset merging are performed on the paths in the set of multiple alternative evacuation paths to reduce unnecessary path output. (6) The processed evacuation path set is input into a hierarchical analysis model based on the fusion entropy weight method. The optimal multiple evacuation paths are screened by comprehensively evaluating the evacuation direction penalty value, evacuation time, congestion, and risk level of each path. The screening process avoids generating only a single path and ensures that multiple optimal paths are included, providing personnel with a variety of evacuation plans that balance efficiency and safety, thereby improving the flexibility and adaptability of evacuation.

2. The network structure model is described based on the equivalent length, congestion, risk level, and availability of evacuation channels according to claim 1, characterized in that The method comprises the following steps: (1) Based on the actual layout and spatial structure of subway stations, a node-edge graph model is established, where nodes represent key locations on evacuation routes and edges represent evacuation channels. The attributes of each edge include equivalent length, congestion, risk level, and channel availability. (2) Using video surveillance cameras and other devices to monitor the flow of people and the impact of fire sources on the paths in real time, the congestion and risk level of each path after a fire occurs are dynamically calculated. If the congestion or risk level of a path exceeds the preset evacuation path selection cost range, the path is determined to be unusable and will be eliminated in the subsequent multi-objective optimization process.

3. The principle for determining the starting point and end point of an evacuation route according to claim 1 is characterized in that The method comprises the following steps: (1) If the location of the fire source and its impact do not endanger the train as an evacuation tool, the nodes and train nodes within a certain range of the fire source location will be used as the starting point for safe evacuation. (2) If the location of the fire source and its impact endanger the safe operation of the train and the train cannot be used as an evacuation tool, the nodes within a certain range of the fire source location will be used as the starting point for safe evacuation. (3) In consideration of the structural characteristics of the subway station, the terminus shall be set as the ground exit and emergency evacuation passage exit of the subway station. (4) When a fire occurs, smoke spread should be considered. If the time it takes for the fire smoke to spread to a certain node or path is less than the time it takes for station personnel to move from the planned evacuation starting point to that node, then the nodes and paths affected by the smoke spread need to be deleted from the safe evacuation path network structure, and new safe evacuation starting points and end points should be determined according to the adjusted network structure.

4. The multi-objective optimization model according to claim 1 is characterized in that the multi-objective optimization model is constructed with the section evacuation direction penalty value, section evacuation time, section congestion and section risk degree as core indicators. The method comprises the following steps: (1) Most subway stations are located underground, and most exits lead to safe areas on the ground. Therefore, upward evacuation is preferred, which can quickly guide people away from underground enclosed spaces and avoid being stranded under limited ventilation conditions. The objective function is set to minimize the penalty value F for the evacuation direction of the path. P , and its calculation formula is as follows. Among them, P ij Represents node v i To node v j Path between ij The penalty value of the direction up; o ij is the direction weight, positive value is the preferred direction, negative value is the non-preferred direction; c ij Select the variable for the path, select 1 and not select 0. ij c ij When ≤0, the path selection does not conform to the priority direction, and the penalty value P ij Increase by 1; when o ij c ij When >0, the path selection complies with the priority direction and the penalty value remains unchanged. (2) Setting the objective function to minimize the path congestion F G The calculation formula is as follows. By minimizing the path congestion, the potential risk of casualties can be greatly reduced, ensuring that all people on site can evacuate to a safe area quickly and orderly. In order to more reasonably calculate the path congestion, the panic coefficient f is introduced. ij The panic index is an important parameter that measures the impact of a crowd's psychological state on evacuation behavior during an emergency, directly influencing changes in congestion. In fire evacuation scenarios, as panic intensifies, individuals' behavior within a confined space becomes more disordered, often leading to significantly increased congestion in densely populated areas. Under high panic indexes, people may be more inclined to choose routes that appear intuitively safe or fast, leading to concentrated traffic on certain sections of road and further exacerbating localized congestion. Among them, C ij Indicates that at node v i To node v j Path between ij The channel capacity on n ij The number of people to be evacuated; ij represents the panic coefficient; k, A and B are all expressed as constants; r ij represents the sum of the radii between people; d ij Indicates the center point distance between people; h ij and t ij They respectively represent the tangent and hairline directions of the force between people. (3) Set the objective function to minimize the path evacuation time F T , and its calculation formula is as follows. By minimizing the path evacuation time, it can ensure that the escape time of personnel is minimized to the greatest extent in an emergency and improve the overall efficiency. Among them, v ij Indicates that at node v i To node v j Path between ij The speed of people on the stairs is generally set by default at 1.20m / s for flat evacuation passages, 0.86m / s for upward movement in stairs, and 0.65m / s for downward movement in stairs. ij For path e ij The equivalent length is calculated as follows: K ij =k ij -k1 L ij =L const (i,j) WITH ij =sinθ×L ij Among them, K ij is the effective width of the path, k ij Refers to the path e ij The inherent width of the path e ij The width affected by the upper obstacle; L ij is the actual length of the path, L const Refers to node v i To node v j The actual distance between nodes is usually a constant between each pair of nodes; Z ij is the difficulty of the path, θ is the inclination angle of the path; α1, α2, and α3 are the influence coefficients of the effective width of the path, the actual length of the path, and the difficulty of the path on the safe evacuation of personnel, respectively, and α1+α2+α3=1. (4) When planning safe evacuation routes in a subway station fire scenario, attention should also be paid to potential threats such as the distance between the planned route and the fire spread path, toxic gases, high temperatures, and smoke, which may cause harm to personnel. Therefore, it is necessary to establish a comprehensive road section risk degree model to ensure that the planned route can effectively avoid potential harm factors and provide a safe evacuation plan for personnel. Set the objective function to minimize the path risk degree F Q , and its calculation formula is as follows. F Q =Q ij F G Among them, n ij Number of people to be evacuated; Q ij is the risk coefficient, and the specific calculation formula is as follows: Among them, q ij For the entire planning path E=(e 12 -e 23 -…-e ij ) and flame spread path E f The average distance, n is the number of road sections, d k For planning section e ij The vertical distance to the nearest flame spreading section. α and β quantify the potential damage to the human body caused by toxic gases and temperature. N represents the product of the toxic gas concentration and the maximum exposure time an individual can safely tolerate at that concentration, providing an important basis for assessing the threat of toxic gases in fires. ρ is the simulated toxic gas concentration in the environment. M is the product of the ambient temperature and the maximum exposure time an individual can safely tolerate at that temperature. T is the simulated ambient temperature, which facilitates temperature analysis and ensures safe evacuation of personnel. (5) In order to ensure that the model obtains a safe and feasible evacuation path during the optimization process, it is necessary to set the evacuation path selection cost interval control and availability-related constraints: Among them, A is a set of connectable node pairs, that is, a set of edges that allow connections, and i,j∈A means that nodes i and j can be connected. fire The time it takes for a fire to start and become a direct threat to people. This means that during each optimization or iteration, the total evacuation time should be less than 0.8 times the time it takes for a fire to start and become a direct threat to people. The congestion on the path should not exceed 0.75, and the risk level of the path should not exceed 0.

5.

5. The method of claim 1 for performing de-looping and subset merging on a plurality of candidate evacuation paths, characterized in that The method comprises the following steps: (1) De-loop the paths, converting the path set into a directed graph and using an adjacency list to store the nodes and their edge relationships. Detect loops by topological sorting. Initialize the in-degree table and remove nodes with an in-degree of 0. If sorting cannot be completed, there is a loop in the graph. Locate the edges or nodes that cause the loop and repair the loop. Delete the redundant edges or nodes that cause loops to ensure that all paths between nodes no longer form loops. Use the DFS method to regenerate all loop-free paths from the repaired graph and record the complete node sequence of each path as output. (2) Subset the paths and merge them. If most of the nodes of two paths are the same, they can be considered redundant paths. Compare the similarity of the two paths. If the repetition rate is higher than 0.9, merge them into one. The similarity calculation formula is as follows:

6. The hierarchical analysis model based on the fusion entropy weight method according to claim 1 selects the optimal multiple evacuation paths, characterized in that The method comprises the following steps: (1) The processed evacuation path set is used as input, and the range normalization method is used to make each indicator dimensionless. (2) Calculate the weight of each goal. By combining the subjective weight obtained by the AHP method and the objective weight obtained by the entropy weight method, the priority problem between goals can be solved to make the decision more reasonable. The combined weight formula is as follows: Among them, the subjective weights are (α1, α2, α3, α4), the objective weights are (β1, β2, β3, β4), and the combined weights are (w1, w2, w3, w4). (3) Each candidate path is evaluated according to the following comprehensive scoring formula: Among them, W is the weight matrix and K is the decision matrix. (4) Sort the paths according to the comprehensive scores, give priority to the paths with higher comprehensive scores, and guide the station personnel to evacuate safely based on this through sound and light information.

Images