A simulation method for transfer time of urban rail transit under bidirectional congestion effect

Abstract

The application belongs to the technical field of rail transit, and specifically discloses a city rail transit transfer time simulation method under bidirectional congestion effect, which comprises the following steps: constructing a node topology graph according to a transfer station building structure, determining boundary conditions according to a passenger flow peak period and train arrival and departure time, and establishing a bottom framework for simulation calculation. The ideal walking speed and congestion parameters of each channel are set through a walking path model. The congestion direction marks are obtained according to the same direction and / or reverse passengers in the channel, and the congestion parameters are determined. A transfer time calculation model is constructed in combination with the bottom framework, the walking path model, the ideal walking speed and the congestion parameters, and the model is iteratively trained through historical data, and finally simulation calculation is carried out. The method has the following advantages: the congestion influence of passengers in the same direction and reverse flow can be accurately simulated, the transfer time under different passenger flows and time periods can be more accurately predicted, and the problem of insufficient precision of traditional models is solved.

Description

Technical Field

[0001] This invention relates to the field of rail transit technology, and more specifically, to a method for simulating urban rail transit transfer time under two-way congestion effects. Background Technology

[0002] The calculation of urban rail transit transfer times relies on static empirical estimations or simplified models. These methods are mainly based on fixed walking speeds and corridor lengths, but fail to consider dynamic factors such as passenger density and individual walking speed differences. Especially during peak hours, the calculated results deviate significantly from the actual situation. In recent years, some studies have improved transfer time calculations by introducing macro-congestion effects or unidirectional flow models. However, these models still fail to effectively capture the interaction and head-on conflicts between two-way passenger flows, especially the conflicts between exiting and transferring passengers, leading to inaccurate identification and assessment of congestion bottlenecks.

[0003] To address this issue, a simulation method for urban rail transit transfer time under two-way congestion is proposed. Summary of the Invention

[0004] The present invention aims to provide a simulation method for urban rail transit transfer time under two-way congestion effect, so as to solve or improve the problem that the above-mentioned technical problems fail to consider the dynamic factors during peak hours and the two-way passenger flow interaction effect, resulting in a large deviation between the calculated transfer time and the actual situation.

[0005] In view of this, the first aspect of the present invention is to provide a method for simulating urban rail transit transfer time under two-way congestion effect.

[0006] The first aspect of the present invention provides a method for simulating urban rail transit transfer time under bidirectional congestion, comprising the following steps: constructing a node topology map based on the building structure of the transfer station, and determining boundary conditions based on the peak passenger flow period and train arrival and departure times of the transfer station; establishing a low-level framework for simulation calculation through the node topology map and the boundary conditions; constructing a walking path model based on multiple relevant variables affecting passenger walking according to the building structure, and setting the ideal walking speed and congestion parameters for passengers in each channel of the walking path model; obtaining the congestion direction marker for each passenger based on all passengers traveling in the same direction and / or in the opposite direction in the channel at the current time, and determining the congestion parameters for passengers in the current channel through the congestion direction marker; constructing a transfer time calculation model through the low-level framework, the walking path model, the ideal walking speed, and the congestion parameters; iteratively training the transfer time calculation model based on historical data, and performing simulation calculations using the trained transfer time calculation model.

[0007] The beneficial effects of this invention compared to the prior art are as follows: By establishing a simulation model that incorporates bidirectional congestion effects, the different congestion impacts experienced by passengers in both unidirectional and bidirectional flow scenarios can be realistically reflected. Through dynamic calculation of congestion direction markers and congestion parameters, the system can accurately simulate passenger transfer behavior in bidirectional flow scenarios and adjust the walking speed of each passenger according to real-time passenger flow density and flow direction, thereby solving the problem that traditional models cannot effectively capture bidirectional congestion effects.

[0008] It can more accurately predict passengers' transfer times under different passenger flows and time periods, so that the calculation results can better reflect the actual transfer experience of passengers in real operation scenarios, thus solving the problem of insufficient accuracy of traditional methods.

[0009] Additional aspects and advantages of embodiments of the invention will become apparent in the following description or may be learned by practice of embodiments of the invention. Attached Figure Description

[0010] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a flowchart of the method steps of the present invention; Figure 2 This is a schematic diagram of the topology of a transfer station node with two lines according to the present invention. Figure 3 This is a passenger arrival distribution diagram for the transfer station according to the present invention; Figure 4 This is a diagram showing the ideal walking speed distribution of passengers within the transfer station passageway according to the present invention. Figure 5 This is a graph showing the relationship between the number of passengers and time in the transfer channel of the present invention; Figure 6 This is a graph showing the relationship between the number of people waiting for the train at each platform and the time, as presented in this invention. Figure 7 This is a comparison chart of simulation results between the present invention and the predicted walking speed scheme; Figure 8 This is a comparison chart of simulation results between the present invention and a scheme that does not consider congestion; Figure 9 This is a schematic diagram of the structure of an electronic device according to the present invention. Detailed Implementation

[0011] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0012] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0013] Please see Figures 1-9 The following describes a method for simulating urban rail transit transfer time under bidirectional congestion effects, an electronic device, and a computer-readable storage medium according to some embodiments of the present invention.

[0014] An embodiment of the first aspect of the present invention proposes a method for simulating urban rail transit transfer times under bidirectional congestion effects. In some embodiments of the present invention, such as... Figures 1-8 As shown, the urban rail transit transfer time simulation method includes: S101. Construct a node topology diagram based on the building structure of the transfer station, and determine the boundary conditions based on the peak passenger flow period and train arrival and departure times of the transfer station; establish the underlying framework for simulation calculation through the node topology diagram and boundary conditions.

[0015] The first step is to construct a node topology diagram based on the architectural structure of the transfer station. This diagram is fundamental to the entire simulation method. By abstracting the various physical areas of the transfer station, such as platforms, passageways, stairs, and escalators, into nodes, and connecting these nodes with edges, we can represent possible passenger walking paths. Each node represents a specific functional area, while each edge represents the path a passenger takes from one area to another. To reflect real-world traffic flow, each edge in the topology diagram must be accompanied by certain physical attributes, such as the width of the passageway, the slope of the stairs, and the speed of the escalator. These attributes directly affect passenger movement speed and transfer time. Therefore, the node topology diagram comprehensively displays the spatial structure and passenger walking paths within the transfer station.

[0016] After constructing the node topology graph, the boundary conditions for the simulation calculation are further determined based on the peak passenger flow periods at transfer stations and the arrival and departure times of trains. By analyzing historical operational data of transfer stations, the passenger flow characteristics during peak periods are determined. Peak periods refer to the morning and evening rush hours, during which passenger flow is high and station congestion is severe. Therefore, the simulation model needs to pay special attention to the dynamic changes in passenger behavior during this period. The arrival and departure times of trains directly determine the time and flow of passengers arriving at transfer stations. Depending on the departure times of different trains, passenger arrival patterns will vary. By obtaining the above train arrival and departure time information, the simulation model can more accurately calculate passenger transfer times. Peak periods and train schedules are key elements of the boundary conditions. By setting these conditions, the simulation calculation can be constrained within a reasonable time and space range, ensuring that the simulation results match the actual operational scenario.

[0017] Based on the constructed node topology and pre-defined boundary conditions, this step also establishes the underlying framework for simulation calculations. This framework is responsible for integrating all input data, the node topology, boundary conditions, etc., and performing efficient calculations. The construction of this framework needs to consider how to dynamically update passenger states during the simulation process and how to calculate passengers' actual walking time, waiting time, etc. To ensure the accuracy of the simulation calculations, the underlying framework adopts a discrete event simulation method, using time steps to progressively simulate passenger behavior within the transfer station. Whenever a new event occurs, such as passenger arrival or train arrival, the passenger state is updated according to the set rules, and their movement path, speed, waiting time, etc., are calculated to obtain real-time transfer time data.

[0018] As can be seen from the above, this step lays a solid foundation for subsequent simulation calculations, ensuring that the simulation model can realistically reflect changes in passenger transfer time under different operating conditions, and providing a scientific basis for refined transfer efficiency analysis.

[0019] Specifically, building the underlying framework of the simulation model of this invention requires clarification and constraint from both spatial and temporal dimensions. In the spatial dimension, a node topology diagram is constructed based on the actual building structure of the transfer station. Functional areas such as stairs, escalators, passageways, and platforms are abstracted as physical nodes, and the connections between nodes correspond to passable paths. In the temporal dimension, this invention defines the boundary conditions for simulation operation. These include: the time range under study (e.g., specific morning and evening peak hours), the arrival and departure times of trains on different lines, and the discrete time step used in the simulation. The train timetable constrains passenger arrival, waiting, and departure behaviors on the platform; the time range limits the scenario interval covered by the simulation; and the discrete time interval determines the accuracy and update frequency of the system's dynamic evolution.

[0020] S102, construct a walking path model based on multiple relevant variables that affect passenger walking according to the building structure, and set the ideal walking speed and congestion parameters for passengers on each channel in the walking path model.

[0021] When constructing the walking path model, multiple architecturally related variables need to be considered comprehensively. These variables include, but are not limited to, the length and width of the passageway, the slope of stairs and escalators, the signage design within the station, the operating speed of escalators and stairs, and the visibility of the surrounding environment. Different architectural elements will have varying degrees of impact on passenger walking speed. Specifically, narrower passageways and stairs may force passengers to slow down, thus increasing transfer time, while wider passageways allow passengers to pass through more quickly. Furthermore, the slope of stairs and escalators, as well as the operating speed of escalators, directly determine passenger walking speed, especially when passageway space is limited, where passenger walking efficiency and transfer time are greatly affected. Therefore, in the walking path model, specific walking time parameters need to be assigned to each passageway based on its geometric characteristics to ensure that the model can simulate real-world passenger behavior.

[0022] The walking path model also needs to set the ideal walking speed for passengers in each aisle. The ideal walking speed refers to the speed at which passengers can walk freely without any obstruction or congestion. This speed is estimated based on statistical data, taking into account differences between different passenger groups, such as those with faster or slower walking paces. In practical applications, the ideal walking speed depends not only on the physical conditions of the aisle but also on factors such as the passenger's destination, urgency, and individual physical strength.

[0023] During peak hours, the interaction between passengers leads to congestion, affecting the actual walking speed of each passenger. The congestion parameter is designed to quantify this impact, especially in densely populated passageways where passengers must avoid each other, significantly reducing the passageway's capacity. To simulate this process, the congestion parameter needs to be dynamically adjusted based on the real-time number and density of passengers in the passageway. In practice, the congestion parameter gradually increases as passengers accumulate in the passageway, causing a gradual decrease in passenger walking speed. Therefore, in the model, the relationship between the congestion parameter and the number of passengers needs to be determined through regression analysis using field surveys or historical data.

[0024] As described above, by comprehensively considering the various variables and parameters, the walking path model constructed in this step can realistically reflect passenger walking behavior within the transfer station and provide important support for subsequent simulation calculations. This model can dynamically adjust passengers' ideal walking speed and congestion effects, ensuring that simulation results fully reflect actual passenger transfer times under different passenger flow and building conditions.

[0025] Specifically, the steps for constructing a walking path model based on multiple relevant variables that affect passenger walking according to building structure include: At least one passageway for passengers to walk through during transfers, as well as a subway line for passengers to transfer to trains, shall be provided within the transfer station.

[0026] Obtain the platforms connecting the passageway and the subway line, and mark the platforms taking into account the direction of the subway line.

[0027] Passenger walking paths are generated using the calibrated platforms and passageways, and a walking path model is constructed based on all passenger walking paths.

[0028] Based on the specific description above, within a transfer station, passageways are the main pathways connecting different areas and platforms. Therefore, the first step is to design and configure each passageway within the transfer station. Transfer stations include various types of passageways with different functions, such as stairs, escalators, and corridors. The characteristics of each passageway can affect passenger walking time. Specifically, longer passageways may require passengers to complete transfers in longer periods, while narrower passageways may cause more severe congestion due to insufficient space, thus affecting passenger walking speed. In this step, the passageways are planned and modeled according to the actual structure of the transfer station, and each passageway is assigned certain physical attributes, such as its width, passage capacity, and maximum walking time.

[0029] Obtaining the platforms connecting passageways and subway lines, and marking the platforms considering the direction of the subway lines, is to ensure that within transfer stations, passengers can smoothly transfer from one platform to another based on the arrival and departure times and directions of trains. When constructing the travel path model, trains traveling in different directions will stop at different platforms, and passenger transfer behavior is influenced by the train's direction. Specifically, passengers may need to choose to use platforms in different directions to catch the next train. To simulate this process, it is necessary to set the platform markings for different directions in the travel path model based on actual operational conditions, and optimize the platform configuration for each direction according to the arrival and departure times of trains to better guide passenger transfers.

[0030] The model generates passenger travel paths using calibrated platforms and passageways. When generating these paths, the optimal route is calculated based on factors such as the order in which passengers enter the platform, their transfer destination, and the train's arrival time. Different passengers may have different transfer needs and route choices; specifically, some passengers may prioritize shorter passageways, while others may choose faster escalators. By considering these different route choices, the model can take into account individual passenger differences and priorities, thereby generating realistic transfer behavior data.

[0031] Based on the walking paths of all passengers, a walking path model for the entire transfer station is constructed. This model not only includes the walking paths of individual passengers but also considers the flow of a large number of passengers within the station. Passengers exist simultaneously within the transfer station, and their flows intertwine. Factors such as the congestion level of passageways, passenger flow at different times, and the walking speed of different passengers are also taken into account to ensure that the generated walking path model can reflect passenger behavior in the actual scenario.

[0032] As can be seen from the above, by setting up channels, calibrating platforms, and generating walking paths, the model can dynamically simulate passenger transfer behavior and reflect the walking speed and transfer efficiency under different passenger flow conditions in real time, providing a model that can highly reproduce passenger flow within urban rail transit transfer stations.

[0033] In any of the above embodiments, the travel path model includes the following formula: in, For passengers The path of travel; For transfer stations in the following directions subway lines The corresponding platform; For transfer stations in the following directions subway lines The corresponding platform; For the line and lines At least one channel between them, and the length of the channel is .

[0034] Specifically, after constructing the underlying structure of the simulation model, dynamic passenger movement rules are modeled, including passenger movement speed distribution and the impact of congestion on movement speed, so that actual passenger transfer behavior can be accommodated in the simulation model. To ensure that the simulation model can highly reproduce actual passenger flow behavior, this invention determines key parameters in a multi-dimensional and refined manner, mainly in two aspects: fitting the ideal passenger movement speed distribution and determining the regression of congestion parameters.

[0035] Before proceeding, it is necessary to define the relevant symbolic variables of the simulation model to describe the calibration and modeling simulation process. The simulation system time discretization is... For transfer stations with subway lines For each line Their directions are respectively For the line direction The train followed Number, its arrival time is Departure time is Define transfer corridors within the transfer station. The channel length is , used for connection and Two platforms. For passengers... Its travel path is Given the starting point The destination is The ideal traveling speed is The actual walking speed is .

[0036] passenger In the path Ideal walking speed If the distribution follows a normal distribution, the mean needs to be estimated. and standard deviation The process involves the following steps: First, free-walking speed data is collected and preprocessed, prioritizing time periods and passageways within urban rail transit transfer stations where passenger density is extremely low and passengers can move freely and continuously, to ensure that the collected speed data is not or minimally affected by congestion. Then, parameter values ​​are estimated based on the maximum likelihood estimation principle, including the mean value. and standard deviation Calculate according to equations (1) and (2) respectively.

[0037] (1) (2) Finally, a goodness-of-fit test is performed to verify the rationality of the normal distribution hypothesis.

[0038] Regression calibration of congestion parameters to solve scale parameters Maximum congestion density Congestion coefficient of passengers traveling in the same direction Congestion coefficient for passengers traveling in the opposite direction The steps are as follows: First, select passageways within urban rail transit transfer stations with varying levels of congestion, including typical scenarios with only one-way high-density passenger flow (simulating same-direction congestion) and simultaneous high-density same-direction and reverse-direction passenger flow (simulating two-way congestion conflict), and collect actual passenger walking speeds. Then, use the least squares method to measure the deviation between the model's predicted and actual values. First initialize the parameters For the parameter with trial values, the maximum congestion density is selected. This can correspond to actual extreme congestion situations; the same-direction congestion coefficient To compare the reverse congestion coefficient The small value reflects the additional inhibitory effect of reverse passenger movement. For parameter estimation, the gradient descent method can be used to iteratively optimize the parameter values. Regarding the scale parameter... Maximum congestion density We can obtain a reference value through regression. , .

[0039] S103: Obtain the congestion direction marker for each passenger based on all passengers traveling in the same direction and / or in the opposite direction within the channel at the current time, and determine the congestion parameters of the passenger within the current channel based on the congestion direction marker.

[0040] Here, the core task of this step is to obtain the congestion direction marker for each passenger. The congestion direction marker is used to identify the congestion situation a passenger encounters in the current passageway, as well as their relative position and direction of travel relative to surrounding passengers. In transfer stations, passengers walk in different directions; therefore, in some specific passageways, there may be both unidirectional and counter-directional passenger flows. This intermingling of bidirectional passenger flows leads to mutual influence among passengers, resulting in congestion. To describe this congestion effect, a specific direction marker is assigned at each time point based on the passenger's flow direction, relative position, and the flow status of surrounding passengers.

[0041] First, it's necessary to determine the current flow direction of passengers within the passageway. For the same passageway, if most passengers are walking in the same direction, it can be considered unidirectional flow; if most passengers are walking in opposite directions, it's considered counter-directional flow. In actual operation, passengers within the passageway may be flowing in both directions. This bidirectional flow will cause congestion, requiring each passenger to be assigned a congestion direction marker to distinguish their interference with other passengers. Each passenger's congestion direction marker is determined based on their relative position and flow direction to other passengers in the current passageway.

[0042] After marking the congestion direction, the next task is to determine the congestion parameters for each passenger in the current passageway based on these markings. Congestion parameters refer to the degree of congestion effect experienced by passengers at a specific time and location, and are related to passenger density within the passageway, passenger flow direction, and the size of the congested area. Congestion parameters within the passageway gradually increase with passenger density, especially in cases of bidirectional passenger flow intermingling, where passengers need to frequently avoid each other, significantly inhibiting walking speed.

[0043] To quantify this impact, congestion parameters are dynamically adjusted based on real-time passenger flow density within the passageway. By analyzing passenger flow in different directions in real time, the congestion parameters for each passenger are updated at each time step, thereby calculating the passenger's actual walking speed. The determination of congestion parameters considers the passenger's relative position, walking direction, and the influence of other passengers on their walking path. Specifically, in the case of unidirectional flow, the congestion parameters are smaller because there is less interference between passengers; while in the case of reverse flow, the congestion parameters are larger because passengers need to avoid each other, increasing travel time.

[0044] As described above, the above steps can simulate passenger behavior in transfer stations under two-way congestion in real time, reflecting the impact of different passenger flow densities and directions on transfer time. Each passenger's congestion parameters are not only related to their location in the passageway, but also dynamically affected by the overall passenger flow situation within the station, thus providing strong data support for subsequent transfer time calculations and station flow optimization.

[0045] Specifically, the steps for obtaining the congestion direction marker for each passenger based on multiple passengers traveling in the same direction and / or opposite direction within the passage at the current moment include: Determine the first directional relationship between the current passenger and the passageway, and the second directional relationship between the current passenger and other passengers in the passageway.

[0046] The congestion direction marker for the current passenger is assigned a value based on the first and second directional relationships.

[0047] Based on the specific description above, the primary directional relationship between the current passenger and the passageway is determined. The goal of this determination is to ascertain the passenger's flow direction within the passageway, specifically whether the passenger is walking in the forward or reverse direction. The direction of the passageway is explicitly designated as a standard reference direction, and the passenger's direction of travel is compared to this standard direction. If the passenger's direction of travel aligns with the reference direction, the primary directional relationship is considered to be in the same direction; if the passenger's direction of travel is opposite to the reference direction, it is considered to be in the opposite direction. This provides a basic flow direction determination for subsequent congestion effect calculations.

[0048] Determine the secondary directional relationship between the current passenger and other passengers in the aisle. This relationship refers to the relative flow direction between the passenger and other passengers traveling together. In real-world scenarios, passengers in the aisle may not simply move in one direction; multiple passengers often walk simultaneously. In such cases, the relative direction and distance between passengers directly affect their actual walking speed. Determining the secondary directional relationship is essentially to clarify the behavioral patterns of passengers when they meet: specifically, whether the current passenger needs to give way to other passengers or can continue walking directly. In this step, based on the relative position and direction of movement of the passengers, if two passengers are walking in the same direction and are close to each other, it indicates a potential risk of collision or congestion between them; if passengers are walking in opposite directions and are close to each other, their encounter may lead to mutual avoidance, further affecting walking speed.

[0049] Based on the above two steps of judgment, the congestion direction marker for the current passenger is assigned a value according to the first and second directional relationships. The value of the congestion direction marker is dynamically adjusted according to the specific situation of the passenger. Specifically, if the current passenger's reference direction is consistent with the direction of the passage and they are flowing in the same direction as other passengers, a lower congestion impact marker is assigned to the passenger, because their walking speed is less affected. Conversely, if the passenger's reference direction is opposite to the direction of the passage and they are flowing in the opposite direction to other passengers, a higher congestion impact marker is assigned to the passenger, meaning that the passenger is experiencing greater congestion. Furthermore, if the current passenger has a strong relative collision or avoidance need with other passengers, the value of the congestion direction marker will be further increased, thus reflecting a greater congestion effect.

[0050] As described above, by determining the directional relationship between the current passenger and the passageway, as well as the relative directional relationship between the passenger and other passengers, a congestion direction marker is assigned to each passenger. This not only reflects the passenger's flow direction within the current passageway but also indicates the potential congestion situation within the passageway. Ultimately, these markers allow for dynamic adjustment of each passenger's walking speed and effectively predict the impact of congestion on transfer time.

[0051] Specifically, the steps for constructing congestion parameters for passengers in the current lane using congestion direction markers include: The same-direction congestion coefficient and the reverse-direction congestion coefficient are respectively set within a preset range.

[0052] The weighting method for the congestion direction markers in the same direction and the opposite direction is determined based on the first directional relationship.

[0053] The congestion parameters of the current passenger's lane are determined by assigning weights to the same-direction congestion coefficient and the reverse-direction congestion coefficient.

[0054] Based on the specific description above, adjustable congestion coefficients for both the same-direction and reverse-direction flows are set within a preset range to differentiate the degree of congestion faced by passengers in different flow directions. The same-direction congestion coefficient represents the degree of speed reduction caused by increased passenger density among passengers flowing in the same direction; the reverse-direction congestion coefficient represents the degree of speed reduction caused by additional friction resulting from mutual avoidance and collisions between two groups of passengers flowing in opposite directions. To simulate these effects, an appropriate numerical range needs to be set for each channel when building the model. The same-direction congestion coefficient will be smaller than the reverse-direction congestion coefficient because there is less interference between passengers traveling in the same direction; while in reverse flow, the need to avoid collisions and the weaving between passengers lead to more severe congestion, resulting in a relatively larger reverse-direction congestion coefficient. This setting process is dynamic and can be adjusted based on actual traffic flow and historical data.

[0055] The weighting of congestion coefficients for both unidirectional and reverse directions is determined based on the first directional relationship. The first directional relationship refers to whether a passenger's direction of travel within the current passageway is unidirectional or reverse relative to the passageway's reference direction. In practice, the passenger's direction of travel determines the type of congestion they encounter. For passengers traveling in the same direction, a lower weight is assigned to the reverse congestion coefficient, as the congestion effect primarily stems from the density of unidirectional passenger flow. Conversely, for passengers traveling in the opposite direction, the reverse congestion coefficient receives a greater weight because avoiding obstacles between passengers in reverse requires more space and time, increasing friction. This weighting is implemented through specific rules and formulas. Based on each passenger's first directional relationship, the weights of their congestion parameters are adjusted in real-time, allowing for dynamic calculation of the degree of congestion impact under different passenger flow conditions and based on the passenger's direction of travel.

[0056] By assigning weights to the same-direction congestion coefficient and the reverse-direction congestion coefficient, the congestion parameters of the current passenger's passage are determined. These congestion parameters represent the overall congestion impact experienced by the passenger in the current passage, influenced not only by passenger flow density but also by passenger direction of travel and avoidance behavior. Weighted calculation of the same-direction and reverse-direction congestion coefficients reflects the actual congestion effect experienced by each passenger in the passage. If passengers in the passage are mostly flowing in the same direction and at a high density, the same-direction congestion coefficient will primarily affect passenger walking speed; if there are more passengers flowing in the opposite direction, the reverse-direction congestion coefficient will dominate the congestion effect, leading to a significant decrease in passenger walking speed. The calculation and real-time adjustment of these congestion parameters can accurately simulate congestion conditions within transfer stations, thus providing a reliable basis for subsequent transfer time calculations.

[0057] As described above, the congestion parameters for each passenger within the passageway are dynamically adjusted and calculated based on the passenger's direction of travel. These parameters not only take into account the density effect between passengers but also fully simulate the situations where passengers need to avoid obstacles in reverse flow.

[0058] S104 constructs a transfer time calculation model through an underlying framework, a travel path model, an ideal travel speed, and congestion parameters; it iteratively trains the transfer time calculation model based on historical data, and performs simulation calculations using the trained transfer time calculation model.

[0059] Here, constructing a transfer time calculation model requires the effective integration of the underlying framework, walking path model, ideal walking speed, and congestion parameters. The underlying framework provides a structured computational platform for the simulation, responsible for organizing the various areas, passageways, staircases, and other physical facilities within the transfer station into a complete simulation network. The walking path model describes the passenger's walking route within the transfer station; based on the building structure and passageway configuration, the model can simulate the walking path of each passenger. The ideal walking speed refers to the speed at which passengers can walk at their maximum capacity in each passageway without congestion; it is determined based on historical data and actual measurements, reflecting passengers' walking capacity in different passageways. The congestion parameter is adjusted according to the current passenger flow density, simulating the slowdown in passenger speed due to congestion during walking. By combining these factors, the transfer time calculation model can simulate passenger walking speed and the transfer process.

[0060] Secondly, after the transfer time calculation model is built, iterative training is performed based on historical data. During simulation, the model's parameters are gradually adjusted and optimized. Historical data refers to information such as passenger flow data, walking speed data, and channel congestion conditions obtained through actual operations. This data provides the model with practical input. During training, by comparing the deviation between simulation results and actual transfer times, relevant parameters in the model, such as walking speed and congestion coefficient, are automatically adjusted to ensure that the value of each parameter is closer to reality. The model training adopts an iterative optimization approach, optimizing the calculation model by minimizing the error between simulation results and actual data.

[0061] Specifically, the model training process includes the following steps: First, historical data such as passenger arrival time, transfer time, and travel routes are obtained and used as input data for initial simulation. Next, the error is calculated based on the difference between the simulation results and actual data, and optimization algorithms are used to adjust the model parameters. This process is repeated until the error between the model's output and the actual observed data is minimized. Through this iterative training method, the transfer time calculation model is gradually optimized, thus providing more accurate prediction results in practical applications.

[0062] After training, the transfer time calculation model can be used for actual simulation calculations. Using the trained model, simulations calculate the transfer time for each passenger based on current passenger flow, train arrival times, and the operational status of transfer stations. During the simulation, passenger walking and waiting times are dynamically updated based on real-time passenger flow data, passenger walking paths, and current congestion levels. This data will provide operations managers with real-time assessments of transfer efficiency, helping them optimize train timetables, adjust passenger flow management strategies, and improve overall transfer efficiency.

[0063] As described above, an efficient and accurate transfer time calculation model is constructed by organically combining the underlying framework, travel path model, ideal travel speed, and congestion parameters. Through iterative training with historical data, the model is continuously optimized to ensure its reliability in practical applications. Ultimately, the trained and optimized model can provide a scientific basis for calculating transfer times and support the optimization of urban rail transit operation and management.

[0064] Specifically, the walking path model and ideal walking speed are used to generate multiple time outputs for passengers in the passageway in the transfer time calculation model.

[0065] As described above, by combining the walking path model with the ideal walking speed, the transfer time calculation model can dynamically generate the time required for each passenger during the transfer process. The walking path model provides passengers with route planning, while the ideal walking speed provides them with the optimal walking speed benchmark. Based on this, the model can dynamically adjust the time output according to the actual congestion situation, passenger flow, and path characteristics, ultimately achieving the simulation calculation of transfer time.

[0066] Specifically, the congestion parameter output unit of the transfer time calculation model is generated based on the congestion direction markers of other passengers in the current passenger's lane and the current passenger's direction relative to the lane.

[0067] As described above, the congestion parameter output unit can generate congestion parameters for each passenger based on real-time passenger flow and their direction of travel. These parameters are then used to calculate the passenger's actual walking time and, combined with other factors such as ideal walking speed and path length, generate the final transfer time prediction. This dynamic congestion-based calculation method ensures that the simulation results more realistically reflect the passenger's actual travel time under different passenger flow densities, transfer times, and congestion conditions.

[0068] Specifically, congestion parameters and ideal travel speeds are used to determine the actual speed output of passengers within the passageway in the transfer time calculation model.

[0069] As can be seen above, the combined effect of congestion parameters and ideal walking speed determines the actual walking speed of passengers in the current passage, thus directly affecting the calculation result of transfer time. It not only considers the physical characteristics of the passage such as width and slope and the ideal walking speed of passengers, but also dynamically reflects the impact of actual passenger flow on passenger walking speed.

[0070] In any of the above embodiments, the transfer time calculation model includes the following formula: (3) (4) (5) (6) (7) Among them, the For passengers The actual travel speed at time t is an output item of the actual speed output section of the transfer time calculation model; the... The ideal walking speed for passengers; The regression calibration of congestion parameters solves the scaling parameter problem; the aforementioned The maximum congestion density; The congestion parameters for the current passengers are output by the congestion parameter output section of the transfer time calculation model, and and These are two subway lines connected by a passageway; and stated For the direction of blockage; the The congestion coefficient in the same direction; The reverse congestion coefficient; The actual speed output of the transfer time calculation model represents the actual time passengers spend passing through the passageway; the natural number... The index variable for solving the right-side travel constraint; The total transfer travel time is the actual speed output of the transfer time calculation model; For passengers Total transfer time; the This refers to the destination information of the passenger's walking route; The minimum waiting time of the actual speed output unit of the transfer time calculation model; This refers to the train's departure time; the stated... This refers to the number of people in the passageway.

[0071] Specifically, the steps for iteratively training the transfer time calculation model based on historical data include: Using time as the iteration step, historical data is input into the transfer time calculation model until the time threshold is reached.

[0072] In iterative training, determine whether the average and standard deviation of the total transfer travel time are acceptable.

[0073] Adjust the parameter set of the transfer time calculation model based on the judgment results. .

[0074] Regarding the specific description above, historical data includes information such as actual transfer times, passenger flow, station layout, and passenger density. This data provides the model with baseline values ​​and references for real-world scenarios. This data is input into the transfer time calculation model to simulate passenger transfer behavior under different conditions. In each iteration, the current transfer travel time is calculated through simulation, and the simulation results are compared with actual data to evaluate the model. This iteration continues until a time threshold is reached, i.e., the preset maximum training period or error requirement. This time threshold is set by the developers based on specific circumstances to ensure that model training can be completed within a reasonable time while avoiding excessively long training times.

[0075] During iterative training, the mean and standard deviation of the total transfer travel time are assessed for their suitability. The mean transfer travel time reflects the average time required for all passengers during their transfers, while the standard deviation measures the volatility of passenger transfer times. These two metrics are used to evaluate the model's stability. If the model's mean transfer time deviates significantly from the actual transfer times in historical data, or if the standard deviation is too large, indicating unstable transfer time predictions, it suggests that the current model's parameters may have issues and cannot effectively reflect reality. Error analysis is performed to identify shortcomings in the model, and the model parameters are adjusted based on the evaluation results.

[0076] As described above, the steps for iteratively training a transfer time calculation model based on historical data involve setting the time step, incorporating historical data for simulation calculations, evaluating and adjusting model parameters, ultimately generating a transfer time calculation model. This model not only provides transfer time predictions but also allows for rapid adjustments based on real-time data and passenger flow changes.

[0077] Specifically, based on the aforementioned walking path model and calibrated congestion parameters, assuming that the congestion markers for passengers in any channel at any given time indicate different walking directions, this invention improves upon Weidmann's Kladek formula and proposes a calculation model for urban rail transit passenger transfer time that considers bidirectional congestion. Specifically, the meaning of each formula in the model is as follows: Equation (3) calculates passengers At any moment The actual travel speed is achieved by multiplying the ideal travel speed by the congestion factor.

[0078] Equation (4) represents the variables in the congestion influencing factors. ,aisle At any moment Passengers The real-time congestion density, the calculation process, here for passengers Indicator congestion signs The difference requires calculating the current total number of passengers traveling in the same direction and the current total number of passengers traveling in the opposite direction, depending on the specific circumstances. , Instructing passengers traveling in the same direction , Instructing passengers traveling in the opposite direction ;when , Instructing passengers traveling in the same direction , Instructing passengers traveling in the opposite direction .

[0079] Equation (5) is used to calculate passengers Through the passage The actual time.

[0080] Equation (6) calculates passengers Total transfer travel time .

[0081] Equation (7) is finally based on the destination information of the passenger's walking path. Calculate its shortest waiting time at the terminal platform. Plus travel time That is, passengers The transfer time. For the transfer time calculation model, the required parameters are... The channel length has been previously calibrated. and passenger walking routes The ideal walking speed was determined in the initial transfer station topology. Samples are obtained from the walking distribution, thus allowing us to obtain any passenger's data. Transfer time.

[0082] This invention provides a simulation method for urban rail transit transfer times under two-way congestion effects. Traditional transfer time calculation methods often rely on static estimations or simplified models, failing to accurately reflect passenger behavior and congestion effects, leading to significant deviations between predicted results and actual conditions. This invention, by introducing two-way passenger flow congestion effects and combining a refined walking path model with dynamic congestion parameters, can accurately simulate passenger transfer times under different passenger flow conditions. Especially during peak hours or under congested conditions, the model can realistically reflect the decrease in passenger walking speed, thus providing more accurate transfer time predictions.

[0083] By constructing a congestion model that considers the interaction of two-way passenger flow, this invention can capture various congestion scenarios that actually occur within transfer stations, including the weaving of passengers flowing in the same and opposite directions, resulting in avoidance and collision effects. Traditional methods often fail to effectively consider the interaction of two-way flow, while this invention can accurately simulate the relative positions and directions of passengers in the passageway under high-density passenger flow environments, thereby accurately calculating the congestion effects and transfer times in various time periods and scenarios.

[0084] Another embodiment of the first aspect of the present invention proposes a training iterative method for a transfer time calculation model. In some embodiments of the present invention, such as... Figures 1-4 As shown, this training iteration method, based on the established passenger transfer time calculation model and calibrated parameters, accurately calculates passenger transfer travel time and waiting time using the Monte Carlo discrete event simulation method, and deeply evaluates the comprehensive impact of two-way congestion on transfer efficiency. The physical meanings of the parameters used in steps S201 to S205 below are the same as those in steps S101 to S104 mentioned above. The relevant steps are shown below: S201, Initialize simulation time parameters Load the topology of the transfer station platforms and passageways to generate the passenger arrival set. For passengers Based on the distribution, the ideal passenger walking speed is generated. and the route Initialize transfer travel time. Transfer waiting time .

[0085] S202, Update passenger boarding and alighting. First, initialize passenger state, for any... In other words, satisfying At that time, a train was on the platform. Stop, if Set up a disembarkation group for the train's arrival time. Otherwise, the disembarkation speed is ,when At that time, delete from the drop-off group. One passenger, namely and replace travel time The total number of passengers in a certain passage at a certain time. The remaining distance of passenger n in the passageway ;when Then proceed to the boarding assembly point. Determine, assuming the given boarding speed is... Then remove from the boarding group One passenger, namely .

[0086] S203, Update passenger actual walking speed. For any walking aisle. Among them passengers Blockage direction marker The passenger can be uniquely determined by formulas (3) and (4). Actual walking speed In particular, when there are fewer passengers, the ideal running speed should be maintained. .

[0087] S204, Update passenger positions in transfer corridors. For any walking corridor... Among them passengers According to speed Update remaining distance ,Right now , place .if Then in Find the next passage , place , , .

[0088] S205, Update time status and indicators. If Then output the travel time. Waiting time Otherwise, place , For any In , then switch to S202.

[0089] The training iteration method for the transfer time calculation model provided by this invention takes the passenger arrival set, the corresponding passenger walking speed, and the walking path as input, and outputs the transfer walking time and waiting time for boarding. Throughout the simulation process, from the system time, it sequentially updates three key states: passenger boarding / alighting, passenger actual walking speed in the passageway, and passenger walking in the passageway.

[0090] Finally, multiple simulations were performed for different generated passenger flows. The total travel time, waiting time, and total transfer time for all passengers in each simulation scenario were calculated, along with key indicators such as mean, standard deviation, distribution pattern, and maximum value. Furthermore, comparative analyses were conducted under different scenarios, such as whether congestion was considered, whether a fixed prediction scenario was used, and different train arrival intervals, to evaluate the simulation indicators and capture the impact of different scenarios.

[0091] Another embodiment of the first aspect of the present invention proposes a specific implementation method for simulating urban rail transit transfer times, taking Chaoyangmen Subway Station in Beijing as the research object. This station is a transfer station for Metro Lines 2 and 6, and its topology is shown in the attached figure. Figure 1 As shown, the key transfer corridor is 107 meters long. The simulation was set for the morning rush hour, from 7:00 AM to 8:00 AM, with a time step of 1 second. During this period, a total of 76 train arrival events in four directions and 7983 passenger transfers were recorded. Detailed train and passenger arrival times and quantities are provided in the appendix. Figure 3 .

[0092] To construct the model parameters, the passenger walking speed under ideal conditions was first calibrated. The frequency distribution of 9052 valid walking samples is shown in the attached figure. Figure 4 As shown, the ideal walking speed is mainly concentrated in the range of 0.9-1.3 m / s, covering approximately two-thirds of the passengers. The speed distribution exhibits a right-skewed (tail-skewed) characteristic, indicating that the dispersion of fast passengers is higher than that of slow passengers. This distribution can be effectively fitted by a normal distribution with a mean of 1.15 and a standard deviation of 0.18. In the actual speed calculation model, the key parameters are determined as follows: scale parameter is 1.913, maximum congestion density is 5.4, same-direction congestion impact factor is 0.24, and reverse-direction congestion impact factor is 0.46. These parameters reflect that the speed reduction effect of reverse passenger flow is greater than that of same-direction passenger flow, and that speed decreases non-linearly with increasing density.

[0093] Based on the above parameter settings, this invention conducted 1000 simulation experiments. The core simulation parameters were set as follows: passenger boarding speed 18 people / second, passenger alighting speed 15 people / second, train stop time 30 seconds, following the principle of alighting first, then boarding. The simulation results are as follows: the average total transfer time was 1,501,778.295 seconds, averaging 188.122 seconds per person; the average passageway walking time was 903,988.014 seconds, averaging 113.239 seconds per person; and the average platform waiting time was 597,790.281 seconds, averaging 74.883 seconds per person. (Appendix) Figure 5The simulation demonstrates the change in the number of passengers in the transfer passage over time. Because passengers arrive in batches with the train, the number of passengers in the passage increases in a step-like manner; however, due to the randomness of individual walking speeds, the decline in the number of passengers is relatively smooth. (Appendix) Figure 6 This shows the changes in queue lengths on platforms in different directions. Passengers arriving at the platform merge into the waiting queues, causing multiple sudden changes in queue lengths and reaching peak values. The peak passenger flow for transferring to Line 2 occurred after 7:30 AM, similar to the adjacent... Figure 4 This aligns with the growth trend in the number of people using the central passage.

[0094] To further verify the superiority of the method of the present invention, it is compared with two simplified schemes: See appendix for a comparison with the fixed-speed scheme. Figure 7 The three horizontal lines in the diagram represent the three time averages calculated by the present invention. The fixed-speed scheme fails to reflect the heterogeneity of passenger speeds; its total transfer time and travel time decrease monotonically as the given speed increases. The fundamental flaw of this scheme is that it incorrectly predicts that passengers on the same train will arrive at the platform at the same time, leading to the distorted conclusion that none of the passengers can board when the train's arrival time is unfavorable. In contrast, the present invention accurately simulates the real-world scenario where some faster-moving passengers successfully transfer.

[0095] See appendix for a comparison with the solution that does not consider congestion. Figure 8 The bar chart results show that, without considering congestion, the average travel time within the corridor is 749,200.662 seconds, or 93.850 seconds per person. This result significantly underestimates the actual travel time within the corridor, proving that ignoring congestion effects will lead to inaccurate predictions, causing some passengers to miss their connecting trains due to insufficient estimated time.

[0096] In summary, the calculation method for considering two-way congestion proposed in this invention can effectively capture the randomness of individual passenger walking speed and the congestion effect caused by the dynamic accumulation of passenger numbers in the passage, providing strong technical support for the formulation of more accurate and reliable rail transit transfer timetables.

[0097] An embodiment of the second aspect of the present invention provides an electronic device. In some embodiments of the present invention, such as... Figure 9 As shown, an electronic device is provided, which may include: a desktop computer, a laptop, a handheld computer, and a cloud server, etc. The electronic device 3 may include, but is not limited to, a processor 301 and a memory 302. Those skilled in the art will understand that... Figure 9 This is merely an example of electronic device 3 and does not constitute a limitation on electronic device 3. It may include more or fewer components than shown, or different components.

[0098] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0099] The memory 302 can be an internal storage unit of the electronic device 3, specifically, a hard disk or RAM of the electronic device 3. The memory 302 can also be an external storage device of the electronic device 3, specifically, a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc., equipped on the electronic device 3. The memory 302 can also include both internal and external storage units of the electronic device 3. The memory 302 is used to store computer programs and other programs and data required by the electronic device.

[0100] An embodiment of the third aspect of the present invention provides a computer-readable storage medium. In some embodiments of the present invention, a computer-readable storage medium is provided that, when executed by processor 301, implements the steps of the above-described method. Therefore, the computer-readable storage medium provided in the third aspect of the present invention has all the technical effects of the above-described steps, which will not be repeated here.

[0101] The above embodiments are only used to illustrate the technical solutions of this disclosure, and are not intended to limit them. Although this disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. The above modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this disclosure, and should all be included within the protection scope of this disclosure.

Claims

1. A simulation method for urban rail transit transfer time under two-way congestion effect, characterized in that, Includes the following steps: A node topology diagram is constructed based on the building structure of the transfer station, and boundary conditions are determined based on the peak passenger flow periods and train arrival and departure times of the transfer station; a low-level framework for simulation calculation is established through the node topology diagram and the boundary conditions. A walking path model is constructed based on multiple relevant variables affecting passenger walking caused by the building structure, and an ideal walking speed and congestion parameters are set for each passage in the walking path model. The steps of constructing the walking path model based on these variables include: setting at least one passage for passenger transfer within the transfer station, and the subway line for passenger transfers; obtaining the platforms connecting the passage and the subway line, and calibrating the platforms considering the direction of the subway line; generating passenger walking paths using the calibrated platforms and passages, and constructing the walking path model based on all passenger walking paths; the walking path model includes the following formula: Among them, the For passengers The travel path; the For transfer stations in the following directions subway lines The corresponding platform; the For transfer stations in the following directions subway lines The corresponding platform; the For the line and lines At least one channel between them, and the length of the channel is ; Based on all passengers traveling in the same direction and / or opposite direction in the channel at the current moment, obtain the congestion direction mark of each passenger, and determine the congestion parameters of the passenger in the current channel through the congestion direction mark; A transfer time calculation model is constructed using the underlying framework, the travel path model, the ideal travel speed, and the congestion parameters. The transfer time calculation model is iteratively trained based on historical data, and simulation calculations are performed using the trained model. The transfer time calculation model includes the following formula: Among them, the For passengers At any moment t The actual travel speed is the output item of the actual speed output unit of the transfer time calculation model; the... The ideal walking speed for passengers; The regression calibration of congestion parameters solves the scaling parameter problem; the aforementioned The maximum congestion density; The congestion parameters of the current passenger are output by the congestion parameter output unit of the transfer time calculation model; and stated For the direction of blockage; the The congestion coefficient in the same direction; The reverse congestion coefficient; The actual speed output of the transfer time calculation model represents the actual time passengers spend passing through the passageway; the natural number... The index variable for solving the right-side travel constraint; The total transfer travel time is the actual speed output of the transfer time calculation model; For passengers Total transfer time; the This refers to the destination information of the passenger's walking route; The minimum waiting time of the actual speed output unit of the transfer time calculation model; This refers to the train's departure time; the stated... This refers to the number of people in the passageway.

2. The urban rail transit transfer time simulation method according to claim 1, characterized in that, The walking path model and the ideal walking speed are used to generate multiple time outputs for passengers in the passageway in the transfer time calculation model.

3. The urban rail transit transfer time simulation method according to claim 1, characterized in that, The congestion parameter output unit of the current passenger in the transfer time calculation model is generated based on the congestion direction markings of other passengers in the current passenger's channel and the current passenger's direction relative to the channel.

4. The urban rail transit transfer time simulation method according to claim 3, characterized in that, The step of obtaining the congestion direction marker for each passenger based on multiple passengers traveling in the same direction and / or opposite direction within the channel at the current moment includes: Determine the first directional relationship between the current passenger and the passageway, and the second directional relationship between the current passenger and other passengers in the passageway; The congestion direction marker for the current passenger is assigned a value based on the first directional relationship and the second directional relationship.

5. The urban rail transit transfer time simulation method according to claim 4, characterized in that, The step of constructing the congestion parameters of passengers in the current lane based on the congestion direction marker includes: The same-direction congestion coefficient and the reverse-direction congestion coefficient are respectively set within a preset range; The weighting method for the congestion direction markers to the same-direction congestion coefficient and the opposite-direction congestion coefficient is determined based on the first directional relationship; The congestion parameters of the current passenger's lane are determined by assigning weights to the same-direction congestion coefficient and the reverse-direction congestion coefficient.

6. The urban rail transit transfer time simulation method according to claim 4, characterized in that, The congestion parameters and the ideal walking speed are used to determine the actual speed output of passengers in the passageway in the transfer time calculation model.

7. The urban rail transit transfer time simulation method according to claim 1, characterized in that, The step of iteratively training the transfer time calculation model based on historical data includes: Using time as the iteration step, the historical data is input into the transfer time calculation model until the time threshold is reached. In iterative training, determine whether the average and standard deviation of the total transfer travel time are acceptable; Adjust the parameter set of the transfer time calculation model based on the judgment results. .

Images