Urban rail transit emergency passenger flow distribution method based on dynamic subsidy optimization

Abstract

The invention discloses an urban rail transit emergency passenger flow distribution method based on dynamic subsidy optimization, and the method comprises the steps: firstly obtaining the sudden interruption event data of a subway station; secondly, calculating the waiting time of each traffic mode, and comprehensively considering the influence of the transport capacity constraint, the passenger tolerance threshold, the ticket price down-regulation policy and the flow dynamic change on the waiting time; thirdly, a multilayer logarithmic linear decision model is adopted to calculate selection probabilities of passengers for different traffic modes, and a ticket price down-regulation scheme is optimized according to a supply-demand relationship; then, through iterative calculation of generalized Nash equilibrium, passenger distribution is dynamically adjusted, so that the supply-demand relationship of each traffic mode tends to be optimally balanced; and finally, a traffic scheduling strategy is adjusted in real time based on an optimization result, passengers are guided to reasonably select a traffic mode, and the overall evacuation efficiency is improved. According to the method, the optimal passenger shunting and ticket price strategy can be formulated according to the real-time passenger flow distribution characteristics, the passenger endurance and the traffic supply capability under the emergency condition of the subway station.

Description

Technical Field

[0001] The present invention relates to the field of traffic management and intelligent scheduling, and in particular to an urban rail transit emergency passenger evacuation method based on dynamic subsidy optimization, which is suitable for rapid passenger evacuation in the event of traffic interruption caused by subway emergencies. Background Art

[0002] As a vital component of urban rail transit systems, subways carry a large number of commuters and are characterized by high density and a fast pace. However, due to the enclosed spaces and limited ventilation of subway stations, emergencies such as equipment failure, natural disasters, terrorist attacks, or extreme weather can lead to large-scale passenger delays and even serious safety accidents. Subway emergencies can paralyze the transportation system and impact the stability of urban operations. Furthermore, due to limited evacuation routes, inappropriate evacuation strategies can easily lead to passenger congestion and disorder, increasing the risk of casualties and property damage. Therefore, research on efficient subway emergency evacuation methods is of great significance for improving urban transportation safety and enhancing public security preparedness.

[0003] Currently, research on subway emergency evacuation primarily focuses on evacuation theory, optimization methods, and modeling and simulation. Existing studies are generally based on macro-evacuation simulations or micro-pedestrian dynamics modeling. By establishing evacuation models, they analyze passenger behavior patterns and evacuation efficiency under different circumstances, providing subway operations management departments with emergency plan optimization recommendations. However, these methods are primarily targeted at subway operators and lack precise guidance on individual passenger behavior, making real-time regulation difficult. In reality, passengers' decision-making behavior in the face of emergencies is influenced by a variety of factors, including waiting time, travel patience, transportation mode selection costs, and fare reduction policies. Therefore, how to quickly guide passengers to make appropriate transportation choices after an emergency occurs, avoid prolonged confinement at subway stations, and improve overall evacuation efficiency remains a core challenge in current emergency management.

[0004] Traditional subway emergency evacuation strategies typically employ fixed models, such as increasing bus shuttle service frequency, temporarily opening shared bike services, or coordinating ride-hailing services to provide a variety of alternative transportation modes. However, these static responses fail to fully account for passengers' individual decision-making characteristics and may lead to overloading of some modes of transportation and underutilization of others, further reducing evacuation efficiency. For example, when bus fares are reduced to encourage passengers to choose public transportation, if waiting times are too long, some passengers may still prefer ride-hailing or shared bikes, resulting in lower-than-expected evacuation efficiency. Furthermore, existing research on optimizing transportation resource scheduling often uses centralized control methods, lacking optimization mechanisms based on individual passenger decision-making behavior, making it difficult to achieve optimal supply-demand matching.

[0005] In recent years, the application of game theory in traffic management has gradually gained attention. By constructing game models to simulate passengers' decision-making processes, it is possible to more accurately predict passenger distribution under different fare reduction and fare policies. However, existing research mainly focuses on optimizing a single mode of transportation (such as ride-hailing or public transportation), lacking joint game optimization models for multiple modes of transportation, making it difficult to achieve dynamic adjustment of supply and demand across modes. Furthermore, most existing game models assume that passengers are completely rational. In emergency evacuation environments, passengers' decision-making behavior is uncertain, and factors such as their patience threshold and perceived waiting time have a significant impact on evacuation outcomes. Therefore, it is necessary to establish a more realistic high-dimensional passenger patience game model to characterize the complex passenger decision-making mechanism and thus optimize the overall evacuation strategy.

[0006] The present invention proposes a method for emergency passenger evacuation of urban rail transit based on dynamic subsidy optimization, combining the passenger patience threshold, the dynamic adjustment mechanism of waiting time and the fare reduction optimization strategy, and realizing the dynamic equilibrium of supply and demand of various modes of transportation through generalized Nash equilibrium calculation. Compared with traditional evacuation methods, this method can more accurately predict passenger behavior and guide passengers to the mode of transportation with the highest evacuation efficiency through optimal fare reduction and fare adjustment strategies, thereby improving overall evacuation efficiency, reducing the impact of emergencies on subway operations, and reducing fare reduction costs, thereby improving resource utilization efficiency. The present invention is suitable for emergency evacuation scenarios after subway emergencies, and provides an intelligent and efficient emergency evacuation decision support solution for urban rail transit systems. Summary of the Invention

[0007] This invention aims to design an urban rail transit emergency passenger flow evacuation method based on dynamic subsidy optimization to address the shortcomings of existing subway emergency evacuation strategies in dynamic passenger decision modeling, optimized transportation mode allocation, and fare reduction regulation. This invention comprehensively considers the dynamic changes in passenger flow after a subway emergency, the carrying capacity of different transportation modes, and the patience threshold of passengers. By constructing a high-dimensional passenger patience game model, it optimizes passengers' transportation mode selection strategies and employs a generalized Nash equilibrium solution to ensure a balance between supply and demand across transportation modes. This invention can provide scientific and intelligent emergency evacuation decision-making support for rail transit operation and management departments, helping operators quickly and efficiently respond to emergencies and traffic disruption risks, reduce passenger detention time, and mitigate safety risks caused by improper evacuation. Furthermore, this method can optimize the allocation of fare reduction resources, making fiscal expenditures more precise and improving the feasibility and cost-effectiveness of emergency response. It has important theoretical and practical significance for the safe operation of urban rail transit systems, the improvement of public emergency management systems, and the improvement of passenger travel safety.

[0008] To achieve the above objectives, the present invention provides an urban rail transit emergency passenger evacuation method based on dynamic subsidy optimization, the specific scheme is as follows:

[0009] When an emergency such as fire, earthquake, or large passenger flow occurs in an urban rail transit station, a high-dimensional passenger patience game model is constructed to optimize passengers' transportation choices after the subway is shut down, thereby improving passenger diversion efficiency.

[0010] The passenger patience game model quantitatively describes the decisions each passenger makes based on their individual circumstances (e.g., waiting time, transportation cost, patience threshold, etc.). Specifically, a passenger's utility function (also known as their cost function) forms the basis for their transportation choice. The passenger's utility function is determined by the following factors and can be written as: Where: T j : Waiting time for selecting transportation mode j; P j : cost / price of transportation mode j; S j : the fare reduction amount of transportation mode j; δ i : Passengers' tolerance to crowding; i It is the sensitivity of individual passengers to traffic changes, reflecting the degree to which each passenger responds to changes in traffic flow. Specifically, it measures how passengers are affected by traffic changes (such as traffic congestion or other people choosing this mode) when choosing a transportation mode. i It is used to quantify the passenger's tolerance for congestion when choosing a mode of transportation. If the traffic volume increases, passengers may feel that the congestion increases and the waiting time is prolonged, which will affect their decision to choose this mode. i It has a great influence on the choice behavior of passengers, especially in the case of high traffic volume and congestion. i It means that passengers are very sensitive to traffic changes and are easily affected by traffic changes. It mainly affects the individual decisions of passengers, which is reflected in the utility function of each passenger in the model. i : The passenger's weight on waiting time (usually related to the passenger's urgency); N j : The number of passengers who choose transportation mode j; is the rate of change of passenger flow for mode j.

[0011] Collect data on sudden disruptions at subway stations, including the number of affected passengers, available connection methods and their supply, and build a passenger flow prediction model based on a dynamic transportation network to predict changes in passenger flow;

[0012] The data collection and modeling mentioned above refers to the system first collecting real-time passenger flow information, transportation mode supply, government regulatory measures and other data after an emergency event (such as equipment failure, fire, terrorist attack, extreme weather, etc.) occurs at a subway station. The data collected includes: the total number of affected passengers N and their distribution; the capacity O of existing transportation modes (buses, online ride-hailing, taxis, shared bicycles, etc.); j The real-time status of the traffic network (road congestion, available trains, etc.) and passenger behavior characteristics (patience threshold, travel habits, etc.) are analyzed. A dynamic traffic modeling framework based on a spatiotemporal network is constructed using historical data and real-time monitoring data to predict passenger flow trends after emergencies and provide input data for optimizing evacuation plans.

[0013] Based on the acquired passenger and available connection data, the waiting time for each mode of transportation is calculated taking into account capacity constraints, passenger patience thresholds, and traffic dynamics.

[0014] The waiting time calculation mentioned above refers to the fact that after an emergency occurs, the waiting time of different modes of transportation is affected by multiple factors such as supply and demand, passenger patience, and fare reductions. Therefore, a dynamic time calculation model is required: Among them: the first reflects the direct relationship between the current number of passengers and the transportation capacity; the second term γ j log(N j ) takes into account the crowding effect caused by the increase in passengers; the third This model reflects the sensitivity of waiting time to changes in passenger flow. Through this model, we can dynamically estimate the waiting time of different modes of transportation under different passenger flow conditions and provide a decision-making basis for passengers' travel choices.

[0015] The log-linear model is used to calculate the probability of passengers choosing different modes of transportation;

[0016] The above-mentioned choice probability calculation refers to the fact that in game theory, the probability of each passenger choosing a certain mode of transportation is determined by the utility of that mode of transportation, and the passenger will choose the mode of transportation that maximizes his utility. Assume that the probability p of each passenger choosing a mode of transportation is i,j is based on the utility function: Where: C i,j is the total cost of passenger i in choosing transportation mode j, including waiting time cost, travel expenses, fare reduction, individual passenger patience threshold, and the cost corresponding to passenger flow changes: λ is a parameter used to adjust the passenger's sensitivity to different transportation modes; P j For the fare, S j is the fare reduction amount; α i represents the passenger's sensitivity to waiting time; δi Reflects passengers' tolerance to crowding; i This formula uses a log-linear model, which describes the relationship between the probability of a passenger's choice and the cost: the probability of choosing a lower-cost option (i.e., one with less waiting time or lower fees) is higher.

[0017] Optimize the fare reduction plan to tilt passengers toward the most efficient evacuation method, and adjust fares at the same time;

[0018] The fare reduction optimization and fare adjustment mentioned above means that in order to optimize the distribution of transportation modes, the government can dynamically adjust the fare reduction S j and fare P j , so that passengers can be inclined towards the optimal evacuation method.

[0019] Adopting the fare reduction optimization model: Among them, ξ j is the dynamic response coefficient of fare reduction to passenger flow changes, reflecting the dynamic impact of passenger flow on fare reduction, α j and β are static adjustment parameters of fare reduction, which are used to describe the law of fare reduction changing with the number of passengers. It is the rate of change of the number of passengers over time, that is, the time partial derivative of passenger flow, which indicates the rate at which passengers choose this mode of transportation at a certain moment. j It consists of two parts: 1) Static part: This item describes the government's benchmark for reducing fares for different modes of transportation: α j : Determines the initial level of fare reduction. Fare reduction policies for different modes of transportation may vary. β: Controls how fare reductions adjust as the number of passengers increases. If β>1, it means that we want to encourage more passengers to use this mode, and the more passengers there are, the higher the fare reduction will be. If β<1, it means that fare reductions for heavily used modes of transportation will be gradually reduced to control the fiscal budget. 2) Dynamic part: This item indicates that the government will adjust the fare reduction response mechanism in real time according to changes in passenger flow: represents the rate of change of the number of passengers over time, if This indicates that there are more and more passengers, and it may be necessary to reduce fare reductions to avoid over-reliance; if This indicates that the number of passengers has decreased, and it may be necessary to increase the price reduction to attract more people to choose this mode of transportation. j ξ is the responsiveness of fare reductions to changes in passenger flow, indicating how the government adjusts fare reductions based on changes in traffic flow. It reflects the dynamic adjustment capability of fare reduction policies, that is, how the government adjusts fare reductions under different traffic conditions to guide passenger choice behavior. jControl the government's adjustment of the fare reduction for transportation mode. Specifically, if the traffic volume of transportation mode j changes (for example, a surge in traffic during peak hours), the government may adjust the fare reduction based on the traffic volume change to guide passengers to choose other transportation modes or avoid overcrowding of a certain mode. j >0, indicating encouragement to relieve congestion: if the flow of a certain mode of transportation increases, the fare reduction will be reduced to avoid excessive congestion; if ξ j <0, indicating an incentive for increased use: if traffic decreases, the fare reduction will increase to attract more passengers to choose this mode.

[0020] Pricing strategy: The government can also adjust the fare P for each mode of transportation j In order to guide more passengers to choose appropriate modes of transportation while ensuring that the transportation modes are not overcrowded, the fare adjustment model is as follows: in, is the base fare, β is the adjustment coefficient, N j is the demand for transportation mode, O j is the maximum capacity of the transportation mode. j >O j When N j <O j When the price of evacuation is high, the fare is reduced to attract more passengers to choose this mode. This strategy can achieve a dynamic balance between supply and demand and optimize the overall evacuation efficiency.

[0021] By iteratively optimizing passenger allocation and calculating the generalized Nash equilibrium, the supply and demand relationship converges to the optimal balance;

[0022] The generalized Nash equilibrium calculation mentioned above refers to the optimization and solution of the generalized Nash equilibrium model to achieve the optimal balance between supply and demand of each mode of transportation. The solution process of the model is an iterative optimization process. Each optimization step updates the passenger's choice probability, waiting time, fare reduction and pricing strategy until the system reaches a stable generalized Nash equilibrium. The optimization process is as follows: a) Initialization: Set the initial waiting time price and fare reductions b) Calculate the selection probability: Calculate the probability of each passenger choosing each mode of transportation based on the current waiting time and price; c) Update the passenger distribution: Based on the selection probability p i,j , update the demand N for each mode of transportation j d) Update waiting time: Update the waiting time for each mode of transportation using the following formula: e) Optimize fare reduction and ticket prices: Adjust fare reduction according to the government's optimization goals and fares Make the system more efficient; f) Convergence judgment: Check passenger distribution and Is it convergent? If it is convergent (i.e. ), then the iteration ends, otherwise returns to step b).

[0023] Finally, the evacuation strategy is executed based on the optimization results, traffic scheduling is adjusted in real time, and passengers are guided to evacuate in an orderly manner.

[0024] The specific embodiments provided by the present invention have the following beneficial effects:

[0025] This paper optimizes emergency evacuation strategies after subway emergencies by constructing a high-dimensional passenger patience game model, enabling passengers to make more rational travel choices based on factors such as wait times, fares, and fare reductions. Compared to traditional static evacuation schemes, this paper employs generalized Nash equilibrium calculations to dynamically optimize the supply and demand relationships among various modes of transportation, ensuring balanced distribution across them. This reduces overload on individual modes and improves overall evacuation efficiency.

[0026] Through a fare reduction optimization mechanism, the present invention can dynamically adjust the supply and demand of different modes of transportation in real time, enabling more precise allocation of fare reduction resources. For example, in situations where public transportation is insufficient and ride-hailing prices are too high, the system can automatically adjust fare reductions to guide some passengers to choose the optimal mode of transportation, reducing wait times and improving the accuracy of fiscal expenditures. Furthermore, this method can reduce the imbalance in transportation modes caused by excessively high fares or insufficient fare reductions, thereby optimizing overall evacuation costs.

[0027] This method is not only applicable to emergency evacuation in subway stations, but can also be extended to scenarios such as urban public transportation management, traffic scheduling for major events, and disaster emergency response, providing a scientific optimization solution for public safety management. By combining mathematical modeling, game optimization, and dynamic control, this method improves the intelligent level of emergency evacuation and has important practical application value for improving the safety, stability, and operational efficiency of urban rail transit systems.

[0028] Other features and advantages disclosed in this embodiment will be described in the subsequent specific implementation manner and are described in detail as follows in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a flow chart of a method for emergency passenger evacuation of urban rail transit based on dynamic subsidy optimization according to an embodiment of the present invention;

[0030] Figure 2 It is an iteration graph of the generalized Nash equilibrium algorithm according to an embodiment of the present invention;

[0031] Figure 3This is a schematic diagram of an auxiliary evacuation passenger service mode for emergency evacuation of passenger flow in urban rail transit according to an embodiment of the present invention. DETAILED DESCRIPTION

[0032] The technical solution of the present invention is further described below with reference to the accompanying drawings and specific embodiments:

[0033] Example 1

[0034] A method for emergency passenger evacuation of urban rail transit based on dynamic subsidy optimization is as follows: Figure 1 The specific implementation steps are as follows:

[0035] (1) Initial parameter setting

[0036] Assume that a fire breaks out at a central subway station (Station A) in a city during the evening rush hour. 12,000 passengers need to be evacuated within one hour. The passengers can choose the following transportation options:

[0037] Table 1 Parameter values

[0038]

[0039] Other parameter settings:

[0040] Passenger sensitivity to waiting time: α B =0.8,α C =1.2,α S =0.5

[0041] Passenger sensitivity to flow changes: ζ B =0.6,ζ C =1.0,ζ S =0.4

[0042] Dynamic adjustment coefficient for fare reduction: ξ B =0.2,ξ C =-0.3,ξ S =0.1

[0043] Fare reduction adjustment parameter: α j =0.5, β=0.7

[0044] Convergence criterion: ε = 10 people, that is, when the number of passengers in each mode of transportation does not change by more than 10 after a certain round of iteration, it is considered to have reached a stable state.

[0045] (2) Calculation process

[0046] This embodiment adopts generalized Nash equilibrium and iterative optimization algorithm to solve the problem. Passenger allocation, waiting time, fare reduction and price are updated in each iteration until the system reaches equilibrium.

[0047] Step 1: Calculate the initial passenger selection probability

[0048] The log-linear decision model is used to calculate the probability of passengers choosing different modes of transportation:

[0049]

[0050] The total cost of the passenger choosing mode j is:

[0051]

[0052] Substituting the initial parameters, we can calculate the initial distribution:

[0053] Public transportation: 6,500 people

[0054] Ride-hailing: 2,500

[0055] Shared bikes: 3,000 people

[0056] Step 2: Calculate supply and demand balance

[0057] Since the online car-hailing capacity (2,000 passengers / hour) has been overutilized, the fare reduction mechanism is automatically adjusted:

[0058]

[0059] Because C =-0.3, the overload of online car-hailing traffic will lead to a decrease in fares, prompting some passengers to switch to public transportation or shared bicycles.

[0060] Step 3: Iteratively optimize fare reductions and prices

[0061] The government dynamically adjusts ticket price reductions:

[0062]

[0063] (3) Iterate until convergence

[0064] Repeat the above steps until:

[0065]

[0066] Finally, a stable distribution is reached:

[0067] Public transportation: 7,200 people

[0068] Ride-hailing: 1,800

[0069] Shared bikes: 3,000 people

[0070] Result Analysis

[0071] Reduced evacuation time: The original waiting time for online ride-hailing vehicles was 10 minutes. Through fare reductions, passengers were diverted to buses and shared bicycles, reducing the average waiting time for online ride-hailing vehicles to 7 minutes and reducing the total evacuation time by 30%.

[0072] Fare reduction optimization: The government reduced the bus fare by 1 yuan per person, and the bus selection rate eventually increased to 60%.

[0073] Example 2

[0074] This embodiment provides an evacuation process for long-distance travel at night with low urgency.

[0075] (1) Initial parameter setting

[0076] Assume that a subway station (Station B) is closed at night due to equipment failure and 8,000 long-distance passengers need to be evacuated. The available transportation options are as follows:

[0077] Table 2 Parameter values

[0078]

[0079] (2) Calculation process

[0080] Applying the same iterative optimization process, adjusting for fare reductions and prices, we end up with the following passenger distribution:

[0081] Public transportation: 4,000 people

[0082] Ride-hailing: 2,800

[0083] Shared bicycles: 1,200 people

[0084] (3) Result analysis

[0085] Fare reduction optimization: The high demand for online ride-hailing services has led to fewer fare reductions. The government has increased fare reductions for buses and shared bikes, guiding some people to choose buses.

[0086] Waiting time optimization: Bus waiting time was reduced from 8 minutes to 6 minutes, while ride-hailing waiting time remained at 12 minutes. Total evacuation time was reduced by 25%.

[0087] Example 3

[0088] This embodiment provides a transportation process with highly intelligent intervention.

[0089] (1) Initial parameter setting

[0090] After a large-scale event, passenger traffic at a subway station (Station C) surged, and capacity was unable to meet the short-term high demand, necessitating the use of alternative modes of transportation for evacuation. The government wanted to optimize evacuation efficiency while reducing the cost of fare reductions. To this end, the government implemented an intelligent fare reduction dynamic allocation system, which automatically adjusts fare reductions based on real-time passenger flow.

[0091] Table 3 Parameter values

[0092]

[0093] (2) Calculation process

[0094] 1) Passenger initial selection

[0095] Public transportation: 6,000 people

[0096] Ride-hailing: 3,000

[0097] Shared bicycles: 2,000 people

[0098] 2) Smart fare reduction system adjustment: Peak period: bus fares are reduced and increased to attract more people to use public transportation, and online ride-hailing fares are reduced and lowered to prevent over-reliance. Off-peak period: shared bike fares are reduced and increased to guide some short-distance passengers.

[0099] Final passenger allocation:

[0100] Public transportation: 6,200 (fare reductions to attract more people)

[0101] Ride-hailing: 2,300 (down 700)

[0102] Bike-sharing: 2,500 (an increase of 500)

[0103] (3) Result analysis

[0104] The smart fare reduction system saves fiscal costs by 18% and reduces waste compared to fixed fare reductions.

[0105] Public transportation utilization rate is improved and passenger experience is optimized.

Claims

1. A method for emergency passenger evacuation of urban rail transit based on dynamic subsidy optimization, characterized in that: The steps include: (1) When an emergency occurs in an urban rail transit station, a high-dimensional passenger patience game model is constructed to optimize passengers' transportation mode choices after the subway is shut down; (2) Collect data on sudden disruption events at subway stations, including the number of affected passengers, available connection methods and their supply, and establish a passenger flow prediction model based on a dynamic transportation network to predict changes in passenger flow; (3) Based on the acquired passenger and available connection data, the waiting time for each mode of transportation is calculated taking into account capacity constraints, passenger patience thresholds, and dynamic changes in traffic flow; (4) Use the log-linear model to calculate the probability of passengers choosing different modes of transportation; (5) Optimize the fare reduction plan to tilt passengers toward the most efficient evacuation method; (6) By iteratively optimizing passenger allocation and calculating the generalized Nash equilibrium, the supply and demand relationship converges to the optimal balance; (7) Finally, the evacuation strategy is executed based on the optimization results, and traffic scheduling is adjusted in real time to guide passengers to evacuate in an orderly manner.

2. The method for emergency passenger evacuation of urban rail transit based on dynamic subsidy optimization according to claim 1 is characterized in that: The high-dimensional passenger patience game model described in step (1) quantifies the basis for passengers to choose transportation modes by calculating the passenger's utility function. The passenger's utility function is expressed as: Among them: U i,j represents the total decision cost of the transportation mode chosen by the passenger, T j : Waiting time for selecting transportation mode j; P j : cost / price of transportation mode j; S j : the fare reduction amount of transportation mode j; δ i : Passengers' tolerance to crowding; i is the sensitivity of individual passengers to traffic flow changes, reflecting the degree of response of each passenger to traffic flow changes, ζ i Used to quantify passengers' tolerance for congestion when choosing a mode of transportation; α i : Passenger's weight on waiting time; N j : The number of passengers who choose transportation mode j; is the rate of change of passenger flow for mode j.

3. The method for emergency passenger evacuation of urban rail transit based on dynamic subsidy optimization according to claim 1 is characterized in that: The specific method of step (2) is to first collect real-time passenger flow information and transportation mode supply information after an emergency occurs at a subway station. The collected data include: the total number of affected passengers N and their distribution; the capacity of existing transportation modes O j ; Real-time status of the traffic network, including road congestion and available trains; Passenger behavior characteristics, including patience threshold and travel habits; Through historical data and real-time monitoring data, a dynamic traffic modeling framework based on the spatiotemporal network is constructed.

4. The method for emergency passenger evacuation of urban rail transit based on dynamic subsidy optimization according to claim 1 is characterized in that: The calculation of waiting time for each mode of transportation described in step (3) refers to the fact that after an emergency occurs, the waiting time of different modes of transportation is affected by factors such as supply and demand, passenger patience, and fare reduction. Therefore, a dynamic time calculation model is required: Among them: the first reflects the direct relationship between the current number of passengers and the transportation capacity; the second term γ j log(N j ) takes into account the crowding effect caused by the increase in passengers; the third It reflects the sensitivity of waiting time to changes in passenger flow; through this dynamic time calculation model, the waiting time of different modes of transportation under different passenger flow conditions can be dynamically estimated, and a decision-making basis can be provided for passengers' travel choices.

5. The method for emergency passenger evacuation of urban rail transit based on dynamic subsidy optimization according to claim 1 is characterized in that: The choice probability calculation described in step (4) refers to the fact that in game theory, the probability of each passenger choosing a certain mode of transportation is determined by the utility of the mode of transportation. Passengers will choose the mode of transportation that maximizes their utility. Assume that the probability of each passenger choosing a mode of transportation is p i,j is based on the utility function: Where: C i,j is the total cost of passenger i in choosing transportation mode j, including waiting time cost, travel expenses, fare reduction amount, individual passenger patience threshold and the cost corresponding to passenger flow changes: λ is an adjustment parameter used to adjust the passenger's sensitivity to different transportation modes; P j For the fare, S j is the fare reduction amount; α i represents the passenger's sensitivity to waiting time; δ i Reflects passengers' tolerance to crowding; i Indicates the degree to which passengers respond to fare reductions.

6. The method for emergency passenger evacuation of urban rail transit based on dynamic subsidy optimization according to claim 1 is characterized in that: The fare reduction scheme described in step (5) is to dynamically adjust the fare reduction amount S in order to optimize the distribution of transportation modes. j and fare P j , so that passengers tend to the optimal evacuation mode. The specific adjustment model is as follows: (6.1) Using the fare reduction optimization model: Among them, ξ j is the dynamic response coefficient of fare reduction to passenger flow changes, reflecting the dynamic impact of passenger flow on fare reduction; α j and β are static adjustment parameters of fare reduction, which are used to describe the law of fare reduction changing with the number of passengers; is the rate of change of the number of passengers over time, that is, the time partial derivative of passenger flow, which indicates the rate at which passengers choose this mode of transportation at a certain moment; the fare reduction S j It consists of two parts: 1) Static part: This term describes the basis for the fare reduction: α j : Determines the initial level of fare reduction, which varies for different modes of transportation; β: Controls how fare reductions adjust as the number of passengers increases. If β > 1, it means that we want to encourage more passengers to use this mode, and the more passengers there are, the higher the fare reduction; if β < 1, it means that fare reductions for heavily used modes of transportation will be gradually reduced; 2) Dynamic part: This item represents a response mechanism to adjust fare reductions in real time based on changes in passenger flow: represents the rate of change of the number of passengers over time, if This indicates that there are more and more passengers, and it may be necessary to reduce fare reductions to avoid over-reliance; if This indicates that the number of passengers has decreased, and that it is necessary to increase the price reduction to attract more people to choose this mode of transportation; j is the response coefficient of fare reduction to changes in passenger flow, indicating how to adjust the size of fare reduction according to changes in the flow of transportation mode. It reflects the dynamic adjustment ability of fare reduction policy, that is, how to adjust fare reduction under different flow conditions to guide passengers' choice behavior. Specifically, if the flow of transportation mode j changes, the fare reduction amount is adjusted according to the flow change to guide passengers to choose other transportation modes or avoid overcrowding of a certain mode. If ξ j >0, indicating encouragement to relieve congestion: if the traffic volume of a certain mode of transportation increases, the fare reduction amount will be reduced to avoid excessive congestion; if ξ j <0, indicating an incentive for increased use: if traffic decreases, the fare reduction amount will increase to attract more passengers to choose this mode; (6.2) Price strategy: Since the price of each mode of transportation is P j Under the premise of ensuring that transportation modes are not overcrowded, more passengers will be guided to choose transportation modes with lower fares, and a fare adjustment model will be established: in, is the base fare, β is the adjustment coefficient, N j is the demand for transportation mode, O j is the maximum capacity of the transportation mode, when N j >O j When N j <O j When the price of evacuation is high, the ticket price is reduced to attract more passengers to choose this mode. This strategy can achieve a dynamic balance between supply and demand and optimize the overall evacuation efficiency.

7. The method for emergency passenger evacuation of urban rail transit based on dynamic subsidy optimization according to claim 1 is characterized in that: The generalized Nash equilibrium calculation described in step (6) refers to the optimization solution using the generalized Nash equilibrium model to achieve the optimal balance between supply and demand of each mode of transportation. The solution process of the model is an iterative optimization process. Each optimization step updates the passenger's choice probability, waiting time, fare reduction and dynamic fare strategy until the system reaches a stable generalized Nash equilibrium. The optimization process is as follows: a) Initialization: Set the initial waiting time T j (0) 、Price P j (0) and fare reductions b) Calculate the probability of selection: Calculate the probability of each passenger choosing each mode of transportation based on the current waiting time and price; c) Update passenger distribution: according to the selection probability p i,j , update the demand N for each mode of transportation j ; d) Update waiting time: Update the waiting time for each mode of transportation using the following formula: e) Optimize fare reduction amount and pricing: adjust the fare reduction amount according to the optimization goal and fare P j (t+1) , making the system more efficient; f) Convergence judgment: Check the passenger distribution N j (t+1) and Is it convergent? If it is convergent (i.e. Then the iteration ends, otherwise return to step b).

8. The method for emergency passenger evacuation of urban rail transit based on dynamic subsidy optimization according to claim 1 is characterized in that: The execution of the final evacuation strategy described in step (7) refers to adjusting the traffic dispatch strategy in real time based on the optimization calculation results, guiding passengers to reasonably choose transportation methods, and using an intelligent evacuation guidance system to push optimal transportation selection suggestions to passengers. Through this strategy, the evacuation speed is improved, the waiting time of passengers is reduced, and the efficiency and safety of emergency evacuation are guaranteed.

Images