Regional multi-modal rail transit passenger flow dynamic allocation method under operation interruption condition

Abstract

The application discloses a kind of regional multi-mode rail transit passenger flow dynamic allocation method under operation interruption condition, including the following contents: one, construct regional multi-mode rail transit timetable extension network;Two, construct multi-mode rail transit network passenger flow allocation model;Three, under the condition of operation interruption, using the station and interval number contained in interruption interval, interruption accident duration, the operation of maintaining operation after occurrence interruption, to multi-mode rail transit network passenger flow allocation model solution, finally obtain each OD pair passenger flow based on effective path passenger flow allocation quantity and its corresponding index.The model constructed by the application can more describe the actual situation distribution of passengers on each path;Based on the condition of operation interruption, it can be compatible with multiple transport modes, which is helpful for emergency collaborative management between multi-mode rail transit;More accurate and effective description of passenger real-time travel trajectory and distribution, which provides theoretical support and guidance for multi-mode rail transit emergency passenger flow allocation.

Description

TECHNICAL FIELD

[0001] The present application relates to a regional multi-mode rail transit passenger flow dynamic allocation method under an operation interruption condition. BACKGROUND

[0002] The future regional rail transit of China is a comprehensive network rail transit system with coordinated development of various rail transit modes. Regional rail transit has no strict definition and is generally considered to include three levels: urban rail transit mainly serving the travel demand of urban central areas; suburban railway mainly serving the passenger transport demand between suburbs and central cities; and intercity and trunk railway mainly serving the passenger transport demand between cities. In the process of changing from single-mode independent operation and management to multi-mode network integrated operation and management of regional rail transit, due to the differences in functional positioning, passenger flow, main technical standards and service objects of various modes of rail transit, the line network scale is becoming increasingly complex, the network transportation organization mode is diversifying, the types and scale of system equipment to be maintained are increasing, and the passenger flow management is becoming more difficult, which leads to an increase in unstable factors affecting the operation safety.

[0003] At the same time, with the change of rail transit network structure and the diversification of travel mode, network operation will inevitably bring the intersection and parallel between lines, which also brings a large number of transfer phenomena between various modes of rail transit. Under the form of regional rail transit network, the transfer passenger flow between various modes selects different travel modes, transfer points and travel paths, which brings different utilities to travelers. The mutual influence between various modes brings many uncertainties to the multi-mode network operation. Therefore, it is necessary to study the passenger flow distribution of multi-mode rail transit network under emergency conditions, analyze the path selection behavior of passengers and the time-varying characteristics of rail transit passenger flow under emergency scenarios of operation interruption, construct the regional rail transit schedule extension network under the condition of operation interruption, and propose a network flow distribution model and solving algorithm to obtain the space-time distribution characteristics of regional rail transit passenger flow. Through the above research, a theoretical basis is laid for the regional rail transit passenger flow distribution under emergency conditions, and technical means are provided for the transportation organization, passenger evacuation and emergency disposal of the operation and management department.

[0004] At present, the related research on rail transit under interruption condition at home and abroad mainly focuses on single mode, especially the research on urban rail transit. The research on passenger flow distribution under the operation interruption condition of multi-mode rail transit is still in its infancy. The network development of multi-mode rail transit not only needs to realize the coordinated operation of trains and the transfer connection of passengers under normal conditions, but also needs to ensure the safe operation of trains, reduce the waiting time of passengers and realize the safe and rapid evacuation of passenger flow under emergency conditions.

[0005] In the study of the impact of rail transit disruption on passenger flow, some scholars have considered the relationship between the location of passenger flow and the disruption interval, classified the passengers affected by the disruption, and determined the size of each type of affected passenger. Some scholars have established an evaluation model for local disruption of urban rail transit to obtain the affected passenger flow at each station. Some scholars have defined the affected passenger flow and constructed a model of passenger travel choice behavior under emergency conditions to predict the redistribution of passenger flow between urban rail stations under emergency conditions. Some scholars have identified the abnormality of the passenger flow entering the station based on the Bayesian prediction method and established a model of the impact of the emergency on passenger flow.

[0006] In the aspect of evacuation of passengers in urban rail transit network during operation disruption, some scholars have studied the influence of operation disruption on urban rail transit network structure and network passenger flow, and given the emergency strategies of urban rail transit system internally and externally. Some scholars have analyzed the influence of sudden event passenger flow transportation, proposed the calculation method of emergency capacity of bus transportation and the site selection method of motor vehicle standby point, and given the emergency linkage implementation measures of bus transportation based on informatization condition. Some scholars have given the starting threshold of bus emergency connection based on the difference between urban rail transit passenger demand and transportation capacity, and considered the occurrence period of emergency event. Some scholars have classified the disrupted passenger flow under the disruption scenario, updated the effective path and reloaded the passenger flow, and realized the multi-path passenger flow random dynamic allocation under the disruption condition. Some scholars have established the multi-path non-equilibrium allocation model under the local disruption of rail transit network, and estimated the passenger flow affected by sudden event by using the historical passenger flow data. Some scholars have researched the modeling of passenger travel behavior and optimization of passenger flow induction under the condition of operation disruption of urban rail transit, and modeled and simulated the passenger travel behavior from the two aspects of station closure and interval disruption. Some scholars have established the path selection algorithm under the conditions of station flow limitation and interval congestion based on the urban rail transit network under uncertain conditions. Some scholars have analyzed the passenger flow characteristics after the operation disruption accident of station or interval, set up the emergency connection of rail transit system based on the intermediate transfer station, and established the optimal feasible path set model of passenger flow evacuation. Some scholars have constructed five kinds of emergency scenes of disruption, analyzed the influence of train operation disruption on passenger flow travel, obtained the passenger flow travel path by using Dijkstra algorithm and k-shortest path algorithm, and determined the passenger flow travel selection based on logit model. Some scholars have analyzed the passenger flow space-time distribution characteristics based on the AFC data under sudden event, and proposed the passenger perceived path decision analysis method. Some scholars have established the topological model of urban rail network based on the dual graph of travel path based on the intercommunication, congestion and operation disruption of urban rail transit network, and given the dynamic distribution of interval disruption passenger flow. Some scholars have analyzed the passenger waiting problem of railway system under the condition of disruption, and given the selection model of passenger waiting or detour behavior. Some scholars have developed the passenger flow induction system of Tokyo subway, considered the predicted recovery time of operation disruption and the expected time consumed on each path, and can give the personalized travel suggestions of passenger such as detour or waiting for operation recovery. Some scholars have analyzed the passenger behavior under the condition of operation disruption of subway station or line based on AFC data, and predicted the traffic volume under emergency condition by using historical passenger flow data.

[0007] In summary, the passenger flow distribution research under the operation interruption condition mainly focuses on urban rail transit and the passenger flow evacuation research between urban rail transit and municipal traffic, and there is little research on the emergency passenger flow distribution between multi-mode rail transit, so the application proposes a regional multi-mode rail transit passenger flow dynamic distribution method under the operation interruption condition from the perspective of the emergency passenger flow distribution between multi-mode rail transit under the operation interruption condition. SUMMARY

[0008] In order to overcome the above-mentioned defects of the prior art, the application proposes a regional multi-mode rail transit passenger flow dynamic distribution method under the operation interruption condition, constructs a timetable extension network of multi-mode regional rail transit under the operation interruption condition based on the timetable extension network theory and considering the affected passenger travel demand and train state under the operation interruption condition, and establishes a stochastic equilibrium distribution model of the timetable extension network, thereby establishing a good theoretical basis for the regional rail transit network emergency management, station or line large passenger flow emergency treatment and the like.

[0009] The technical scheme adopted by the application to solve the technical problems is as follows: a regional multi-mode rail transit passenger flow dynamic distribution method under the operation interruption condition, comprising the following contents:

[0010] I. constructing a regional multi-mode rail transit timetable extension network;

[0011] II. constructing a multi-mode rail transit network passenger flow distribution model;

[0012] III. under the operation interruption condition, using the station and interval number contained in the interruption interval, the interruption accident duration and the operation measure maintained after the interruption, the multi-mode rail transit network passenger flow distribution model is solved, and finally the passenger flow distribution amount of each OD pair based on the effective path and the corresponding index are obtained.

[0013] Compared with the prior art, the application has the following positive effects:

[0014] 1) real-time: the data output in the application can make a dynamic response according to the real-time change of the actual situation input, and a new dynamic passenger flow distribution path can be generated according to the real-time passenger flow quantity, congestion waiting time and the like to adapt to the real-time operation demand, and meanwhile, since the train operation arc, train stopping arc and transfer arc are added, the travel path and waiting time of each passenger can be clearly and accurately described.

[0015] 2) flexibility: the application considers multiple rail transit modes, and can adjust the input rail transit mode category according to different actual situations to meet the demand of different scenes, thereby ensuring that the application has high flexibility in different occasions. Meanwhile, the timetable extension network meeting different conditions can be generated according to the adjustment of the line and station operation condition.

[0016] 3) Applicability: The present application considers the train overcrowding and passenger stay on the train during train overload due to the strict capacity limit of the train in rail transit, and the impact of the fare of the path, as well as the randomness of the passenger travel process, and constructs a traffic flow distribution using the idea of stochastic user equilibrium (SUE) distribution for the timetable expansion network. BRIEF DESCRIPTION OF DRAWINGS

[0017] The present application will be described by way of example and with reference to the accompanying drawings, in which:

[0018] Figure 1 Flow chart for passenger flow distribution under interruption;

[0019] Figure 2 Schematic diagram of multi-modal rail transit network transformation;

[0020] Figure 3 Schematic diagram of multi-modal rail transit timetable expansion network;

[0021] Figure 4 Flow chart of multi-modal rail transit stochastic user equilibrium distribution algorithm. DETAILED DESCRIPTION

[0022] A regional multi-modal rail transit passenger flow dynamic distribution method under operation interruption, the present application will be specifically described below in combination with the accompanying drawings, as shown in the figure, the method mainly includes two parts, the construction of the timetable expansion network, the construction of the multi-modal rail transit stochastic user equilibrium distribution model, including the following contents: Figure 1

[0023] I. Construction stage of multi-modal rail transit timetable expansion network

[0024] (1) Multi-modal rail transit topology network transformation

[0025] The present application aims to distribute passenger flow based on multi-modal rail transit timetable expansion network under operation interruption, the following will describe the present application taking four rail transit stations as an example. As shown in the figure, a, b, c, and d all represent the road network topology layer of rail transit stations, the train between two points only runs between the two points, and there is no cross-line running. Considering multiple transportation modes, including urban rail transit and other rail transit modes (such as intercity railway and urban railway, etc.) except urban rail transit. Figure 2

[0026] ​​The road network topology layer is used to describe the overall topology and morphology in the network system. It is formed by the connection of track lines, track stations and the relationship between stations. The rail transit physical network is defined as a specific connectable and abstract graph composed of a series of edges and vertices by using the corresponding relationship of network structure elements. Since there are multiple transportation modes and transfers between transportation modes in the multi-mode service network, in order to reasonably represent the transfer between transportation modes, the multi-mode rail transit network needs to be transformed accordingly. The principles of network transformation are as follows:

[0027] 1) If there are multiple transportation modes between two nodes, add corresponding lines between the two points, one line corresponding to one transportation mode.

[0028] 2) If there is a transfer process in a node in the multi-mode service network, the node splitting method is used to handle it. That is, each transportation mode is separated at the transfer node, and each transportation mode is represented by a new node. The transfer between transportation modes is represented by the connection between new nodes.

[0029] The multi-mode service network transformation schematic diagram is shown in Figure 2 The nodes on the graph are abstracted from rail transit operating stations, and the edges on the graph are abstracted from operating lines according to the operating direction. The "nodes" and "edges" of this layer network correspond to "stations" and "inter-station sections" respectively. The stations correspond to the actual road network stations. The stations are connected through the edges, and all the edges correspond to the collection of two station sections in the actual road network. In order to further describe the connection relationship of the nodes on the physical road network, the "platform", "track section" and "transfer channel" in the physical network are all represented by nodes and connection lines. The transfer station is split into multiple transfer nodes, which are connected by transfer arcs. The transfer node a with multiple modes is split into ac and au nodes, each representing a transportation mode, representing the intercity train transfer point and the urban rail train transfer point respectively. Similarly, the transfer node b is split into bs and bu nodes, representing the city train transfer point and the urban rail train transfer point respectively. Through this modeling method, the multi-mode rail transit network is actually constructed into a connected graph, which builds the bottom layer network model for the time table expansion network and passenger flow distribution in the following text.

[0030] (2) Time table expansion network construction

[0031] In order to uniformly describe the space-time expansibility of the nested rail transit physical network and train operation network, the application describes the space-time expansibility of the network by constructing a multi-mode rail transit timetable expansion network model. The basis for constructing the multi-mode rail transit timetable expansion network is to associate and expand the space lines and nodes of the topological network in the time dimension according to the arrival and departure times of the trains in the train timetable. The nodes of the static network are associated by the running routes and arrival and departure times of the trains, and the timetable expansion network is formed according to the train number.

[0032] Defining [0, T] as the operation interval of the multi-mode rail transit system, the application takes the minimum departure interval of the rail transit as a unit time interval, and assumes that the decision cycle is n unit time periods. In the timetable expansion network, the time interval [0, T] can be represented as a series of discrete time points {1t, 2t, …, nt}, where t = T / n. In order to simplify the representation, t can be ignored, so that in the Figure 3 In can be directly represented as {1, 2, …, n}. It is assumed that all times (in-car time, arrival time, departure time) can be obtained by t unit time, such as Figure 3 As shown in the figure, the horizontal direction represents stations a, b, c, and d, and the vertical axis represents discrete time points. At the same time, in order to represent the actual situation of passenger transfer, the Figure 2 Multi-mode rail transit network conversion schematic diagram, in Figure 3 Each time node is split, and the stations and transfer points are connected by transfer arcs.

[0033] The specific process of the multi-mode rail transit network timetable expansion is as follows:

[0034] Step 1: The time dimension is expanded according to all the physical station nodes of each train number on the timetable of each line, that is, the arrival and departure time labels of the trains are added to the physical nodes, the physical nodes are expanded by the time labels of different trains, and the arrival and departure stations of the trains are represented by expanding the physical nodes;

[0035] Step 2: The different expanded nodes are connected by the arrival and departure times of the trains in the timetable, and the arc sequence formed is the train running route. Through the train running route, the trains of the same train number on the same line are connected to the departure time and arrival time of the adjacent station nodes, and the interval running arc is formed.

[0036] Step 3: The arrival and departure times of the trains of different lines at the same physical station are connected by transfer arcs.

[0037] It is worth noting that since most of the existing rail stations do not have conflicts in receiving and sending trains, it is assumed that all train stations do not have conflicts in receiving trains, that is, different directions of incoming trains can be received at the same time. Figure 3Different arc categories represent different kinds of rail transit modes, but the length and slope of each arc do not represent the running time distance and speed, only represent the departure time of the train at a certain point and the arrival time at the corresponding point, and the arc direction in the figure conforms to the objective time change trend.

[0038] The application is based on effective path for timetable expansion, in order to ensure the loop-free property and connectivity property of the effective timetable expansion path, the effective path is defined as follows:

[0039] 1) The passenger cannot get off the train on the same line and then transfer to another line and then transfer back to this line during the trip, that is, the running arc belonging to the same line interval cannot appear discontinuously.

[0040] 2) The passenger cannot transfer twice or more times using the same transfer station during the trip, that is, the transfer arcs belonging to the same transfer station cannot appear at the same timetable expansion path at the same time.

[0041] 3) In general, the passenger will not take a "return trip" during the trip, that is, the timetable expansion nodes belonging to the same station cannot appear at the same timetable expansion path at the same time.

[0042] II. Multi-modal rail transit stochastic equilibrium distribution model construction stage

[0043] The multi-modal rail transit stochastic equilibrium distribution model construction includes two parts: the generalized cost function of the timetable expansion path and the giving of the timetable expansion network stochastic equilibrium distribution model. The timetable expansion path is obtained through the effective timetable expansion k-path search, and in the rail transit timetable expansion network, the network topology structure, the train reachable condition of line connection and the transfer time are the main factors considered in the timetable k-short path search.

[0044] (1) Generalized cost function based on timetable expansion path

[0045] The expanded network architecture is represented by the graph G(N, A, L η,i , T), wherein N represents the set of expanded station nodes; A represents the arc set; L η,i represents the set of i trains on the route L belonging to the route η, wherein η represents the route type, η = 1, 2, 3 for intercity line, city area line and subway line; and T represents the set of discrete times of trains at station nodes. r, s respectively represent the timetable expansion nodes on the network, r, s correspond to the two endpoints of the train running interval, and t1, t2 can be respectively represented as the arrival time of the train at the interval endpoints r, s. Therefore, the generalized cost of the timetable expansion path is composed of the following three parts:

[0046] 1) Passenger in-vehicle time cost This refers to the broad cost incurred by passengers on the train, including the time passengers spend on the train, T. IVT and the function of in-vehicle congestion The passenger time on the extended timetable route includes the time and cost of all sections of travel and stops. Therefore, for... The train stopped at And by Departure Arrival Passenger time in the car It can be calculated using the following formula, where Indicates that the train is in the running arc Time cost, Indicates that the train stops at [station name]. Time cost:

[0047]

[0048] Due to train capacity limitations, as the number of passengers increases, the crowding inside the train continuously intensifies until it reaches the maximum occupancy rate. Passenger discomfort gradually increases with the train's occupancy rate, leading to additional costs related to travel time. These costs rise sharply when there are no seats available. An amplification factor related to the flow rate on the timetable extension arc is used. This indicates situations where additional charges are incurred for on-board time:

[0049]

[0050] In the formula, express Throughflow on the timetable extended arc, C l Z represents the train's approved passenger capacity, ω represents the train's occupancy rate, and Z represents the train's full load factor. l This represents the number of seats on the train. In the formula, α... l γ and y are passenger perception factors. Since the train occupancy rate changes with the number of passengers getting on and off at stations, it is a dynamically updated variable.

[0051] Then consider the cost of passenger time on the train when it is crowded. It can be calculated using the following formula:

[0052]

[0053] 2) Passenger transfer fees This represents the time cost for a passenger transferring from line l to line l' at a timetable extension node r on a timetable-based extended route. Transfer costs are divided into single-system transfer costs and multi-system transfer costs, and include the transfer time cost. And the transfer penalty fee σ. In addition, when considering the transfer costs for the entire extended timetable route, each additional transfer incurs an extra fee beyond the transfer time, called the transfer penalty fee. The passenger's transfer cost can then be expressed by the following formula:

[0054]

[0055] In the formula, These represent the time cost for single-system and multi-system transfers at the timetable extension node r, respectively.

[0056] 3) Passenger fare expenditure Because different rail transit modes have different fare prices, the fare cost for passengers traveling within the same origin-destination (OD) mode will also be different depending on the type of rail transit they choose. The fare cost will change for each route if a passenger chooses any route or any train within the specified travel time. Therefore, the impact of fare costs on passenger route choices needs to be considered.

[0057]

[0058] In the formula, Indicates the running arc The running distance, ρ η This represents the fare rate for route segment η.

[0059] On the timetable extension network, within the allocated time period Δt, Generalized cost of the k-th timetable extension path The function is as follows:

[0060]

[0061] In the formula, The generalized cost representing the timetable extension arc. This represents a generalized cost function with capacity constraints on the timetable extension path. This indicates the association parameter between the timetable extension arc and the timetable extension path. The value is 1 when the timetable extension arc is within the timetable path, and 0 otherwise. β i This indicates the weight of each cost.

[0062] (2) Stochastic equilibrium assignment model of timetable extended network

[0063] In the process of passenger travel, the travel cost of passenger in the network is not only a random variable, but also related to the passenger flow of the path. In urban rail transit, due to the capacity limit of train, the passenger's stay on the train caused by the train congestion and overloading due to too large passenger flow will affect the cost of the path. Therefore, the traffic flow distribution model is used to build the traffic flow distribution of the timetable expansion network by using the idea of stochastic user equilibrium (SUE) distribution. In the state of stochastic user equilibrium, all selected paths between an OD pair do not have the same actual impedance, and the distribution flow on each selected path is equal to the product of the traffic volume between the OD pair and the selection probability of the path.

[0064]

[0065] wherein, denotes the kth timetable expansion path between o i d j denotes the distribution flow of the kth timetable expansion path between o denotes the passenger flow between o i d j denotes the selection probability of the path, since the selection probability is related to the perceived generalized cost of the path, and the perceived generalized cost of the path is related to the generalized cost of the timetable expansion arc and is a random variable, which is the SUE condition. When the network reaches the state of stochastic user equilibrium, the flow and the cost on the path satisfy the following relationship:

[0066]

[0067] wherein, θ>0 is a parameter for measuring the randomness of the network, representing the random characteristics of the cost perception of the passenger under the familiarity of the road network, and the larger θ represents the more familiar the passenger is to the timetable expansion path between the OD pair, and the smaller the randomness is. When θ→∞, the SUE condition is approximately the user equilibrium (UE) condition. According to formula (6), formula (8) can be converted into:

[0068]

[0069] The network dynamic distribution problem of passenger flow can be converted into the network passenger flow loading process by using the method of timetable expansion to expand the time dimension of the urban rail transit network. Therefore, the stochastic user equilibrium distribution optimization model of the timetable expansion network under the condition of train capacity limit is constructed as follows:

[0070]

[0071]

[0072] wherein, ​denotes the allocated flow on the kth schedule path, K rs,l denotes the train capacity of line l, v rs denotes the flow of schedule expansion arc a(r, s), denotes the set of valid paths between OD pair o

[0073] After the above model is constructed, the application of the model under the condition of operation interruption is performed by inputting the station and section numbers contained in the interruption section, the duration of the interruption accident, the operation measures maintained after the interruption occurs, including the train operation loop, the maximum passenger capacity of the train and the departure interval, to finally obtain the passenger flow allocation of each OD pair based on the valid path and the corresponding indicators. Assuming that the road network G has an interruption operation in the section (a, b) between T1-T2, the passenger flow classification rules are as follows:

[0074] Type 1: passenger flow entering the road network before T1 and passing through the interruption section (a, b) during T1-T2;

[0075] Type 2: passenger flow entering the road network during T1-T2 and the valid path between OD pairs does not contain the interruption section under normal circumstances;

[0076] Type 3: passenger flow entering the road network during T1-T2 and the valid path between OD pairs contains the interruption section under normal circumstances.

[0077] The specific algorithm steps are shown in Figure 4 :

[0078] Step 1: road network construction and initialization.

[0079] According to the interruption event, the interruption section, the interruption duration and the operation organization measures under the interruption, the initial road network topology structure is changed, the nodes and arcs related to the section (a, b) are deleted, and the schedule expansion network under the condition of operation interruption is constructed by the method of schedule expansion. Initialization, for all schedule expansion arcs in the road network the arc flow v rs = 0.

[0080] Step 2: calculate the travel cost of each section.

[0081] According to the definition of valid path in the foregoing, the k-shortest path algorithm (edge deletion method) based on Dijkstra algorithm is called to obtain the valid path set of OD pair o i d j and calculate the travel cost of each valid path

[0082] Step 3: calculate the valid path alternative probability.

[0083] The improved logit model​ wherein OD pair o i d j average impedance of all paths, calculated for OD pair o i d j alternative probability of each effective path

[0084] Step four: loading passenger flow.

[0085] For type 2 passenger flow, the passenger flow is directly loaded, and for type 1 and type 3 passenger flow, it is determined whether the passenger flow is reachable, if the passenger flow is not reachable, the passenger flow is out of the station, and if the passenger flow is reachable, the passenger flow is loaded according to the effective path set. Combined with the path alternative probability calculation formula (10), the path distribution flow and arc distribution flow of each step are obtained, and the path passenger flow, arc flow and path time are stored.

[0086] Step five: convergence judgment.

[0087] Convergence judgment condition:

[0088] ε is a preset convergence error, generally a sufficiently small positive number, if the convergence judgment condition is not met, return to step two, let n=n+1, update the road network travel cost, continue to distribute the passenger flow of n+1 node, otherwise go to step six.

[0089] Step six: result output.

[0090] The algorithm ends, and the passenger flow distributed on each effective path in the required period is counted The path flow and arc flow are used to calculate and count the interval line passenger flow, transfer passenger flow and other passenger flow indexes.

[0091] In summary, the present application can achieve the following beneficial effects:

[0092] 1) Based on the passenger flow distribution characteristics under the operation interruption state, by constructing a random equilibrium distribution model based on the timetable extended network, the constraints of train capacity on passenger flow distribution in the operation interruption network and the randomness of passenger effective path selection under congestion are considered, and the actual distribution of passengers on each effective path is more accurately described.

[0093] 2) Under the operation interruption condition, it is compatible with multiple transportation modes, and is helpful for emergency collaborative management among multi-modal rail transit;

[0094] 3) At the time of operation interruption, combined with the line interruption condition and passenger path selection, the real-time travel trajectory and distribution of passengers are more accurately and effectively described, which provides theoretical support and guidance for multi-modal rail transit emergency passenger flow distribution.

Claims

1. A regional multi-modal rail transit passenger flow dynamic allocation method under an operation interruption condition, characterized in that: Comprise the following contents: I. Construct regional multi-mode rail transit timetable extension network; II. Construct multi-mode rail transit network passenger flow distribution model: (1) Establish a generalized cost function based on timetable extension path: wherein denotes the generalized cost of a timetable extended arc, denotes the flow on a timetable extended arc ; denotes the association parameter of a timetable extended arc to a timetable extended path, which is 1 when the timetable extended arc is in the timetable path, otherwise 0; β i denotes the weight of each cost, denote the passenger in-vehicle time cost, the passenger interchange cost and the passenger fare expenditure, respectively; denotes the generalized cost function with capacity constraints on a timetable extended path; wherein: 1) the passenger's time in the vehicle fee is calculated as follows: Wherein: wherein, denotes the flow on the timetable extended arc , where: denotes the timetable extended nodes on the network, r, s correspond to the two endpoints of a train operating section, t1, t2 denote the time when the train arrives at the section endpoints r, s, respectively; C l denotes the authorized passenger capacity of the train, ω denotes the train full load rate, Z l denotes the number of seats on the train, α l and γ are passenger perception factors; denotes the time cost of the train stopping at ; 2) the passenger transfer fee is calculated by the following equation: In the formula, respectively represent the time cost of single-mode and multi-mode transfer at the schedule expansion node r; σ represents the transfer penalty cost. 3) the passenger fare expenditure is calculated as follows: wherein denotes the running distance of the running arc denotes the running distance of the running arc η denotes the fare rate of the route section η,i denotes the set of i train classes on the route L belonging to the route section (2) Establish a random equilibrium distribution model of timetable extension network: where A denotes the set of arcs, crs(x) denotes the generalized cost of running arc x, and 0 denotes the cost perception stochastic characteristic of passengers under the familiarity of the road network, denotes the assigned flow on the kth schedule path, K rs,l denotes the train capacity of line l, v rs denotes the flow of the schedule expansion arc a(r, s), denotes the set of effective paths between OD pairs, denotes the association parameter of the schedule expansion arc and the schedule expansion path; wherein: In the formula, represents o i d j passenger flow between, represents the effective path alternative probability; III. Under the condition of operation interruption, use the stations and interval numbers contained in the interruption interval, the duration of the interruption accident, and the operation measures maintained after the interruption, solve the random equilibrium distribution model of the timetable extension network, and finally obtain the passenger flow distribution of each OD pair based on the effective path and the corresponding indicators: Step 1, network construction and initialization, let the iteration number n = 1; Step 2, calculate the travel cost of each effective path; Step 3, calculate the effective path selection probability; Step 4, load the passenger flow on the effective path set according to the passenger flow classification type, calculate and store the path distribution flow and arc distribution flow at each step; Step 5, convergence judgment: if the convergence judgment condition is not met, return to step 2, let n = n + 1, update the road network travel cost, and continue to distribute the passenger flow of the node, otherwise go to step 6; Step 6, calculate the passenger flow on each effective path in the required period, and calculate the interval line passenger flow and transfer passenger flow indicators using path flow and arc flow.

2. The regional multi-modal rail transit passenger flow dynamic distribution method under operation interruption condition according to claim 1, characterized in that: The method for constructing a regional multi-mode rail transit timetable extension network is: (1) Network transformation of multi-mode rail transit network: 1) If there are multiple transportation modes between two nodes, add corresponding connecting lines between the two points, one connecting line corresponding to one transportation mode; 2) If there is a transfer process at a node in the multi-mode service network, use node splitting to handle it: separate each transportation mode at the transfer node, and represent each transportation mode with a new node and the connecting arc between the new nodes to represent the transfer of the transportation mode; (2) Timetable extension of the multi-mode rail transit network after network transformation: Step 1, time dimension expansion according to all physical station nodes of each train stop on the timetable: add train arrival and departure time labels to physical nodes, expand physical nodes through different train time labels, and represent train arrival and departure stations through expanded physical nodes; Step 2, connect different expanded nodes through train arrival and departure times to form train operation paths, and connect the departure time and arrival time of the same train on the same line through train operation paths to form interval operation arcs; Step 3, connect the arrival and departure times of different line trains at the same physical station through transfer arcs.

3. The regional multi-modal rail transit passenger flow dynamic distribution method under operation interruption condition according to claim 1, characterized in that: The effective path candidate probability is calculated as follows where denotes the OD pair o i d j average impedance of all paths; 0 denotes the cost perception stochastic characteristic of passengers under the familiarity of road network; denotes the set of effective paths between OD pair.

4. The regional multi-modal rail transit passenger flow dynamic distribution method under operation interruption condition according to claim 3, characterized in that: The convergence judgment condition is: Wherein, ε is a preset convergence error.

Images