Urban rail transit traffic distribution method and system in interconnection and intercommunication direct express train operation mode

Abstract

The invention provides an urban rail transit traffic distribution method and system in an interconnection and intercommunication direct express operation mode. The method comprises the following steps: reconstructing a line network structure; calculating K short effective paths; reconstructing direct express vehicle passenger flow distribution under the network; common passenger flow distribution under a conventional network; and superposition of passenger flow distribution results. Based on a line network structure, an operation mode and a space-time path of a train, and time and place of passenger flow entering and leaving stations, a passenger flow-traffic flow coupling relation in a travel process is fully considered, a space-time distribution state of passenger flow is restored, high-precision, fine-grained and multi-dimensional passenger flow statistical data is output, urban rail transit passenger flow distribution is realized, and the urban rail transit passenger flow distribution efficiency is improved. And effective support is provided for daily passenger flow analysis of operating enterprises.

Description

Technical Field

[0001] The present invention belongs to the field of rail transit planning and management, and in particular, relates to an urban rail transit flow distribution method and system under an interconnected express train operation mode. Background Art

[0002] Networked operation is the development trend of urban rail transit, and the interconnected express train operation mode is one of the specific manifestations of networked operation. Among them, interconnection refers to the connection of track paths between lines, and express trains refer to cross-line operation of trains and stop at major stations. Compared with ordinary trains, passengers of interconnected express trains do not need to get off and transfer, and the transfer connection is more efficient. On the other hand, they can reach the destination station faster, saving travel time. Taking Chongqing Rail Transit as an example, the interconnected express trains of Line 4-Loop Line-Line 5 are currently running all day, passing through 29 stations (transfer stations are not counted repeatedly), stopping at 12 stations, and skipping 17 stations. The whole journey saves more than 30 minutes, and the maximum train load rate during peak hours is close to 100%. Intuitive data shows that the realization of the interconnected express train operation mode has not only improved the transfer efficiency and enhanced the service capacity of urban rail transit, but also provided an advanced and practical direction for the development of networked operations.

[0003] However, the development of networking requires the continuous refinement of passenger flow control, and enterprise managers need to master the spatiotemporal distribution of passenger flow under the interconnected express train operation mode. Express trains have the characteristics of fast and direct access. Passengers at covered stations are more inclined to take express trains. Passengers taking interconnected express trains not only do not need to get off and transfer, but can also take express trains and then transfer to ordinary trains on the same line to reach their destination faster, breaking through the traditional idea of ​​flow distribution without transfer on the same line. The traditional urban rail transit flow distribution method can effectively solve the passenger flow analysis problem under the conventional operation mode, but it has exposed certain limitations in the face of the emerging interconnected express train operation mode, which is mainly reflected in the following three aspects: first, there is a lack of segmentation of express train passengers; second, the construction of express train advantage paths is ignored, and the impact of express train (including express trains at major stations) operation advantages on passengers is not highlighted; third, the transfer passenger flow statistics strategy is used to ignore the impact of no transfer.

[0004] In view of the interconnected express train operation mode, in order to improve the shortcomings of traditional methods, it is necessary to segment passengers based on passenger flow OD data, and fully consider the selection advantages of express trains when allocating passengers, and adjust the passenger flow statistics strategy based on the characteristics of express trains that do not require passengers to get off and transfer, so as to improve the accuracy of the temporal and spatial distribution of passenger flow and provide effective passenger flow data support for operating companies.

[0005] The patent document "Time Domain Optimization Method for Urban Rail Transit Based on Passenger Flow Spatiotemporal Distribution Deduction" (CN115689085A) introduces the concepts of characteristic passenger flow and characteristic direction, which becomes the link for coupling vehicle flow and driving, and realizes accurate and flexible compilation. However, it is aimed at transfer situations, does not consider the construction of advantageous paths, and uses transfer passenger flow statistics strategy.

[0006] The patent document "A method, system, device and storage medium for optimizing cross-line operations of rail transit" (CN113822461A) proposes a method for optimizing cross-line operations of one-way direct rail transit, but it selects a suitable cross-line rail operation plan by studying the costs of passenger travel and corporate operations under different rail operation plans, without considering the nature of passengers and the construction of direct express advantage paths.

[0007] Although the patent documents "A method, system, equipment and storage medium for optimizing cross-line operations of rail transit" (CN202110849045) and the patent document "A method for compiling a multi-year integrated route plan considering the timing of line network construction" (CN202311148575) propose relevant technologies for cross-line operations or passenger flow distribution of urban rail transit, there are no modeling examples of interconnected operating lines, and physical paths are used as the basis for transfers. Therefore, it is impossible to describe the relationship between interconnected operating lines and ordinary operating lines and distinguish the selection priority between express trains and ordinary trains, and ignore the cross-line direct travel and transfer travel of express trains. Summary of the invention

[0008] In view of the defects in the prior art, the object of the present invention is to provide a method and system for urban rail flow distribution under an interconnected express train operation mode.

[0009] According to the present invention, a method for distributing urban rail traffic in an interconnected express train operation mode includes:

[0010] Step S1: Reconstruct the network structure to obtain a directed network G;

[0011] Step S2: Calculate K shortest effective paths in the directed network;

[0012] Step S3: Allocate direct express passenger flow under the directed network G according to K short effective paths;

[0013] Step S4: Allocate the remaining ordinary passenger flow under the conventional network;

[0014] Step S5: According to the direct express passenger flow and the remaining ordinary passenger flow, the passenger flow allocation results are superimposed to obtain the spatiotemporal distribution state.

[0015] Preferably, in step S1, according to the stations that the interoperable express trains pass through, a virtual express train line is added on the basis of the conventional network to reconstruct the network structure of the urban rail transit.

[0016] The directed network G0 (L0, A0, N0) represents the conventional network of urban rail transit;

[0017] Among them, L0, A0, and N0 represent the lines, sections, and station collections in the conventional network, respectively.

[0018] The section includes the internal connection section of the transfer station generated in the transfer direction, and the directed network G (L, A, N) represents the reconstructed line network structure;

[0019] Among them, L = L0 ∪ L1, A = A0 ∪ A1, N = N0 ∪ N1;

[0020] L1, A1, and N1 represent the virtual line, virtual section, and virtual station set formed by the stopover of the express train respectively;

[0021] L, A, and N represent the reconstructed line, section, and station set, respectively;

[0022] The maximum index of N is n, where n represents the total number of stations;

[0023] The interval running time of the virtual line is less than the interval running time of the corresponding section of the line in the conventional network, and is obtained from the direct express train schedule.

[0024] In the step S3, based on the reconstructed network structure, the interconnected express train passenger flow is allocated in combination with the space-time path.

[0025] The comprehensive path cost c of the kth short circuit k and passenger flow OD comprehensive cost threshold c threshold The calculation formula is:

[0026] c k =f×t k , c threshold =min{c min ×(1+a),c min ×b};

[0027] Among them, f represents the transfer penalty coefficient, which is related to the line network structure and the number of transfers;

[0028] t k represents the travel cost of the kth short circuit;

[0029] c min Represents the comprehensive cost of the shortest path;

[0030] a represents the relative threshold of comprehensive cost, a∈(0,1);

[0031] b represents the absolute threshold of comprehensive cost, unit is min, usually b∈[10,30];

[0032] The MNL model is used to calculate the probability of each path being selected and matched with the timetable.

[0033] The probability Z of the kth short circuit being selected k The calculation formula is:

[0034] V k =exp[-(c k -c min ) 2 / (2σ 2 ))],Z k =V k / (∑ k∈K V k );

[0035] Among them, V k represents the utility value of the kth short circuit;

[0036] K is the total number of paths;

[0037] σ represents the probability adjustment coefficient.

[0038] Preferably, in step S2, the A* algorithm and the queue optimization algorithm are used to calculate the K shortest effective paths (K≤5) between the passenger flow ODs of the conventional network and the reconstructed line network structure, the circuitous paths are eliminated according to the recorded stations passed by the paths, and the effective paths are selected in combination with the comprehensive path costs and their selection probabilities are calculated, including:

[0039] Step S2.1: Initialize the directed network G(L,A,N) and the reverse directed network

[0040] in, Represents the section of reverse reconstruction.

[0041] Step S2.2: According to Combined with the queue optimization algorithm, the reverse shortest path (d∈N) between the terminal d and the station w in the passenger flow OD is calculated, and the cost is h(w), where h(w) represents the shortest path cost from the station w to the terminal d.

[0042] Step S2.3: Let queue W (W∈N) be the queue of subsequent node stations of starting point O, initialize f(w)=∞, w∈W, and sort in ascending order according to f(w), where f(w) is the evaluation function of the A* algorithm.

[0043] Step S2.4: Calculate the cost g(w) between the starting point O and the subsequent node station w based on G(L,A,N), and let f(w)=g(w)+h(w).

[0044] Among them, if it is determined that w = d, then delete the node station w from the queue W, add the subsequent node stations of the node station w, w = w + 1, let f(w) = ∞, output the k-th shortest path, and k = k + 1; if w < d, then continue to execute step S2.4;

[0045] Judge the condition of k = K. If so, the K-shortest paths of the passenger flow OD are output; if not, sort f(w) in ascending order and continue to execute step S2.4.

[0046] Preferably, in step S4, based on the conventional network, search for the spatio-temporal path based on the passenger flow OD and the timetable, and allocate the remaining ordinary passenger flow in combination with the spatio-temporal path.

[0047] The selection of the spatio-temporal path includes:

[0048] p lastest Step: Search for the latest valid spatio-temporal path p of the passenger's final arrival line according to the passenger flow OD lastest ;

[0049] p earliest Step: Search for the earliest valid spatio-temporal path p of the passenger's departure line according to the passenger flow OD earliest ;

[0050] Set search step: According to p lastest and p earliest Search for the set of valid spatio-temporal paths.

[0051] The p lastest step includes: Step S4.1.1: Let i = 1, t1 = oT, t2 = oT j ;

[0052] Among them, i represents the i-th station;

[0053] t1 represents the time when the passenger swipes the card to get off or transfers and gets off at the current spatio-temporal arc;

[0054] oT represents the passenger's departure time;

[0055] j represents the j-th station;

[0056] represents the i-th station in the reverse set of stations passed by the OD path, and the maximum index is m;

[0057] t2 represents the in-station walking time of the passenger at the current spatio-temporal arc;

[0058] oT j represents the out-station walking time of station j.

[0059] Step S4.1.2: When min|t1 - t2 - aT r,j | and When the train r is arrive The spacetime arc of p is added lastest ;

[0060] Among them, S r represents the set of stations where train r stops;

[0061] aT r,j represents the arrival time of train r at station j.

[0062] Step S4.1.3: If i+1=m, complete p lastest In the space-time arc of passenger behavior, execute p earliest Steps: If i+1≠m, then i=i+1, t1=dT r,j , t2=tT j , repeat step S4.1.2;

[0063] Among them, dT r,j represents the departure time of train r at station j;

[0064] T j represents the transfer travel time at station j.

[0065] The p earliest The steps include: Step S4.2.1: let i=1, t1=eT, t2=eT j ;

[0066] Among them, eT represents the time of entering the station;

[0067] Indicates the i-th station in the forward set of stations that the path passes through, with a maximum index of m;

[0068] eT j represents the entry and travel time of station j.

[0069] Step S4.2.2: When min|dT r,j -t1-t2|and When the train r is arrive The spacetime arc of p is added earliest .

[0070] Step S4.2.3: If i+1=m, complete p earliest In the passenger behavior space-time arc, execute the step set search step; if i+1≠m, then i=i+1, t1=aT r,j , t2=tT j , repeat step S4.2.2.

[0071] In the set search step, if p lastest =p earliest , then the effective space-time path set of passengers P = {p lastest}={p earliest}; if p lastest ≠p earliest , then the effective space-time path set of passengers P = [p lastest ,p earliest ].

[0072] Preferably, in step S5, the reconstructed line network structure and the passenger flow distribution result of the conventional network are superimposed to generate the spatiotemporal distribution state of the line network passenger flow, including:

[0073] Step S5.1: Initialize the passenger state, sort the passenger set Q (q∈Q) in ascending order according to the arrival time, and let x q =1,y q =1;

[0074] Where Q represents the set of passengers in the road network, q∈Q;

[0075] x q represents the space-time path index selected by passenger q, 0 <x q <P q ;

[0076] y q Indicates the route index of the space-time path selected by passenger q, 0 <y q <l q,p ;

[0077] l q,p represents the space-time arc of the lth train segment selected by passenger q on space-time path p, l q,p ∈L q,p ;

[0078] L q,p Represents the set of train space-time arcs that passenger q chooses for space-time path p.

[0079] Step S5.2: Initialize time parameters, set simulation time t as the operation start time, time step t step .

[0080] Step S5.3: Update the space-time state according to the passenger space-time path, let i = x q , j = y q , t1=sT q,i,j ;

[0081] Among them, sT q,p,l represents the starting time of the lth train space-time arc in passenger q’s space-time path p.

[0082] Step S5.4: Update the train congestion, let e q,p,l =v q,p,l / b q,p,l ;

[0083] Among them, e q,p,l represents the train congestion of the lth train space-time arc in the space-time path p of passenger q;

[0084] v q,p,l represents the train passenger capacity of the lth train space-time arc in the space-time path p of passenger q;

[0085] b q,p,l Represents the train capacity of the lth train space-time arc in the space-time path p of passenger q.

[0086] Step S5.5: Traverse the passenger set Q. If t>t1 exists, execute step S5.6; if t>t1 does not exist, execute step S5.8.

[0087] Step S5.6: If the train load factor is less than the expected load factor e q,p,l <e exp , or the train load factor is greater than the expected load factor and is the latest reachable space-time path e q,p,l ≥e exp And x q =P q , then the passenger gets on the bus v q,p,l =v q,p,l +1, execute step S5.7; if the train load factor is greater than the expected load factor and it is not the latest reachable time-space path e q,p,l ≥e exp And x q <P q , then x q =x q +1, return to step S5.3.

[0088] Step S5.7: If x q = l q,p , then exclude q from Q; if x q ≠l q,p , then y q =y q +1;

[0089] Among them, L represents the total number of routes, l∈L.

[0090] Step S5.8: Update the simulation time, set t = t + t step If the simulation time t is greater than or equal to the operation end time, the simulation ends; if the simulation time t is greater than the operation end time, return to step S5.3.

[0091] According to the present invention, a city rail transfer flow system under an interconnected express train operation mode is provided, comprising:

[0092] Module M1: Reconstruct the network structure to obtain the directed network G;

[0093] Module M2: Calculate K shortest valid paths in a directed network;

[0094] Module M3: Allocate direct express passenger flow under the directed network G according to K shortest effective paths;

[0095] Module M4: Allocate the remaining ordinary passenger flow under the conventional network;

[0096] Module M5: Based on the direct express passenger flow and the remaining ordinary passenger flow, the passenger flow allocation results are superimposed to obtain the spatiotemporal distribution state.

[0097] Preferably, in the module M1, according to the stations where the express trains are interconnected, a virtual express train line is added on the basis of the conventional network to reconstruct the network structure of the urban rail transit.

[0098] The directed network G0 (L0, A0, N0) represents the conventional network of urban rail transit;

[0099] Among them, L0, A0, and N0 represent the lines, sections, and station collections in the conventional network, respectively.

[0100] The section includes the internal connection section of the transfer station generated in the transfer direction, and the directed network G (L, A, N) represents the reconstructed line network structure.

[0101] Among them, L = L0 ∪ L1, A = A0 ∪ A1, N = N0 ∪ N1;

[0102] L1, A1, and N1 represent the virtual line, virtual section, and virtual station set formed by the stopover of the express train respectively;

[0103] L, A, and N represent the reconstructed line, section, and station set, respectively;

[0104] The maximum index of N is n, where n represents the total number of stations;

[0105] The interval running time of the virtual line is less than the interval running time of the corresponding section of the line in the conventional network, and is obtained from the direct express train schedule.

[0106] In the module M3, based on the reconstructed network structure, the interconnected express passenger flow is allocated in combination with the space-time path.

[0107] The comprehensive path cost c of the kth short circuit k and passenger flow OD comprehensive cost threshold c threshold The calculation formula is:

[0108] c k =f×t k , c threshold =min{c min ×(1+a),c min ×b};

[0109] Among them, f represents the transfer penalty coefficient, which is related to the line network structure and the number of transfers;

[0110] t k represents the travel cost of the kth short circuit;

[0111] c min Represents the comprehensive cost of the shortest path;

[0112] a represents the relative threshold of comprehensive cost, a∈(0,1);

[0113] b represents the absolute threshold of comprehensive cost, unit is min, usually b∈[10,30].

[0114] The MNL model is used to calculate the probability of each path being selected and matched with the timetable.

[0115] The probability Z of the kth short circuit being selected k The calculation formula is:

[0116] V k =exp[-(c k -c min ) 2 / (2σ 2 ))],Z k =V k / (∑ k∈K V k );

[0117] Among them, V k represents the utility value of the kth short circuit;

[0118] K is the total number of paths;

[0119] σ represents the probability adjustment coefficient.

[0120] Preferably, the module M2 uses the A* algorithm and the queue optimization algorithm to calculate the K shortest effective paths (K≤5) between the passenger flow ODs of the conventional network and the reconstructed line network structure, eliminates the detour paths according to the recorded stations passed by the paths, and selects the effective paths in combination with the comprehensive path costs and calculates their selection probabilities, including:

[0121] Module M2.1: Initialize the directed network G(L,A,N) and the reverse directed network

[0122] in, Indicates the cross-section of reverse reconstruction.

[0123] Module M2.2: According to Combine the queue optimization algorithm to calculate the reverse shortest path from the end point d to the station w in the passenger flow OD (d ∈ N), with the cost h(w), where h(w) represents the shortest path cost from the station w to the end point d.

[0124] Module M2.3: Let the queue W (W ∈ N) be the queue of subsequent node stations of the starting point O, initialize f(w) = ∞, w ∈ W, and sort in ascending order of f(w), where f(w) is the evaluation function of the A* algorithm.

[0125] Module M2.4: Calculate the cost g(w) of the starting point O and the subsequent node station w according to G(L, A, N), and let f(w) = g(w) + h(w).

[0126] Among them, judge if w = d, then delete the node station w from the queue W, add the subsequent node stations of the node station w, w = w + 1, let f(w) = ∞, output the k-th shortest path, k = k + 1; if w < d, then repeatedly trigger Module M2.4;

[0127] Judge the condition of k = K. If so, the output of the K-shortest path of the passenger flow OD is completed; if not, sort in ascending order of f(w) and repeatedly trigger Module M2.4.

[0128] Preferably, in the module M4, based on the conventional network, search for the spatio-temporal path based on the passenger flow OD and the timetable, and allocate the remaining ordinary passenger flow in combination with the spatio-temporal path.

[0129] The selection of the spatio-temporal path includes:

[0130] p lastest Module: Search for the latest valid spatio-temporal path p of the passenger's final arrival line according to the passenger flow OD lastest ;

[0131] p earliest Module: Search for the earliest valid spatio-temporal path p of the passenger's departure line according to the passenger flow OD earliest ;

[0132] Set search module: According to p lastest and p earliest Search for the set of valid spatio-temporal paths.

[0133] The p lastest module includes: Module M4.1.1: Let i = 1, t1 = oT, t2 = oT j ;

[0134] Among them, i represents the i-th station;

[0135] t1 represents the time when the passenger swipes the card or transfers to another bus at the current time-space arc;

[0136] oT represents the passenger’s departure time;

[0137] j represents the jth station;

[0138] It represents the i-th station in the reverse set of stations that the OD path passes through, and the maximum index is m;

[0139] t2 represents the passenger’s travel time in the station in the current space-time arc;

[0140] oT j represents the departure travel time of station j.

[0141] Module M4.1.2: When min|t1-t2-aT r,j | and When the train r is arrive The spacetime arc of p is added lastest ;

[0142] Among them, S r represents the set of stations where train r stops;

[0143] aT r,j represents the arrival time of train r at station j.

[0144] Module M4.1.3: If i+1=m, complete p lastest The space-time arc of passenger behavior triggers p earliest module; if i+1≠m, then i=i+1, t1=dT r,j , t2=tT j , trigger module M4.1.2;

[0145] Among them, dT r,j represents the departure time of train r at station j;

[0146] T j represents the transfer travel time at station j.

[0147] The p earliest The modules include: Module M4.2.1: Let i = 1, t1 = eT, t2=eT j ;

[0148] Among them, eT represents the time of entering the station;

[0149] Indicates the i-th station in the forward set of stations that the path passes through, with a maximum index of m;

[0150] eT j represents the entry and travel time of station j.

[0151] Module M4.2.2: When min|dT r,j -t1-t2|and When the train r is arrive The spacetime arc of p is added earliest .

[0152] Module M4.2.3: If i+1=m, complete p earliest The passenger behavior space-time arc triggers module M4.3; if i+1≠m, then i=i+1, t1=aT r,j , t2=tT j , trigger module M4.2.2.

[0153] In the set search module, if p lastest =p earliest , then the effective space-time path set of passengers P = {p lastest}={p earliest}; if p lastest ≠p earliest , then the effective space-time path set of passengers P = [p lastest ,p earliest ].

[0154] Preferably, in module M5, the reconstructed line network structure and the passenger flow distribution result of the conventional network are superimposed to generate the spatiotemporal distribution state of the line network passenger flow, including:

[0155] Module M5.1: Initialize the passenger state, sort the passenger set Q(q∈Q) in ascending order according to the arrival time, and let x q =1,y q =1;

[0156] Where Q represents the set of passengers in the road network, q∈Q;

[0157] x q represents the space-time path index selected by passenger q, 0 <x q <P q ;

[0158] y q Indicates the route index of the space-time path selected by passenger q, 0 <y q <l q,p ;

[0159] l q,p represents the space-time arc of the lth train segment selected by passenger q on space-time path p, l q,p ∈Lq,p ;

[0160] L q,p Represents the set of train space-time arcs that passenger q chooses for space-time path p.

[0161] Module M5.2: Initialize time parameters, set simulation time t as the start time of operation, time step t step .

[0162] Module M5.3: Update the space-time state according to the passenger space-time path, let i = x q , j = y q , t1=sT q,i,j ;

[0163] Among them, sT q,p,l represents the starting time of the lth train space-time arc in passenger q’s space-time path p.

[0164] Module M5.4: Update train congestion, let e q,p,l =v q,p,l / b q,p,l ;

[0165] Among them, e q,p,l represents the train congestion of the lth train space-time arc in the space-time path p of passenger q;

[0166] v q,p,l represents the train passenger capacity of the lth train space-time arc in the space-time path p of passenger q;

[0167] b q,p,l Represents the train capacity of the lth train space-time arc in the space-time path p of passenger q.

[0168] Module M5.5: Traverse the passenger set Q. If t>t1 exists, trigger module M5.6; if t>t1 does not exist, trigger module M5.8.

[0169] Module M5.6: If the train load factor is less than the expected load factor e q,p,l <e exp , or the train load factor is greater than the expected load factor and is the latest reachable space-time path e q,p,l ≥e exp And x q =P q , then the passenger gets on the bus v q,p,l =v q,p,l +1, trigger module M5.7; if the train load factor is greater than the expected load factor and is not the latest reachable space-time path e q,p,l ≥e exp And x q <P q , then x q =x q+1, trigger module M5.3.

[0170] Module M5.7: If x q = l q,p , then exclude q from Q; if x q ≠l q,p , then y q =y q +1;

[0171] Among them, L represents the total number of routes, l∈L.

[0172] Module M5.8: Update simulation time, set t = t + t step If the simulation time t is greater than or equal to the operation end time, the simulation ends; if the simulation time t is greater than the operation end time, module M5.3 is triggered.

[0173] Compared with the prior art, the present invention has the following beneficial effects:

[0174] 1. The present invention is based on the network structure, the operation mode and space-time path of the train, as well as the time and place of passenger flow entering and leaving the station. When allocating traffic, the present invention fully considers the selection advantage of the express train and the coupling relationship between passenger flow and vehicle flow during the travel process, and highlights the advantage of the express train in terms of fast and direct access.

[0175] 2. Aiming at the interoperable express train operation mode, the present invention subdivides passengers on the basis of passenger flow OD data, allocates passenger flow in steps, and simulates the spatiotemporal distribution of express passenger flow and ordinary passenger flow in turn, thereby simulating diversified travel under the conditions of interoperable express trains and improving the accuracy of the spatiotemporal distribution of passenger flow.

[0176] 3. The present invention adjusts the passenger flow statistics strategy according to the characteristic of express trains that passengers do not need to get off or transfer, outputs high-precision, fine-grained and multi-dimensional passenger flow statistics, realizes the passenger flow distribution of urban rail transit, and provides effective support for the daily passenger flow analysis of operating companies. BRIEF DESCRIPTION OF THE DRAWINGS

[0177] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0178] Figure 1 This is a schematic diagram of the technical route of the urban rail transit traffic distribution method.

[0179] Figure 2 This is a schematic diagram of the reconstructed network based on the stops of express trains.

[0180] Figure 3 This is a flowchart of searching for K shortest paths using the A* algorithm.

[0181] Figure 4It is a three-dimensional schematic diagram of the space-time path based on passenger flow OD and timetable.

[0182] Figure 5 It is a two-dimensional schematic diagram of the space-time path based on passenger flow OD and timetable.

[0183] Figure 6 Schematic diagram of passenger flow allocation process in heterogeneous networks. DETAILED DESCRIPTION

[0184] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0185] Interconnection lines are a network operation concept that will not change the scale of line network construction. It is a cross-line operation achieved by combining stations on existing ordinary lines. The purpose of the present invention is to overcome the defects of the prior art and provide a method for allocating passenger flow in urban rail transit under the interconnection express train operation mode. The technical route is as follows: Figure 1 As shown, the following steps are included:

[0186] Step S1: reconstructing the wire mesh structure;

[0187] Specifically, according to the stations where the express trains pass through, virtual express train lines are added on the basis of the conventional network to reconstruct the urban rail transit network structure;

[0188] In more preferred examples, the network structure is reconstructed based on the express train stop stations, and the conventional network of urban rail transit is represented by a directed network G0 (L0, A0, N0);

[0189] Among them, L0, A0, and N0 represent the line, section, and station set in the conventional network respectively;

[0190] The stations of the virtual express line need to correspond to the stations where the express train stops. Since the stations of the virtual express line can connect to the regular line, it is necessary to build the transfer relationship between all the stations where the express train stops and the stations on the regular line. The section includes the internal connection section of the transfer station generated by the transfer direction, thus forming a connected directed network. The reconstructed network structure is represented by the directed network G(L,A,N);

[0191] Among them, L = L0 ∪ L1, A = A0 ∪ A1, N = N0 ∪ N1;

[0192] L1, A1, and N1 represent the virtual line, virtual section, and virtual station set formed by the stopover of the express train respectively;

[0193] L, A, and N represent the reconstructed line, section, and station set, respectively;

[0194] The maximum index of N is n, where n represents the total number of stations. The reconstructed network diagram is as follows: Figure 2 shown.

[0195] The interval running time of the virtual line of the express train is less than the interval running time of the corresponding section of the regular line, and should be obtained directly from the express train timetable. The starting and ending stations of the virtual line are determined by the starting and ending points of the interconnected train route, and the intermediate stations of the virtual line are determined by the interconnected express train stop stations. The virtual line stations and the ordinary line stations form virtual transfer stations. This ensures that after adding the virtual line, the entire line network is a connected network.

[0196] Step S2: Calculate K short effective paths;

[0197] Specifically, the K shortest effective paths (K≤5) between each passenger flow OD in the conventional network and the reconstructed network are calculated respectively;

[0198] In more preferred examples, K shortest paths are searched based on travel time costs. In urban rail transit networks, interval operating time and stop time are key factors that directly affect travel efficiency. Express trains have the advantages of convenience and speed, which are mainly reflected in the shortening of interval operating time. For passengers, it can shorten the travel time. Therefore, it is necessary to quantify the characteristics of virtual line express trains in the modeling process. Urban rail transit passengers are highly sensitive to travel time, and express trains further amplify this feature. In order to highlight the selection advantage of express trains, the interval operating time and stop time are used as the path cost of express trains. It is stipulated that virtual lines only run interconnected express trains, and conventional lines only run ordinary trains. The time parameters of express trains and ordinary trains represent the path impedance of virtual lines and conventional lines respectively, thereby distinguishing the length of the running time of express trains and ordinary trains, and highlighting the advantages of express trains in fast and direct access.

[0199] The A* (A-Star) algorithm and the queue optimization (Shortest Path Faster Algorithm) algorithm are used to calculate the K shortest paths of the network, and then the circuitous paths are directly eliminated according to the recorded stations passed by the paths. Finally, the effective paths are selected based on the comprehensive path costs and their selection probabilities are calculated.

[0200] by Figure 3 For example, the A* algorithm (A-Star algorithm) can quickly calculate the K shortest paths of a directed network, including the following steps:

[0201] Step S2.1: Initialize the directed network G(L,A,N) and the reverse directed network in Indicates the cross-section of reverse reconstruction.

[0202] Step S2.2: According to Combined with the queue optimization algorithm, calculate the reverse shortest path from the end point d to other stations w in the network for a certain passenger flow OD (d ∈ N), with the cost being h(w), where h(w) represents the shortest path cost from station w to the end point d.

[0203] Step S2.3: Let the queue W (W ∈ N) be the queue of subsequent node stations of the starting point O. Initialize f(w) = ∞, w ∈ W, and sort in ascending order of f(w). f(w) is the evaluation function of the A* algorithm and can be understood as the cost of the path passing through station w between passenger flow OD in the scenario of K - shortest path search.

[0204] Step S2.4: Calculate the cost g(w) of the starting point O and other subsequent node stations w according to G(L, A, N), let f(w) = g(w) + h(w), and judge that if w < d, then delete the node station w from the queue W, add the subsequent node stations of the node station w, w = w + 1, let f(w) = ∞, and continue to execute Step S2.4;

[0205] If w = d (i.e., the current node has extended to the end point), then output the k - th shortest path, k = k + 1, and judge the condition of k = K. If so, the K - shortest paths of the passenger flow OD are output; if not, sort in ascending order of f(w) and continue to execute Step S2.4.

[0206] Step S3: Allocate the through - express passenger flow under the reconstructed network;

[0207] The through - express trains with interconnection and interoperability enrich the transfer behaviors of passengers. Both cross - line direct and in - line transfer break through the traditional transfer mode of operation. Therefore, perform step - by - step flow allocation and simulate the spatio - temporal distribution of through - express passenger flow and ordinary passenger flow in sequence. Specifically, on the basis of the reconstructed network, combine spatio - temporal path allocation to allocate the through - express passenger flow with interconnection and interoperability;

[0208] In more preferred examples, under the reconstructed network, since the transportation speed of the through - express train is much greater than that of the ordinary train, when the cost of the shortest path is much less than the costs of other k (k > 1) shortest paths, passengers are more inclined to choose the shortest path for travel. Therefore, passengers who meet the conditions for taking the through - express train will be allocated in this stage, and then effective paths are screened based on the comprehensive path cost. At the same time, considering that the virtual line of the through - express train greatly reduces the travel time, generally, passengers will actively exclude circuitous paths when traveling in the urban rail transit system, that is, it is impossible to return from the front station to the departure - direction station. Therefore, the interference of the virtual line path of the through - express train should be excluded during the screening of effective paths.

[0209] However, there is a special situation in the interconnected express train operation mode, that is, some passengers take the express train to the stop station of the destination line, and then transfer to the ordinary train of the line to reach the destination. This special travel scenario cannot be reproduced using the conventional network. However, under the reconstructed network, the virtual line of the express train is one of the extended lines of the line network, and is in parallel with other conventional lines. It can effectively simulate the competitive relationship between the express train and the ordinary train, thereby reproducing the special travel scenario.

[0210] If the comprehensive cost of the other k (k>1) short circuits is greater than the passenger flow OD comprehensive cost threshold, then the path is an invalid path. The comprehensive cost of the k-th short circuit path c k and passenger flow OD comprehensive cost threshold c threshold The calculation formula is as follows:

[0211] c k =f×t k ;

[0212] c threshold =min{c min ×(1+a),c min ×b};

[0213] Among them, f represents the transfer penalty coefficient, which is related to the line network structure and the number of transfers;

[0214] t k represents the travel cost of the kth short circuit;

[0215] c min Represents the comprehensive cost of the shortest path;

[0216] a represents the relative threshold of comprehensive cost, a∈(0,1);

[0217] b represents the absolute threshold of comprehensive cost, unit is min, usually b∈[10,30].

[0218] Based on the effective paths between passenger flow OD, the MNL (Multinominal Logit Model) model can be used to calculate the selection probability of each path. However, the selection probability only represents the passenger's preference, and the actual travel path needs to be matched with the timetable. The selection probability Z of the kth short circuit k The calculation formula is as follows:

[0219] V k =exp[-(c k -c min ) 2 / (2σ 2 ))];

[0220] Z k =V k / (∑k∈K V k );

[0221] Among them, V k represents the utility value of the kth short circuit;

[0222] K is the total number of paths;

[0223] σ represents the probability adjustment coefficient.

[0224] Step S4: Allocate normal passenger flow under the conventional network;

[0225] Specifically, on the basis of the conventional network, the remaining ordinary passenger flow is allocated in combination with the time-space path;

[0226] In more preferred examples, under the conventional network, the passengers and express train schedules that have been allocated under the reconstruction network are no longer considered, that is, the allocation of ordinary passengers under the ordinary operation mode is completed. Based on the search for the spatiotemporal path between the passenger flow OD and the schedule, the effective path between the above passenger flow OD can only represent a physically accessible path in space. However, urban rail transit is "people follow the train", so it is necessary to take a train with a fixed time to reach the destination. The physically accessible path and the train schedule jointly affect the passenger's travel spatiotemporal path.

[0227] Figure 4 and Figure 5 The three-dimensional schematic diagram and the two-dimensional schematic diagram respectively describe the travel space-time path of passengers who transfer once.

[0228] Among them, A, B, C, D, E, and F are spatial positions, representing the gate of the departure station, the platform of the departure station, the platform of the transfer station for changing outbound lines, the platform of the transfer station for changing inbound lines, the platform of the terminal station, and the gate of the terminal station. The space-time path is composed of continuous space-time arcs, as shown by space-time arcs ① to ⑦. ①, ②, ④, ⑤, and ⑦ represent six types of passenger behaviors, namely, entering the station, waiting at the starting point, running for transfer, waiting for transfer, and getting off the station, which are called behavioral space-time arcs, and they vary from passenger to passenger; ③, ⑥, and ⑧ are collectively referred to as train space-time arcs, which are determined by the interval running time and stop time in the timetable. The influence of train congestion is considered when allocating traffic. When the train congestion reaches a certain limit, passengers may not be able to take the current train due to train congestion and need to queue at the platform to wait for the next train. The corresponding train space-time arc ⑧ represents the accessible train that passengers miss due to congestion and other reasons during the waiting process. The steps for space-time path selection are as follows:

[0229] p lastest Steps: Search for the latest effective space-time path of the terminal route according to the passenger flow OD, including:

[0230] Step S4.1.1: Let i = 1, t1 = oT, t2=oT j ;

[0231] Among them, i represents the i-th station;

[0232] t1 represents the time when the passenger swipes the card or transfers to another bus at the current time-space arc;

[0233] oT represents the passenger’s departure time;

[0234] j represents the jth station;

[0235] It represents the i-th station in the reverse set of stations that the OD path passes through, and the maximum index is m;

[0236] t2 represents the passenger’s travel time in the station in the current space-time arc;

[0237] oT j represents the departure travel time of station j.

[0238] Step S4.1.2: When min|t1-t2-aT r,j | and Take the train from arrive The spacetime arc of p is added lastest ;

[0239] Among them, S r represents the set of stations where train r stops;

[0240] aT r,j represents the arrival time of train r at station j;

[0241] p lastest Indicates the effective space-time path where the passenger arrives at the latest route.

[0242] Step S4.1.3: If i+1=m, complete p lastest In the passenger behavior space-time arc, execute step S4.2; if i+1≠m, then i=i+1, t1=dT r,j , t2=tT j , repeat step S4.1.2.

[0243] Among them, dT r,j represents the departure time of train r at station j;

[0244] T j represents the transfer travel time at station j.

[0245] p earliest Steps: Search for the earliest effective space-time path of the originating route based on the passenger flow OD, including:

[0246] Step S4.2.1: Let i = 1, t1 = eT, t2=eT j ;

[0247] Among them, eT represents the time of entering the station;

[0248] Indicates the i-th station in the forward set of stations that the path passes through, with a maximum index of m;

[0249] eT j represents the entry and travel time of station j.

[0250] Step S4.2.2: When min|dT r,j -t1-t2|and Take the train from arrive The spacetime arc of p is added earliest ;

[0251] Among them, p earliest Indicates the earliest valid space-time path of the passenger's origin route.

[0252] Step S4.2.3: If i+1=m, complete p earliest If the passenger behavior space-time arc is in the middle, execute step S4.3; otherwise, i=i+1, t1=aT r,j , t2=tT j , repeat step S4.2.2.

[0253] Set search steps: According to p lastest and p earliest Search for a set of valid space-time paths;

[0254] If p lastest =p earliest , then the effective space-time path set P = {p lastest}={p earliest}; if p lastest ≠p earliest , then the effective space-time path set P = [p lastest ,p earliest ].

[0255] Among them, P represents the set of effective space-time paths of passengers.

[0256] According to the competition and connection between virtual lines and ordinary lines, the trains that passengers take are selected in combination with passenger flow OD and entry and exit time. The time and space path of travel is constructed through the combination of express trains and ordinary trains. The reconstructed line network is used to restore special transfer behaviors in the matching link of people and vehicles. In this way, diversified travel under the conditions of interconnected express trains is simulated, and the accuracy of passenger flow allocation is further improved.

[0257] Step S5: superimpose passenger flow distribution results.

[0258] Specifically, the passenger flow distribution results of the reconstructed network and the conventional network are superimposed to generate the spatiotemporal distribution status of the line network passenger flow.

[0259] In more preferred examples, under the reconstructed network, the passenger flow distribution result is not limited to express trains. Instead, it is the impact of the express train passenger flow travel chain on the line network passenger flow distribution under the reconstructed network. The spatiotemporal distribution state of the line network passenger flow needs to be superimposed with the distribution results of the express train passenger flow under the reconstructed network and the ordinary passenger flow under the conventional network. A virtual route is constructed based on the stops and operating time of the express train. It is possible to compare the path selection probability with the ordinary route in the same spatial dimension, and allocate passenger flow in a heterogeneous network based on the operating mode. In general, passengers who are eligible to take an interconnected express train will give priority to taking an express train so that they can reach their destination faster. Therefore, the express train passenger flow can be allocated under the reconstructed network first, and then the ordinary passenger flow can be allocated under the conventional network. Finally, the distribution results can be superimposed to obtain the passenger flow distribution state. Figure 6 Take this as an example, the specific steps are as follows:

[0260] Step S5.1: Initialize the passenger state, sort the passenger set Q (q∈Q) in ascending order according to the arrival time, and let x q =1,y q =1.

[0261] Where Q represents the set of passengers in the road network, q∈Q;

[0262] x q represents the space-time path index selected by passenger q, 0 <x q <P q ;

[0263] y q Indicates the route index of the space-time path selected by passenger q, 0 <y q <l q,p ;

[0264] l q,p represents the space-time arc of the lth train segment selected by passenger q on space-time path p, l q,p ∈L q,p ;

[0265] L q,p Represents the set of train space-time arcs that passenger q chooses for space-time path p.

[0266] Step S5.2: Initialize time parameters. Let simulation time t be the operation start time, time step t step .

[0267] Step S5.3: Update the space-time state according to the passenger space-time path, let i = x q , j = y q , t1=sT q,i,j .

[0268] Among them, sT q,p,l represents the starting time of the lth train space-time arc in passenger q’s space-time path p.

[0269] Step S5.4: Update the train congestion, let e q,p,l =v q,p,l / b q,p,l .

[0270] Among them, e q,p,l represents the train congestion of the lth train space-time arc in the space-time path p of passenger q;

[0271] v q,p,l represents the train passenger capacity of the lth train space-time arc in the space-time path p of passenger q;

[0272] b q,p,l Represents the train capacity of the lth train space-time arc in the space-time path p of passenger q.

[0273] Step S5.5: Traverse the passenger set Q. If t>t1 exists, execute step S5.6; if t>t1 does not exist, execute step S5.8.

[0274] Step S5.6: If the train load factor is less than the expected load factor, that is, e q,p,l <e exp , or the train load rate is greater than the expected load rate and is the latest reachable space-time path, that is, e q,p,l ≥e exp And x q =P q , then the passenger gets on the bus v q,p,l =v q,p,l +1, execute step S5.7; if the train load factor is greater than the expected load factor and it is not the latest reachable time-space path, that is, e q,p,l ≥e exp And x q <P q , then x q =x q +1, return to step S5.3.

[0275] Step S5.7: If x q = l q,p , then exclude q from Q; if x q ≠l q,p , then y q =y q +1.

[0276] Among them, L represents the total number of routes, l∈L.

[0277] Step S5.8: Update the simulation time, set t = t + t step If the simulation time t is greater than or equal to the operation end time, the simulation ends; if the simulation time t is greater than the operation end time, return to step S5.3.

[0278] The present invention also provides an urban rail transit passenger flow distribution system under an interconnected express train operation mode. The urban rail transit passenger flow distribution system under the interconnected express train operation mode can be realized by executing the process steps of the urban rail transit passenger flow distribution method under the interconnected express train operation mode, that is, those skilled in the art can understand the urban rail transit passenger flow distribution method under the interconnected express train operation mode as a preferred implementation mode of the urban rail transit passenger flow distribution system under the interconnected express train operation mode.

[0279] Those skilled in the art know that, in addition to realizing the system and its various devices, modules, and units provided by the present invention in a purely computer-readable program code, it is entirely possible to realize the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a hardware component, and the devices, modules, and units included therein for realizing various functions can also be regarded as structures within the hardware component; the devices, modules, and units for realizing various functions can also be regarded as both software modules for realizing the method and structures within the hardware component.

[0280] According to the present invention, a city rail transfer flow system under an interconnected express train operation mode is provided, comprising:

[0281] Module M1: Reconstruct the network structure to obtain the directed network G;

[0282] Module M2: Calculate K shortest valid paths in a directed network;

[0283] Module M3: Allocate direct express passenger flow under the directed network G according to K shortest effective paths;

[0284] Module M4: Allocate the remaining ordinary passenger flow under the conventional network;

[0285] Module M5: Based on the direct express passenger flow and the remaining ordinary passenger flow, the passenger flow allocation results are superimposed to obtain the spatiotemporal distribution state.

[0286] In more preferred examples, in the module M1, according to the stations where the express trains are interconnected, virtual express train lines are added on the basis of the conventional network to reconstruct the network structure of the urban rail transit.

[0287] The directed network G0 (L0, A0, N0) represents the conventional network of urban rail transit.

[0288] Among them, L0, A0, and N0 represent the lines, sections, and station collections in the conventional network, respectively.

[0289] The section includes the internal connection section of the transfer station generated in the transfer direction, and the directed network G (L, A, N) represents the reconstructed line network structure.

[0290] Among them, L=L0∪L1, A=A0∪A1, N=N0∪N1.

[0291] L1, A1, and N1 represent the virtual line, virtual section, and virtual station collection formed by the stops of the express train respectively.

[0292] L, A, and N represent the reconstructed line, section, and station set, respectively.

[0293] The maximum index of N is n, where n represents the total number of stations;

[0294] The interval running time of the virtual line is less than the interval running time of the corresponding section of the line in the conventional network, and is obtained from the direct express train schedule.

[0295] In the module M3, based on the reconstructed network structure, the interconnected express passenger flow is allocated in combination with the space-time path.

[0296] The comprehensive path cost c of the kth short circuit k and passenger flow OD comprehensive cost threshold c threshold The calculation formula is:

[0297] c k =f×t k , c threshold =min{c min ×(1+a),c min ×b};

[0298] Among them, f represents the transfer penalty coefficient, which is related to the line network structure and the number of transfers;

[0299] t k represents the travel cost of the kth short circuit;

[0300] c min Represents the comprehensive cost of the shortest path;

[0301] a represents the relative threshold of comprehensive cost, a∈(0,1);

[0302] b represents the absolute threshold of comprehensive cost, unit is min, usually b∈[10,30].

[0303] The MNL model is used to calculate the probability of each path being selected and matched with the timetable.

[0304] The probability Z of the kth short circuit being selected k The calculation formula is:

[0305] V k =exp[-(c k -c min ) 2 / (2σ 2 ))],Z k =V k / (∑ k∈K V k );

[0306] Among them, V k represents the utility value of the kth short circuit;

[0307] K is the total number of paths;

[0308] σ represents the probability adjustment coefficient.

[0309] In more preferred examples, the module M2 uses the A* algorithm and the queue optimization algorithm to calculate the K shortest effective paths (K≤5) between the passenger flow ODs of the conventional network and the reconstructed line network structure, eliminates the detour paths according to the recorded stations passed by the paths, and selects the effective paths in combination with the comprehensive path costs and calculates their selection probabilities, including:

[0310] Module M2.1: Initialize the directed network G(L,A,N) and the reverse directed network

[0311] in, Represents the section of reverse reconstruction.

[0312] Module M2.2: According to Combined with the queue optimization algorithm, the reverse shortest path (d∈N) between the terminal d and the station w in the passenger flow OD is calculated, and the cost is h(w), where h(w) represents the shortest path cost from the station w to the terminal d.

[0313] Module M2.3: Let queue W (W∈N) be the queue of subsequent node stations starting from O, initialize f(w)=∞, w∈W, and sort in ascending order according to f(w), where f(w) is the evaluation function of the A* algorithm.

[0314] Module M2.4: Calculate the cost g(w) between the starting point O and the subsequent node station w based on G(L,A,N), and let f(w) = g(w) + h(w).

[0315] Among them, if it is determined that w < d, then the node station w is deleted from the queue W, the subsequent node stations of the node station w are added, w = w + 1, f(w) = ∞ is set, and the module M2.4 is repeatedly triggered;

[0316] If w = d, then the k-th shortest path is output, and k = k + 1;

[0317] Judge the condition of k = K. If so, the output of the K shortest paths of the passenger flow OD is completed; if not, f(w) is sorted in ascending order, and the module M2.4 is repeatedly triggered.

[0318] In more preferred examples, in the module M4, based on the conventional network, the space-time path is searched based on the passenger flow OD and the timetable, and the remaining ordinary passenger flow is allocated in combination with the space-time path.

[0319] The selection of the space-time path includes:

[0320] p lastest Module: Search for the latest valid space-time path p of the passenger's final arrival line according to the passenger flow OD lastest ;

[0321] p earliest Module: Search for the earliest valid space-time path p of the passenger's departure line according to the passenger flow OD earliest ;

[0322] Set search module: According to p lastest and p earliest Search for the set of valid space-time paths.

[0323] The p lastest module includes: Module M4.1.1: Let i = 1, t1 = oT, t2 = oT j ;

[0324] Among them, i represents the i-th station;

[0325] t1 represents the time when the passenger swipes the card to get off or transfers at the current space-time arc;

[0326] oT represents the passenger's departure time;

[0327] j represents the j-th station;

[0328] represents the i-th station in the reverse set of stations passed by the OD path, and the maximum index is m;

[0329] t2 represents the in-station walking time of the passenger at the current space-time arc;

[0330] oT j represents the out-station walking time of station j.

[0331] Module M4.1.2: When min|t1-t2-aT r,j | and When the train r is arrive The spacetime arc of p is added lastest ;

[0332] Among them, S r represents the set of stations where train r stops;

[0333] aT r,j represents the arrival time of train r at station j.

[0334] Module M4.1.3: If i+1=m, complete p lastest The space-time arc of passenger behavior triggers p earliest module; if i+1≠m, then i=i+1, t1=dT r,j , t2=tT j , trigger module M4.1.2;

[0335] Among them, dT r,j represents the departure time of train r at station j;

[0336] T j represents the transfer travel time at station j.

[0337] The p earliest The modules include: Module M4.2.1: Let i = 1, t1 = eT, t2=eT j ;

[0338] Among them, eT represents the time of entering the station;

[0339] Indicates the i-th station in the forward set of stations that the path passes through, with a maximum index of m;

[0340] eT j represents the entry and travel time of station j.

[0341] Module M4.2.2: When min|dT r,j -t1-t2|and When the train r is arrive The spacetime arc of p is added earliest .

[0342] Module M4.2.3: If i+1=m, complete p earliest The passenger behavior space-time arc triggers module M4.3; if i+1≠m, then i=i+1, t1=aT r,j , t2=tT j , trigger module M4.2.2.

[0343] In the set search module, if p lastest =p earliest , then the effective space-time path set of passengers P = {p lastest}={p earliest}; if p lastest ≠p earliest , then the effective space-time path set of passengers P = [p lastest ,p earliest ].

[0344] In more preferred embodiments, the module M5 superimposes the reconstructed line network structure and the passenger flow distribution result of the conventional network to generate the temporal and spatial distribution state of the line network passenger flow, including:

[0345] Module M5.1: Initialize the passenger state, sort the passenger set Q(q∈Q) in ascending order according to the arrival time, and let x q =1,y q =1;

[0346] Where Q represents the set of passengers in the road network, q∈Q;

[0347] x q represents the space-time path index selected by passenger q, 0 <x q <P q ;

[0348] y q Indicates the route index of the space-time path selected by passenger q, 0 <y q <l q,p ;

[0349] l q,p represents the space-time arc of the lth train segment selected by passenger q on space-time path p, l q,p ∈L q,p ;

[0350] L q,p Represents the set of train space-time arcs that passenger q chooses for space-time path p.

[0351] Module M5.2: Initialize time parameters, set simulation time t as the start time of operation, time step t step .

[0352] Module M5.3: Update the space-time state according to the passenger space-time path, let i = x q , j = y q , t1=sT q,i,j ;

[0353] Among them, sT q,p,l represents the starting time of the lth train space-time arc in passenger q’s space-time path p.

[0354] Module M5.4: Update train congestion, let e q,p,l =v q,p,l / b q,p,l ;

[0355] Among them, e q,p,l represents the train congestion of the lth train space-time arc in the space-time path p of passenger q;

[0356] v q,p,l represents the train passenger capacity of the lth train space-time arc in the space-time path p of passenger q;

[0357] b q,p,l Represents the train capacity of the lth train space-time arc in the space-time path p of passenger q.

[0358] Module M5.5: Traverse the passenger set Q. If t>t1 exists, trigger module M5.6; if t>t1 does not exist, trigger module M5.8.

[0359] Module M5.6: If the train load factor is less than the expected load factor e q,p,l <e exp , or the train load factor is greater than the expected load factor and is the latest reachable space-time path e q,p,l ≥e exp And x q =P q , then the passenger gets on the bus v q,p,l =v q,p,l +1, trigger module M5.7; if the train load factor is greater than the expected load factor and is not the latest reachable space-time path e q,p,l ≥e exp And x q <P q , then x q =x q +1, trigger module M5.3.

[0360] Module M5.7: If x q = l q,p , then exclude q from Q; if x q ≠l q,p , then y q =y q +1;

[0361] Among them, L represents the total number of routes, l∈L.

[0362] Module M5.8: Update simulation time, set t = t + t stepIf the simulation time t is greater than or equal to the operation end time, the simulation ends; if the simulation time t is greater than the operation end time, module M5.3 is triggered.

[0363] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A method for urban rail transit flow distribution under the interconnected express train operation mode, characterized in that: Including: Step S1: Reconstruct the network structure to obtain a directed network G; Step S2: Calculate the K shortest effective paths of the directed network; Step S3: Allocate the through-express passenger flow under the directed network G according to the K shortest effective paths; Step S4: Allocate the remaining ordinary passenger flow under the conventional network; Step S5: According to the through-express passenger flow and the remaining ordinary passenger flow, superimpose the passenger flow allocation results to obtain the spatio-temporal distribution state.

2. The urban rail transfer flow method under the interconnected express train operation mode according to claim 1 is characterized in that: In the said Step S1, according to the stations passed by the through-express trains for interconnection, add virtual lines for through-express trains on the basis of the conventional network to reconstruct the network structure of urban rail transit; The directed network G0(L0, A0, N0) represents the conventional network of urban rail transit; Wherein, L0, A0, and N0 respectively represent the line, section, and station sets in the conventional network; The said section includes the internal connection section of the transfer station generated by the transfer direction, and the directed network G(L, A, N) represents the reconstructed network structure; Wherein, L = L0 ∪ L1, A = A0 ∪ A1, N = N0 ∪ N1; L1, A1, and N1 respectively represent the virtual line, virtual section, and virtual station sets formed by the stops of the through-express trains; L, A, and N respectively represent the reconstructed line, section, and station sets; The maximum index of N is n, and n represents the total number of stations; The running time of the interval of the said virtual line is less than the running time of the corresponding section of the line in the conventional network, and is obtained from the through-express train timetable; In the said Step S3, on the basis of the reconstructed network structure, allocate the through-express passenger flow for interconnection and intercommunication in combination with the spatio-temporal path; The comprehensive path cost c of the kth short circuit k and passenger flow OD comprehensive cost threshold c threshold The calculation formula is: c k =f×t k ,c threshold =min{c min ×(1+a),c min ×b}; Wherein, f represents the transfer penalty coefficient, which is related to the network structure and the number of transfers; t k represents the travel cost of the kth short circuit; c min Represents the comprehensive cost of the shortest path; a represents the relative threshold of the comprehensive cost, a ∈ (0, 1); b represents the absolute threshold of the comprehensive cost, with the unit of min, and usually b ∈ [10, 30]; Use the MNL model to calculate the selection probability of each path, and match it in combination with the timetable; The probability Z of the kth short circuit being selected k The calculation formula is: In k =exp[-(c k -c min ) 2 / (2σ 2 ))],From k =V k / (∑ k∈K In k ); Among them, V k represents the utility value of the kth short circuit; K is the total number of paths; σ represents the probability adjustment coefficient.

3. The urban rail transfer flow method under the interconnected express train operation mode according to claim 1 is characterized in that: In the said Step S2, use the A* algorithm and the queue optimization algorithm to calculate the K shortest effective paths (K ≤ 5) between each passenger flow OD of the conventional network and the reconstructed network structure, eliminate the circuitous paths according to the recorded path passing stations, screen the effective paths in combination with the comprehensive cost of the paths and calculate their selection probabilities, including: Step S2.1: Initialize the directed network G(L,A,N) and the reverse directed network in, represents the reverse reconstructed section; Step S2.2: According to Combined with the queue optimization algorithm, the reverse shortest path (d∈N) between the terminal d and the station w in the passenger flow OD is calculated, and the cost is h(w), where h(w) represents the shortest path cost from the station w to the terminal d; Step S2.3: Let the queue W (W ∈ N) be the subsequent node station queue of the starting point O, initialize f(w) = ∞, w ∈ W, and sort it in ascending order of f(w), where f(w) is the evaluation function of the A* algorithm; Step S2.4: Calculate the cost g(w) of the starting point O and the subsequent node station w according to G(L, A, N), and let f(w) = g(w) + h(w); Wherein, judge if w = d, then delete the node station w from the queue W, add the subsequent node stations of the node station w, w = w + 1, let f(w) = ∞, output the k-th shortest path, k = k + 1; if w < d, then continue to execute Step S2.4; Judge the condition of k = K. If so, the K shortest paths of the passenger flow OD are output; if not, sort in ascending order of f(w) and continue to execute Step S2.

4.

4. The urban rail transfer flow method under the interconnected express train operation mode according to claim 1 is characterized in that: In the step S4, on the basis of the conventional network, the space-time path is searched based on the passenger flow OD and the timetable, and the remaining ordinary passenger flow is allocated in combination with the space-time path; The selection of the space-time path includes: p lastest Steps: Search for the effective space-time path p with the latest passenger arrival time according to the passenger flow OD lastest ; p earliest Steps: Search for the earliest effective space-time path p of the passenger departure route based on passenger flow OD earliest ; Set search steps: According to p lastest and p earliest Search for a set of valid space-time paths; The p lastest The steps include: Step S4.1.1: let i=1, t1=oT, t2=oT j ; Among them, i represents the i-th station; t1 represents the time when the passenger swipes the card or transfers to another bus at the current time-space arc; oT represents the passenger’s departure time; j represents the jth station; It represents the i-th station in the reverse set of stations that the OD path passes through, and the maximum index is m; t2 represents the passenger’s travel time in the station in the current space-time arc; oT j represents the departure travel time of station j; Step S4.1.2: When min|t1-t2-aT r,j | and When the train r is arrive The spacetime arc of p is added lastest ; Among them, S r represents the set of stations where train r stops; aT r,j represents the arrival time of train r at station j; Step S4.1.3: If i+1=m, complete p lastest In the space-time arc of passenger behavior, execute p earliest Steps: If i+1≠m, then i=i+1, t1=dT r,j , t2=tT j , repeat step S4.1.2; Among them, dT r,j represents the departure time of train r at station j; T j represents the transfer travel time at station j; The p earliest The steps include: Step S4.2.1: let i=1, t1=eT, t2=eT j ; Among them, eT represents the time of entering the station; M i OD Indicates the i-th station in the forward set of stations that the path passes through, with a maximum index of m; eT j represents the entry and travel time of station j; Step S4.2.2: When When the train r is arrive The spacetime arc of p is added earliest ; Step S4.2.3: If i+1=m, complete p earliest In the passenger behavior space-time arc, execute the step set search step; if i+1≠m, then i=i+1, j=M i OD , t1=aT r,j , t2=tT j , repeat step S4.2.2; In the set search step, if p lastest =p earliest , then the effective space-time path set of passengers P = {p lastest }={p earliest }; if p lastest ≠p earliest , then the effective space-time path set of passengers P = [p lastest ,p earliest ].

5. The urban rail transfer flow method under the interconnected express train operation mode according to claim 1 is characterized in that: In step S5, the reconstructed line network structure and the passenger flow distribution result of the conventional network are superimposed to generate the temporal and spatial distribution state of the line network passenger flow, including: Step S5.1: Initialize the passenger state, sort the passenger set Q (q∈Q) in ascending order according to the arrival time, and let x q =1,y q =1; Where Q represents the set of passengers in the road network, q∈Q; x q represents the space-time path index selected by passenger q, 0 <x q <P q ; y q Indicates the route index of the space-time path selected by passenger q, 0 <y q <l q,p ; l q,p represents the space-time arc of the lth train segment selected by passenger q on space-time path p, l q,p ∈L q,p ; L q,p The set of space-time arcs representing the train that passenger q chooses for space-time path p; Step S5.2: Initialize time parameters, set simulation time t as the operation start time, time step t step ; Step S5.3: Update the space-time state according to the passenger space-time path, let i = x q , j = y q , t1=sT q,i,j ; Among them, sT q,p,l represents the starting time of the lth train spacetime arc in the spacetime path p of passenger q; Step S5.4: Update the train congestion, let e q,p,l =v q,p,l / b q,p,l ; Among them, e q,p,l represents the train congestion of the lth train space-time arc in the space-time path p of passenger q; v q,p,l represents the train passenger capacity of the lth train space-time arc in the space-time path p of passenger q; b q,p,l represents the train capacity of the lth train space-time arc in the space-time path p of passenger q; Step S5.5: traverse the passenger set Q, if t>t1 exists, execute step S5.6; if t>t1 does not exist, execute step S5.8; Step S5.6: If the train load factor is less than the expected load factor e q,p,l <e exp , or the train load factor is greater than the expected load factor and is the latest reachable space-time path e q,p,l ≥e exp And x q =P q , then the passenger gets on the bus v q,p,l =v q,p,l +1, execute step S5.7; if the train load factor is greater than the expected load factor and it is not the latest reachable time-space path e q,p,l ≥e exp And x q <P q , then x q =x q +1, return to step S5.3; Step S5.7: If x q = l q,p , then exclude q from Q; if x q ≠l q,p , then y q =y q +1; Where L represents the total number of routes, l∈L; Step S5.8: Update the simulation time, set t = t + t step If the simulation time t is greater than or equal to the operation end time, the simulation ends; if the simulation time t is greater than the operation end time, return to step S5.

3.

6. An urban rail transit flow distribution system under the interconnected express train operation mode, characterized in that: include: Module M1: Reconstruct the network structure to obtain the directed network G; Module M2: Calculate K shortest valid paths in a directed network; Module M3: Allocate direct express passenger flow under the directed network G according to K shortest effective paths; Module M4: Allocate the remaining ordinary passenger flow under the conventional network; Module M5: Based on the direct express passenger flow and the remaining ordinary passenger flow, the passenger flow allocation results are superimposed to obtain the spatiotemporal distribution state.

7. The urban rail transfer flow system under the interconnected express train operation mode according to claim 6 is characterized in that: In the module M1, according to the stations that the interconnected express trains pass through, a virtual express train line is added on the basis of the conventional network to reconstruct the network structure of the urban rail transit; The directed network G0 (L0, A0, N0) represents the conventional network of urban rail transit; Among them, L0, A0, and N0 represent the line, section, and station set in the conventional network respectively; The section includes the internal connection section of the transfer station generated in the transfer direction, and the directed network G (L, A, N) represents the reconstructed line network structure; Among them, L = L0 ∪ L1, A = A0 ∪ A1, N = N0 ∪ N1; L1, A1, and N1 represent the virtual line, virtual section, and virtual station set formed by the stopover of the express train respectively; L, A, and N represent the reconstructed line, section, and station set, respectively; The maximum index of N is n, where n represents the total number of stations; The interval running time of the virtual line is less than the interval running time of the corresponding section of the line in the conventional network, which is obtained from the direct express train schedule; In the module M3, based on the reconstructed network structure, the interconnected express passenger flow is allocated in combination with the time-space path; The comprehensive path cost c of the kth short circuit k and passenger flow OD comprehensive cost threshold c threshold The calculation formula is: c k =f×t k ,c threshold =min{c min ×(1+a),c min ×b}; Among them, f represents the transfer penalty coefficient, which is related to the line network structure and the number of transfers; t k represents the travel cost of the kth short circuit; c min Represents the comprehensive cost of the shortest path; a represents the relative threshold of comprehensive cost, a∈(0,1); b represents the absolute threshold of comprehensive cost, unit is min, usually b∈[10,30]; Use the MNL model to calculate the probability of each path being selected and match it with the timetable; The probability Z of the kth short circuit being selected k The calculation formula is: In k =exp[-(c k -c min ) 2 / (2σ 2 ))],From k =V k / (∑ k∈K In k ); Among them, V k represents the utility value of the kth short circuit; K is the total number of paths; σ represents the probability adjustment coefficient.

8. The urban rail transit flow distribution system under the interconnected express train operation mode according to claim 6 is characterized in that: The module M2 uses the A* algorithm and the queue optimization algorithm to calculate the K shortest effective paths (K≤5) between the passenger flow ODs of the conventional network and the reconstructed line network structure, removes the detour paths according to the recorded paths passing through the stations, selects the effective paths based on the comprehensive path costs, and calculates their selection probabilities, including: Module M2.1: Initialize the directed network G(L,A,N) and the reverse directed network in, represents the reverse reconstructed section; Module M2.2: According to Combined with the queue optimization algorithm, the reverse shortest path (d∈N) between the terminal d and the station w in the passenger flow OD is calculated, and the cost is h(w), where h(w) represents the shortest path cost from the station w to the terminal d; Module M2.3: Let the queue W (W ∈ N) be the queue of subsequent node stations starting from the origin O. Initialize f(w) = ∞, w ∈ W, and sort it in ascending order of f(w). f(w) is the evaluation function of the A* algorithm; Module M2.4: Calculate the cost g(w) between the origin O and the subsequent node station w according to G(L, A, N), and let f(w) = g(w) + h(w); Among them, if it is judged that w = d, then delete the node station w from the queue W, add the subsequent node stations of the node station w, w = w + 1, let f(w) = ∞, and output the k-th shortest path, k = k + 1; if w < d, then repeatedly trigger Module M2.4; Judge the condition of k = K. If so, the K-shortest paths of the passenger flow OD are output; if not, sort f(w) in ascending order and repeatedly trigger Module M2.

4.

9. The urban rail transit flow distribution system under the interconnected express train operation mode according to claim 6 is characterized in that: In the said Module M4, based on the conventional network, search for the spatio-temporal path based on the passenger flow OD and the timetable, and allocate the remaining ordinary passenger flow in combination with the spatio-temporal path; The selection of the said spatio-temporal path includes: p lastest Module: Search for the effective space-time path p with the latest passenger arrival time according to passenger flow OD lastest ; p earliest Module: Search for the earliest effective space-time path p of the passenger's departure route based on passenger flow OD earliest ; Collection search module: According to p lastest and p earliest Search for a set of valid space-time paths; The p lastest The modules include: Module M4.1.1: Let i=1, t1=oT, t2=oT j ; Among them, i represents the i-th station; t1 represents the time when the passenger swipes the card to leave the station or transfers and gets off at the current spatio-temporal arc; oT represents the time when the passenger leaves the station; j represents the j-th station; It represents the i-th station in the reverse set of stations that the OD path passes through, and the maximum index is m; t2 represents the in-station walking time of the passenger at the current spatio-temporal arc; oT j represents the departure travel time of station j; Module M4.1.2: When min|t1-t2-aT r,j | and When the train r is arrive The spacetime arc of p is added lastest ; Among them, S r represents the set of stations where train r stops; aT r,j represents the arrival time of train r at station j; Module M4.1.3: If i+1=m, complete p lastest The space-time arc of passenger behavior triggers p earliest module; if i+1≠m, then i=i+1, t1=dT r,j , t2=tT j , trigger module M4.1.2; Among them, dT r,j represents the departure time of train r at station j; T j represents the transfer travel time at station j; The p earliest The modules include: Module M4.2.1: Let i = 1, t1 = eT, t2=eT j ; Among them, eT represents the in-station walking time; Indicates the i-th station in the forward set of stations that the path passes through, with a maximum index of m; eT j represents the entry and travel time of station j; Module M4.2.2: When min|dT r,j -t1-t2|and When the train r is arrive The spacetime arc of p is added earliest ; Module M4.2.3: If i+1=m, complete p earliest The passenger behavior space-time arc triggers module M4.3; if i+1≠m, then i=i+1, t1=aT r,j , t2=tT j , trigger module M4.2.2; In the set search module, if p lastest =p earliest , then the effective space-time path set of passengers P = {p lastest }={p earliest }; if p lastest ≠p earliest , then the effective space-time path set of passengers P = [p lastest ,p earliest ].

10. The urban rail transfer flow system under the interconnected express train operation mode according to claim 6, characterized in that: In Module M5, superimpose the reconstructed line network structure and the passenger flow distribution result of the conventional network to generate the spatio-temporal distribution state of the line network passenger flow, including: Module M5.1: Initialize the passenger state, sort the passenger set Q(q∈Q) in ascending order according to the arrival time, and let x q =1,y q =1; Among them, Q represents the set of network passengers, q ∈ Q; x q represents the space-time path index selected by passenger q, 0 <x q <P q ; y q Indicates the route index of the space-time path selected by passenger q, 0 <y q <l q,p ; l q,p represents the space-time arc of the lth train segment selected by passenger q on space-time path p, l q,p ∈L q,p ; L q,p The set of space-time arcs representing the train that passenger q chooses for space-time path p; Module M5.2: Initialize time parameters, set simulation time t as the start time of operation, time step t step ; Module M5.3: Update the space-time state according to the passenger space-time path, let i = x q , j = y q , t1=sT q,i,j ; Among them, sT q,p,l represents the starting time of the lth train spacetime arc in the spacetime path p of passenger q; Module M5.4: Update train congestion, let e q,p,l =v q,p,l / b q,p,l ; Among them, e q,p,l represents the train congestion of the lth train space-time arc in the space-time path p of passenger q; v q,p,l represents the train passenger capacity of the lth train space-time arc in the space-time path p of passenger q; b q,p,l represents the train capacity of the lth train space-time arc in the space-time path p of passenger q; Module M5.5: Traverse the passenger set Q. If there exists t > t1, trigger Module M5.6; if there does not exist t > t1, then trigger Module M5.8; Module M5.6: If the train load factor is less than the expected load factor e q,p,l <e exp , or the train load factor is greater than the expected load factor and is the latest reachable space-time path e q,p,l ≥e exp And x q =P q , then the passenger gets on the bus v q,p,l =v q,p,l +1, trigger module M5.7; if the train load factor is greater than the expected load factor and is not the latest reachable space-time path e q,p,l ≥e exp And x q <P q , then x q =x q +1, trigger module M5.3; Module M5.7: If x q = l q,p , then exclude q from Q; if x q ≠l q,p , then y q =y q +1; Among them, L represents the total number of passing lines, l ∈ L; Module M5.8: Update simulation time, set t = t + t step If the simulation time t is greater than or equal to the operation end time, the simulation ends; if the simulation time t is greater than the operation end time, module M5.3 is triggered.

Images