A Simulation Evaluation Modeling Method for the Flow Limitation Scheme of an Urban Rail Transit Network

Abstract

A simulation evaluation modeling method for the passenger flow limiting scheme of an urban rail transit network, comprising the following steps: Step 1: Design the mesoscopic simulation scheduling principle according to the purpose of the passenger flow limiting scheme; Step 2: Model the abstract intelligent agent group; Step 3: Control the passenger arrival event; Step 4: Establish an evaluation index system for the passenger flow limiting scheme. The simulation evaluation modeling method for the passenger flow limiting scheme of the urban rail transit network of the present invention improves the efficiency and cost of modeling and subsequent model transformation, effectively reduces the computational redundancy of experiments; can display the simulation status of the current passenger flow limiting scheme through an intuitive and vivid network passenger flow density map, realizing the visualization of station and train data during the simulation process; the generated data results can accurately describe the differences in various evaluation indicators under various passenger flow control measures, and can effectively help the urban rail company complete the evaluation and management of the passenger flow limiting scheme, solving the problem of evaluating the passenger flow limiting scheme of the urban rail transit network.

Description

Technical Field

[0001] The present invention relates to the technical field of evaluation of rail transit passenger flow limiting schemes. Background Art

[0002] In recent years, major cities across the country have been actively developing urban rail transit. For cities with existing rail lines, the uneven spatio-temporal distribution of passenger flow is the main current situation faced by rail transit operating companies. Implementing a passenger flow limiting scheme is a common means of passenger flow control. A good passenger flow limiting scheme can balance various contradictions such as safety, fairness of riding, and train capacity. To achieve effective control of passenger flow in the urban rail transit operation network, scholars have conducted a series of studies on the design of passenger flow limiting schemes. Due to different perspectives and objectives in studying passenger flow control means, the related research on the design of passenger flow limiting schemes is roughly divided into two levels: micro stations and macro line networks. However, with the continuous deepening of the system engineering concept, the research on line network-level passenger flow limiting schemes has gradually attracted attention. From the current relevant studies, the research on line network-level passenger flow limiting schemes mainly focuses on the controllability determination method of the passenger flow network, the coordinated control of train operation plans and passenger flow limiting schemes at each station, the division of passenger flow limiting periods, reducing the passenger's entry delay time, and ensuring the safety of the passenger flow limiting scheme. Whether the formulated passenger flow limiting scheme is scientific often needs to be verified and evaluated by means of simulation modeling. However, there are few relevant R & D cases for the simulation evaluation of specific line network passenger flow limiting schemes. The main difficulties are reflected in that the number of stations, passengers, and trains at the line network level is often huge and their characteristics are diverse, the logic of transfer passenger flow in the line network is relatively complex, the calculation amount of the simulation model is extremely large, it is difficult to track the data of the flow process of each passenger individual in the line network, and even if the simulation evaluation of the passenger flow limiting scheme for a specific rail network is realized, when there are renovations or new lines of facilities such as stations and lines in the network, the model may face a huge amount of process reengineering. Summary of the Invention

[0003] The present invention aims to effectively solve the problems of huge numbers of stations, passengers, and trains and complex transfer logic in line network-level simulation, is applicable to any urban rail transit network, and can effectively reduce the calculation redundancy of passenger data flowing in a complex urban rail transit network and the model transformation cost.

[0004] The technical solution adopted by the present invention to achieve the above object is: A simulation evaluation modeling method for an urban rail transit line network passenger flow limiting scheme, comprising the following steps:

[0005] Step 1: Design a mesoscopic simulation scheduling principle according to the purpose of the passenger flow limiting scheme. The simulation logic model for passenger flow control is:

[0006] a i represents the starting moment of a certain passenger flow limiting period or the ending moment of the previous passenger flow limiting period; b i represents the arrival moment of a certain train; t iIndicates the arrival time of a certain passenger outside the station; t m Indicates that the number of passengers inside the station has reached the upper limit of the flow-limiting plan; x 1 Is a 1 To a 2 The flow-limiting number of passengers at the station corresponding to the time period; x 2 Is a 2 To a 3 The flow-limiting number of passengers at the station corresponding to the time period; among them, t m Passengers arriving after need to wait outside the station. Passengers can board the train as long as there are passengers arriving at the station during the train's stop time. If b i Occurs at t m Before, all the platform passengers of the arriving train board the train, and the station does not generate waiting time from a i To b i If b i Occurs at t m After, all the platform passengers of the arriving train board the train, and at the same time, the passengers waiting outside the station start to be released into the station one after another until the number of passengers entering the station reaches the flow-limiting number of passengers at the station. The station generates waiting time from a i To b i If a 2 At time x 2 <x 1 And n > x 2 , it is defaulted that the next time the station releases passengers to enter the station is at b i At the time and later. The conditions for the station to allow passengers to enter at a 2 At the time and later are:

[0007]

[0008] Step 2: Model the abstract intelligent agent group:

[0009] Step 2-1: Establish a station intelligent agent group. Take the abstract station intelligent agent as the most basic unit for constructing the physical space model of the track network, provide an environment for data exchange between passengers and train intelligent agents, realize the exchange of passenger data between stations, and complete the abstract movement of passengers in space and time in the track network;

[0010] Step 2-2: Establish a passenger intelligent agent group. Use the train arrival event to control the transfer of passenger data from the boarding station to the alighting station. When the train arrives at a certain station, release the data of the passengers getting off at that station and continue to complete the subsequent in-station logic. In this way, reduce the calculation amount of passenger data in the network model, and at the same time realize the tracking of individual passengers in the network;

[0011] Step 2-3: Establish a train agent group. Without considering the transfer and inheritance of all the data of passenger agents to train agents after the train stops at stations, simplify the data exchange process between passengers and trains as follows: Update the train statistical indicators and define the train numbers taken by passengers according to the status data related to train arrivals, boarding and alighting numbers at stations, etc.

[0012] Step 3: Control of passenger arrival events:

[0013] Step 3-1: Based on the established abstract agent group, design simulation-driven events.

[0014] Step 3-2: Complete the initial definition of specific passenger arrival events in the top-level process logic of the model.

[0015] Step 4: Establish an evaluation index system for the passenger flow restriction plan:

[0016] Step 4-1: Evaluate the fairness of the rail network. Statistically calculate the average waiting time of all passengers arriving outside the station for each station, and judge the fairness of the passenger flow restriction plan by comparing the mean and variance of the average waiting times of each station.

[0017] Waiting time of the i-th passenger waiting outside the k-th passenger flow restriction station in the model:

[0018] T k (i) = t k2 (i) - t k1 (i) (2)

[0019] Average waiting time of passengers at the k-th passenger flow restriction station in the model:

[0020]

[0021] Average waiting time of passenger flow restriction stations in the network:

[0022]

[0023] Variance of waiting time of passenger flow restriction stations in the network:

[0024]

[0025] Among them: m represents the number of passenger flow restriction stations in the model; n k represents the total number of passengers waiting outside the k-th passenger flow restriction station in the model; t k1 (i) represents the arrival time of the i-th passenger waiting outside the k-th passenger flow restriction station; t k2 (i) represents the entry time of the i-th passenger waiting outside the k-th passenger flow restriction station;

[0026] Step 4-2: Evaluate the profitability of the passenger transport system. Select the passenger turnover as the evaluation index. Calculate the passenger turnover of the network based on each passenger's OD data. Use the mileage between stations as vector data, and the passenger agent accumulates the mileage between stations from the boarding station to the alighting station in the direction of travel. For transfer passengers, the calculation of their turnover needs to be accumulated multiple times as described above for multiple sections of the journey on different lines.

[0027] The turnover mileage of the g-th passenger at the top level of the model within the rail network:

[0028]

[0029] The passenger turnover within the rail network:

[0030]

[0031] Where: g represents the sequential number of the passenger set at the top level of the model; h represents the number of passengers in the passenger set at the top level of the model; a g represents the first boarding section of the g-th passenger; b g represents the last boarding section of the g-th passenger; l j represents the mileage of the j-th boarding section of the g-th passenger;

[0032] Step 4-3: Evaluate the safety of the train's riding environment. The basis for measuring the train's safety is represented by the actual passenger capacity of the train: The actual passenger capacity of the train at the departure time of the k-th station:

[0033] N k = N k-1 + X k - Y k (8)

[0034] The maximum passenger capacity of the train throughout the journey:

[0035] N H = max{N 1 , N 2 ,......, N m} (9)

[0036] Where: X k represents the number of passengers boarding the train at the k-th station; Y k represents the number of passengers alighting the train at the k-th station.

[0037] In step 2-1, according to the different attributes of each station in the network, variables such as the current number of passengers and the number of people waiting outside the station at different stations are defined based on the uniqueness of the individual primary codes within the agent group, as well as the passenger action logic process. Thus, all stations in the network can be regarded as an agent group, realizing personalized control over the flow process of passengers with complex OD characteristics in the network, as well as the statistics and summary of evaluation indicators for the large number of stations in the line network. The event scheduling of the passenger action logic within a single station agent is achieved through train events in the top-level agent.

[0038] In step 2-2, the single-passenger data processing logic is as follows: The passenger generates and is given a number, starting point, and direction of travel. It is judged whether the number of people inside the station corresponding to the arrival station has reached the upper limit. If it has not reached the upper limit, the passenger is incorporated into the passenger set corresponding to the platform according to the direction of travel. It is judged whether the train has arrived. If the train has arrived, the passenger data is incorporated into the passenger set corresponding to the station according to the alighting station. It is judged whether to transfer at the next station where the train has arrived. If transferring, the passenger data is incorporated into the passenger set corresponding to the platform according to the direction of travel of the train. If not transferring, the passenger data is removed from the current platform passenger set.

[0039] In step 2-3, the single-train agent data processing logic is as follows: The train generates and is given a number. The train stops at stations and updates the in-vehicle passenger capacity data according to the boarding and alighting passengers. It is judged whether the current passenger volume is greater than the historical data. If it is greater than the historical data, the maximum passenger capacity of the train is set equal to the current passenger capacity. If it is not greater than the historical data, it is judged whether the train has reached the terminal station. If it has reached the terminal station, it ends. If it has not reached the terminal station, the train generates and adds the boarding passenger data at this station to the passenger sets of each arrival station.

[0040] In step 3-1, the flow-limiting time period is divided according to the time periods with different passenger arrival rates. The corresponding means that follow the Poisson distribution are set in different time periods. The non-stationary Poisson distribution within all the simulated time periods is approximately replaced by stationary Poisson distributions with different means using the method of "replacing the curve with a straight line".

[0041] In step 3-2, according to the mean of the Poisson distribution followed by the arrival of passengers at each station, the object generator of a single source is controlled to generate passengers. At the generation moment of each passenger, the passenger ID is assigned in sequence, the spatial position of the starting station is returned, the data of the terminal station and transfer stations are defined, and the whole-course turnover mileage is automatically calculated. Through the generated unique passenger ID, the initial definition of specific passenger arrival events is completed in the top-level process logic of the model.

[0042] In step 4-1, the evaluation model statistically calculates the waiting time outside the station for all passengers in the network at the inbound node, stores all the passenger waiting time data in the order of the passenger agents in the sets within each station agent, calculates the average waiting time in units of stations, and finally calculates the average waiting time and variance of the network.

[0043] The simulation evaluation modeling method for the passenger flow limiting scheme of urban rail transit network of the present invention improves the efficiency and cost of modeling and subsequent model transformation, effectively reduces the computational redundancy of experiments, and can display the simulation status of the current passenger flow limiting scheme through an intuitive and vivid network passenger flow density map, realizing the visualization of station and train data during the simulation process. The generated data results can accurately describe the differences in various evaluation indicators under various passenger flow control measures, and can effectively help urban rail companies complete the evaluation and management of the passenger flow limiting scheme, solving the problem of evaluating the passenger flow limiting scheme of urban rail transit network. Brief Description of the Drawings

[0044] Figure 1 It is a schematic diagram of passenger flow limiting for a single-direction rail transit line of the present invention.

[0045] Figure 2 It is a schematic diagram of the simulation logic of passenger flow control of the present invention.

[0046] Figure 3 It is a schematic diagram of data exchange of an abstract agent group of the present invention.

[0047] Figure 4 It is a logic flow chart of single-passenger data processing of the present invention.

[0048] Figure 5 It is a logic flow chart of single-train agent data processing of the present invention.

[0049] Figure 6 It is a schematic diagram of the operating state of the simulation evaluation model of the passenger flow limiting scheme of Dalian urban rail transit network at a certain moment. Detailed Embodiments

[0050] A simulation evaluation modeling method for the passenger flow limiting scheme of urban rail transit network of the present invention has an implementation environment of a simulation environment based on AnyLogic software and an operating system of Microsoft Windows 7 or above. The development tool uses Visualstudio 2013, and the development language uses Java language. This modeling method includes 4 steps: designing the mesoscopic simulation scheduling principle according to the purpose of the passenger flow limiting scheme, modeling the abstract agent group, classifying and controlling passenger arrival events, and establishing an evaluation index system for the passenger flow limiting scheme, which are specifically as follows:

[0051] Step 1: Design the mesoscopic simulation scheduling principle according to the purpose of the passenger flow limiting scheme:

[0052] Step 1-1: Analyze the purpose of the passenger flow limiting scheme. Taking a single-direction line in the urban rail transit network as an example, the description of the passenger flow limiting scheme problem is as shown in the appendix Figure 1 where, S 1 is the starting station, S i is the transfer station, S nis the terminal station; the controllable variables include the train safety capacity and the platform flow limit, the unit can be the number of passengers or the ratio; the decision variable is the optimal passenger flow or flow limit rate (the ratio of flow limit to station demand) of station i within the flow limit period △t (generally 15min or 30min). It can be seen that for a certain running direction, when the passenger flow getting off at each station of the rail transit line is small and the passenger flow getting on is large, the macroscopic passenger flow is accumulated and spread along the line with the train as the carrier, and the safety risks brought by the passenger flow congestion also spread along the line, and the transportation efficiency will also change accordingly. Therefore, the ultimate goal of the flow limit scheme design is to provide a flow limit scheme for each station on the rail transit line based on the safety, efficiency and fairness of the ride.

[0053] Step 1-2: Design the simulation scheduling principle of the flow limiting scheme. In the process of implementing the flow limiting scheme, different station passenger flow states have different ways of scheduling system events in the model. Taking a certain flow limiting period of a station in the line network as an example, the simulation logic model of its passenger flow control is as shown in the attached figure. Figure 2 As shown,

[0054] a i Indicates the start time of a current limiting period or the end time of the previous current limiting period;

[0055] b i represents the arrival time of a train, t i Indicates the time a passenger arrives outside the station;

[0056] t m Indicates that the number of passengers in the station has reached the upper limit of the flow control plan;

[0057] x 1 for a 1 to a 2 The number of people allowed at the station corresponding to the time period;

[0058] x 2 for a 2 to a 3 The number of people allowed at the station corresponding to the time period;

[0059] Among them, t m Passengers arriving later need to wait outside the station. Passengers arriving during the train's stop time can board the train. i Occurs at m Before, all passengers arriving at the train platform are on board, and the station is a i To b i No waiting time is generated; if b i Occurs at m After that, all passengers on the platform who have arrived at the train will board the train, and passengers waiting outside the station will be allowed to enter the station one by one until the number of passengers entering the station reaches the station's flow limit. The station will then be closed.i to b i Generate a waiting time; if a 2 Time x 2 < x 1 and n > x 2 , then the default time for the next release of passengers into the station at the station is at b i Time and later. At a 2 The conditions for the station to allow passengers to enter at time and later are:

[0060]

[0061] According to the above passenger flow control logic, the future timetable of the simulation model consists of two basic events: the arrival of each passenger and the arrival of the train. According to the types of movement logic of passengers in the network, the basic event of passenger arrival can be subdivided into four types: the arrival of upward transfer passengers, the arrival of upward non-transfer passengers, the arrival of downward transfer passengers, and the arrival of downward non-transfer passengers. The arrival of passengers at each station follows a non-stationary Poisson distribution with different means. The running time and stop time of each train are based on the train operation diagram. When the simulation clock advances to the future event occurrence time, the main program of the simulation system will update the system state according to the event type, conduct data exchange for each agent related to the event, and advance the simulation clock to the next future event. For example, at time b i when a train arrives, the waiting passengers on the platform will be cleared, and data exchange of passengers needs to be conducted between the train and the station, and between stations; at a 2 time, the arrival rate of passengers at all stations in the urban rail network will be refreshed, and the flow-limiting number of each station will be re-assigned.

[0062] Step 2: Model the abstract agent group:

[0063] According to the discrete event scheduling principle in Step 1, model the interaction of passengers, stations, and trains. For the mesoscopic simulation evaluation model of the line network flow-limiting scheme, it is not necessary to pay too much attention to the details of passengers' activities in the station and the line network, but only need to pay attention to the accuracy and effectiveness of relevant data statistics in the flow-limiting process. Therefore, the data exchange of the simulation evaluation model is simplified to a model structure with passenger agents and train agents as physical flow objects and each station agent in the rail network as the basis for abstract data storage and exchange. The data exchange logic of the abstract agent group is as shown in the appendix Figure 3 as follows.

[0064] Step 2-1: Establish a station agent group. The abstract station agent is used as the most basic unit for constructing the physical space model of the rail network, providing an environment for data exchange between passengers and train agents, realizing the exchange of passenger data between stations, and completing the abstract movement of passengers in space and time in the rail network. According to the different attributes of each station in the network, variables such as the current number of passengers and the number of people waiting outside the station at different stations, as well as the passenger action logic process, are defined based on the uniqueness of the individual primary key within the agent group. Thus, all stations in the network can be regarded as an agent group, realizing personalized control over the flow process of passengers with complex OD characteristics in the network, as well as the statistics and summary of evaluation indicators for the large number of stations in the line network. Although the station agent does not belong to the top-level agent in the model, the event scheduling of the passenger action logic within a single station agent can be achieved through the train events in the top-level agent.

[0065] Step 2-2: Establish a passenger agent group. The movement of passenger agents in the station is simplified to a process of changing the data storage location according to the station state and the arrival of trains. That is, the train arrival event is used to control the transfer of passenger data from the boarding station to the alighting station. When the train arrives at a certain station, the data corresponding to the passengers getting off at that station is released to continue with the subsequent in-station logic. In this way, the computational amount of passenger data in the network model is reduced, and at the same time, the tracking of individual passengers in the network is realized. The specific processing logic flow of individual passenger data is as shown in the appendix Figure 4 As shown, the processing logic of individual passenger data is as follows: The passenger generates and is given a number, starting point, and direction of travel. It is judged whether the number of people within the station corresponding to the arrival station has reached the upper limit. If it has not reached the upper limit, the passenger is incorporated into the passenger set of the corresponding platform according to the direction of travel. It is judged whether the train has arrived. If the train has arrived, the passenger data is incorporated into the passenger set of the corresponding station according to the alighting station. It is judged whether the passenger transfers at the alighting station. If transferring, the passenger data is incorporated into the passenger set of the corresponding platform according to the train direction. If not transferring, the passenger data is removed from the current platform passenger set.

[0066] Step 2-3: Establish a train agent group. Ignore the details of passengers getting on and off the train agent, that is, do not consider the transfer and inheritance of all the data of passenger agents to the train agent after the train stops at the station. The data exchange process between passengers and trains is simplified to: updating the train statistical indicators and defining the train number taken by the passengers according to the state data such as train stops and the number of passengers getting on and off related to the station. The processing logic of individual train agent data is as shown in the appendix Figure 5As shown in the figure, the data processing logic of a single train agent is as follows: The train generates and assigns a number, stops at stations, and updates the passenger capacity data in the train according to the boarding and alighting passengers. It judges whether the current number of passengers is greater than the historical data. If it is greater than the historical data, the maximum passenger capacity of the train is set equal to the current passenger capacity. If it is not greater than the historical data, it judges whether the train has reached the terminal station. If it has reached the terminal station, the process ends. If it has not reached the terminal station, the train generates and adds the data of the passengers boarding at this station to the passenger sets of each arrival station.

[0067] Step 3: Passenger arrival event control method.

[0068] Step 3-1: Based on the established abstract agent group, design a simulation-driven event, that is, the occurrence of the passenger arrival event. Since the flow-limiting period is divided according to different passenger arrival rates (generally following a non-stationary Poisson distribution) in time segments, the mean values of the corresponding Poisson distributions are set in different time segments, and the non-stationary Poisson distribution in all simulated time segments is approximately replaced by stationary Poisson distributions with different mean values using the method of "replacing the curve with a straight line".

[0069] Step 3-2: Since the probability of random events occurring simultaneously in the time dimension is 0, according to the mean values of the Poisson distributions followed by the passenger arrivals at each station, control the single-source object generator to generate passengers. Assign a passenger ID in sequence at the generation moment of each passenger, return the spatial position of its starting station, define data such as the terminal station and transfer station, and automatically calculate the total round-trip mileage. Through the generated unique passenger ID, complete the initial definition of the specific passenger arrival event in the top-level process logic of the model.

[0070] Step 4: Establish an evaluation index system for the flow-limiting scheme.

[0071] Step 4-1: Evaluate the fairness of the rail network. Statistically calculate the average waiting time of all passengers arriving outside the station for each station, and judge the fairness of the flow-limiting scheme by comparing the mean and variance of the average waiting times of each station. Therefore, the evaluation model needs to count the off-station waiting times of all passengers in the network at the entry nodes of the stations, store all the passenger waiting time data in the order of the passenger agents in the agent sets of each station, calculate the average waiting time for each station, and finally calculate the average waiting time and variance of the network.

[0072] The waiting time of the i-th off-station waiting passenger at the k-th flow-limiting station in the model:

[0073] T k (i) = t k2 (i) - t k1 (i)(2)

[0074] The average waiting time of passengers at the k-th flow-limiting station in the model:

[0075]

[0076] Average waiting time of flow-limiting stations in the network:

[0077]

[0078] Variance of waiting time of flow-limiting stations in the network:

[0079]

[0080] Wherein:

[0081] m represents the number of flow-limiting stations in the model;

[0082] n k represents the total number of passengers waiting outside the k-th flow-limiting station in the model;

[0083] t k1 (i) represents the arrival time of the i-th passenger waiting outside the k-th flow-limiting station;

[0084] t k2 (i) represents the boarding time of the i-th passenger waiting outside the k-th flow-limiting station;

[0085] Step 4-2: Evaluate the effectiveness of the passenger transport system. Select the passenger turnover volume as the evaluation index, and calculate the passenger turnover volume of the network based on each passenger OD data. Take the inter-station mileage as vector data, and the passenger agent accumulates the inter-station mileage from the boarding station to the alighting station in the riding direction in sequence. For transfer passengers, the calculation of their turnover volume needs to be accumulated multiple times as above according to multiple riding sections of different lines.

[0086] Turnover mileage of the g-th passenger in the top layer of the model within the rail network:

[0087]

[0088] Passenger turnover volume within the rail network:

[0089]

[0090] Wherein:

[0091] g represents the sequential number of the passenger set in the top layer of the model;

[0092] h represents the number of passengers in the passenger set in the top layer of the model;

[0093] a g represents the first riding section of the g-th passenger;

[0094] b g represents the last riding section of the g-th passenger;

[0095] l j represents the mileage of the j-th boarding section of the g-th passenger.

[0096] Step 4-3: Evaluate the safety of the train's riding environment. The basis for measuring the train's safety can be represented by the actual passenger capacity of the train. During the modeling process, the congestion level is converted into passenger capacity for statistics and calculation. Statistically calculate the maximum passenger capacity of the train throughout the route, and compare the maximum passenger capacities of all train trips at each time period as the basis for evaluating the safety index of the flow-limiting plan.

[0097] The actual passenger capacity of the train at the departure time of the k-th station:

[0098] N k = N k-1 + X k - Y k (8)

[0099] The maximum passenger capacity of the train throughout the journey:

[0100] N H = max{N 1 , N 2 ,......, N m} (9)

[0101] Where:

[0102] X k represents the number of passengers boarding the train at the k-th station

[0103] Y k represents the number of passengers getting off the train at the k-th station

[0104] The implementation process of the flow-limiting plan for the urban rail transit network is complex, and the situations of each station in the network are different. The modeling work is complex and cumbersome, with a large amount of calculation. The present invention designs the dispatching principle of flow-limiting plan simulation, proposes a modeling method based on an abstract intelligent agent group, improves the efficiency and cost of modeling and subsequent model transformation, and effectively reduces the operation redundancy of experiments. The Dalian urban rail transit network flow-limiting plan simulation evaluation model developed based on the above theoretical method, as shown in the appendix Figure 6 can display the simulation status of the current flow-limiting plan through an intuitive and vivid network passenger flow density map, realizing the visualization of station and train data during the simulation process; the generated data results can accurately describe the differences in various evaluation indicators under various passenger flow control measures, and can effectively help the urban rail company complete the evaluation and management of the flow-limiting plan. This not only solves the problem of evaluating the flow-limiting plan for the urban rail transit network, but also has certain reference significance for the simulation modeling problems related to urban rail transit operation organization.

[0105] The present invention is described by way of examples, and those skilled in the art will appreciate that various changes or equivalent replacements can be made to these features and examples without departing from the spirit and scope of the present invention. Additionally, under the teachings of the present invention, these features and examples can be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application belong to the protection scope of the present invention.

Claims

1. A simulation evaluation modeling method for the passenger flow control scheme of an urban rail transit network, characterized in that: It includes the following steps: Step 1: Design the mesoscopic simulation scheduling principle according to the purpose of the passenger flow control scheme. The simulation logic model of passenger flow control is: a i represents the start time of a certain current-limiting period or the end time of the previous current-limiting period; b i represents the arrival time of a certain train; t i represents the time when a certain passenger arrives outside the station; t m represents that the number of passengers inside the station has reached the upper limit of the current-limiting plan; x 1 is a 1 to a 2 the corresponding number of passengers limited by the station during the period; x 2 is a 2 to a 3 the corresponding number of passengers limited by the station during the period; among them, t m passengers arriving after that need to wait outside the station. Passengers can board the train as long as there are passengers arriving at the station during the train's stop time at the station. If b i occurs before t m then all the platform passengers of the arriving train board the train, and the station does not generate waiting time from a i to b i If b i occurs after t m then all the platform passengers of the arriving train board the train, and at the same time, the passengers waiting outside the station start to be released into the station one after another until the number of passengers entering the station reaches the number of passengers limited by the station. The station generates waiting time from a i to b i If at time a 2 the number of passengers x 2 <x 1 and n > x 2 , where n is the number of passengers entering the station, then it is default that the next time for the station to release passengers into the station is at time b i and later. The conditions for the station to allow passengers to enter at time a 2 and later are as follows: Step 2: Model the abstract intelligent agent group: Step 2-1: Establish a station intelligent agent group. Take the abstract station intelligent agent as the most basic unit for constructing the physical space model of the rail network, provide an environment for data exchange between passengers and train intelligent agents, realize the exchange of passenger data between stations, and complete the abstract movement of passengers in time and space in the rail network; Step 2-2: Establish a passenger intelligent agent group. Use the train arrival event to control the transfer of passenger data from the boarding station to the alighting station. When the train arrives at a certain station, release the data of the passengers getting off at that station and continue to complete the subsequent in-station logic. In this way, reduce the calculation amount of passenger data in the network model, and at the same time realize the tracking of individual passengers in the network; Step 2-3: Establish a train intelligent agent group. Without considering the transfer and inheritance of all passenger intelligent agent data to the train intelligent agent after the train stops at the station, simplify the data exchange process between passengers and trains as: update the train statistical indicators and define the train numbers taken by passengers according to the status data such as train stops, the number of passengers getting on and off at the station; Step 3: Control of passenger arrival events: Step 3-1: Based on the established abstract intelligent agent group, design the simulation-driven event; Step 3-2: Complete the initial definition of specific passenger arrival events in the top-level process logic of the model; Step 4: Establish an evaluation index system for the passenger flow control scheme: Step 4-1: Evaluate the fairness of the rail network. Statistically calculate the average waiting time of all passengers arriving outside the station for each station, and judge the fairness of the passenger flow control scheme by comparing the mean and variance of the average waiting times of each station. The waiting time of the i-th passenger waiting outside the station at the k-th passenger flow control station in the model: T k (i) = t k2 (i) - t k1 (i) (2) The average waiting time of passengers at the k-th passenger flow control station in the model: The average waiting time of passenger flow control stations in the network: The variance of the waiting time of passenger flow control stations in the network: Where: m represents the number of flow-limiting stations in the model; n k represents the total number of passengers waiting outside the k-th flow-limiting station in the model; t k1 (i) represents the arrival time of the i-th passenger waiting outside the k-th flow-limiting station; t k2 (i) represents the entry time of the i-th passenger waiting outside the k-th flow-limiting station; Step 4-2: Evaluate the effectiveness of the passenger transport system. Select the passenger turnover as the evaluation index, calculate the passenger turnover of the network based on the OD data of each passenger, take the mileage between stations as vector data, and the passenger intelligent agent accumulates the mileage between stations from the boarding station to the alighting station in the direction of travel in turn; for transfer passengers, the calculation of their turnover needs to be accumulated multiple times as above according to multiple sections of the journey on different lines. The turnover mileage of the g-th passenger in the rail network at the top level of the model: The passenger turnover in the rail network: Where: g represents the sequential number of the passenger set at the top layer of the model; h represents the number of passengers in the passenger set at the top layer of the model; a g represents the first boarding section of the g-th passenger; b g represents the last boarding section of the g-th passenger; l j represents the mileage of the j-th boarding section of the g-th passenger; Step 4-3: Evaluate the safety of the train riding environment. The basis for measuring the train safety is represented by the actual passenger capacity of the train: The actual passenger capacity of the train at the departure time of the k-th station: N k = N k-1 + X k - Y k (8) The maximum passenger capacity of the train throughout the journey: N H = max{N 1 , N 2 ,......, N m} (9) Where: X k represents the number of passengers boarding the train at the k-th station; Y k represents the number of passengers alighting from the train at the k-th station.

2. A simulation evaluation modeling method for the passenger flow control scheme of an urban rail transit network according to claim 1, characterized in that: In step 2-1, according to the different attributes of each station in the network, variables such as the current number of passengers and the number of people waiting outside the station at different stations are defined based on the uniqueness of the individual primary codes within the agent group, as well as the passenger action logic process. Thus, all stations in the network can be regarded as an agent group, realizing personalized control over the flow process of passengers with complex OD characteristics in the network, as well as the statistics and summary of evaluation indicators for a large number of stations in the line network. Through train events in the top-level agent, event scheduling of the passenger action logic within a single station agent is achieved.

3. A simulation evaluation modeling method for an urban rail transit network flow-limiting scheme according to claim 1, characterized in that: In step 2-2, the single-passenger data processing logic is as follows: The passenger generates and is given a number, starting point, and riding direction, and determines whether the number of people inside the station corresponding to the arrival station has reached the upper limit. If it has not reached the upper limit, the passenger is incorporated into the passenger set corresponding to the platform according to the riding direction, and it is determined whether the train has arrived. If the train has arrived, the passenger data is incorporated into the passenger set corresponding to the station according to the alighting station. It is determined whether to transfer at the alighting station. If transferring, the passenger data is incorporated into the passenger set corresponding to the platform according to the driving direction. If not transferring, the passenger data is removed from the current platform passenger set.

4. A simulation evaluation modeling method for an urban rail transit network flow-limiting scheme according to claim 1, characterized in that: In step 2-3, the single-train agent data processing logic is as follows: The train generates and is given a number, the train stops at the station and updates the in-vehicle passenger capacity data according to the boarding and alighting passengers, and determines whether the current passenger volume is greater than the historical data. If it is greater than the historical data, the maximum passenger capacity of the train is set equal to the current passenger capacity. If it is not greater than the historical data, it is determined whether the train has reached the terminal station. If it has reached the terminal station, it ends. If it has not reached the terminal station, the train generates and adds the boarding passenger data at this station to the passenger sets of each arrival station.

5. A simulation evaluation modeling method for an urban rail transit network flow-limiting scheme according to claim 1, characterized in that: In step 3-1, the flow-limiting period is divided according to time periods with different passenger arrival rates, and the corresponding means of the Poisson distribution are set in different periods. The non-stationary Poisson distribution within all simulated periods is approximately replaced by stationary Poisson distributions with different means using the method of "replacing the curve with a straight line".

6. A simulation evaluation modeling method for an urban rail transit network flow-limiting scheme according to claim 1, characterized in that: In step 3-2, according to the mean of the Poisson distribution followed by the arrival of passengers at each station, a single-source object generator is controlled to generate passengers. At the generation moment of each passenger, the passenger ID is sequentially assigned, the spatial position of the starting station is returned, the data of the terminal station and transfer station are defined, and the whole journey turnover mileage is automatically calculated. Through the generated unique passenger ID, the initial definition of specific passenger arrival events is completed in the top-level process logic of the model.

7. A simulation evaluation modeling method for an urban rail transit network flow-limiting scheme according to claim 1, characterized in that: In the said step 4-1, the evaluation model calculates the off-station waiting time of all passengers in the network at the inbound nodes, stores all the passenger waiting time data in the order of the passenger agents in each station agent set, calculates the average waiting time for each station, and finally calculates the average waiting time and variance of the network.

Images