A passenger flow induction method for urban rail transit network considering multiple disposal measures

Abstract

The present invention provides a method for inducing passenger flow in an urban rail transit network taking into account a variety of disposal measures. The method includes: according to the classification of emergencies and corresponding disposal measures, the types of passengers affected are obtained, and the average induction acceptance of different types of OD is initialized; for train interruption, passenger clearance and return, station passenger flow regulation, passengers in affected sections, and passengers departing from and arriving at the interruption section, a set of recommended travel plans for passengers is generated, and the travel cost of each path is calculated; the path recommendation plan is initialized, the path selection probability, the number of passengers on the path and the fitness value are calculated based on the Logit model, the passenger flow distribution of the road network is updated, and a new recommendation plan is generated using the whale optimization algorithm based on Pareto optimality to determine the best travel plan. The method of the present invention takes into account the different effects of various disposal measures on passengers, as well as the effects of induction on passenger flow distribution and travel costs, in view of emergencies, to achieve dynamic induction of passenger flow in the road network.

Description

Technical Field

[0001] The present invention relates to the technical field of urban rail transit passenger flow management, and in particular to an urban rail transit road network passenger flow induction method considering multiple disposal measures. Background Art

[0002] Urban rail transit is characterized by energy saving, land saving, large capacity, safety and convenience. It is a green and environmentally friendly transportation system, especially suitable for large and medium-sized cities. In addition to first-tier cities such as Beijing, Shanghai, and Guangzhou, the scale of urban rail transit construction in Shenzhen, Chengdu, and Nanjing is also growing continuously, with the proportion of passenger traffic increasing year by year, and network construction gradually taking shape. With the improvement of the urban rail transit network, passengers' travel is more convenient, but the urban rail transit system attracts a large number of passengers, which puts higher requirements on operation management and passenger travel services.

[0003] In the urban rail transit network, problems such as uneven passenger flow distribution, large-scale delays due to emergencies, and train failures due to bad weather often occur, which not only affects the travel efficiency of passengers, but also brings challenges to the operation and management of urban rail transit. How to combine the actual operation of urban rail transit, consider dynamic disposal measures, and propose a more intelligent and efficient passenger flow induction method has become an urgent problem to be solved in the current urban rail transit field.

[0004] At present, the passenger flow induction methods in the existing technology often focus on the recommendation of detour routes, and do not provide targeted induction measures for various disposal measures under emergencies, such as train interruption, passenger clearance and return, station passenger flow regulation and forced stop of train sections. The existing passenger flow induction methods are based on fixed routes and timetables, lack flexibility and real-time performance, and are difficult to respond quickly to sudden changes in passenger flow status, resulting in a large deviation between the induction strategy and actual demand. At the same time, the existing passenger flow induction methods often ignore the personalized needs and travel habits of passengers, resulting in poor induction effects. Summary of the invention

[0005] An embodiment of the present invention provides a method for inducing passenger flow in an urban rail transit network taking into account a variety of disposal measures, so as to effectively induce passenger flow under different network conditions.

[0006] In order to achieve the above object, the present invention adopts the following technical scheme.

[0007] A method for inducing passenger flow in an urban rail transit network considering multiple disposal measures, comprising:

[0008] According to the classification of emergencies and corresponding disposal measures, the affected types of passengers are obtained, and the average induced acceptance of different types of OD is initialized;

[0009] For train interruptions, passenger clearance and reversal, station passenger flow control, affected sections, and passengers departing from and arriving at interrupted sections, a solution space corresponding to different emergencies is established to generate a set of recommended travel plans for passengers. The travel cost of each path in the recommended travel plan set is simulated and calculated based on the current road network passenger flow distribution.

[0010] Initialize the path recommendation plan, calculate the path selection probability, the number of passengers on the path and the fitness value based on the Logit model according to the travel cost of each path in the travel plan alternative set and the average induced acceptance of different types of OD, update the road network passenger flow distribution, and use the whale optimization algorithm based on Pareto optimality to generate a new recommendation plan, and determine the best travel plan for passengers through iteration.

[0011] Preferably, the method further comprises:

[0012] The recommendation results are obtained through post-induction evaluation, and the average induction acceptance of different types of starting point to end point OD is updated based on the support vector machine regression SVR model.

[0013] Preferably, the classification of emergencies and corresponding handling measures to assess the types of passengers affected includes:

[0014] When a vehicle failure, power supply failure, platform screen door failure, communication failure, line failure, signal failure or an abnormally large passenger flow event occurs, corresponding handling measures are set up, including section interruption and route change, partial train route change, train delay or station passenger flow control; the affected passenger types are divided into the following four categories: passengers who need to return due to train interruption, passengers affected by station passenger flow control, passengers passing through the affected section, and passengers departing from and arriving at the interrupted section.

[0015] Preferably, for the train interruption, passenger clearance and return, station passenger flow regulation, affected sections of the route, and passengers departing from and arriving at the interruption section, a solution space corresponding to different emergencies is established, a recommended travel plan alternative set for passengers is generated, and the travel cost of each path in the recommended travel plan alternative set is simulated and calculated according to the current road network passenger flow distribution, including:

[0016] S21: For passengers who have to return due to train interruptions, the travel plan alternatives include waiting for the next train to board, taking other routes or changing transportation methods. The current road network passenger flow distribution is simulated to calculate the travel costs corresponding to the alternative plans.

[0017] S22: For passengers who are subject to station passenger flow control, the travel plan alternative set is the routes under different passenger flow control, detours to other routes or other transportation modes. The different passenger flow control modes include inbound passenger flow control, transfer passenger flow control or train transfer skipping. Combined with the current road network passenger flow distribution, simulation is performed to calculate the travel cost corresponding to the alternative plan;

[0018] S23: For passengers passing through the affected section, the travel plan alternative set is the route that the passengers originally planned to pass through the affected section, other detour routes or other transportation methods, combined with the current road network passenger flow distribution, to simulate and calculate the travel cost corresponding to the alternative plan;

[0019] S24: For passengers departing from and arriving at the interruption section, the travel plan alternative set includes passengers waiting for the emergency to end, taking other routes or choosing other modes of transportation. According to the expected interruption duration and the current road network passenger flow distribution, a simulation is performed to calculate the travel cost corresponding to the alternative plan.

[0020] Preferably, the calculation of the passenger travel cost for train interruption, clearing and returning in step S21 includes the following processing steps:

[0021] Step 1: Based on the updated train operation plan, create a train set that includes the last N trains that arrive at the station and do not need to clear passengers and return;

[0022] Step 2: Sort the passengers according to their current queue positions;

[0023] Step 3: Traverse the train set. When the nth train that does not need to clear passengers and turn around arrives, determine the number of passengers that can get on the train based on the number of passengers getting off the station and the remaining capacity constraint of the train. At this time, the passenger waiting time is the difference between the train arrival time and the passenger arrival time, and the passenger delay time is the passenger waiting time. In this case, the passenger travel time cost is the sum of the historical travel time under normal circumstances and the passenger delay time of the path.

[0024] Step 4: Update the passenger queue positions and re-sort them according to the confirmed number of boarding passengers;

[0025] Step 5: Repeat steps 3 and 4 until all passengers are on board.

[0026] Preferably, in step S22, when the passenger flow of the station exceeds the station carrying capacity, the passenger flow control is adopted. If it is a transfer station, the transfer flow limiting measures are adopted to calculate the travel cost of passengers under different passenger flow controls. The specific processing steps include:

[0027] Step 1: Determine the number of people queuing in different directions according to the passengers' travel or transfer destinations. In each queue of the direction of travel, arrange the passengers in order according to their current queue positions, where the queue of passengers going up is S and the queue of passengers going down is L;

[0028] Step 2: Refer to the historical control duration data of the entry or transfer passenger flow control, combine the dispatcher's real-time observation and estimation value, and comprehensively estimate the release duration of the entry or transfer passenger flow control;

[0029] Step 3: Establish the set of arriving trains after the entry or transfer passenger flow control is completed, where the set of upbound trains that have arrived the most recently n times is N, and the set of downbound trains that have arrived the most recently m times is M, and the train set I = N∪M;

[0030] Step 4: Traverse the train set I. When the i-th train arrives, calculate the number of passengers who can get on the train based on the number of passengers getting off the station obtained by simulation and the remaining capacity constraint of the train. At this time, the passenger travel time cost is the sum of the passenger flow control time, the queuing time and the historical travel time under normal circumstances of the path. The calculation formula is:

[0031] T total =T con +T wait―on +T his

[0032] Among them, T total is the passenger travel time cost, T con is the passenger flow control time, T wait―on The time for passengers to wait in line for boarding, T his is the historical travel time of the route under normal circumstances;

[0033] Step 5: Update the passenger queue position according to the number of simulated boarding passengers. If the train belongs to the train set N, update the queue S; if the train belongs to the train set M, update the queue L.

[0034] Step 6: Repeat steps 4 and 5 until all passengers are on board.

[0035] Preferably, in step S22, in view of the large number of passengers at the transfer station, a train transfer skipping measure is adopted to calculate the passenger travel cost under the control measure, and the specific processing steps include:

[0036] Step 1: Determine whether the treatment measure type is a train transfer skipping measure. If so, execute steps 8 to 14. If not, execute steps 2 to 7.

[0037] Step 2: Determine the number of people queuing in different directions according to the passengers' travel or transfer destinations. In each queue of the direction of travel, arrange the passengers in order according to their current queue positions, where the queue of passengers going in the upward direction is S, and the queue of passengers going in the downward direction is L;

[0038] Step 3: Refer to the historical control duration data of the entry or transfer passenger flow control, combine the dispatcher's real-time observation and estimation value, and comprehensively estimate the release duration of the entry or transfer passenger flow control;

[0039] Step 4: Establish the set of arriving trains after the entry or transfer passenger flow control is completed, where the set of upbound trains that have arrived the most recently d times is D, and the set of downbound trains that have arrived the most recently e times is E, and the train set I = D∪E;

[0040] Step 5: Traverse the train set I. When the i-th train arrives, calculate the number of passengers who can get on the train based on the number of people getting off the station obtained by simulation and the remaining capacity constraint of the train. At this time, the passenger delay time is the sum of the passenger flow control time and the queuing waiting time, and the calculation formula is:

[0041] T delay =T con +T wait―on

[0042] Among them, T delay is the passenger delay time, T con is the passenger flow control time, T wait―on Line up passengers and wait for their boarding time.

[0043] Step 6: Update the passenger queue position according to the number of simulated boarding passengers. If the train belongs to the train set N, update the queue S; if the train belongs to the train set M, update the queue L.

[0044] Step 7: Repeat steps 5 and 6 until all passengers are on board;

[0045] Step 8: Initialize station data, assume that the station that adopts the train transfer skipping measure is station A, and the next station of this station is station B;

[0046] Step 9: According to the adjusted train operation diagram, establish a train set F that includes the most recent f trains that pass through the passenger's station and do not stop at the transfer skip station, and a train set G that includes the most recent g trains that pass through the passenger's station and stop at the transfer skip station, and the train set I = F∪G;

[0047] Step 10: Sort the passengers according to their current queue positions on the platform or on the train;

[0048] Step 11: Determine the passenger type. Passenger types are divided into the following three types:

[0049] ① Passengers who are already on the train and choose to get off at station k along the way and wait for the train that stops at station A. The delay time for such passengers is the sum of the time from getting off the train and finding the queue to the time of waiting in line to get on the train. The calculation formula is:

[0050]

[0051] In the formula, For passengers to get off at station k and find the time when the queue starts, is the waiting time for passengers to board the bus at station k;

[0052] ② Passengers who are already on the train and choose to arrive at Station B and wait for the train to return to Station A. The delay time for this type of passengers is the sum of the travel time of the train between Station A and Station B, the time the passengers get off the train at Station B and find the queue to start queuing, the time they queue at Station B to wait for boarding, and the travel time between Station B and Station A. The calculation formula is:

[0053]

[0054] Where, T A―B is the travel time of the train between Station A and Station B. For passengers to get off at station B and find out when the queue starts, The time passengers spend waiting in line at station B;

[0055] ③ Passengers waiting at the station. The delay time of these passengers is equal to the time they queue up and wait for the train at station l.

[0056] The travel time cost of each type of passenger under passenger flow control is the sum of the passenger delay time and the historical travel time of the route under normal circumstances. The calculation formula is:

[0057] T total =T delay +T his

[0058] Step 12: Traverse the i-th train in the train set I. If train i belongs to the set M, calculate through simulation T A―B , Combined with the number of passengers getting off at the current station and the remaining capacity constraint of the train, determine the number of passengers who can get on the train at the current station. At this time, the affected passengers cannot get on the train, and update the queue order of the unaffected passengers after they get on the train; if train i belongs to set N, combined with the number of passengers getting off at the station and the remaining capacity constraint of the train, determine the number of passengers who can get on the train at the current station. At this time, all passengers can get on the train;

[0059] Step 13: Update the passenger queue positions and re-sort them according to the simulated number of boarding passengers;

[0060] Step 14: Repeat steps 11 to 13 until all affected passengers get off at station A.

[0061] Preferably, the estimated path delay time, the risk of forced stop in the subsequent section of the route, and the passenger delay time are calculated in step S23, and the calculation method is as follows:

[0062] The estimated path delay time is related to the length of time the train stops at a station or section, and the calculation formula is:

[0063]

[0064] In the formula, Estimate the duration of the forced stop of train number i, T event T is the estimated impact time of the emergency, train is the train’s stop time at the platform, N stat is the number of platforms between the current position and the affected section, V is the train speed, D k,o is the travel distance between the current station k and the affected station o.

[0065] The formula for calculating the risk of forced stop in the subsequent section of the route is:

[0066]

[0067] In the formula, The risk of forced stop of train number i in the subsequent section. D is the estimated forced stop time of the preceding train when an emergency occurs. i,i―1 is the distance between the two vehicles;

[0068] The formula for calculating passenger delay time is:

[0069]

[0070] The travel time cost of passengers waiting for the emergency to end is the sum of the passenger delay time and the historical travel time of the path under normal circumstances. The calculation formula is:

[0071] T total =T delay +T his

[0072] Preferably, the initialization route recommendation scheme calculates the route selection probability, the number of passengers on the route and the fitness value based on the Logit model according to the travel cost of each route in the travel plan alternative set and the average induced acceptance of different types of OD, updates the passenger flow distribution of the road network, generates a new recommendation scheme using the whale optimization algorithm based on Pareto optimality, and determines the best travel plan for passengers through iteration, including:

[0073] S31: For the travel mode alternatives corresponding to different types of emergencies, the initial path recommendation scheme is determined according to the difference in path travel costs, the path selection probability is calculated using the Logit model, and the distribution of passenger flow on the alternative paths is calculated in combination with the induced acceptance of the type of OD;

[0074] S32: Match the delay time of different types of passengers with the number of passengers of each type under the current road network passenger flow distribution, and calculate the minimum total passenger delay time. The calculation formula is:

[0075]

[0076] Where U is the set of all event types, P is the set of all passengers, and p u For passengers of type u, T u is the average delay time of type u events, and Z is the total delay time of passengers;

[0077] Calculate the difference between the recommended cost calculated based on the current road network passenger flow distribution and the simulated cost obtained by simulating the passenger flow after the recommended redistribution. The formula for minimizing the difference is:

[0078]

[0079] In the formula, R is the set of all recommended paths, is the recommended cost calculated for path r, is the simulation cost of path r obtained through simulation, and E is the difference between the recommended cost and the simulation cost;

[0080] S33: according to the initial path recommendation scheme and the initial fitness value, as well as the number of passengers on each path, the passenger flow distribution of the road network is updated, a partial solution is generated by a non-dominated genetic algorithm according to the whale optimization algorithm based on Pareto optimality, and a better solution set I is screened out by using the Pareto optimality theory, which is used to diversify the whale positions in the whale optimization algorithm;

[0081] The initial path recommendation scheme is used as the initial position of the whale, and the solutions in the I set are proportionally replaced with some non-optimal solutions in the whale population. The replacement ratio changes dynamically with the number of iterations, and the whale position is updated;

[0082] Simulate according to the passenger flow distribution of the road network, update the travel costs of paths, calculate the path selection probability, the number of passengers on the path, and the fitness through the new whale position mapping recommendation scheme;

[0083] Compare the size of the total passenger delay time Z in the fitness of each round with the difference degree E between the recommended cost and the simulation cost through the non-dominated relationship, update the passenger flow distribution of the road network, retain the optimal solution of each round, and gradually approach the global optimal solution by continuously iterating and updating the whale position;

[0084] Determine the best travel plan for passengers according to the optimal solution based on the whale optimization algorithm.

[0085] Preferably, calculating the path selection probability using the Logit model in S31, and combining the induced acceptance degree of this type of OD, calculating the distribution of passenger flow on the alternative paths, including:

[0086] Suppose there are three paths A, B, and C in the alternative set of travel modes for this OD. The travel time costs of paths A, B, and C under the current passenger flow distribution of the road network are a, b, and c respectively, and the path recommended by the initial path recommendation scheme is A. Then for the passengers on path B, assuming its induced acceptance degree is h (0 < h < 1), then h proportion of the passengers on path B will pay attention to the recommendation and make a choice decision between paths A and B. The selection probabilities of these passengers making a choice decision between paths A and B are respectively:

[0087]

[0088] In the formula, θ is the time parameter, and a, b, and c are the travel time costs of paths A, B, and C under the current passenger flow distribution of the road network, that is, the average travel time of this path, with the unit of seconds.

[0089] Similarly, the selection probabilities of passengers on path C between paths A and C are respectively and

[0090] Suppose the number of passengers passing through paths A, B, and C per unit time is m, n, and z respectively. This OD type is a long-distance one with the destination station being the business district. According to the actual investigation, initialize the induced acceptance degree of this type of OD as h (0 < h < 1). Then the distribution of passenger flow on the alternative paths is respectively:

[0091]

[0092] In the formula, N A 、N B 、N C are the number of passengers on paths A, B, and C respectively.

[0093] It can be seen from the technical solutions provided by the above embodiments of the present invention that the method of the present invention provides common support for the induction of passenger flow in urban rail transit network, provides technical conditions for scientific operation organization, and improves the level of network operation services. It can solve the problems that the traditional passenger flow induction method does not consider the targeted induction based on disposal measures, the induction scheme based on fixed lines and timetables cannot reflect the real-time network status, and ignores the personalized needs of passengers and post-induction evaluation.

[0094] Additional aspects and advantages of the present invention will be given in part in the following description, which will become obvious from the following description, or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0096] Figure 1 A processing flow chart of a method for inducing passenger flow in an urban rail transit network taking into account multiple disposal measures provided by an embodiment of the present invention;

[0097] Figure 2 A schematic diagram of the classification of emergency response measures and corresponding travel plans provided in an embodiment of the present invention;

[0098] Figure 3 A flow chart of calculating the travel cost of passengers when a train is interrupted, cleared and turned back, provided in an embodiment of the present invention;

[0099] Figure 4 A flowchart for calculating passenger travel costs under different passenger flow control measures provided by an embodiment of the present invention;

[0100] Figure 5 A schematic diagram of the delay of the affected section of the route provided by an embodiment of the present invention;

[0101] Figure 6 A flow chart of a method for calculating an optimal travel plan provided by an embodiment of the present invention;

[0102] Figure 7 An example diagram of the path costs of various treatment measures provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0103] The embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

[0104] It will be understood by those skilled in the art that, unless expressly stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or coupling. The term "and / or" used herein includes any unit and all combinations of one or more associated listed items.

[0105] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless defined as herein.

[0106] To facilitate understanding of the embodiments of the present invention, several specific embodiments will be further explained below with reference to the accompanying drawings, and each embodiment does not constitute a limitation on the embodiments of the present invention.

[0107] The embodiment of the present invention fully considers the travel needs of affected passengers caused by various disposal measures under emergencies, as well as the passengers' acceptance of the recommended plans after receiving the induction information, to recommend effective urban rail transit network passenger flow induction plans, more accurately realize passenger flow induction under different network conditions, and provide strong support for operational decisions under emergencies.

[0108] The processing flow chart of a method for inducing passenger flow in an urban rail transit network considering multiple disposal measures provided by an embodiment of the present invention is as follows: Figure 1 As shown, the processing steps include the following:

[0109] Step S1: According to the classification of emergencies and corresponding disposal measures, the affected types of passengers are obtained and the average induced acceptance of different types of OD is initialized;

[0110] Step S2: For train interruption, passenger clearance and return, station passenger flow control, affected sections, and passengers departing from and arriving at the interruption section, a solution space corresponding to different emergencies is established, and a recommended travel plan set for passengers is generated. The travel cost of each path is simulated and calculated according to the current road network passenger flow distribution, and classified induction is performed;

[0111] Step S3: Initialize the path recommendation plan. According to the path travel cost, calculate the path selection probability, the number of passengers on the path and the fitness value based on the Logit model and the average induced acceptance of different types of OD, update the road network passenger flow distribution, and use the whale optimization algorithm based on Pareto optimality to generate a new recommendation plan. Determine the best travel plan through iteration.

[0112] Step S4: Obtain recommendation results through post-induction evaluation, and update the average induction acceptance of different types of OD (Origin to Destination) based on the SVR (Support Vector Regression) model to provide data support for subsequent induction.

[0113] Specifically, the above step S1 includes: Figure 2 The present invention is a schematic diagram of the classification of emergency response measures and corresponding travel plans, specifically including: when emergencies such as vehicle failure, power failure, screen door failure, communication failure, line failure, signal failure or abnormally large passenger flow occur, it may cause the section to be unable to pass trains for a period of time, the number of trains passing through the section to decrease, and a large passenger flow to occur. The corresponding response measures include section interruption and route change, part of the train route change, train delays during operation or station passenger flow control. According to the above response measures, the types of passengers affected are divided into the following four categories: passengers who return due to train interruption, passengers affected by station passenger flow control, passengers passing through the affected section, and passengers departing from and arriving at the interrupted section.

[0114] Specifically, the above step S2 includes:

[0115] S21: For passengers who have to return due to train interruptions, the travel plan alternatives include waiting for the next train to board, taking other routes or changing transportation methods. The current road network passenger flow distribution is simulated to calculate the travel costs corresponding to the alternative plans.

[0116] S22: For passengers who are subject to station passenger flow control, the travel plan alternative set is the routes under different passenger flow control, detours to other routes or other transportation modes. The different passenger flow control modes include inbound passenger flow control, transfer passenger flow control or train transfer skipping. Combined with the current road network passenger flow distribution, simulation is performed to calculate the travel cost corresponding to the alternative plan;

[0117] S23: For passengers passing through the affected section, the travel plan alternative set is the route that the passengers originally planned to pass through the affected section, other detour routes or other transportation methods, combined with the current road network passenger flow distribution, to simulate and calculate the travel cost corresponding to the alternative plan;

[0118] S24: For passengers departing from and arriving at the interruption section, the travel plan alternative set includes passengers waiting for the emergency to end, taking other routes or choosing other modes of transportation. According to the expected interruption duration and the current road network passenger flow distribution, a simulation is performed to calculate the travel cost corresponding to the alternative plan.

[0119] The travel costs corresponding to the above alternatives are generated using the whale algorithm for global optimization.

[0120] The flowchart of calculating the passenger travel cost of returning due to train interruption in S21 is as follows: Figure 3 As shown, the processing steps include the following:

[0121] Step 1: Based on the updated train operation plan, create a train set that includes the last N trains that arrive at the station and do not need to clear passengers and return;

[0122] Step 2: Sort the passengers according to their current queue positions;

[0123] Step 3: Traverse the train set. When the nth train that does not need to clear passengers and turn around arrives, determine the number of passengers that can get on the train based on the number of passengers getting off the station and the remaining capacity constraint of the train. At this time, the passenger waiting time is the difference between the train arrival time and the passenger arrival time, and the passenger delay time is the passenger waiting time. In this case, the passenger travel time cost is the sum of the historical travel time under normal circumstances and the passenger delay time of the path.

[0124] Step 4: Update the passenger queue positions and re-sort them according to the confirmed number of boarding passengers;

[0125] Step 5: Repeat steps 3 and 4 until all passengers are on board.

[0126] In S22, when the passenger flow of the station is large and exceeds the station carrying capacity, the passenger flow control is adopted. If it is a transfer station, transfer flow limiting measures may be adopted to calculate the travel cost of passengers under different passenger flow controls. The specific steps include:

[0127] Step 1: Determine the number of people queuing in different directions according to the passengers' travel or transfer destinations. In each queue of the direction of travel, arrange the passengers in order according to their current queue positions, where the queue of passengers going up is S and the queue of passengers going down is L;

[0128] Step 2: Refer to the historical control duration data of the entry or transfer passenger flow control, combine the dispatcher's real-time observation and estimation value, and comprehensively estimate the release duration of the entry or transfer passenger flow control;

[0129] Step 3: Establish the set of arriving trains after the entry or transfer passenger flow control is completed, where the set of upbound trains that have arrived the most recently n times is N, and the set of downbound trains that have arrived the most recently m times is M, and the train set I = N∪M;

[0130] Step 4: Traverse the train set I. When the i-th train arrives, calculate the number of passengers that can get on the train based on the number of people getting off the station obtained by simulation and the remaining capacity constraint of the train. At this time, the passenger travel time cost is the sum of the passenger flow control time, the queuing time and the historical travel time under normal circumstances of the path. The calculation formula is:

[0131] T total =T con +T wait―on +T his

[0132] Among them, T total is the passenger travel time cost, T con is the passenger flow control time, T wait―on The time for passengers to wait in line for boarding, T his is the historical travel time of the route under normal circumstances.

[0133] Step 5: Update the passenger queue position according to the number of simulated boarding passengers. If the train belongs to the train set N, update the queue S; if the train belongs to the train set M, update the queue L.

[0134] Step 6: Repeat steps 4 and 5 until all passengers are on board.

[0135] In S22, in view of the large number of passengers at the transfer station, a train transfer skipping measure is adopted to calculate the passenger travel cost under the regulation measure. The specific processing steps include:

[0136] Step 1: Determine whether the treatment measure type is a train transfer skipping measure. If so, execute steps 8 to 14. If not, execute steps 2 to 7.

[0137] Step 2: Determine the number of people queuing in different directions according to the passengers' travel or transfer destinations. In each queue of the direction of travel, arrange the passengers in order according to their current queue positions, where the queue of passengers going in the upward direction is S, and the queue of passengers going in the downward direction is L;

[0138] Step 3: Refer to the historical control duration data of the entry or transfer passenger flow control, combine the dispatcher's real-time observation and estimation value, and comprehensively estimate the release duration of the entry or transfer passenger flow control;

[0139] Step 4: Establish the set of arriving trains after the entry or transfer passenger flow control is completed, where the set of upbound trains that have arrived the most recently d times is D, and the set of downbound trains that have arrived the most recently e times is E, and the train set I = D∪E;

[0140] Step 5: Traverse the train set I. When the i-th train arrives, calculate the number of passengers who can get on the train based on the number of people getting off the station obtained by simulation and the remaining capacity constraint of the train. At this time, the passenger delay time is the sum of the passenger flow control time and the queuing waiting time, and the calculation formula is:

[0141] T delay =T con +T wait―on

[0142] Among them, T delay is the passenger delay time, T con is the passenger flow control time, T wait―on Line up passengers and wait for their boarding time.

[0143] Step 6: Update the passenger queue position according to the number of simulated boarding passengers. If the train belongs to the train set N, update the queue S; if the train belongs to the train set M, update the queue L.

[0144] Step 7: Repeat steps 5 and 6 until all passengers are on board;

[0145] Step 8: Initialize station data, assume that the station that adopts the train transfer skipping measure is station A, and the next station of this station is station B;

[0146] Step 9: According to the adjusted train operation diagram, establish a train set F that includes the most recent f trains that pass through the passenger's station and do not stop at the transfer skip station, and a train set G that includes the most recent g trains that pass through the passenger's station and stop at the transfer skip station, and the train set I = F∪G;

[0147] Step 10: Sort the passengers according to their current queue positions on the platform or on the train;

[0148] Step 11: Determine the passenger type. Passenger types are divided into the following three types:

[0149] ① Passengers who are already on the train and choose to get off at station k along the way and wait for the train that stops at station A. The delay time for such passengers is the sum of the time from getting off the train and finding the queue to the time of waiting in line to get on the train. The calculation formula is:

[0150]

[0151] In the formula, For passengers to get off at station k and find the time when the queue starts, is the waiting time for passengers to board the bus at station k;

[0152] ② Passengers who are already on the train and choose to arrive at Station B and wait for the train to return to Station A. The delay time for this type of passengers is the sum of the travel time of the train between Station A and Station B, the time the passengers get off the train at Station B and find the queue to start queuing, the time they queue at Station B to wait for boarding, and the travel time between Station B and Station A. The calculation formula is:

[0153]

[0154] Where, T A―B is the travel time of the train between Station A and Station B. For passengers to get off at station B and find out when the queue starts, The time passengers spend waiting in line at station B;

[0155] ③ Passengers waiting at the station. The delay time of these passengers is equal to the time they queue up and wait for the train at station l.

[0156] The travel time cost of each type of passenger under passenger flow control is the sum of the passenger delay time and the historical travel time of the route under normal circumstances. The calculation formula is:

[0157] T total =T delay +T his

[0158] Step 12: Traverse the i-th train in the train set I. If train i belongs to the set M, calculate through simulation T A―B , Combined with the number of passengers getting off at the current station and the remaining capacity constraint of the train, determine the number of passengers who can get on the train at the current station. At this time, the affected passengers cannot get on the train, and update the queue order of the unaffected passengers after they get on the train; if train i belongs to set N, combined with the number of passengers getting off at the station and the remaining capacity constraint of the train, determine the number of passengers who can get on the train at the current station. At this time, all passengers can get on the train;

[0159] Step 13: Update the passenger queue positions and re-sort them according to the simulated number of boarding passengers;

[0160] Step 14: Repeat steps 11 to 13 until all affected passengers get off at station A.

[0161] The flowchart of calculating the passenger travel cost under different passenger flow control measures in S22 is as follows: Figure 4 As shown, the process is as follows:

[0162] Step 1: Determine whether the treatment measure type is a train transfer skipping measure. If so, execute steps 8 to 14. If not, execute steps 2 to 7.

[0163] Step 2: Determine the number of people queuing in different directions according to the passengers' travel or transfer destinations. In each queue of the direction of travel, arrange the passengers in order according to their current queue positions, where the queue of passengers going in the upward direction is S, and the queue of passengers going in the downward direction is L;

[0164] Step 3: Refer to the historical control duration data of the entry or transfer passenger flow control, combine the dispatcher's real-time observation and estimation value, and comprehensively estimate the release duration of the entry or transfer passenger flow control;

[0165] Step 4: Establish the set of arriving trains after the entry or transfer passenger flow control is completed, where the set of upbound trains that have arrived the most recently d times is D, and the set of downbound trains that have arrived the most recently e times is E, and the train set I = D∪E;

[0166] Step 5: Traverse the train set I. When the i-th train arrives, calculate the number of passengers who can get on the train based on the number of people getting off the station obtained by simulation and the remaining capacity constraint of the train. At this time, the passenger delay time is the sum of the passenger flow control time and the queuing waiting time, and the calculation formula is:

[0167] T delay =T con +T wait―on

[0168] Among them, T delay is the passenger delay time, T con is the passenger flow control time, T wait―on Line up passengers and wait for their boarding time.

[0169] Step 6: Update the passenger queue position according to the number of simulated boarding passengers. If the train belongs to the train set N, update the queue S; if the train belongs to the train set M, update the queue L.

[0170] Step 7: Repeat steps 5 and 6 until all passengers are on board;

[0171] Step 8: Initialize station data, assume that the station that adopts the train transfer skipping measure is station A, and the next station of this station is station B;

[0172] Step 9: According to the adjusted train operation diagram, establish a train set F that includes the most recent f trains that pass through the passenger's station and do not stop at the transfer skip station, and a train set G that includes the most recent g trains that pass through the passenger's station and stop at the transfer skip station, and the train set I = F∪G;

[0173] Step 10: Sort the passengers according to their current queue positions on the platform or on the train;

[0174] Step 11: Determine the passenger type. Passenger types are divided into the following three types:

[0175] ① Passengers who are already on the train and choose to get off at station k along the way and wait for the train that stops at station A. The delay time for such passengers is the sum of the time from getting off the train and finding the queue to the time of waiting in line to get on the train. The calculation formula is:

[0176]

[0177] In the formula, For passengers to get off at station k and find the time when the queue starts, is the waiting time for passengers to board the bus at station k;

[0178] ② Passengers who are already on the train and choose to arrive at Station B and wait for the train to return to Station A. The delay time for this type of passengers is the sum of the travel time of the train between Station A and Station B, the time the passengers get off the train at Station B and find the queue to start queuing, the time they queue at Station B to wait for boarding, and the travel time between Station B and Station A. The calculation formula is:

[0179]

[0180] Where, T A―B is the travel time of the train between Station A and Station B. For passengers to get off at station B and find out when the queue starts, The time passengers spend waiting in line at station B;

[0181] ③ Passengers waiting at the station. The delay time of these passengers is equal to the time they queue up and wait for the train at station l.

[0182] The travel time cost of each type of passenger under passenger flow control is the sum of the passenger delay time and the historical travel time of the route under normal circumstances. The calculation formula is:

[0183] T total =Tdelay +T his

[0184] Step 12: Traverse the i-th train in the train set I. If train i belongs to the set M, calculate through simulation T A―B , Combined with the number of passengers getting off at the current station and the remaining capacity constraint of the train, determine the number of passengers who can get on the train at the current station. At this time, the affected passengers cannot get on the train, and update the queue order of the unaffected passengers after they get on the train; if train i belongs to set N, combined with the number of passengers getting off at the station and the remaining capacity constraint of the train, determine the number of passengers who can get on the train at the current station. At this time, all passengers can get on the train;

[0185] Step 13: Update the passenger queue positions and re-sort them according to the number of simulated boarding passengers;

[0186] Step 14: Repeat steps 11 to 13 until all affected passengers get off at station A.

[0187] In S23, the estimated route delay time, the risk of forced stop in the subsequent section of the route, and the passenger delay time are calculated in the following manner:

[0188] The estimated path delay time is related to the length of time the train stops at a station or section, and the calculation formula is:

[0189]

[0190] In the formula, Estimate the duration of the forced stop of train number i, T event T is the estimated impact time of the emergency, train is the train’s stop time at the platform, N stat is the number of platforms between the current position and the affected section, V is the train speed, D k,o is the travel distance between the current station k and the affected station o.

[0191] The formula for calculating the risk of forced stop in the subsequent section of the route is:

[0192]

[0193] In the formula, The risk of forced stop of train number i in the subsequent section. D is the estimated forced stop time of the preceding train when an emergency occurs. i,i―1 is the distance between the two vehicles.

[0194] The formula for calculating passenger delay time is:

[0195]

[0196] The travel time cost for passengers waiting for the end of an emergency is the sum of the passenger delay time and the historical travel time under normal circumstances for this route. The calculation formula is as follows:

[0197] T total = T delay + T his

[0198] Specifically, Figure 6 is the flowchart of the calculation method for the optimal travel plan of the present invention. The step S3 includes:

[0199] S31: For the alternative travel mode sets corresponding to different emergency types, according to the travel cost of the route, through the initialized route recommendation plan, use the Logit model to calculate the route selection probability, and combine the induced acceptance degree of this type of OD to calculate the distribution of passenger flow on the alternative routes. The calculation process is as follows:

[0200] ① Suppose there are three routes A, B, and C in total in the alternative travel mode set of this OD. The travel time costs of routes A, B, and C under the current road network passenger flow distribution are a, b, and c respectively. The route recommended by the initial route recommendation plan is A. Then for the passengers on route B, assuming its induced acceptance degree is h (0 < h < 1), then h proportion of the passengers on route B will pay attention to the recommendation and make a choice decision between routes A and B. The choice probabilities of these passengers making a choice decision between routes A and B are respectively:

[0201]

[0202] In the formula, θ is the time parameter, and a, b, and c are the travel time costs of routes A, B, and C under the current road network passenger flow distribution, that is, the average travel time of this route, with the unit of seconds.

[0203] Similarly, the choice probabilities of the passengers on route C between routes A and C are respectively and

[0204] ② Suppose the number of passengers passing through routes A, B, and C per unit time is m, n, and z respectively. This type of OD is a long-distance one, and the destination station is a business district. According to the actual investigation situation, the induced acceptance degree of this type of OD is initialized as h (0 < h < 1). Then the distribution of passenger flow on the alternative routes is respectively:

[0205]

[0206]

[0207] In the formula, NA 、N B 、N C are the number of passengers on paths A, B, and C respectively.

[0208] S32: The fitness calculation aims to minimize the total passenger delay time and minimize the difference between the recommended cost and the simulated cost. The calculation basis for minimizing the total passenger delay time is that the delay time of different types of passengers matches the number of passengers of each type under the current road network passenger flow distribution. The calculation formula is:

[0209]

[0210] Where U is the set of all event types, P is the set of all passengers, and p u For passengers of type u, T u is the average delay time of type u events, and Z is the total delay time of passengers.

[0211] Minimize the difference between the recommended cost and the simulated cost by comparing the difference between the recommended cost calculated based on the current road network passenger flow distribution and the simulated cost obtained by simulating the passenger flow after the recommended redistribution. The calculation formula is:

[0212]

[0213] In the formula, R is the set of all recommended paths, is the recommended cost calculated for path r, is the simulation cost of path r obtained through simulation, and E is the difference between the recommended cost and the simulation cost.

[0214] S33: The initial path recommendation scheme and initial fitness value, as well as the number of passengers on each path, are obtained through the above calculations. The road network passenger flow distribution is updated. New recommendation schemes are continuously generated according to the whale optimization algorithm based on Pareto optimality. The new round of path travel costs are simulated through the updated road network passenger flow distribution. The path selection probability and the number of passengers on the path also change with the new round of recommendation schemes and costs. The fitness value and road network passenger flow distribution are updated, and the optimal solution is obtained by iteration. The specific steps are as follows:

[0215] ① Generate partial solutions through non-dominated genetic algorithm, and use Pareto optimality theory to screen out the better solution set I, which is used to diversify the whale positions in the whale optimization algorithm;

[0216] ② The initial path recommendation scheme is used as the initial position of the whale, and the solutions in the I set are proportionally replaced with some non-optimal solutions in the whale population. The replacement ratio changes dynamically with the number of iterations, and the whale position is updated;

[0217] ③ According to the simulation of passenger flow distribution on the road network, the path travel cost is updated, and the path selection probability, the number of passengers on the path and the fitness are calculated through the new whale position mapping recommendation plan;

[0218] ④ Compare the total passenger delay time Z and the difference E between the recommended cost and the simulated cost in each round of fitness through non-dominated relations, update the passenger flow distribution of the road network, retain the optimal solution of each round, and gradually approach the global optimal solution through continuous iteration and updating of the whale position;

[0219] ⑤ According to the optimal solution based on the whale optimization algorithm, determine the best travel plan for passengers and recommend the best travel plan for passengers.

[0220] Specifically, the method of updating the induced acceptance of different types of ODs by the SVR model in S4 includes:

[0221] S41: The feedback results of the passengers who participated in the induction are used as the evaluation object. Through the online evaluation of the APP, the passengers’ evaluation of the induction information is collected, and the evaluation content includes the accuracy of the induction information, the impact on travel decisions, etc.

[0222] S42: Conduct statistical analysis on the collected evaluation data, calculate the induced acceptance of each OD, and evaluate the overall effectiveness of the recommended plan.

[0223] S43: Based on the passenger travel distance, OD is subdivided into long-distance and short-distance categories. Based on the attributes of the destination station, it is divided into five categories: commercial area, business area, residential area, mixed area and others. The travel distance and the attributes of the destination station are used as feature values, and the corresponding induced acceptance of OD is used as the target value. The input is input into the SVR model for training to predict the average induced acceptance of different types of OD as the initial induced acceptance of the next round of induction.

[0224] Figure 5 This is a schematic diagram of the delay of the affected section of the route of the present invention, including relevant demonstrations of forced stops of trains in case of emergencies. Train 1 and Train 2 were originally scheduled to travel from Station 1 to Station 6. At time t1, an emergency occurred at Station 5. At this time, Train 1 arrived at Station 3 and began to make an emergency stop, and Train 2 arrived at Station 2 and began to make an emergency stop. Since Train 2 is far away from the station where the emergency occurred, the forced stop time is shorter than that of Train 1. At time t2, Train 2 ended its forced stop. Soon after, at time t3, Train 1 ended its forced stop. When Train 2 arrived at Station 4, Train 1 was running in the section between Station 4 and Station 5. In order to ensure that Train 2 maintained a safe driving distance from the preceding train (Train 1), Train 2 made another forced stop at Station 4 and ended its forced stop at time t5.

[0225] Figure 7This is an example diagram of the path costs of various disposal measures of the present invention, including the cost calculation process of train operation interruption, passenger clearance and return, and train transfer and skipping during station passenger flow control. For the train interruption and clearing and turning back, assume that passengers 1-100 get off from clearing train 1 at 9:00, and passengers 101-200 get off from clearing train 2 at 9:05, and establish a set of the last 4 trains that arrived at the station and did not clear and turn back; train 1 arrived at the station at 9:07, 20 people got off, and the remaining train capacity was 114 people, then the first 114 people can get on the train, and the waiting time for passengers 1-100 is 7 minutes, and the waiting time for passengers 101-114 is 2 minutes; train 2 arrived at the station at 9:10, 5 people got off, and the remaining train capacity was 80 people, then the first 80 people in the new sequence can get on the train, and the waiting time for passengers 1-80 is 5 minutes; train 3 arrived at the station at 9:12, 15 people got off, and the remaining train capacity was 100 people, then the first 100 people in the new sequence can get on the train, and the waiting time for passengers 1-6 is 7 minutes, at which time the number of people in the queue is 0, and the calculation ends. For the train transfer skipping measures in the passenger flow control of stations, it is assumed that Station 4 is a train transfer skipping station, where trains M2 and train N1 stop at Station 4, train M1 skips at Station 4, passengers 1 and 2 are on train M1, and passenger 3 waits at Station 3; passenger 1 gets off at Station 3, queues up for train M2, and takes this train to Station 4. The delay time is the sum of the time from getting off the train M1 and finding the queue to start queuing and the time waiting in line at Station 3; passenger 2 gets off at Station 5, queues up for train N1, and takes this train back to Station 4. The delay time is the sum of the time from Station 4 to Station 5, the time from getting off the train M1 and finding the queue to start queuing, the time waiting in line at Station 5, and the train travel time from Station 5 to Station 4; passenger 3 waits at Station 3. Since train M1 does not stop at Station 4, passenger 3 does not get on the train, waits in line, and takes train M2 to Station 3. The delay time is the time waiting for train M2 at Station 3.

[0226] In summary, the method of the present invention classifies emergencies and corresponding handling measures, considers the different impacts of various handling measures on passengers, and the impact of induction on passenger flow distribution and travel costs, and realizes dynamic induction of road network passenger flow.

[0227] The method of the present invention can accurately analyze the influence of OD categories on the acceptance of passenger induction through post-induction evaluation, and realize accurate regulation of passenger flow induction. The method of the present invention can comprehensively consider the dynamic changes of passenger flow caused by various disposal measures and induction of the road network, realize accurate induction of passenger flow, improve the operating efficiency of the transportation system, and reduce the travel cost of passengers in emergencies, which has important practical application value.

[0228] Those skilled in the art can understand that the accompanying drawings are only schematic diagrams of an embodiment, and the modules or processes in the accompanying drawings are not necessarily required to implement the present invention.

[0229] It can be known from the description of the above implementation methods that those skilled in the art can clearly understand that the present invention can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present invention or certain parts of the embodiments.

[0230] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The device and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0231] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for inducing passenger flow in an urban rail transit network taking into account multiple disposal measures, characterized in that: include: According to the classification of emergencies and corresponding disposal measures, the affected types of passengers are obtained and the induced acceptance of different types of OD is initialized; For train interruptions, passenger clearance and reversal, station passenger flow control, affected sections, and passengers departing from and arriving at interrupted sections, a solution space corresponding to different emergencies is established to generate a set of recommended travel plans for passengers. The travel cost of each path in the recommended travel plan set is simulated and calculated based on the current road network passenger flow distribution. Initialize the route recommendation plan, calculate the route selection probability, the number of passengers on the route and the fitness value based on the Logit model according to the travel cost of each route in the travel plan alternative set and the induced acceptance of different types of OD, update the passenger flow distribution of the road network, generate a new recommendation plan based on the Pareto optimal whale optimization algorithm, and determine the best travel plan for passengers through iteration; The solution space corresponding to different emergencies is established for the train interruption, passenger clearance and return, station passenger flow regulation, affected sections, and passengers departing from and arriving at the interruption section, and a recommended travel plan alternative set for passengers is generated. The travel cost of each path in the recommended travel plan alternative set is simulated and calculated according to the current road network passenger flow distribution, including: S21: For passengers who have to return due to train interruptions, the travel plan alternatives include waiting for the next train to board, taking other routes or changing transportation methods. The current road network passenger flow distribution is simulated to calculate the travel costs corresponding to the alternative plans. S22: For passengers who are subject to station passenger flow control, the travel plan alternative set is the routes under different passenger flow control, detours to other routes or other transportation modes. The different passenger flow control modes include inbound passenger flow control, transfer passenger flow control or train transfer skipping. Combined with the current road network passenger flow distribution, simulation is performed to calculate the travel cost corresponding to the alternative plan; S23: For passengers passing through the affected section, the travel plan alternative set is the route that the passengers originally planned to pass through the affected section, other detour routes or other transportation methods, combined with the current road network passenger flow distribution, to simulate and calculate the travel cost corresponding to the alternative plan; S24: For passengers departing from and arriving at the interruption section, the travel plan alternative set includes passengers waiting for the emergency to end, taking other routes, or choosing other transportation modes. According to the expected interruption duration and the current road network passenger flow distribution, simulation is performed to calculate the travel cost corresponding to the alternative plan; The initialization route recommendation scheme is based on the travel cost of each route in the travel plan alternative set and the induced acceptance of different types of OD, and the path selection probability, the number of passengers on the path and the fitness value are calculated based on the Logit model, the passenger flow distribution of the road network is updated, and a new recommendation scheme is generated using the whale optimization algorithm based on Pareto optimality. The best travel plan for passengers is determined through iteration, including: S31: For the travel mode alternative sets corresponding to different types of emergencies, determine the initial path recommendation scheme according to the difference degree of path travel costs, calculate the path selection probability using the Logit model, and combine the induced acceptance degree of this type of OD to calculate the distribution of passenger flow on the alternative paths; S32: Match the delay time of different types of passengers with the number of passengers of each type under the current road network passenger flow distribution, and calculate the minimum total passenger delay time. The calculation formula is: Where U is the set of all event types, P is the set of all passengers, and p u For passengers of type u, T u is the average delay time of type u events, and Z is the total delay time of passengers; Calculate the difference degree between the recommended cost calculated through the current road network passenger flow distribution and the simulation cost obtained through simulation after the passenger flow is redistributed according to the recommendation. The calculation formula for minimizing the difference degree is: In the formula, R is the set of all recommended paths, is the recommended cost calculated for path r, is the simulation cost of path r obtained through simulation, and E is the difference between the recommended cost and the simulation cost; S33: Update the road network passenger flow distribution according to the initial path recommendation scheme, the initial fitness value, and the number of passengers on each path. Generate partial solutions through the non-dominated genetic algorithm based on the whale optimization algorithm using the Pareto optimal theory, and screen out the relatively optimal solution set for the whale positions in the diverse whale optimization algorithm; Take the initial path recommendation scheme as the initial whale position, replace some non-optimal solutions in the whale population with the solutions in the relatively optimal solution set according to a certain proportion, and the replacement proportion changes dynamically with the number of iterations to update the whale position; Simulate according to the road network passenger flow distribution, update the path travel cost, map the recommendation scheme through the new whale position, and calculate the path selection probability, the number of passengers on the path, and the fitness; Compare the size of the total passenger delay time Z and the difference degree E between the recommended cost and the simulation cost in each round of fitness through the non-dominated relationship, update the road network passenger flow distribution, retain the optimal solution in each round, and gradually approach the global optimal solution by continuously iterating and updating the whale position; Based on the optimal solution of the whale optimization algorithm, determine the best travel plan for passengers. The calculation of the path selection probability using the Logit model and the combination of the induced acceptance degree of this type of OD in S31 to calculate the distribution of passenger flow on the alternative paths includes: Suppose there are three paths A, B, and C in the travel mode alternative set of this OD. The travel time costs of paths A, B, and C under the current road network passenger flow distribution are a, b, and c respectively. The path recommended by the initial path recommendation scheme is A. Then for the passengers on path B, assuming their induced acceptance degree is h (0 < h < 1), then h proportion of the passengers on path B will pay attention to the recommendation and make a choice decision between paths A and B. The selection probabilities of these passengers making a choice decision between paths A and B are: In the formula, θ is the time parameter; Similarly, the choice probabilities of passengers on path C between paths A and C are and Suppose the number of passengers passing through paths A, B, and C per unit time is m, n, and z respectively. This type of OD is a long-distance one with the destination station being the business district. Initialize the induced acceptance degree of this type of OD as h (0 < h < 1) according to the actual investigation. Then the distribution of passenger flow on the alternative paths is: Where N A 、N B 、N C are the number of passengers on paths A, B, and C respectively.

2. The method according to claim 1, characterized in that The method further includes: Obtain the recommendation result of the best travel plan for passengers through the post-induction evaluation, and update the induced acceptance degree of different types of OD based on the support vector machine regression SVR model.

3. The method according to claim 1, characterized in that The obtaining of the affected types of passengers according to the emergencies and the corresponding disposal measures classification includes: When a vehicle failure, power supply failure, platform screen door failure, communication failure, line failure, signal failure or an abnormally large passenger flow event occurs, corresponding handling measures are set up, including section interruption and route change, partial train route change, train delay or station passenger flow control; the affected passenger types are divided into the following four categories: passengers who need to return due to train interruption, passengers affected by station passenger flow control, passengers passing through the affected section, and passengers departing from and arriving at the interrupted section.

4. The method according to claim 1, characterized in that: The calculation of the passenger travel cost for train interruption, clearing and returning passengers in step S21 includes the following processing steps: Step 1: Based on the updated train operation plan, create a train set that includes the last N trains that arrive at the station and do not need to clear passengers and return; Step 2: Sort the passengers according to their current queue positions; Step 3: Traverse the train set. When the n'th train that does not need to clear passengers and turn around arrives, determine the current number of passengers that can get on the train based on the number of passengers getting off the station and the remaining capacity constraint of the train. At this time, the passenger waiting time is the difference between the train arrival time and the passenger arrival time, and the passenger delay time is the passenger waiting time. In this case, the passenger travel time cost is the sum of the historical travel time under normal circumstances and the passenger delay time of the path. Step 4: Update the passenger queue positions and re-sort them according to the confirmed number of boarding passengers; Step 5: Repeat steps 3 and 4 until all passengers are on board.

5. The method according to claim 1, characterized in that In step S22, when the passenger flow of the station exceeds the station carrying capacity, the passenger flow control is adopted. If it is a transfer station, the transfer flow limiting measures are adopted to calculate the travel cost of passengers under different passenger flow controls. The specific processing steps include: Step 1: Determine the number of people queuing in different directions according to the passengers' travel or transfer destinations. In each queue of the direction of travel, arrange the passengers in order according to their current queue positions, where the queue of passengers going up is S and the queue of passengers going down is L; Step 2: Refer to the historical control duration data of the entry or transfer passenger flow control, combine the dispatcher's real-time observation and estimation value, and comprehensively estimate the release duration of the entry or transfer passenger flow control; Step 3: Establish the set of arriving trains after the entry or transfer passenger flow control is completed, where the set of upbound trains that have arrived the most recently n' times is N, and the set of downbound trains that have arrived the most recently m' times is M, and the train set I = N∪M; Step 4: Traverse the train set I. When the i-th train arrives, calculate the number of passengers who can get on the train based on the number of passengers getting off the station obtained by simulation and the remaining capacity constraint of the train. At this time, the passenger travel time cost is the sum of the passenger flow control time, the queuing time and the historical travel time under normal circumstances of the path. The calculation formula is: T total =T con +T wait―on +T his Among them, T total is the passenger travel time cost, T con is the passenger flow control time, T wait―on The time for passengers to wait in line for boarding, T his is the historical travel time of the route under normal circumstances; Step 5: Update the passenger queue position according to the number of simulated boarding passengers. If the train belongs to the train set N, update the queue S; if the train belongs to the train set M, update the queue L. Step 6: Repeat steps 4 and 5 until all passengers are on board.

6. The method according to claim 1, characterized in that In step S22, in view of the large number of passengers at the transfer station, a train transfer skipping measure is adopted to calculate the passenger travel cost under the regulation measure. The specific processing steps include: Step 1: Determine whether the treatment measure type is a train transfer skipping measure. If so, execute steps 8 to 14. If not, execute steps 2 to 7. Step 2: Determine the number of people queuing in different directions according to the passengers' travel or transfer destinations. In each queue of the direction of travel, arrange the passengers in order according to their current queue positions, where the queue of passengers going in the upward direction is S, and the queue of passengers going in the downward direction is L; Step 3: Refer to the historical control duration data of the entry or transfer passenger flow control, combine the dispatcher's real-time observation and estimation value, and comprehensively estimate the release duration of the entry or transfer passenger flow control; Step 4: Establish the set of arriving trains after the entry or transfer passenger flow control is completed, where the set of upbound trains that have arrived the most recently d times is D, and the set of downbound trains that have arrived the most recently e times is E, and the train set I = D∪E; Step 5: Traverse the train set I. When the i-th train arrives, calculate the number of passengers who can get on the train based on the number of people getting off the station obtained by simulation and the remaining capacity constraint of the train. At this time, the passenger delay time is the sum of the passenger flow control time and the queuing waiting time, and the calculation formula is: T delay =T con +T wait―on Among them, T delay is the passenger delay time, T con is the passenger flow control time, T wait―on Queue passengers for boarding time; Step 6: Update the passenger queue position according to the number of simulated boarding passengers. If the train belongs to the train set D, update the queue S; if the train belongs to the train set E, update the queue L. Step 7: Repeat steps 5 and 6 until all passengers are on board; Step 8: Initialize station data, assume that the station that adopts the train transfer skipping measure is station A', and the next station of this station is station B'; Step 9: According to the adjusted train operation diagram, establish a train set F that includes the most recent f trains that pass through the passenger's station and do not stop at the transfer skip station, and a train set G that includes the most recent g trains that pass through the passenger's station and stop at the transfer skip station, and the train set I = F∪G; Step 10: Sort the passengers according to their current queue positions on the platform or on the train; Step 11: Determine the passenger type. Passenger types are divided into the following three types: ① For passengers who are already on the train and choose to get off at station k along the way and wait for the train that stops at station A', the delay time for such passengers is the sum of the time from getting off the train and finding the queue to the time of waiting in line to get on the train. The calculation formula is: In the formula, For passengers to get off at station k and find the time when the queue starts, is the waiting time for passengers to board the bus at station k; ② For passengers who are already on the train and choose to arrive at station B' and wait for the train to return to station A', the delay time for such passengers is the sum of the travel time of the train between stations A' and B', the time when the passengers get off at station B' and find the queue to start queuing, the time waiting in line at station B' to get on the train, and the travel time between stations B' and A'. The calculation formula is: Where, T A―B is the travel time of the train between station A' and station B'. For passengers to get off at stop B' and find out when the queue starts, The time that passengers wait in line at station B' for boarding; ③ Passengers waiting at the station. The delay time of these passengers is equal to the time they queue up and wait for the train at station l. The travel time cost of each type of passenger under passenger flow control is the sum of the passenger delay time and the historical travel time of the route under normal circumstances. The calculation formula is: T total =T delay +T his Step 12: Traverse the i-th train in the train set I. If train i belongs to the set M, calculate through simulation T A―B , Combined with the number of passengers getting off at the current station and the remaining capacity constraint of the train, determine the number of passengers who can get on the train at the current station. At this time, the affected passengers cannot get on the train, and update the queue order of the unaffected passengers after they get on the train; if train i belongs to set N, combined with the number of passengers getting off at the station and the remaining capacity constraint of the train, determine the number of passengers who can get on the train at the current station. At this time, all passengers can get on the train; Step 13: Update the passenger queue positions and re-sort them according to the simulated number of boarding passengers; Step 14: Repeat steps 11 to 13 until all affected passengers get off at station A'.

7. The method according to claim 1, characterized in that In step S23, the estimated route delay time, the risk of forced stop in the subsequent section of the route, and the passenger delay time are calculated in the following manner: The estimated path delay time is related to the length of time the train stops at a station or section, and the calculation formula is: In the formula, Estimate the duration of the forced stop of train number i, T event N is the estimated impact time of the emergency, stat is the number of platforms between the current position and the affected section, and V is the train running speed; The formula for calculating the risk of forced stop in the subsequent section of the route is: In the formula, The risk of forced stop of train number i in the subsequent section. D is the estimated forced stop time of the preceding train when an emergency occurs. i,i―1 is the distance between the two vehicles; The formula for calculating passenger delay time is: The travel time cost of passengers waiting for the emergency to end is the sum of the passenger delay time and the historical travel time of the path under normal circumstances. The calculation formula is: T total =T delay +T his T his is the historical travel time of the route under normal circumstances.

Images