Train timetable and train bottom scheduling collaborative optimization method based on flexible marshalling mode

Abstract

The invention provides a train timetable and train bottom scheduling collaborative optimization method based on a flexible marshalling mode. The method comprises the steps of obtaining high-speed railway line conditions and passenger flow demand data in an operation period; taking minimization of total passenger travel time as a target, based on route conditions and passenger flow demand data, generating a current optimal solution through a demand-oriented greedy heuristic algorithm, the current optimal solution including an initial train timetable and an initial train bottom scheduling scheme; and performing iterative optimization on the current optimal solution through a variable neighborhood search algorithm within a preset iteration frequency range, and performing disturbance and local search operation on the current optimal solution so as to determine a target train timetable and a target train bottom scheduling scheme which meet the requirement of the optimal benefit of the railway system. According to the method, the optimal scheme for obtaining the time table and the vehicle bottom scheduling in a short time is realized, so that the comprehensive improvement of the high-speed railway operation in three dimensions of service quality, resource utilization rate and operation economy is realized.

Description

Technical Field

[0001] This invention relates to the field of transportation technology, and in particular to a method for collaborative optimization of train timetables and rolling stock scheduling based on flexible formation modes. Background Technology

[0002] As high-speed rail moves towards intelligent operation, accurately matching dynamically changing passenger flow demands with efficient utilization of limited transportation resources has become a core challenge in high-speed rail operation planning. Traditional fixed-formation trains often face the dual dilemma of peak-hour capacity shortages and off-peak resource idleness when dealing with uneven passenger flow in time and space. Flexible formation technology, with its ability to dynamically adjust train formation lengths and optimize capacity allocation in real time, is becoming a key breakthrough in solving this problem. It not only achieves "on-demand formation" through physical coupling, improving transportation efficiency and resource utilization, but also deeply integrates with cutting-edge technologies such as intelligent scheduling and automatic driving, injecting core momentum into the evolution of the high-speed rail system towards "safety, efficiency, greenness, and intelligence," thus attracting significant industry attention.

[0003] The coordination between fluctuating passenger demand and railway transport services has attracted significant attention from both academia and industry. Adjusting train formation is a common approach. Currently, there are two main train formation modes: fixed formation and flexible combination. In the fixed formation mode, the number of train carriages remains constant and does not change with fluctuations in passenger demand. This mode is favored due to its ease of operation and has become the primary formation mode. However, it fails to consider the differences in passenger demand across time and space. During periods of low passenger demand, trains carry a large number of empty seats; while during peak demand periods, seating capacity is significantly insufficient. Unlike the fixed formation mode, the flexible combination mode allows trains to adjust the number of carriages at the line's terminus. This adjustment is achieved by coupling trains together or splitting longer trains into shorter ones. Trains adjust the number of carriages according to time-varying passenger demand. During peak periods, trains increase the number of carriages by coupling with other trains; while during off-peak periods, the number of carriages is reduced by splitting formations. It is worth noting that the flexible combination mode primarily addresses the imbalance of passenger demand over time and cannot effectively correct spatial differences in passenger demand distribution. For example, when passenger demand varies significantly across different sections of a railway line, flexible train formation modes struggle to address the spatial imbalance in passenger demand. Therefore, this invention proposes a flexible train formation mode that further considers splitting or suspending carriages at intermediate stations along the line, and optimizes the utilization rate of transportation resources by combining flexible stopping, turnaround, and overtaking strategies. Train timetables and rolling stock scheduling are crucial issues in high-speed railway operation and management. This invention provides a method for the collaborative optimization of high-speed railway train timetables and rolling stock scheduling.

[0004] Currently, algorithms for solving the optimization problems of high-speed railway train timetables and rolling stock coordination fall into two categories: one relies on commercial optimization software, which establishes relatively simple mathematical optimization models and uses commercial software such as CPLEX and GROBI to solve them. This type of method cannot quickly solve optimization problems at the train timetable level, thus limiting its practical application. The other type uses heuristic algorithms to solve optimization models. These models can quickly obtain a feasible train timetable and rolling stock solution, making them more suitable for real-world applications. Currently, designing reasonable and effective heuristic algorithms to solve the optimization problems of train timetables and rolling stock coordination is still in the exploratory stage, requiring more technological support to meet the needs of real-world scenarios.

[0005] In summary, during the intelligent operation phase of high-speed railways, accurately matching dynamic passenger flow demand with efficient utilization of transportation resources is the core challenge of high-speed railway operation planning. Traditional fixed-formation train configurations struggle to address the spatial and temporal imbalances in passenger flow. Flexible combination technology, with its ability to dynamically adjust train length and optimize capacity allocation in real time, has become a key breakthrough. It not only improves transportation efficiency and resource utilization but also deeply integrates with cutting-edge technologies to drive the evolution of the high-speed railway system, attracting significant industry attention. However, it's important to note that while current flexible combination models can alleviate passenger flow imbalances on a temporal scale, they are insufficient to effectively address the spatial distribution differences in passenger demand. Furthermore, current train timetable and rolling stock scheduling coordination schemes typically utilize mathematical optimization models, such as mixed-integer programming models and linear optimization models. These optimization models cannot quickly find optimal solutions in large-scale operational scenarios. Additionally, solving existing mathematical optimization models heavily relies on commercial optimization solvers and computer hardware, resulting in time-consuming calculations and significant hardware costs. Therefore, it is impossible to quickly and stably obtain coordinated design schemes for train timetables and rolling stock scheduling, hindering the ability of high-speed railway operators to respond swiftly. Summary of the Invention

[0006] This invention provides a method for coordinated optimization of train timetables and rolling stock scheduling based on flexible formation modes, in order to overcome the shortcomings of the prior art.

[0007] According to a first aspect of the present invention, a method for collaborative optimization of train timetables and rolling stock scheduling based on flexible formation modes is provided, comprising the following steps: S110, acquiring passenger flow demand data during the operation period and the conditions of the high-speed railway line, wherein the passenger flow demand data includes the origin and destination of passenger flow, the earliest boarding time, the latest boarding time, and the number of passengers; S120, with the objective of minimizing the total travel time of passengers, generating a current optimal solution based on the line conditions and the passenger flow demand data using a demand-oriented greedy heuristic algorithm, wherein the current optimal solution includes an initial train timetable and an initial rolling stock scheduling scheme; S130, iteratively optimizing the current optimal solution within a preset number of iterations using a variable neighborhood search algorithm, wherein the iterative optimization includes perturbing and local search operations on the current optimal solution to determine a target train timetable and a target rolling stock scheduling scheme that satisfies the optimal benefits for the railway system, wherein the current optimal solution is updated in each iteration.

[0008] Optionally, the step of generating the current optimal solution based on the route conditions and the passenger flow demand data using a demand-oriented greedy heuristic algorithm includes: S11, determining the train scheduling scheme: by sequentially scheduling each train... u ∈ U Insert train sequence The system assigns an even number of train services to each train set, restricts train services to turning back only at the line's terminus, and determines the train service sequence. and turnaround station sequence S12. Compile train timetables: By setting the arrival / departure times of the first train at both station 1 and station |R| to 10 ... Subsequent locomotives will be set to depart at the minimum interval. Depart in sequence to determine departure time sequence By passing each car bottom m ∈ M The train formation scheme is determined so that shunting operations are only performed on the first and last trains at their initial stations, and no skipping or overtaking operations are performed at any station. Stopping Plan and crossover matrix S13. Adjust train timetable: Sequence of train services. Turnaround station sequence Train formation scheme Stopping Plan Information converted into train timetables and The system determines whether the current train timetable meets the departure interval constraint and the maintenance time constraint. If it does, the current train timetable is feasible; otherwise, if the departure interval constraint is not met, the departure time of the train sets is adjusted until the departure interval constraint is met; if the maintenance time constraint is not met, the train schedule is adjusted from the train task sequence. Remove trains that are outside their operating hours; repeat the above steps until all departure interval constraints and maintenance time constraints are met, then update the departure time series. Value; S14, Carriage Allocation: Allocate carriages to the terminal station of the line. r ( r ={1,| R The carriages are evenly distributed among the departing train sets according to the train number and task sequence. Update the number of carriages for the remaining train services, and then update the train formation matrix. S15. Passenger Allocation: The train scheduling scheme and train timetable determined through steps S11-S14 are used as the initial train scheduling scheme and the initial train timetable. Based on the initial train scheduling scheme and the initial train timetable, the passenger allocation scheme is optimized by solving the passenger allocation model. S16. Output Current Optimal Solution: The initial train scheduling scheme and the initial train timetable are output as the current optimal solution.

[0009] Optionally, in step S13, the following method is used to determine whether the current train timetable meets the departure interval constraint and maintenance time constraint: by judging the arrival and departure times of different trains at the station, it is determined whether the current train timetable meets the departure interval constraint: if the arrival and departure time interval of any two trains at the same station is not less than [a certain value]... If the current train timetable meets the departure interval constraint, then the current train timetable does not meet the departure interval constraint; otherwise, the current train timetable does not meet the departure interval constraint. This is achieved by determining whether the arrival and departure times of all trains fall within the train operating hours. Within this period, determine whether the current train timetable meets the maintenance time constraint: if the arrival and departure times of all trains fall within the train operating hours. If the current train timetable meets the maintenance time constraint, then the current train timetable does not meet the maintenance time constraint; otherwise, the current train timetable does not meet the maintenance time constraint.

[0010] Optionally, the target train timetable and target rolling stock scheduling scheme that satisfies the optimal benefits of the railway system are determined through the following steps: S41, Input the current optimal solution including the initial train timetable and the initial rolling stock scheduling scheme, with the preset algorithm iteration number as N_iteration, and let n=1, where n is the iteration number index; S42, Perform a perturbation operation on the current optimal solution to obtain several first new solutions, and find the first new solution that satisfies the optimal benefits of the railway system from the several first new solutions to update the current optimal solution, thus obtaining the optimal solution after the first update; S43, Perform a local search operation on the optimal solution after the first update to find the second new solution that satisfies the optimal benefits of the railway system, thus updating the optimal solution after the first update, thus obtaining the optimal solution after a second update; S44, Let n=n+1, and determine whether n is greater than N_iteration. If n is greater than N_iteration, output the optimal solution after the second update as the current optimal solution; otherwise, execute step S42 until the algorithm converges.

[0011] Optionally, in step S42, the disturbance operation includes four types of disturbance operators, wherein: the first type of disturbance operator acts on the train task sequence. The specific operations include: randomly canceling the last few train trips of a certain train set, and completely swapping the task sequence. Two lines of data, partial exchange task sequence Two lines of data; the second type of perturbation operator is applied to the turnaround station sequence. The specific steps include: randomly selecting a train number. u Define array Storage train number u For stations that can turn around, randomly select one from an array. Station turnaround; the third type of disturbance operator applies to the station stopping scheme. The specific steps include: randomly selecting a train number. u Define the set of stopping schemes that change only one station as Train number u The stop plan has been changed to a set of stop plans. One of the stopping schemes; the fourth type of disturbance operator applied to train formation. and The specific operations include: randomly selecting train formations. B One of the train numbers u station r As the disturbance location; if the value of the disturbance location is 1, then set it to 0, and set... Z Train numbers u station r The value is 0; otherwise, set it to 1, and set... Z Train numbers u stationr The number varies within train formation and station restrictions.

[0012] Optionally, in step S43, the execution probabilities of the four types of perturbation operators are calculated as follows:

[0013] in, Indicates the perturbation operator i In the n Execution probability at the next iteration Indicates the perturbation operator i In the j The improvement rate of the objective function in the next iteration.

[0014] Optionally, in step S43, the local search operation includes two types of local search operators, wherein: the first type of local search operator acts on the vehicle's departure time. For any car underside m The local search range for departure time is . and They represent the bottom of the car. m The departure times of the preceding and following car bodies; the second type of local search operator acts on the overtaking matrix. Randomly select a train number u and the station r As the perturbation location. Define a new array. Stored at the station r The departure time is later than the train number u Train numbers. Select the array sequentially. A train number v Let the train number u The departure time is Variations within a range. Among them... w Train number is train number v At the station r Subsequent train services.

[0015] Optionally, the execution probabilities of the two types of local search operators are calculated as follows:

[0016] in, Represents the local search operator i In the n Execution probability at the next iteration Represents the local search operator i In the j The improvement rate of the objective function in the next iteration, wherein the optimal benefit of the railway system is set as the objective function.

[0017] According to a second aspect of the present invention, an electronic device is provided, including a processor and a memory storing a program. The program includes instructions that, when executed by the processor, cause the processor to perform the steps performed by the method of the first aspect described above.

[0018] According to a third aspect of the present invention, a computer storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method of the first aspect described above.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: The train timetable and rolling stock scheduling method based on flexible formation mode provided by this invention is a heuristic algorithm of variable neighborhood search. Based on the provided train timetable and rolling stock scheduling scheme, users can plan new train timetables and rolling stock scheduling schemes according to the current high-speed railway lines and passenger demand. Moreover, this method can obtain high-quality train timetables and rolling stock scheduling schemes in a short time, which is beneficial to the adjustment and optimization of the train operation diagram in the later stage, and helps to achieve a comprehensive improvement in the three dimensions of high-speed railway operation: service quality, resource utilization and operation economy. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0021] Figure 1 This is a flowchart of the steps of the method of the present invention.

[0022] Figure 2 The flowchart shows the demand-driven greedy heuristic algorithm for the method of this invention.

[0023] Figure 3 This is a schematic diagram of the loss sharing strategy used in the method of the present invention. Detailed Implementation

[0024] To provide a clearer understanding of the technical features, objectives, and effects of the embodiments of the present invention, specific implementation methods of the embodiments of the present invention will now be described with reference to the accompanying drawings.

[0025] In this document, “exemplary” means “serving as an example, illustration or description”, and any illustrations or implementations described herein as “exemplary” should not be construed as a more preferred or advantageous technical solution.

[0026] To enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art should fall within the protection scope of the present invention.

[0027] The specific implementation of the embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0028] See Figure 1 As shown, in the scenario of sudden interruption of urban rail transit, the train timetable and rolling stock scheduling collaborative optimization method based on flexible formation mode provided by this invention includes the following steps: Step S110: Obtain high-speed railway line conditions, passenger flow demand data during operating hours, and related parameters.

[0029] First, determine the conditions of the high-speed railway line. For example, the conditions of a high-speed railway line include the set of stations R={1,...,|R|}, and the travel time between sections of the line. and stop time .

[0030] In one embodiment, the upward direction of the railway line is defined as the direction in which the station number increases, and the downward direction is the direction in which the station number decreases.

[0031] In addition, other important parameters extracted include passenger demand data and other characteristic parameters obtained through processing route conditions and passenger demand data, such as calculating the total travel time of passengers. Passenger demand data includes passengers' origin, earliest and latest departure times, and the number of passengers.

[0032] Step S120: The initial train timetable and train scheduling scheme are obtained by using a demand-oriented greedy heuristic algorithm.

[0033] The demand-driven greedy heuristic algorithm first determines the train sequence to be executed by the train sets, setting a headway constraint for each train set's departure. Train sets are evenly distributed across each set according to the number of carriages at the departure station. This operation guarantees the generation of a train timetable and train set scheduling scheme. However, some trains may still not meet the headway constraint. Simultaneously, some trains may have operating times exceeding the train operating hours, thus failing to meet the train maintenance time constraint. Based on this, the demand-driven greedy heuristic algorithm can adjust the train set departure times to satisfy the minimum headway constraint and cancel some trains to satisfy the train maintenance time constraint. This ultimately yields an initial feasible solution, i.e., the initial train timetable adjustment scheme.

[0034] See Figure 2 The initial train timetable and initial rolling stock scheduling scheme are obtained using a demand-driven greedy heuristic algorithm, which includes the following steps: Step S11: Determine the train scheduling plan. Sequentially schedule each train number... u ∈ U Insert train sequence And ensure that the number of train services assigned to each train set is even. Restrict train services to only turn back at the line's terminus. At this point, the train service sequence... and turnaround station sequence It can be determined.

[0035] It should be understood that in step S11, the set U This represents the set of train numbers.

[0036] Step S12, compile the train timetable. Set the arrival / departure times of the first train at station 1 and station |R| to be... Subsequent locomotives will depart at the minimum interval. Trains depart in sequence. This determines the departure time sequence. Under each car m ∈ M Shunting operations are only performed on the first and last trains at their initial stations, and skipping or overtaking operations are not performed at any station. The train formation scheme is determined accordingly. Stopping Plan and crossover matrix .

[0037] It should be understood that in step S12, the train's operating hours are described as follows: .in, and These represent the start and end times of daily railway operations, respectively. (Set) M This indicates a collection of items under the vehicles.

[0038] Step S13, adjust the train timetable. Sequence the train numbers. Turnaround station sequence Train formation scheme Stopping Plan Information converted into train timetables and The system determines whether the current train timetable meets the departure interval constraint and the maintenance time constraint. If it does, the current train timetable is feasible; otherwise, if the departure interval constraint is not met, the departure time of the train sets is adjusted until the departure interval constraint is met; if the maintenance time constraint is not met, the train schedule is adjusted from the train task sequence. Remove trains that are outside their operating hours; repeat the above steps until all departure interval constraints and maintenance time constraints are met, then update the departure time series. The value of .

[0039] Step S14, Car Assignment. Assign the terminal station of the line. r ( r ={1,| R The carriages are evenly distributed among the departing train sets. This is based on the train's task sequence. Update the number of carriages for the remaining train services, and then update the train formation matrix. .

[0040] It should be understood that in step S14, at the line's terminus station (i.e., station 1 and station | R |) The number of carriages on the departing train is .in, This indicates the maximum number of carriages that can be assembled from the train's structure. Indicates at the station r The initial number of carriages.

[0041] Step S15, Passenger Allocation. The train scheduling scheme and train timetable determined in steps S11-S14 are used as the initial train scheduling scheme and the initial train timetable. Based on the initial train scheduling scheme and the initial train timetable, the passenger allocation scheme is optimized by solving the passenger allocation model (M1).

[0042] Step S16: Output the initial train scheduling scheme and the initial train timetable as the current optimal solution.

[0043] Specifically, in Figure 2In this embodiment, an initial train timetable and initial rolling stock scheduling scheme are first obtained using a greedy strategy (steps S11-S12). However, this solution is not guaranteed to be feasible, so it is necessary to determine whether it satisfies the locomotive headway constraint and the vehicle maintenance time constraint. If satisfied, the solution is an initial feasible solution, and the algorithm terminates. Otherwise, trains need to be canceled or the departure time of the rolling stock needs to be adjusted until the train timetable and rolling stock scheduling scheme satisfy the locomotive headway constraint and the vehicle maintenance time constraint. At this point, the algorithm terminates, and the train timetable and rolling stock scheduling scheme that satisfies the locomotive headway constraint and the vehicle maintenance time constraint at this time are output as the current optimal solution.

[0044] Step S130: Use the variable neighborhood search algorithm to determine the optimal train timetable and rolling stock scheduling scheme, i.e., determine the target train timetable and target rolling stock scheduling scheme.

[0045] Since step S130 only obtains an initial train timetable and an initial rolling stock scheduling scheme, this initial feasible solution, although it is the current optimal solution, cannot guarantee that it is the optimal train timetable and the optimal rolling stock scheduling scheme.

[0046] Based on this, the present invention further proposes a heuristic algorithm based on variable neighborhood search, which improves the current train timetable and train scheduling scheme by gradually optimizing.

[0047] For example, the following scenario assumptions are made regarding the applicability of this algorithm: (1) The train's travel time and stop time are fixed during operation.

[0048] (2) When designing train timetables and rolling stock scheduling schemes, it is necessary to ensure the constraints of train headway and train maintenance time.

[0049] Specifically, step S130 involves perturbing and performing local search operations on the current solution, updating the current optimal solution in each iteration until the maximum number of iterations is reached. For example, the current solution is the optimal solution obtained in the previous iteration. Information about the current solution includes: train schedule and task sequence. and turnaround station sequence Departure time series Train formation scheme Stopping Plan and crossover matrix Train formation matrix wait.

[0050] In one embodiment, the perturbation operation is implemented by applying a cancellation or early turnaround strategy to specific trains in the current solution. This perturbation operation can significantly alter the structure of the current optimal solution, preventing the search from getting trapped in local optima.

[0051] In one embodiment, local search improves the current optimal solution locally by changing the departure time of a specific train within a safe headway.

[0052] See Figure 3 This is an example of improving the initial train timetable and rolling stock scheduling scheme based on the variable neighborhood search algorithm to determine the optimal scheme, specifically including the following steps: Step S41: Input the initial solution (initial train timetable and initial train scheduling scheme) as the current optimal solution. The preset number of algorithm iterations is N_iteration, let n=1, where n is the iteration index; For example, the number of algorithm iterations is the condition for the algorithm to terminate. The minimum value is usually set to 100. The more iterations there are, the more likely the optimal solution will be found, but the search time will also increase.

[0053] Step S42: Perform a perturbation operation on the current optimal solution to obtain several first new solutions. Find the first new solution that satisfies the optimal benefit of the railway system from the several first new solutions to update the current optimal solution and obtain the optimal solution after the first update. It should be understood that in the description of this invention, the terms "first" and "second" are used only for convenience in describing different components or names, and should not be construed as indicating or implying a sequential relationship, relative importance, or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first" and "second" may explicitly or implicitly include at least one of that feature.

[0054] For example, the perturbation operation includes four types of perturbation operators, wherein: The first type of perturbation operator is applied to the train task sequence. The specific operations include: randomly canceling the last few train trips of a certain train set, and completely swapping the task sequence. Two lines of data, partial exchange task sequence Two lines of data; The second type of perturbation operator applies to the turnaround station sequence. The specific steps include: randomly selecting a train number. u Define array Storage train number u For stations that can turn around, randomly select one from an array. Turning back at the station; The third type of perturbation operator is applied to the stopping scheme. The specific steps include: randomly selecting a train number. u Define the set of stopping schemes that change only one station as Train number u The stop plan has been changed to a set of stop plans. One of the stop options; The fourth type of disturbance operator is applied to train formation. and The specific operations include: randomly selecting train formations. B One of the train numbers u station r As the disturbance location; if the value of the disturbance location is 1, then set it to 0, and set... Z Train numbers u station r The value is 0; otherwise, set it to 1, and set... Z Train numbers u station r The number varies within train formation and station restrictions.

[0055] These four types of perturbation operators are executed with a certain probability, and the probability of each type of operator is calculated by the following formula:

[0056] In the formula, Indicates the perturbation operator i In the n The execution probability at the next iteration. Operator i In the j The improvement rate of the objective function in the next iteration.

[0057] Step S43: Perform a local search operation on the optimal solution after the first update to find a second new solution that satisfies the optimal benefit of the railway system, so as to update the optimal solution after the first update and obtain the optimal solution after the second update; For example, local search operations include two types of local search operators, wherein: The first type of local search operator applies to the departure time of the vehicle. For any car underside m The local search range for departure time is . and They represent the bottom of the car. m The departure times of the preceding and following train sets; The second type of local search operator applies to the cross-row matrix. Randomly select a train number u and the station r As the perturbation location. Define a new array. Stored at the station r The departure time is later than the train number u Train numbers. Select the array sequentially. A train number v Let the train number uThe departure time is Variations within a range. Among them... w Train number is train number v At the station r Subsequent train services.

[0058] The execution probabilities of the two types of local search operators are calculated as follows:

[0059] in, Represents the local search operator i In the n Execution probability at the next iteration Represents the local search operator i In the j The improvement rate of the objective function in the next iteration, where the optimal benefit of the railway system is set as the objective function.

[0060] The following will describe in detail the implementation examples of the demand-oriented greedy heuristic algorithm and the variable neighborhood search algorithm.

[0061] I. Examples of Demand-Oriented Greedy Heuristic Algorithms In flexible train formation, trains can change their formation at intermediate technical stations by unpacking or suspending carriages. However, considering the train formation capacity and station operational capacity, operational safety constraints must be considered in the design of train timetables and rolling stock scheduling. Specifically, this application proposes a demand-oriented greedy heuristic algorithm to design a train timetable and rolling stock scheduling scheme that satisfies the constraints of locomotive headway and train maintenance time.

[0062] (1) Storage of train timetables and rolling stock information Demand-driven greedy heuristic algorithms aim to obtain initially feasible train timetables and rolling stock scheduling schemes through a greedy heuristic strategy. Therefore, storing train timetables and rolling stock information is fundamental for perturbation operations and local search operations. For example, collecting train task sequences... Store the specific train numbers executed by all train sets. Further, the perturbation operation is performed on the train number task sequence. turnaround station sequence Stopping plan and train formation and The perturbation operation is implemented, and the local search is achieved by applying the departure time under the vehicle. and crossover matrix Implementation. Train timetables can be described using the information provided in the decision-making process.

[0063] (2) Passenger allocation Passenger allocation refers to assigning passengers to various trains in the current train timetable. This application uses a method of solving a mathematical model, namely the passenger allocation model M1, to load passengers into the mathematical model. The passenger allocation model M1 is as follows:

[0064]

[0065] In the formula, integer parameters They represent passengers respectively. p Ticket revenue and passenger time cost. 0-1 parameters. Indicates train number u Is it during the time period? w Arriving passengers p The destination station; 1 if it is, 0 otherwise. Integer parameter. Indicates passenger p During the period w The average travel time to reach the destination station. (Integer variable) Indicates passenger p Select train number u The number of people. and They represent passengers respectively. p The origin and destination points and the number of people. 0-1 parameters. Indicates train number u and passengers p The value is 1 if the travel directions are consistent, and 0 otherwise. (0-1 parameter) Indicates train number u Are you at the station? r The value is 1 if the station is stopped, otherwise it is 0. (0-1 parameter) Indicates that passengers have arrived at the station. r Whether the person is still inside the vehicle; 1 if yes, 0 otherwise. Integer parameter. Indicates train number u Arrival at the station r The number of carriages in the previous train formation. Integer parameter. Indicates train number u At the station r The number of carriages to be unpacked / suspended. This mathematical optimization model can be solved quickly using any commercial software such as Gurobi or Cplex.

[0066] (3) Measurement indicators The effectiveness of train timetables and rolling stock scheduling schemes must consider not only passenger ticket revenue but also the railway operator's operating costs and passenger travel costs. In one embodiment, the following indicators are used to measure the profitability of train timetables and rolling stock scheduling schemes:

[0067] Where, integer parameters These represent the costs of train operation within the section, turnaround operations, and shunting operations, respectively. (0-1 variables) Indicates train number u Are you at the station? r Train number to be operated after turnaround operation v If yes, the value is 1; otherwise, it is 0. This variable is determined by the train task sequence. and turnaround station sequence This was calculated. (0-1 variables) Indicates train number u Are you at the station? r If a shunting operation is performed, the value is 1; otherwise, it is 0.

[0068] In summary, the demand-oriented greedy heuristic algorithm provided by this invention obtains an initial feasible train timetable and train scheduling scheme by continuously searching for new train timetables and train scheduling schemes, and by adjusting train departure times and canceling some train services.

[0069] II. Examples of Variable Neighborhood Search Algorithms This paper proposes a demand-driven greedy heuristic algorithm to provide an initial feasible train timetable and rolling stock scheduling scheme for the variable neighborhood search algorithm. However, this initial feasible train timetable and rolling stock scheduling scheme cannot be guaranteed to be the optimal solution. Therefore, this application further provides a heuristic algorithm based on the variable neighborhood search algorithm to improve the current train timetable and rolling stock scheduling scheme by progressively optimizing it.

[0070] (1) Disturbance operation Perturbation operations significantly alter the structure of the current optimal solution, preventing the search from getting trapped in local optima. The perturbation operation method includes the following steps: Step S51: Set N_neighbor, let n=1, and proceed to step S52.

[0071] Step S52: Generate a random number rand between 0 and 1. If rand ≤ Proceed to step S53; if <rand≤ Proceed to step S54; if <rand≤ Proceed to step S55. If rand > Proceed to step S56.

[0072] For example, the probability of generating random numbers between 0 and 1 follows a uniform distribution between 0 and 1.

[0073] Step S53, randomly select a car undersidem Perform one of the following operations on it: cancel the last few trains, or completely swap the task sequence. Two lines of data, partial exchange task sequence Given two lines of data, let n = n + 1, then proceed to step S57.

[0074] Step S54, randomly select a car underside m Train number to be operated u Define an array Storage train number u For stations that can turn around, randomly select one from an array. If the train turns back at a station, let n = n + 1, and proceed to step S57.

[0075] Step S55: Randomly select a train number. u Define the set of stopping schemes that change only one station as Train number u The stop plan has been changed to a set of stop plans. One of the stopping schemes. Let n = n + 1, and go to step S57.

[0076] Step S56: Randomly select train formations B One of the train numbers u station r As the location of the disturbance 。 If the value is 1, then set it to 0, and set... Z Train numbers u station r The value is 0. Otherwise, set it to 1, and set... Z Train numbers u station r The quantity varies within the constraints of train formation and station. Let n = n + 1, and go to step S57.

[0077] Step S57: Determine if n is greater than N_neighbor. If so, output all new train timetable schemes and rolling stock scheduling schemes, and update the disturbance probability. Otherwise, proceed to step S52.

[0078] It should be noted that the disturbance operation will eventually generate N_neighbor neighborhood solutions (train timetables and rolling stock scheduling schemes), and the solution with the highest railway profit among the neighborhood solutions will be the current optimal solution.

[0079] (2) Local search Local search specifically includes the following steps: Step S61: Set N_neighbor, let n=1, and proceed to step S62.

[0080] Step S62: Generate a random number rand between 0 and 1. If rand ≤ Proceed to step S63; otherwise, proceed to step S64.

[0081] For example, the probability of generating random numbers between 0 and 1 follows a uniform distribution between 0 and 1.

[0082] Step S63, randomly select a car underside m The local search range for departure time is . and They represent the bottom of the car. m The departure times of the preceding and following car units. Let n = n + 1, then go to step S65.

[0083] For example, randomly selecting the vehicle chassis m Under each car m The probability of selection follows a uniform distribution.

[0084] Step S64, randomly select a train number. u and the station r As the perturbation location. Define a new array. Stored at the station r The departure time is later than the train number u Train numbers. Select the array sequentially. A train number v Let the train number u The departure time is Variations within a range. Among them... w Train number is train number v At the station r The subsequent train numbers. Let n = n + 1, go to step S65.

[0085] Step S65: Determine if n is greater than N_neighbor. If so, output all new train timetable schemes and rolling stock scheduling schemes, and update the disturbance probability. Otherwise, proceed to step S62.

[0086] After a local search operation, the train timetable and rolling stock scheduling scheme that maximizes railway profits are obtained as the current optimal solution. To reach the maximum number of iterations for the complete algorithm, this current optimal solution is used as the perturbation operation for the next iteration.

[0087] It should be noted that there may be situations where train formation is not feasible. This invention ensures the feasibility of train formation through a loss-sharing strategy. The loss-sharing strategy is specifically stated as follows: After executing the disturbance operator, if the train formation matrix of train number u∈U at station r∈R ( If it becomes infeasible, then it is necessary to determine all train numbers earlier than train number u∈U. And the related losses will be evenly distributed among these train services. For example... Figure 3 In (a), after the train formation adjustment, train number 3 needs to add 2 carriages, but station 2 lacks available carriages, making the train formation plan infeasible. By applying a loss-sharing strategy, the carriage shortage for train number 3 is evenly distributed between train numbers 1 and 2. Specifically, train number 1 adds 1 carriage, while train number 2 adds 1 fewer carriage. Therefore, Figure 3 (b) shows that the train formation scheme in the revised train timetable becomes feasible.

[0088] In summary, existing train timetable and rolling stock coordination optimization methods rely on commercial optimization software and have long computation times, making them unsuitable for large-scale high-speed railway operation optimization. To ensure train operation safety and transportation efficiency, this invention provides a train timetable and rolling stock scheduling coordination optimization based on flexible formation modes, employing three combined strategies to design complete daily train timetable and rolling stock scheduling schemes (flexible stopping, flexible turnaround strategy, and overtaking strategy). Furthermore, to obtain the optimal train timetable and rolling stock scheduling scheme, this invention further proposes a heuristic algorithm based on variable neighborhood search, which can obtain train timetable and rolling stock scheduling schemes in a shorter time, providing technical support for high-speed railway operation managers and offering greater guidance value in practice.

[0089] As another example, the present invention also provides a train timetable and rolling stock scheduling collaborative optimization system based on flexible formation modes, comprising: The acquisition module is used to acquire passenger flow demand data during the high-speed railway line conditions and operating hours. The passenger flow demand data includes the origin and destination of the passenger flow, the earliest departure time, the latest departure time, and the number of passengers. The initial solution generation module is used to generate the current optimal solution based on the route conditions and passenger flow demand data, with the goal of minimizing the total travel time of passengers. The current optimal solution includes the initial train timetable and the initial train scheduling scheme. The iterative optimization module is used to iteratively optimize the current optimal solution within a preset number of iterations using a variable neighborhood search algorithm. The iterative optimization includes perturbing the current optimal solution and performing local search operations to determine the target train timetable and target rolling stock scheduling scheme that satisfies the optimal benefits of the railway system. The current optimal solution is updated in each iteration.

[0090] The system in this embodiment is used to implement the corresponding methods in the foregoing multiple method embodiments and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0091] As another example, embodiments of the present invention also provide an electronic device, which is an example of a hardware device that can be applied to various aspects of the present invention. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0092] The electronic device may include a processor, a communications interface, memory, and a communications bus.

[0093] The processor, communication interface, and memory communicate with each other via a communication bus. The communication interface is used to communicate with other electronic devices or servers.

[0094] The processor is used to execute programs, specifically the relevant steps in the above method embodiments.

[0095] Specifically, the program may include program code, which includes computer operation instructions.

[0096] The processor may be a CPU, an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The one or more processors included in a smart device may be of the same type, such as one or more CPUs; or they may be of different types, such as one or more CPUs and one or more ASICs.

[0097] Memory is used to store programs. Memory may include high-speed RAM, and may also include non-volatile memory, such as at least one disk drive.

[0098] The program, when executed by a processor, is used to cause an electronic device to perform the method of the present invention.

[0099] Furthermore, the specific implementation of each step in the program can be found in the corresponding descriptions of the steps and units in the above method embodiments, and will not be repeated here. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices and modules described above can be referred to the corresponding process descriptions in the foregoing method embodiments, and will not be repeated here.

[0100] This invention also provides a computer storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the methods of the various embodiments of this invention. The corresponding process descriptions in the foregoing method embodiments can be referred to, and will not be repeated here.

[0101] The methods described above according to embodiments of the present invention can be implemented in hardware, firmware, or as software or computer code that can be stored in a recording medium (such as a CD-ROM, RAM, floppy disk, hard disk, or magneto-optical disk), or as computer code originally stored on a remote recording medium or a non-transitory machine-readable medium and subsequently stored on a local recording medium, downloaded via a network. Thus, the methods described herein can be processed by software stored on a recording medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware (such as an ASIC or FPGA). It is understood that the computer, processor, microprocessor controller, or programmable hardware includes storage components (e.g., RAM, ROM, flash memory, etc.) capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods described herein. Furthermore, when a general-purpose computer accesses code used to implement the methods shown herein, the execution of the code transforms the general-purpose computer into a dedicated computer for executing the methods shown herein.

[0102] Specific embodiments of the invention have now been described. Other embodiments are within the scope of the appended claims. In some cases, the actions described in the claims can be performed in a different order and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing can be advantageous.

[0103] It should be understood that although this specification is described according to various embodiments, not every embodiment contains only one independent technical solution. This way of describing the specification is only for clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

[0104] Finally, it should be noted that the above embodiments are only used to illustrate the embodiments of the present invention, and are not intended to limit the embodiments of the present invention. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the embodiments of the present invention. Therefore, all equivalent technical solutions also fall within the scope of the embodiments of the present invention, and the patent protection scope of the embodiments of the present invention should be defined by the claims.

Claims

1. A method for coordinated optimization of train timetables and rolling stock scheduling based on flexible formation modes, comprising the following steps: S110. Obtain passenger flow demand data during the high-speed railway line conditions and operating hours. The passenger flow demand data includes passenger flow origin and destination, earliest boarding time, latest boarding time, and number of passengers. S120. With the goal of minimizing the total travel time of passengers, based on the route conditions and the passenger flow demand data, the current optimal solution is generated through a demand-oriented greedy heuristic algorithm. The current optimal solution includes the initial train timetable and the initial train scheduling scheme. S130. Within a preset number of iterations, the current optimal solution is iteratively optimized using a variable neighborhood search algorithm. The iterative optimization includes perturbing and local search operations on the current optimal solution to determine the target train timetable and target rolling stock scheduling scheme that satisfies the optimal benefits of the railway system. The current optimal solution is updated in each iteration.

2. The method according to claim 1, characterized in that, The process of generating the current optimal solution based on the route conditions and passenger demand data using a demand-oriented greedy heuristic algorithm includes: S11. Determine the train scheduling plan: By sequentially scheduling each train... u ∈ U Insert train sequence The system assigns an even number of train services to each train set, restricts train services to turning back only at the line's terminus, and determines the train service sequence. and turnaround station sequence ; S12. Compile train timetables: By setting the arrival / departure time of the first train at both station 1 and station |R| to 10 ... Subsequent locomotives will be set to depart at the minimum interval. Depart in sequence to determine departure time sequence By passing each car bottom m ∈ M The train formation scheme is determined so that shunting operations are only performed on the first and last trains at their initial stations, and no skipping or overtaking operations are performed at any station. Stopping Plan and crossover matrix ; S13. Adjust train timetable: Sequence of train services Turnaround station sequence Train formation scheme Stopping Plan Information converted into train timetables and The system determines whether the current train timetable meets the departure interval constraint and the maintenance time constraint. If it does, the current train timetable is feasible; otherwise, if the departure interval constraint is not met, the departure time of the train sets is adjusted until the departure interval constraint is met; if the maintenance time constraint is not met, the train schedule is adjusted from the train task sequence. Remove trains that are outside their operating hours; repeat the above steps until all departure interval constraints and maintenance time constraints are met, then update the departure time series. The value; S14. Carriage Allocation: Allocating carriages to the line's terminal station. r ( r ={1,| R The carriages are evenly distributed among the departing train sets according to the train number and task sequence. Update the number of carriages for the remaining train services, and then update the train formation matrix. ; S15. Passenger allocation: The train scheduling scheme and train timetable determined through steps S11-S14 are used as the initial train scheduling scheme and the initial train timetable. Based on the initial train scheduling scheme and the initial train timetable, the passenger allocation scheme is optimized by solving the passenger allocation model. S16. Output the current optimal solution: Output the initial train scheduling scheme and the initial train timetable as the current optimal solution.

3. The method according to claim 2, characterized in that, In step S13, the following method is used to determine whether the current train timetable meets the departure interval constraint and maintenance time constraint: By determining the arrival and departure times of different trains at the stations, it is determined whether the current train timetable meets the departure interval constraint: If the time interval between the arrival and departure of any two trains at the same station is not less than If so, the current train timetable satisfies the departure interval constraint; Otherwise, the current train timetable does not meet the departure interval constraint; By determining whether the arrival and departure times of all trains fall within the train operating hours. Within this process, determine whether the current train timetable meets the maintenance time constraint: If the arrival and departure times of all trains fall within the train operating hours If the current train timetable meets the maintenance time constraint, then the current train timetable meets the maintenance time constraint. Otherwise, the current train timetable does not meet the maintenance time constraint.

4. The method according to claim 1, characterized in that, The following steps are used to determine the target train timetable and target rolling stock scheduling scheme that best satisfies the benefits of the railway system: S41. Input the current optimal solution including the initial train timetable and the initial rolling stock scheduling scheme. The preset number of algorithm iterations is N_iteration, let n=1, where n is the iteration index; S42. Perform a perturbation operation on the current optimal solution to obtain several first new solutions. Find the first new solution that satisfies the optimal benefit of the railway system from the several first new solutions to update the current optimal solution and obtain the optimal solution after the first update. S43. Perform a local search operation on the optimal solution after the first update to find a second new solution that satisfies the optimal benefit of the railway system, so as to update the optimal solution after the first update and obtain the optimal solution after the second update. S44. Let n = n + 1, and determine whether n is greater than N_iteration. If n is greater than N_iteration, output the optimal solution after the latest update as the current optimal solution; otherwise, execute step S42 until the algorithm converges.

5. The method according to claim 4, characterized in that, In step S42, the perturbation operation includes four types of perturbation operators, wherein: The first type of perturbation operator is applied to the train task sequence. The specific operations include: randomly canceling the last few train trips of a certain train set, and completely swapping the task sequence. Two lines of data, partial exchange task sequence Two lines of data; The second type of perturbation operator applies to the turnaround station sequence. The specific steps include: randomly selecting a train number. u Define array Storage train number u For stations that can turn around, randomly select one from an array. Turning back at the station; The third type of perturbation operator is applied to the stopping scheme. The specific steps include: randomly selecting a train number. u Define the set of stopping schemes that change only one station as Train number u The stop plan has been changed to a set of stop plans. One of the stop options; The fourth type of disturbance operator is applied to train formation. and The specific operations include: randomly selecting train formations. B One of the train numbers u station r As the disturbance location; if the value of the disturbance location is 1, then set it to 0, and set... Z Train numbers u station r The value is 0; otherwise, set it to 1, and set... Z Train numbers u station r The number varies within train formation and station restrictions.

6. The method according to claim 5, characterized in that, In step S43, the execution probabilities of the four types of perturbation operators are calculated as follows: in, Indicates the perturbation operator i In the n Execution probability at the next iteration Indicates the perturbation operator i In the j The improvement rate of the objective function in the next iteration.

7. The method according to claim 6, characterized in that, In step S43, the local search operation includes two types of local search operators, wherein: The first type of local search operator applies to the departure time of the vehicle. For any car underside m The local search range for departure time is , and They represent the bottom of the car. m The departure times of the preceding and following train sets; The second type of local search operator applies to the cross-row matrix. Randomly select a train number u and the station r As the perturbation location. Define a new array. Stored at the station r The departure time is later than the train number u Select the train numbers in the array sequentially. A train number v Let the train number u The departure time is Variations within the range, of which w Train number is train number v At the station r Subsequent train services.

8. The method according to claim 7, characterized in that, The execution probabilities of the two types of local search operators are calculated as follows: in, Represents the local search operator i In the n Execution probability at the next iteration Represents the local search operator i In the j The improvement rate of the objective function in the next iteration, wherein the optimal benefit of the railway system is set as the objective function.

9. An electronic device, characterized in that, include: processor; Memory for stored programs; The program includes instructions that, when executed by the processor, cause the processor to perform the steps of the method as described in any one of claims 1-8.

10. A computer storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1-8.

Images