Subway train operation optimization method, device and storage medium based on multi-unit

Abstract

The present invention relates to a subway train operation optimization method, device and storage medium based on multiple train sets. The method is aimed at the operation of subway trains under large and small route modes. By constructing a double-layer operation plan optimization model including an upper-layer multiple train set operation plan optimization model and a lower-layer multiple train set deployment plan adaptation model, the subway train operation optimization plan is obtained to complete the subway train operation optimization under large and small route modes. Compared with the prior art, the present invention has the significant advantages of improving the matching degree between transportation capacity and passenger flow demand and reducing the waste of transportation capacity resources by combining route and multiple train sets.

Description

Technical Field

[0001] The present invention relates to the technical field of urban rail transit, and in particular to a method, a device and a storage medium for optimizing the operation of a subway train based on multiple train sets. Background Art

[0002] The imbalance of passenger flow distribution in urban subway trains is mainly due to the imbalance of passenger flow time and passenger flow space. The multi-marshaling mode is an operation organization technology that designs and dispatches trains with different marshaling lengths in advance by the vehicle base to address the differences in passenger flow in different time periods or sections of subway trains, and does not disassemble or reconnect them during operation.

[0003] Since the multi-marshaling mode solves the imbalance in passenger flow time in subway train operation, it has limitations in solving the imbalance in passenger flow space. Therefore, it is necessary to combine long and short routes to adapt to the needs of different passenger flow spaces and achieve the adaptation of transport capacity and passenger flow demand in time and space.

[0004] Document 1: China's authorized invention patent CN103793760B discloses a method for optimizing the transfer allocation ratio of passenger flow within a multi-modal integrated transportation hub. Its main function is to maximize user travel satisfaction by considering factors such as user requirements for transfer walking and waiting time, and comfort of transfer passages. It does not provide favorable factors that match transport capacity with passenger flow, and cannot actually solve the problem of subway train operation. Summary of the invention

[0005] The purpose of the present invention is to provide a subway train operation optimization method, device and storage medium based on multi-units, based on travel passenger flow, combined with large and small route modes, to establish a double-layer model of multi-units, adopt a nested genetic algorithm to solve, obtain the optimal solution of the train operation plan, and realize the reasonable allocation of subway train operation resources.

[0006] The technical solution to achieve the purpose of the present invention is:

[0007] A subway train operation optimization method based on multiple train sets. The method aims at the operation of subway trains under large and small route modes. By constructing a double-layer operation plan optimization model including an upper-layer multi-train operation plan optimization model and a lower-layer multi-train deployment plan adaptation model, the subway train operation optimization plan is obtained, and the subway train operation optimization under the large and small route modes is completed.

[0008] Furthermore, if the imbalance coefficient of the line section during the peak period is greater than the set threshold, the subway train operation belongs to the large and small route mode.

[0009] Furthermore, a nested genetic algorithm is used to solve the double-layer operation scheme optimization model to obtain a subway train operation optimization scheme.

[0010] Furthermore, the optimization model of the upper multi-unit train operation plan is:

[0011] min Z 1 =λ 1 T w ′+λ 2 (V L ′+T r ′)

[0012]

[0013] Where: Z 1 represents the objective function of the optimization model of the upper multi-unit train operation plan; when e=1, it represents a large route, and when e=2, it represents a small route; T w ′ represents the normalized passenger travel cost; V L ′ represents the normalized train travel distance; T r ′ represents the normalized train running time; r represents the line section passed by the OD pair; f 1 Indicates the frequency of departure on major routes; f 2 Indicates the frequency of departure on a short route; f min Indicates the minimum departure frequency, f max Indicates the maximum departure frequency; J 1 Indicates the section running within the large route; J 2 Indicates the section where large and small routes run on the same line; J 1,r express OD pair set when J 12,r Indicates that r first passes through J 1 Then through J 2 OD pair set when J 11,r Indicates that r passes through J in sequence 1 , J 2 , J 1 OD pair set when J 2,r express OD pair set when J 21,r Indicates that r first passes through J 2 Then through J 1 OD pair set when Q e,r represents the cross-sectional passenger flow on the intersection r; n e represents the number of trains running on route e; θ e represents the train load factor on route e; θ max Indicates the maximum load factor of the train on route e; U e Indicates the number of vehicles on the route e; U max represents the maximum number of vehicles on the route e; C erepresents the number of vehicles on route e; N represents the total number of stations on the route; S 0 Indicates the starting station number of the short route, S 1 Indicates the terminal station number of the short route.

[0014] Furthermore, the adaptation model of the lower-level multi-unit train deployment plan is:

[0015]

[0016]

[0017] Where: Z 2 I represents the objective function of the lower-level multi-unit train deployment plan adaptation model; zz Indicates the train tracking interval; n max Indicates the maximum number of train formations; q ij Indicates the OD pair per unit time ( i,j ) is the one-way passenger flow within the line section passed by; I′ represents the departure interval between the small-route train and the large-route train when the small-route train is in front and the large-route train is behind; k represents the ratio of the number of departure trains of the small-route to the number of departure trains of the large-route.

[0018] Furthermore, the calculation formula of the unbalanced coefficient of the line section during the peak period is as follows:

[0019]

[0020]

[0021] Where: η s Indicates the unbalanced coefficient of line section during peak hours; d m(m+1) represents the cross-sectional passenger flow from station m to station m+1 in one direction during peak hours; q ij represents the one-way passenger flow in the route section that the OD pair (i, j) passes through per unit time; N represents the total number of stations on the large route; d max Indicates the maximum cross-sectional passenger flow in one direction during peak hours.

[0022] Furthermore, the nested genetic algorithm is used for solving the problem. The specific operation process is as follows:

[0023] S10: Initialize parameters, the parameters include:

[0024] Set the maximum number of population sizes for the upper-layer multi-unit train operation plan optimization model, randomly generate the departure frequency and departure ratio of the population individuals, set the maximum value of the upper-layer evolutionary generation and default the initial value of the upper-layer evolutionary generation counter to 0;

[0025] Set the maximum number of population sizes for the lower-level multi-unit train deployment plan adaptation model, the number of randomly generated units, and the population individuals of the departure interval, set the maximum value of the lower-level evolutionary generation, and default the initial value of the lower-level evolutionary generation counter to 0;

[0026] S20: Substitute the population of the initialized upper multi-unit train operation plan optimization model as known variables into the lower multi-unit train deployment plan adaptation model for genetic operation:

[0027] Obtain the objective function of the lower-layer multi-unit train deployment plan adaptation model, calculate the fitness value of the number of units and the departure interval, and perform genetic operations on it. When the number of iterations of the lower-layer multi-unit train deployment plan adaptation model reaches the preset lower-layer evolutionary generation maximum value, obtain the optimal solution of the lower-layer multi-unit train deployment plan adaptation model;

[0028] S30: Substitute the optimal solution of the lower-level multi-unit train deployment plan adaptation model as a known variable into the upper-level multi-unit train operation plan optimization model for genetic operation:

[0029] Obtain the objective function of the upper multi-unit train operation plan optimization model, calculate the fitness values ​​of the departure frequency and the departure ratio, and perform genetic operations on them to obtain the current feasible solution of the upper multi-unit train operation plan optimization model;

[0030] S40: When the number of iterations of the upper-layer multi-unit train operation plan optimization model reaches the preset upper-layer evolutionary algebraic maximum value, the current feasible solution of the upper-layer multi-unit train operation plan optimization model is the optimal solution of the operation plan optimization model, and the nested genetic algorithm solution is terminated and the iterative loop is stopped.

[0031] In addition, the present invention provides a computer-readable storage medium, wherein: a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the above-mentioned subway train operation optimization method based on multiple sets are implemented.

[0032] The present invention also provides a multi-unit subway train operation optimization device, including a memory for storing a computer program and a processor for implementing the steps of the above-mentioned multi-unit subway train operation optimization method when executing the computer program.

[0033] Compared with the prior art, the present invention has the following significant advantages: it adopts a multi-marshaling mode under large and small route modes, thereby balancing the problems of subway line section full load rate and departure interval time; it makes the deployment of transport capacity more precise; it improves the matching degree between transport capacity and passenger flow demand, thereby reducing the waste of transport capacity resources and reducing the operating costs of enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1It is a schematic diagram of the operating steps of the subway train operation optimization method based on multiple sets of the present invention.

[0035] Figure 2 It is a schematic diagram of the operation of a subway train on large and small routes of the present invention.

[0036] Figure 3 It is a schematic diagram of passenger flow classification types of subway train operation in the upward direction of the present invention.

[0037] Figure 4 It is a schematic diagram of balanced departure of large and small group subway trains according to the present invention.

[0038] Figure 5 It is a schematic diagram of the non-balanced departure of large and small-sized subway trains according to the present invention.

[0039] Figure 6 It is a schematic diagram of the flow of the nested genetic algorithm of the present invention. DETAILED DESCRIPTION

[0040] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0041] As shown in Question 1, the present invention provides a method for optimizing the operation of a subway train based on multiple train sets, comprising the following steps:

[0042] Obtain historical train operation data to determine the passenger flow and section imbalance coefficient of the line section during peak hours;

[0043] When the line section imbalance coefficient during the peak period is greater than the threshold, the line section is in an unbalanced state and the large and small cross-section mode is adopted;

[0044] Determine the turnaround station of the route and set up small route sections;

[0045] With the goal of minimizing the weighted sum of passenger travel costs and enterprise operating costs, an operation plan optimization model is constructed. The operation plan optimization model includes an upper-level multi-unit train operation plan optimization model and a lower-level multi-unit train deployment plan adaptation model.

[0046] The nested genetic algorithm is used to find the optimal solution for running the optimization scheme model.

[0047] Among them, the train operation history data includes train operation data, line data and passenger flow data: the train operation data includes the number of train formations, the travel time of different trains in each section and the original timetable data; the line data includes station information and station network importance; the passenger flow data is the subway transportation passenger flow; the importance of the station network is judged by the node betweenness of the station. When there are stations with large node betweenness in non-small route sections, the network service level of subway transportation is reduced.

[0048] Specifically, an operating day is divided into two periods, peak and off-peak. Based on the passenger flow statistics during the peak period, the cross-sectional passenger flow and cross-sectional imbalance coefficient of the line during the peak period are obtained. The calculation formula is:

[0049]

[0050]

[0051] Where: η s Indicates the unbalanced coefficient of line section during peak hours; d m(m+1) represents the cross-sectional passenger flow from station m to station m+1 in one direction during peak hours; q ij represents the one-way passenger flow in the route section that the OD pair (i, j) passes through per unit time; N represents the total number of stations on the large route; d max Indicates the maximum cross-sectional passenger flow in one direction during peak hours.

[0052] The line section imbalance coefficient η during peak hours s For analysis, when η s When it is greater than 1.8, the line section is in a very unbalanced state, and a large and small route mode is adopted for optimization. Considering the fairness demand of passengers on the line for the level of riding service, stations are set within the small route section, and the calculated value of the cross-sectional passenger flow is greater than 1 / 2 of the maximum cross-sectional passenger flow in one direction.

[0053] like Figure 2 As shown in the figure, the turning stations of the large and small routes are determined. The train departs from station 1 to station N in the upward direction, and vice versa in the downward direction. The large route train operates through the entire line, that is, the train goes from station 1 to station N and then turns back; the small route train operates in part of the section, that is, the train goes from station S 0 To site S 1 Then turn back; the departure frequency of the large route is f 1 , the departure frequency of the small route is f 2 , the dispatch ratio is 1:k, that is, f 2 =kf 1 And k is an integer multiple; the number of train sets on a large route is n 1 , the number of short-distance train sets is n 2 ; The section with only large roads is J 1 The section where the large and small routes run on the same line is J 2 .

[0054] Specifically, the passenger travel cost includes the passenger waiting time, and the calculation formula is as follows:

[0055]

[0056]

[0057] Where: I 0 Indicates the departure interval of trains on major routes; T w represents the passenger travel cost; represents the passenger flow from station p to station q in the upward direction, It represents the passenger flow from station p to station q in the downward direction.

[0058] Specifically, the enterprise operating cost includes the train running distance and train running time, and the specific calculation formula is as follows:

[0059] C VT =V L +T r

[0060]

[0061]

[0062] Where: C VT Represents the enterprise operating cost; V L Indicates the distance traveled by the train; T r represents the train running time; e represents the train route (e=1 is a large route, e=2 is a small route); f e represents the departure frequency of route e; n e is the number of trains running on route e; t x The duration of a single operating period is 1 hour; represents the length of the intersection e in the upward direction, Indicates the length of the intersection e in the downstream direction; t z f is the train turnaround time; represents the speed of intersection e in the uplink and downlink directions, Indicates the speed of route e in the downward direction. "+" indicates that the train is running in the upward direction; "-" indicates that the train is running in the downward direction.

[0063] like Figure 3 As shown in the figure, under the large-small route mode, the passenger flow shared by the large-route train and the small-route train is different. Taking the uplink direction as an example, the passenger flow is classified into six types; among them, the four types of Ⅰ, Ⅱ, Ⅲ, and Ⅵ can only take the large-route train, and the two types of Ⅳ and Ⅴ can take the large-small route train. The specific calculation formula is as follows:

[0064]

[0065]

[0066] Where: Q e,rrepresents the cross-sectional passenger flow on the intersection r; β 1 is the proportion of passengers who can only take long-distance trains, β 2 is the proportion of passengers who take the large and small trains without preference (taking the first train to arrive); r represents the line section between station i and station j; Indicates the ratio of departure intervals between small-route trains and large-route trains.

[0067] Specifically, J. 1 Indicates the section running in the large route, J 2 Indicates the section where large and small routes run on the same line. The passenger flow classification is as follows: J 1,r for OD pair set (interval r is the interval that the OD pair passes through; in the uplink direction, for example, when J 1 From Station 1 to Station S 0 When the section of the station is r, the passenger flow in section r includes types I, II, and III; when J 1 For S 1 When the section is from the station to the Nth station, the passenger flow in section r includes types III, V, and VI); J 12,r For r, first pass through J 1 After J 2 OD pair set (including the second and third passenger flows); J 11,r For r, first pass through J 1 and J 2 After J 1 OD pair set (Type III passenger flow); J 2,r for OD pair set (type IV passenger flow); J 21,r For r, first pass through J 2 After J 1 OD pair set (Type V passenger flow).

[0068] like Figure 4-5 As shown in the figure, the average waiting time of passengers in the small route section can be calculated by the departure interval ratio between the small route trains and the large route trains, where:

[0069] When the train is in balanced departure state:

[0070]

[0071]

[0072]

[0073] Where: I′ represents the departure interval between the small-route train and the large-route train (sequential relationship: the small-route train is in front and the large-route train is behind); t wrepresents the average waiting time of passengers in a short route section; α represents the average waiting time coefficient of passengers in a short route section.

[0074] When the train is in an unbalanced departure state:

[0075]

[0076]

[0077]

[0078]

[0079] Where: I″ is the departure interval time between the large-route train and the small-route train (sequential relationship: the small-route train is at the back and the large-route train is at the front), and I″≠I′.

[0080] Specifically, the passenger travel cost T is calculated using the linear weighting method. w , Train travel distance V L and train running time T r The normalization method is used to convert multiple targets into a single target. The specific calculation formula is as follows:

[0081]

[0082]

[0083]

[0084] Specifically, a weight coefficient is introduced to characterize the importance of passenger travel costs and enterprise operating costs, and the weight coefficient λ is introduced 1 Indicates the importance of passenger travel costs and introduces the weight coefficient λ 2 Indicates the importance of enterprise operating costs; set λ 1 +λ 2 =1,λ 1 and λ 2 The value of is determined by the enterprise based on the actual situation of the route; when the passenger satisfaction of the route is poor, the weight coefficient is set to λ 1 >λ 2 , when the line has a situation where the enterprise operating cost is too high, the weight coefficient is set to λ 1 <λ 2 .

[0085] Specifically, an optimization model for the operation plan of upper multi-unit trains is established, and the calculation formula is as follows:

[0086] min Z 1 =λ 1 Tw ′+λ 2 (V L ′+T r ′)

[0087]

[0088] Where: Z 1 represents the objective function of the optimization model of the upper multi-unit train operation plan; when e=1, it represents a large route, and when e=2, it represents a small route; T w ′ represents the normalized passenger travel cost; V L ′ represents the normalized train travel distance; T r ′ represents the normalized train running time; r represents the line section passed by the OD pair; f 1 Indicates the frequency of departure on major routes; f 2 Indicates the frequency of departure on a short route; f min Indicates the minimum departure frequency, f max Indicates the maximum departure frequency; Q e,r represents the cross-sectional passenger flow on the intersection r; n e represents the number of trains running on route e; θ e represents the train load factor on route e; θ max Indicates the maximum load factor of the train on route e; U e Indicates the number of vehicles on the route e; U max represents the maximum number of vehicles on the route e; C e represents the number of vehicles on route e; N represents the total number of stations on the route; S 0 Indicates the starting station number of the short route, S 1 Indicates the terminal station number of the short route.

[0089] Based on the upper-layer multi-unit train operation plan optimization model, in order to take into account the interests of both passengers and operating companies and ensure the matching degree between the deployed capacity and passenger flow demand, a lower-layer multi-unit train deployment plan adaptation model is built. The calculation formula is as follows:

[0090]

[0091]

[0092] Where: Z 2 I represents the objective function of the lower-level multi-unit train deployment plan adaptation model; zz Indicates the train tracking interval; n max Indicates the maximum number of train formations; q ijIt represents the one-way passenger flow in the line section passed by the OD pair (i, j) per unit time; I′ represents the departure interval between the small-route train and the large-route train when the small-route train is in front and the large-route train is behind; k represents the ratio of the number of departure trains of the small-route to the number of departure trains of the large-route.

[0093] In this embodiment, the AFC system (full name is Automatic Fare Collection System, which refers to the subway transportation automatic ticket collection system) is used to obtain q ij .

[0094] like Figure 6 As shown in the figure, the specific operation process of nested genetic algorithm solution is as follows:

[0095] S10: Initialize parameters, the parameters include:

[0096] Set the maximum number of population size pop_size_1 of the upper multi-unit train operation plan optimization model, and set the evolutionary algebraic counter δ according to the population individuals that randomly generate the departure frequency and departure ratio. 1 =0, maximum evolutionary generation max_generation_1, crossover probability Mutation probability

[0097] Set the maximum number of population size of the lower-level multi-unit train deployment plan adaptation model pop_size_2, and randomly generate population individuals with the number of units and departure intervals, and set the evolutionary algebra counter δ 2 =0, maximum evolutionary generation max_generation_2, crossover probability Mutation probability

[0098] S20: Substitute the population of the initialized upper multi-unit train operation plan optimization model as known variables into the lower multi-unit train deployment plan adaptation model for genetic operation:

[0099] The penalty function method is used to obtain the objective function of the lower-level multi-unit train deployment plan adaptation model, the fitness values ​​of the number of units and the departure interval are calculated, and genetic operations are performed on them. When the number of iterations of the lower-level multi-unit train deployment plan adaptation model reaches the preset lower-level evolutionary generation maximum value max_generation_2, the optimal solution of the lower-level multi-unit train deployment plan adaptation model is obtained, including the optimal number of units and departure interval time.

[0100] S30: Substituting the optimal solution of the lower-layer multi-unit train deployment plan adaptation model as a known variable into the upper-layer multi-unit train operation plan optimization model for genetic operation:

[0101] The penalty function is used to calculate the objective function of the upper-layer multi-unit train operation plan optimization model, the fitness values ​​of the departure frequency and the departure ratio are calculated, and genetic operations are performed on them to obtain the current feasible solution of the upper-layer multi-unit train operation plan optimization model, including the departure frequency and the departure ratio of the current feasible solution.

[0102] Determine the termination condition of the nested genetic algorithm. When the number of iterations of the upper multi-unit train operation plan optimization model reaches the maximum evolutionary generation max_generation_1, the current feasible solution of the upper multi-unit train operation plan optimization model is the optimal solution of the operation plan optimization model, and the nested genetic algorithm solution is terminated and the iterative cycle is stopped.

[0103] The genetic operations mentioned above include selection classification operation, crossover strategy operation and mutation operator operation.

[0104] The present application also provides a subway train operation optimization device based on multiple sets, including: a memory and a processor; a computer program is stored on the memory, and when the computer program is executed by the processor, the above-mentioned subway train operation optimization method based on multiple sets is implemented.

[0105] The present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned subway train operation optimization method based on multiple train sets are implemented. The computer-readable storage medium may include: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0106] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0107] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for optimizing subway train operation based on multiple train sets, characterized in that: The method aims at the operation of subway trains under the large-small route mode. By constructing a double-layer operation plan optimization model including an upper-layer multi-marshaling train operation plan optimization model and a lower-layer multi-marshaling train deployment plan adaptation model, the subway train operation optimization plan is obtained to complete the subway train operation optimization under the large-small route mode. If the imbalance coefficient of the line section during the peak period is greater than the set threshold, the subway train operation belongs to the large-small route mode; wherein: The optimization model of the upper multi-unit train operation plan is: Where: Z1 represents the objective function of the upper multi-unit train operation plan optimization model; e=1 represents a large route, and e=2 represents a small route; represents the normalized passenger travel cost; represents the normalized train travel distance; represents the normalized train running time; Indicates the line section that the OD pair passes through; Indicates the frequency of departure on major routes; Indicates the frequency of departure on a small route; Indicates the minimum departure frequency, Indicates the maximum departure frequency; Indicates the section running within the large route; Indicates the section where large and small routes run on the same line; express OD pair set at time; express First pass Then go through OD pair set at time; express Pass by , , OD pair set at time; express OD pair set at time; express First pass Then go through OD pair set at time; Indicates on the road Cross-sectional passenger flow; Indicates the number of train sets running on route e; Indicates the train load rate on route e; Indicates the maximum load factor of trains on route e; Indicates the number of vehicles in use on route e; Indicates the maximum number of vehicles used on route e; Indicates the number of vehicles on the intersection e; Indicates the total number of stations on the major routes; Indicates the starting station number of the small route. Indicates the terminal station number of the short route; The adaptation model of the lower-level multi-unit train deployment plan is: Where: represents the objective function of the lower-level multi-unit train deployment plan adaptation model; Indicates the train tracking interval; Indicates the maximum number of train formations; Indicates the OD pair per unit time One-way passenger flow within the route section; It indicates the departure interval between the small-circuit train and the large-circuit train when the small-circuit train is in front and the large-circuit train is behind; Indicates the ratio of the number of trains departing from small routes to the number of trains departing from large routes; The calculation formula of the unbalanced coefficient of the line section during the peak period is as follows: Where: Indicates the unbalanced coefficient of line section during peak hours; Indicates one-way travel from the station during peak hours To site The cross-sectional passenger flow; Indicates the OD pair per unit time One-way passenger flow within the route section; Indicates the total number of stations on the major routes; Indicates the maximum cross-sectional passenger flow in one direction during peak hours.

2. The subway train operation optimization method based on multiple train sets according to claim 1 is characterized in that: The nested genetic algorithm is used to solve the double-layer operation scheme optimization model to obtain the subway train operation optimization scheme.

3. The subway train operation optimization method based on multiple train sets according to claim 2 is characterized in that: The nested genetic algorithm solves the problem, and the specific operation process is as follows: S10: Initialize parameters, the parameters include: Set the maximum number of population sizes of the upper multi-unit train operation plan optimization model, randomly generate the departure frequency and departure ratio of the population individuals, set the maximum value of the upper evolutionary generation and default the initial value of the upper evolutionary generation counter to 0; Set the maximum number of population sizes of the lower-layer multi-unit train deployment plan adaptation model, the number of randomly generated units and the population individuals of the departure interval, set the maximum value of the lower-layer evolutionary generation and default the initial value of the lower-layer evolutionary generation counter to 0; S20: Substituting the initialized population of the upper multi-unit train operation plan optimization model as known variables into the lower multi-unit train deployment plan adaptation model for genetic operation: Obtain the objective function of the lower-layer multi-unit train deployment plan adaptation model, calculate the fitness value of the number of vehicles in the train set and the departure interval, and perform genetic operations on them. When the number of iterations of the lower-layer multi-unit train deployment plan adaptation model reaches the preset lower-layer evolutionary generation maximum value, obtain the optimal solution of the lower-layer multi-unit train deployment plan adaptation model; S30: Substituting the optimal solution of the lower-layer multi-unit train deployment plan adaptation model as a known variable into the upper-layer multi-unit train operation plan optimization model for genetic operation: Obtaining the objective function of the upper multi-unit train operation plan optimization model, calculating the fitness values ​​of the departure frequency and the departure ratio, and performing genetic operations on them to obtain the current feasible solution of the upper multi-unit train operation plan optimization model; S40: When the number of iterations of the upper-layer multi-unit train operation plan optimization model reaches the preset upper-layer evolutionary algebraic maximum value, the current feasible solution of the upper-layer multi-unit train operation plan optimization model is the optimal solution of the operation plan optimization model, and the nested genetic algorithm solution is terminated and the iterative loop is stopped.

4. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the subway train operation optimization method based on multiple sets as claimed in any one of claims 1 to 3 are implemented.

5. A multi-unit subway train operation optimization device, characterized in that: include: Memory for storing computer programs; A processor is used to implement the steps of the subway train operation optimization method based on multiple sets as described in any one of claims 1 to 3 when executing the computer program.

Images