Flexible train marshalling working diagram compiling method considering subway supply and demand matching

Abstract

The invention discloses a flexible train marshalling working diagram compilation method considering subway supply and demand matching, and belongs to the technical field of traffic transportation. According to passenger flow requirements changing along with time, coupling and uncoupling of different train units can be achieved at some stations. In view of the use of various train units, the coupling and uncoupling of the train units must follow the specific reorganization rules of marshalling stations. Wherein the mixed integer linear programming model aims to optimize a train bottom marshalling and train timetable, and also incorporates passenger distribution constraints to evaluate the number of passengers in train service or the number of passengers detained in a station, and reduce the number of passengers detained, the number of train marshalling and the operation cost. According to the method, part of calculation and constraint conditions are relaxed in an approximation mode so as to improve the calculation speed, a comprehensive optimization problem is decomposed into two sub-problems through a two-stage heuristic method, the two sub-problems are solved through a heuristic search method, and the operation cost can be reduced while the travel requirements of passengers can be met as much as possible.

Description

Technical Field

[0001] The present invention relates to the field of transportation technology, and in particular to a method for compiling a flexible train marshaling operation diagram taking into account subway supply and demand matching. Background Art

[0002] Urban rail transit is an indispensable mode of transportation in many cities. It is deeply favored by the majority of travelers for its safe, reliable and punctual service characteristics. The planning research on urban rail transit can be roughly divided into two categories: cost-oriented and demand / service-oriented train operation planning. In the past few decades, cost-oriented train operation planning has attracted attention first. Its core is to reduce operating costs such as labor costs, energy consumption, vehicle purchase and maintenance. In recent years, service-oriented planning has become increasingly popular, focusing more on improving passenger service levels, such as maximizing the total demand attracted by trains, minimizing passenger waiting / travel time, and avoiding passenger congestion. Many studies have reached similar conclusions: compared with a timetable with fixed departure intervals, a single stop mode and a fixed vehicle composition, a timetable with flexible departure intervals, flexible stop modes and flexible vehicle composition can provide better services to passengers and save more operating costs. Summary of the invention

[0003] The purpose of the present invention is to provide a method for compiling a flexible train marshaling operation diagram considering the matching of subway supply and demand, aiming to determine a train schedule and a vehicle dispatching plan considering a flexible vehicle marshaling strategy. Different from the traditional fixed train marshaling mode, the flexible vehicle marshaling strategy means that the train marshaling can be flexibly adjusted at selected stations according to the different passenger demands along the rail transit line, thereby meeting the passenger demands of different regions and time periods and further improving the vehicle utilization rate.

[0004] To achieve the above object, the present invention provides a method for compiling a flexible train marshaling operation diagram considering subway supply and demand matching, comprising the following steps:

[0005] S1. Obtain passenger flow information in urban rail transit and basic data for train operation diagram planning; the time-varying passenger flow data mentioned here refers to the data on the uneven distribution and time correlation of passenger arrivals in a specific time period due to the random nature of passenger arrivals, including: urban rail transit single-day automatic ticket vending and checking system data, etc.; the basic data for the preparation of the train operation diagram includes: train disassembly / coupling time, section running time, vehicle type, train stop time, maximum / minimum train running interval time, maximum / minimum return time;

[0006] S2. Establishing a train schedule optimization model under the flexible marshaling mode based on the basic data of the train operation diagram, solving it using the GUORBI commercial solver, and generating a flexible marshaling train schedule;

[0007] S3. Establish a collaborative optimization model of passenger flow distribution and vehicle turnover plan based on the flexible train schedule, use the GUORBI commercial solver to solve, and generate a vehicle turnover plan and a passenger flow distribution plan;

[0008] S4, recording the data related to the flexible marshaling train timetable, the vehicle bottom turnover plan and the passenger flow distribution plan, and generating an objective function evaluation index;

[0009] S5. Based on the evaluation index, through cyclic iterative optimization, the iterative target optimization value that meets the termination condition and the flexible marshaling train operation diagram compilation result are selected.

[0010] Preferably, according to the actual operation diagram data, under the condition that the interval running time and the stop time of the train service are fixed, a timetable optimization model under the flexible marshaling mode is established, which specifically includes:

[0011] The vehicle marshaling of a train service may be coupled and uncoupled during operation, and the vehicle marshaling of a train when it arrives at the terminal station may be different from the vehicle marshaling when it leaves the starting station. To facilitate calculation, the present invention decomposes a complete train service into multiple train journeys according to the number of selected marshaling stations and establishes a series of constraints to complete the conversion between train services and train journeys, so as to subsequently calculate the capacity of the train service in each section.

[0012] The variables in the timetable optimization part are the running intervals and reorganization times between train services. The algorithm starts with the initial departure locomotive and gradually reduces the maximum allowed number of locomotives to avoid passengers being stranded on the platform. The initial value of the maximum departure locomotive is set according to the actual maximum locomotive law (this is usually given by the subway company's operating guidelines to ensure a minimum service frequency for passengers). Usually, the maximum departure time during peak hours and in busy directions is smaller than during off-peak hours and in off-busy directions.

[0013] This type of timetable optimization method under flexible marshaling mode can optimize train arrival / departure times and vehicle marshaling to better meet time-dependent and unbalanced passenger demands.

[0014] Preferably, the schedule optimization model in the flexible grouping mode is a mixed integer programming model, and the specific modeling process is as follows:

[0015] Establish the constraints required for each stage of the model, including: train service running interval constraints, train service arrival and departure time constraints, train service reorganization time constraints at the marshaling yard, train service return time constraints, and establish a mapping between train departure time and departure period;

[0016] The objective function of the timetable optimization model is set to maximize the running interval between adjacent train services.

[0017] Preferably, the flexible marshaling technology takes into account that in addition to the coupling / uncoupling operations of trains at the starting and ending stations of urban rail transit lines (the starting and ending stations are linked to the yard), the coupling / uncoupling operations can also be performed at intermediate stations with multiple tracks. This greatly increases the flexibility of train marshaling, so that the generated train operation diagram can better meet the time-dependent and unbalanced passenger needs, and at the same time can effectively reduce the train operating costs.

[0018] Preferably, the passenger flow distribution and vehicle turnover plan collaborative optimization model established based on the flexible train schedule optimization result has the following specific modeling process:

[0019] The constraints required for each stage of the model are established, including: vehicle turnover constraints, marshaling yard capacity constraints, train service adaptation constraints, train service marshaling constraints, mapping constraints between train services and each train journey, and passenger flow allocation constraints.

[0020] According to the mathematical properties of the constraint conditions, the nonlinear constraint is converted into a linear constraint;

[0021] The objective functions of the timetable optimization model are set, including: minimizing the number of stranded passengers, minimizing the purchase cost of train units, and minimizing the operating cost of trains. Priorities are determined for these three objective functions: the first priority is to meet the travel needs of passengers, the second priority is to reduce the purchase cost of train units, and the third priority is to reduce the operating cost of trains.

[0022] Therefore, the present invention adopts the above-mentioned structure to prepare a flexible train marshaling operation diagram considering subway supply and demand matching, which has the following beneficial effects:

[0023] The present invention generates a train service marshaling plan and timetable that matches the time-space uneven distribution of constant passenger flow by establishing and solving a MILP model, so as to reduce the number of stranded passengers, the number of train units and the operating cost. In order to effectively solve the comprehensive optimization problem, this study proposes an approximation method and a two-stage heuristic algorithm, and combines the hybrid solution of the GUROBI commercial solver. The approximate method approximately relaxes some calculations and constraints to improve the calculation speed. The two-stage heuristic method divides the comprehensive optimization problem into two sub-problems, and uses a heuristic search method to solve the two sub-problems. In addition, when modeling, the equipment and facilities of each station on the line are fully considered, combined with the characteristics of passenger flow distribution, while ensuring the feasibility of the train bottom turnover plan, it can quickly generate a flexible marshaling operation diagram that meets the existing facilities and equipment of the line and can match the passenger flow. It can meet the operating company's requirements for the compilation of a flexible marshaling operation diagram to accurately deploy transportation capacity, reduce operating costs and improve service quality, which has very important theoretical guiding significance for the application of flexible marshaling in actual scenarios.

[0024] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A flow chart of a method for compiling a flexible train marshaling operation diagram in an urban rail transit line taking into account supply and demand matching provided by the present invention;

[0026] Figure 2 A diagram showing the relationship between train services and train journeys to which the present invention is applicable;

[0027] Figure 3 The figure is a schematic diagram of the process of coupling and uncoupling of train services to which the present invention is applicable. DETAILED DESCRIPTION

[0028] The technical solution of the present invention is further described below through the accompanying drawings and embodiments.

[0029] Unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0030] Example

[0031] like Figure 1-3 As shown, the present invention provides a method for compiling a flexible train marshaling operation diagram considering subway supply and demand matching, which is specifically as follows:

[0032] S1: Initialization, including initialization iteration number u=0, maximum train running interval Number of people stranded Reshuffling time of adjacent train journeys at marshaling yards Objective function weight value J1 *(0) =J2 *(0) =J3 *(0) =0,L (0) =0, etc., and input relevant train operation data.

[0033] S2: Establish a two-stage heuristic algorithm.

[0034] S3: Update the number of iterations u=u+1, update the train reorganization time and update the maximum train running interval.

[0035] S4: Establish a train schedule optimization model M2. The specific model is as follows:

[0036] In order to facilitate the calculation of the reorganization and arrival and departure of train services at each station, each complete train service i is cut into different train journeys g according to the number of marshaling stations.

[0037] Among them, to maximize the adjacent train (d g,χ and ) is the objective function, and the expression is as follows:

[0038]

[0039] The constraints are as follows:

[0040]

[0041]

[0042] Among them, m g,χ The train journey of station x before train journey g, g, g′ is the index of the train journey, T pair is the set of connected train trip pairs, (g,g′) is a set of connected train trip pairs, (g,g′)∈T pair ,s g-g′ It means that the train journeys g and g′ are connected at the marshaling yard s, s g-g′ ∈T pair , χ is the index of the terminal, S f is the set of terminal stations, K is the set of stations, K = {1, 2, ..., |K|}. k is the index of the station, k∈K. T s,v is the set of train journeys starting from station s in direction v, the set of train journeys ending at terminal χ, The set of train trips with the starting station being the terminal station x. D is the set of directions, D = {down, up}, v is the index of the direction, v∈D. g,s The arrival time of train trip g at station s (a g,s′ , a g,χ Similarly), d g,s The departure time of train trip g at station s, t0,t f is the service start and end time, t g-g' is the adaptation time between train journey g and train journey g′, is the minimum and maximum train running interval at station s, r g The running time for the train journey g is specified, dw g,χ is the stop time of train journey g at station x (dw g,s Similarly), m g,s The train journey of the station s before the train journey g, M is a very large positive number; η g,g′,χ is a binary variable. If train trip g and train trip g′ connect at station x (g is before g′), then η g,g′,χ =1. is the minimum and maximum return time of the train service at station χ, and H is the set of time intervals. is an auxiliary variable, equal to (d i,k -d i-1,k )·α i,k,h α i,k,h is a binary variable, indicating whether the time when train service i leaves station k is [t h-1 ,th ] time period. i,k is the departure time of train service i at station k. o is the index of station k in the marshaling station set S, if there is f(k,S) = 1. g i,o is the train journey corresponding to train service i when it leaves station o. is the operating time of the train service between stations k and k+1, dw k is the stop time of the train service at station k. v is the set of train services in direction v, and i is the index of the train service. h Indicates t h time.

[0043] The objective function (Formula (1)) is to maximize the adjacent train (d g,χ and ) and their departure interval cannot exceed the maximum departure interval specified by the station. (As in constraint (2)). Subject to actual operating conditions, the departure intervals of adjacent train services must be kept within a certain range. As shown in constraint (3), formula (4) ensures that the running time of each train journey between adjacent stations meets the specified running time r g Formulas (5) and (7) ensure that each train journey requires a fixed stop time at each station. In order to allow passengers to get on and off the train normally, formula (5) indicates that additional stop time needs to be considered at the marshaling yard to allow trains to reorganize. g-g' ,like Figure 3 Formula (6) is to ensure that the departure time of each train service must be within the experimental time period. Formulas (8)-(10) ensure that the return time a of the train service g',χ -d g,χ -t g-g' (where t g-g' Adjust the time for trains) within a reasonable range Formula (11)-(12) is (d i,k -d i-1,k )·α i,k.h The linearized formula of . The departure time of each train service at each station can be calculated by constraint (13). Formulas (14)-(17) are used to determine which time period the departure time of the two-way train services at their respective departure stations belongs to, so as to calculate the passenger flow in different time periods.

[0044] In this model, we assume that the running time and stop time of train services in each section are constant. Therefore, the variables in this part only include the train running interval and the reorganization time between each train journey.

[0045] S5: Use the GUROBI commercial solver to solve the timetable optimization model in S4 and solve the train turnaround parameters and parameters related to passenger flow distribution and And output parameters;

[0046] S6: Input the train turnaround related parameters and passenger flow distribution related parameters obtained in S5 into the passenger flow distribution and vehicle bottom turnover plan collaborative optimization model M3.

[0047] S7: Establish a collaborative optimization model M3 for passenger flow distribution and vehicle turnover plan. The specific model is as follows:

[0048] The objective function is to minimize the number of stranded passengers J1, minimize the fixed cost of the car bottom J2 (i.e. the cost required to purchase the car bottom), and minimize the operating cost of the car bottom (reducing the number of times the car bottom is used through flexible grouping) J3:

[0049] min J1,J2,J3(18);

[0050] Each train trip g can only have one car body type p (Formula (19)):

[0051]

[0052] And the length of its marshaling cannot exceed the platform length of the station (Formula (20)):

[0053]

[0054] Formula (21) calculates the number of train units in the current train journey:

[0055]

[0056] Formulas (22) and (23) ensure that the floor reconfiguration between adjacent train trips satisfies the reconfiguration rules of the station:

[0057]

[0058]

[0059] Formula (24) calculates the number of train unit deviations between adjacent train trips:

[0060]

[0061] η g,g',χ is a binary variable, indicating whether the train trip g is connected to the return train trip g' after turning back at station x. Formulas (25)-(26) ensure that each outbound train trip can only be connected to one return train trip after turning back:

[0062]

[0063] Formula (27) indicates that if the train journey g arrives at station x and the preceding train journey without reversal connects with it, then the bottom of the train comes from the garage and returns to the garage δ g',χ :

[0064]

[0065] Formula (28) If a train journey g arrives at station x and there is no subsequent train journey connecting with it, then the bottom of the train needs to return to the depot τ g,χ :

[0066]

[0067] Formulas (29)-(30) ensure that the connection between adjacent train trips after reversing at the reversing station complies with the station marshaling rules:

[0068]

[0069] The train unit quantity deviation between adjacent return train trips at the terminal station can be calculated by formula (33), where formula (31)-(32) is β g,p,g',p',χ ·η g,g',χ The linearization constraints are:

[0070]

[0071] Regarding the station vehicle capacity, first of all, it is necessary to ensure that the number of train units used on the line does not exceed the maximum number of available train units N. u (As constrained by (34)):

[0072]

[0073] In addition, the initial number of train units at each station in each direction and the number of train units remaining at the station after each train journey leaves the station cannot exceed the upper limit of train units specified in each direction. (as constrained by (35)-(36)), formula (37) calculates the train journey g at the ordinary marshaling station The stock quantity of train units at the depot after leaving is as follows:

[0074]

[0075] Formula (38) indicates that for the terminal station in one direction (v = up, if χ = s f ; v = down, if x = s0), only when the train journey chooses to return to the depot at this station Only then will the vehicle inventory change at the station be affected. The expression is as follows:

[0076]

[0077] Formulas (39)-(40) are for N g,u ·τ g,χ The linearization process is:

[0078]

[0079]

[0080] Similarly, for a starting point in one direction (v = down, if χ = s f ; v = up, if x = s0), when the bottom of the train journey g comes from the garage or from the return trip and adapted at this station (Z g,u ) will affect the vehicle inventory changes at the station:

[0081]

[0082] (42)-(43) are the nonlinear part N about train service departure g,u ·δ g,χ The linearization process is:

[0083]

[0084] In reality, the time spent on coupling and uncoupling of train services is different. Therefore, the reorganization status of the train service at this station should be determined according to the deviation of the number of train sets between two adjacent train trips, and the reorganization time should be calculated respectively, such as formulas (44)-(51). The specific expression is as follows:

[0085]

[0086]

[0087] Among the constraints related to passenger flow allocation, formula (52) is based on the arrival rate of each pair of OD(k,k') passengers at each time:

[0088]

[0089] The number of OD passengers for each pair arriving between the time when the last train service left the station and the time when the current train service leaves is accumulated. Constraints (53)-(54) are linearized constraints:

[0090]

[0091] Formula (55) is the total number of arriving passengers at the station:

[0092]

[0093] The number of waiting passengers includes the number of newly arrived passengers and the number of stranded passengers who were unable to take the previous train service (Formula (56)):

[0094]

[0095] The number of boarding passengers cannot exceed the total number of waiting passengers (constraint (57)):

[0096]

[0097] To ensure that the remaining capacity of the arriving trains cannot be exceeded (constraint (58)):

[0098]

[0099] The difference between waiting passengers and boarding passengers is the number of stranded passengers:

[0100]

[0101] The proportion of different ODs is calculated based on historical passenger flow data. Passengers with different ODs get on the bus in proportion, so the number of passengers getting off the bus can also be calculated in proportion:

[0102]

[0103] The number of passengers on each train is only related to the passengers getting on and off the train:

[0104]

[0105] Formula (62) is related to the calculation of the remaining capacity of train services. This value is only related to the number of people on board and the number of people getting off the train:

[0106]

[0107] Finally, since the train service may be reorganized at the marshaling yard, which will affect the capacity of the train service, it is necessary to determine the capacity of the train service in different sections through formula (63), which is as follows:

[0108]

[0109] S8: Use the GUROBI commercial solver to solve the vehicle turnover plan and passenger flow distribution plan respectively, and output the objective function value J1 *(u) ,J2 *(u) ,J3 *(u), Number of stranded passengers Coupling or uncoupling time

[0110] S9: Based on the output results of step S5 and step S8, calculate the optimization target value L of this iteration (u) =J1 *(u) +J2 *(u) +J3 *(u) .

[0111] S10: The loop termination is determined based on the solution generation and the number of loops. The determination conditions are as follows: (u) -L (u -1) |≤ε, that is, to determine the objective function value L under the current number of iterations (u) The objective function value L obtained from the previous iteration (u -1) Is the absolute value of the difference less than a minimum value ε?

[0112] Where K is the set of stations, K = {1, 2, ..., |K|}. k, k′, k″ are the indices of the stations, k, k′, k″∈K. s, s′ are the indices of the stations, s, s′∈S, s0,s f are the starting and ending stations of the line. U is a set of train types, u is the index of the train type, u∈U. P is a set of possible marshaling sets, p is the index of a train marshaling, p∈P; ε is a set of adaptation rules, ε={O,X,abO,aXb}, e is the index of the adaptation rule, e∈ε. Γ e is a set of marshaling pairs (p, p′)(p, p′∈P) such that the train marshaling changes from p to p′ after being reorganized according to the reorganization rule e∈ε. D is the set of directions, D={down,up}, v is the index of the direction, v∈D. I v is the set of train services in direction v, i is the index of the train service, i∈I v . T v is the set of train journeys in direction v. g,g′ are the indices of train journeys, g,g′∈T v . T pair is the set of connected train journey pairs. (g,g′) is a set of connected train journey pairs, (g,g′)∈T pair .s g-g′ To connect train trips g and g′ at marshaling yard s, s g-g′ ∈T pair .e s is the traffic diversion rule used at station s. f is the set of terminal stations, S f ={s0,s f}. χ is the index of the terminal station, χ∈S f .n p,u is the number of train units of type u in marshaling p. g,s,v represents the set of train trips that depart from station s and are scheduled to depart before train trip g in direction v. represents the set of incoming station train trips that are scheduled to arrive at the terminal station x before train trip g (including g). N represents the set of train trips that are scheduled to arrive at the terminal x before train trip g (including g). u is the total number of available U-train units. c ,t uc Indicates the train service coupling and uncoupling time. is the minimum and maximum turnaround time for a train serving at station χ. k,k',h For in [t h-1 ,t h ] period, the arrival rate of passengers from station k to station k'. k”,k is the proportion of passengers at station k” whose destination is station k. ζ is a positive time interval. c p is the standard passenger capacity of vehicle formation p. f(k,S) is the function for detecting whether station k is a marshaling station. ξ is the maximum full load rate allowed for the train. dw k is the stop time of the train service at station k. is the operating time of the train service between stations k and k+1. u is the cost of purchasing train unit u. π g,p The cost of using car set p for train journey g. is a binary variable, If a train unit is unpacked when the train journey g is converted to g', otherwise is a binary variable, If a train unit is coupled when the train journey g is converted to g', otherwise

[0113] x g,p is a binary variable indicating whether combination p is used for train journey g; N g,u is an integer variable representing the number of U-train units used in the train journey g; l p is the length of group g; L g is the length of the shortest platform corresponding to the train journey g; β g,p,g′,p′ is a binary variable, indicating whether the train journey g is the formation p, and whether the train journey g′ is the formation p′. g,u is an integer variable, indicating the number of trains that are connected before the train journey g (Zg,u >0) or decompilation (Z g,u <0). g,g′,χ is a binary variable. If train trip g and train trip g′ connect at station x (g is before g′), then η g,g′,χ =1. g,χ is a binary variable. If the train trip g leaves from station x, then δ g,χ =1. τ g,χ is a binary variable. If the train trip g enters the depot at station x, then τ g,χ =1. is an auxiliary variable, equal to β g,p,g′,p′ ·η g,g′,χ . is the number of U-type train units stored at station s in direction v after the train journey g departs. is the number of initial u-train units stored at station s in the v direction. is an auxiliary variable, equal to N g,u ·τ g,χ . is an auxiliary variable, equal to N g,u ·δ g,χ . is the number of passengers who arrive at station k and go to k' between the time when train service i-1 and i leave. i,k,h is a binary variable, indicating whether the time when train service i leaves station k is [t h-1 ,t h ] time period. is an auxiliary variable, equal to (d i,k -d i-1,k )·α i,k,h . represents the number of passengers arriving at station k between the time when train service i-1 and the time when train i leaves. represents the number of people waiting at station k to take train service i. represents the number of passengers who boarded train service i at station k and left the station. It represents the number of passengers remaining in the train when train service i arrives at station k and the passengers have already got off. It represents the number of passengers in the train after train service i leaves station k. The number of passengers getting off at station k for train service i. represents the number of stranded passengers in the station after train service i leaves station k. represents the maximum passenger capacity of train service i after leaving station k. i,k represents the departure time of train service i at station k.

[0114] 1) The objective function value L at the current number of iterations (u) The objective function value L obtained from the previous iteration (u-1) If the absolute value of the difference is less than ε, jump to step S3;

[0115] 2) If the above termination conditions are not met, exit the loop and output all current solutions.

[0116] Therefore, the present invention adopts the above-mentioned method for compiling a flexible train formation operation diagram that takes into account the matching of subway supply and demand. It can be applicable to scenarios where the temporal and spatial distribution of passenger flow is uneven and the transportation capacity does not match the demand. It flexibly changes the train formation length at specific stations to match the passenger flow demand in different regions and time periods, while improving the vehicle utilization rate and ensuring the feasibility of the vehicle bottom turnover plan, providing a reference basis for operating companies in the practical application of flexible formation technology.

[0117] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for compiling a flexible train marshaling operation diagram considering subway supply and demand matching, characterized by: The following steps are involved: S1. Obtain passenger flow information in urban rail transit and basic data for train operation diagram planning; S2. Establish a train schedule optimization model under the flexible marshaling mode based on the basic data of the train operation diagram, use the GUORBI commercial solver to solve, and generate a flexible marshaling train schedule; S3. Establish a collaborative optimization model of passenger flow distribution and car base turnover plan based on the flexible train schedule, use the GUORBI commercial solver to solve, and generate the car base turnover plan and passenger flow distribution plan; S4, recording the data related to the flexible marshaling train timetable, the vehicle bottom turnover plan and the passenger flow distribution plan, and generating the objective function evaluation index; S5. Based on the evaluation index, through cyclic iterative optimization, the iterative target optimization value that meets the termination conditions and the result of the flexible marshaling train operation diagram compilation are selected.

2. The method for compiling a flexible train marshaling operation diagram considering subway supply and demand matching according to claim 1, characterized in that: In step S1, the passenger flow information in urban rail transit includes: the data of the urban rail transit automatic ticket vending and checking system for a single day; the basic data for compiling the train operation diagram includes: train disassembly / coupling time, section running time, vehicle type, train stop time, maximum / minimum train running interval time, maximum / minimum return time.

3. The method for compiling a flexible train marshaling operation diagram considering subway supply and demand matching according to claim 1, characterized in that: In step S2, the schedule optimization model under the flexible marshaling mode is established as follows: Breaking down a complete train service into multiple train journeys based on the number of marshaling yards selected; Establish a series of constraints to complete the mapping between train services and train journeys and calculate the capacity of train services in each section; The variables in the timetable optimization part are the running intervals and reorganization times between train services; The optimal starting head is found from the initial starting head, and the maximum allowed value of the head is gradually reduced. The initial value of the maximum starting head is set according to the actual maximum head law.

4. The method for compiling a flexible train marshaling operation diagram considering subway supply and demand matching according to claim 1, characterized in that: In step S3, the timetable optimization model under the flexible grouping mode is a mixed integer programming model, and the specific modeling process is as follows: Establish the constraints required for each stage of the model, including: train service running interval constraints, train service arrival and departure time constraints, train service reorganization time constraints at the marshaling yard, train service return time constraints, and establish the mapping between train departure time and departure period; The objective function of the timetable optimization model is set, and the objective function includes maximizing the running interval between adjacent train services.

5. The method for compiling a flexible train marshaling operation diagram considering subway supply and demand matching according to claim 1, characterized in that: In step S4, a passenger flow distribution and car bottom turnover plan collaborative optimization model is established based on the flexible train schedule optimization result. The specific modeling process is as follows: Establish the constraints required for each stage of the model, including: vehicle turnover constraints, marshalling yard capacity constraints, train service adaptation constraints, train service marshalling constraints, mapping constraints between train services and train journeys, and passenger flow distribution constraints. According to the mathematical properties of the constraints, the nonlinear constraints are transformed into linear constraints; Set the objective function of the schedule optimization model, including: Minimize the number of stranded passengers, minimize the purchase cost of train units, and minimize the operating cost of the train.

Images