High-speed rail passenger and freight train working diagram optimization method, system, equipment, medium and product

Abstract

The invention discloses a high-speed rail passenger and freight train working diagram optimization method, system and device, a medium and a product, and the method comprises the steps: firstly determining an optimization target, including minimizing the total travel time of a freight train, the operation deviation of the passenger train and the cargo delivery delay time; thirdly, constructing a mathematical model, and defining decision variables, an objective function and constraint conditions to form a high-speed railway passenger and freight train working diagram optimization model; on the basis of the model, a rolling time domain decomposition strategy is adopted, and a global scheduling problem is segmented into a local optimization problem of a plurality of time windows. And for each time window, obtaining a local optimal solution by using a linear programming solver so as to ensure that the dispatching of each train accords with the set constraint. Then, global search is carried out on the local solutions through a simulated annealing algorithm, the departure time and the operation plan of the train are optimized, and local optimum is avoided. And finally, a variable neighborhood local optimization method is applied, and the quality of the whole scheme is further improved by finely adjusting the solution.

Description

Technical Field

[0001] This invention belongs to the field of railway transportation and logistics optimization technology, specifically relating to a method, system, equipment, medium and product for optimizing the timetable of high-speed passenger and freight trains. Background Technology

[0002] High-speed rail boasts significant advantages such as high speed, high punctuality, and large transport capacity, and has shown important application prospects in the field of high-value-added freight transportation in recent years. Relying on high-speed rail to transport fresh produce, express parcels, and other goods with high time-sensitivity requirements can significantly improve logistics efficiency and service quality. The train timetable is the core of railway transport organization, and its scientific compilation directly affects the allocation of transport resources and transport timeliness. However, existing high-speed rail timetable compilation methods are mainly geared towards passenger transport and are difficult to directly apply to the operational needs of freight trains. With the continuous expansion and improvement of the high-speed rail network, freight transport demands are becoming more diversified and personalized, placing higher demands on the compilation of high-speed rail freight train timetables, such as… Figure 1 As shown.

[0003] In existing technologies, conventional railway freight train timetables are primarily designed for transporting bulk commodities such as coal and steel. These commodities are less sensitive to transit time, and trains typically require marshalling and reassembly at technical stations during operation. In contrast, high-speed railway freight transport mostly utilizes fixed-formation EMU trains, primarily transporting high-value-added, time-sensitive goods, generally without mid-journey marshalling and reassembly. Furthermore, high-speed railway networks prioritize passenger trains, with most passenger trains operating directly. Timetable design focuses on meeting passenger travel time and train frequency requirements, lacking consideration for the characteristics of freight train transport.

[0004] In high-speed rail freight transport, transfer connections and time window constraints are crucial factors influencing timetable design. Transfer connections ensure smooth connections between different lines and trains, expanding the transport network's coverage and improving flexibility and accessibility. Time window constraints require train schedules to match cargo delivery deadlines to guarantee cargo quality and meet customer timeliness requirements. However, existing timetable optimization research still falls short in comprehensively considering transfer connections and time window factors, lacking optimization models and algorithms that can balance passenger train priority and freight timeliness.

[0005] In summary, existing technologies suffer from the following problems: Timetable generation primarily focuses on passenger transport, lacking methods tailored to the characteristics of high-speed rail freight transport. Existing methods fail to effectively combine freight transfer connections with time window constraints, making it difficult to guarantee the efficiency and punctuality of freight transport. Furthermore, there is a lack of collaborative optimization models and algorithms that balance the priority of passenger train operation with the timeliness requirements of freight transport.

[0006] Therefore, there is an urgent need for a high-speed railway timetable optimization method that can comprehensively consider transfer connections, time window constraints, and coordinate the operation needs of passenger and freight trains, so as to improve freight efficiency and transportation timeliness and meet the requirements of modern efficient logistics development. Summary of the Invention

[0007] To address the technical problems existing in the background art, this invention aims to provide a method, system, equipment, medium, and product for optimizing high-speed railway passenger and freight train timetables. Specifically targeting the collaborative optimization problem of high-speed railway passenger and freight train timetables, it innovatively incorporates factors such as transfer connections and freight time windows, constructs a multi-objective integer programming model, and designs a novel rolling time-domain-simulated annealing-variable neighborhood search (RH-SA-VNS) train timetable optimization algorithm. Finally, based on actual numerical examples, the effectiveness of the proposed model and algorithm is verified.

[0008] To solve the technical problem, the technical solution of the present invention is as follows:

[0009] A method for optimizing the timetable of high-speed passenger and freight trains, the method comprising:

[0010] The optimization objective, decision variables, objective function and constraints are determined, and a high-speed railway passenger and freight train operation timetable optimization model is constructed. Based on the model and the freight train operation plan, the global problem is decomposed into time-related local subproblems using a rolling time-domain strategy. The linear programming solver is called to solve and splice the subproblems one by one to generate an initial feasible solution S0, which is then used as the initial current solution S.

[0011] Starting from the current solution S, the simulated annealing iteration process begins. In each iteration, a neighborhood operator is selected from four types of neighborhood structures using a roulette wheel approach to perturb S. For each time window affected by the perturbation, the linear programming solver is called again to obtain a local solution. and local solutions By splicing the solutions together, a candidate solution S′ is formed.

[0012] The acceptance and updating of solutions are controlled according to the Metropolis criterion: if S′ is better than S, then S is updated to S′; if it is worse than S, then it is accepted with a preset probability, otherwise it remains unchanged; at the same time, if the current solution is better than the historical best solution, then the best solution S* is updated.

[0013] After multiple iterations, the algorithm continuously perturbs, reconstructs, and optimizes the current solution, gradually improving the global solution and finally outputting the optimized passenger and freight train operation diagram S*.

[0014] Furthermore, determining the objective function specifically includes:

[0015] A multi-objective optimization framework is constructed, and objective functions are set from three dimensions: freight efficiency, passenger transport stability, and cargo timeliness.

[0016] The total travel time for newly added freight trains is the shortest:

[0017] (1)

[0018] Equation (1) calculates the number of new freight trains arriving at the originating station. To the final stop The total travel time is obtained by summing the arrival and departure time differences. This indicates that freight train i has arrived at its destination. Arrival time, This indicates that freight train i originates from the starting station. Departure time For the collection of new freight trains;

[0019] Existing buses with the least deviation:

[0020] (2)

[0021] To assemble existing passenger trains, This is the set of stations that train i passes through; Let be the departure time of train i at station j; Let i be the planned departure time of the existing passenger train i at station j; Equation (2) includes three parts: the departure time deviation of the existing passenger train at the starting station, the running time deviation between stations, and the stopping time deviation. The overall deviation is measured by summing the absolute values ​​and the goal is to minimize it.

[0022] Minimize delivery delays:

[0023] (3)

[0024] in:

[0025] The delivery delay penalty function for goods g depends on the actual delivery time. With respect to the allowed time window boundaries, For the aggregation of cargo flow, The actual delivery time of the goods flow g; Equation (3) is based on the actual delivery time of the goods. With time window boundaries, including the earliest allowed time and latest permitted time Given the relationship, calculate the delay penalty and sum it;

[0026] Finally, a weighted summation method is used to integrate the three sub-objectives into a single objective function. By assigning different weights, the optimization requirements of freight efficiency, passenger stability, and cargo timeliness are balanced, as detailed below:

[0027] (4)

[0028] in, , , These are the weighting coefficients for the three sub-objectives.

[0029] Furthermore, the determination of the constraints specifically includes:

[0030] (1) Constraints related to the time of arrival of goods

[0031] (5)

[0032] Constraint (5) represents the flow of goods. Delivery time From the station to its destination train The arrival times are determined by a weighted summation. As an indicator parameter, it represents the cargo flow specified in the cargo transportation plan. Whether by train Transportation; this constraint ensures that the delivery time of goods is consistent with the actual arrival time of the transporting train, providing a basis for calculating delivery delays;

[0033] (2) Number of stops

[0034] (6)

[0035] Constraint (6) stipulates that trains must stop at pre-set stations, while allowing additional stops to improve the flexibility and feasibility of timetable preparation;

[0036] (3) Stop time constraints

[0037] (7)

[0038] (8)

[0039] Constraints (7) and (8) indicate the train If at the station If the train stops at a station, the stopping time must be within the minimum stopping time. and maximum stop time Between these parameters, the stopping time for freight trains includes the time for loading and unloading goods, while the stopping time for passenger trains must meet the basic time requirements for passenger boarding and alighting, thus constraining the passage parameters. The 0-1 characteristic ensures that the station does not stop ( = 0), the train's stop time is 0;

[0040] (4) Interval running time constraints

[0041] (9)

[0042] (10)

[0043] Constraints (9) and (10) limit the train From the station to The interval running time of +1 must be within the minimum interval running time. With the maximum interval running time Within the range; among which, Additional time for train departure Additional time for stopping, by comparing with stop variables , The correlation between stops and the total travel time of the section is characterized.

[0044] (5) Departure time constraints

[0045] (11)

[0046] (12)

[0047] Constraint (11) indicates that for existing passenger trains Its departure times at each station No earlier than the originally planned time And the latest deviation from the original plan shall not exceed the allowable threshold. To ensure the punctuality of passenger services; constraint (12) indicates that for newly added freight trains Its starting station The departure time must be limited to the minimum allowable time. With maximum allowed time between;

[0048] (6) Transfer connection time constraints

[0049] (13)

[0050] (14)

[0051] Constraints (13) and (14) indicate that when a transfer connection exists, the connection time between the two trains must be within the minimum transfer time. With maximum transit time Between these, to ensure the smooth operation of cargo loading, unloading, and transfer; when there is no continuity, the constraint is automatically relaxed by using a large constant M;

[0052] (7) Train interval constraints

[0053] (15)

[0054] (16)

[0055] (17)

[0056] Constraints (15)-(17) ensure the same station The departure interval between any two trains shall not be less than the minimum safe departure interval. The arrival interval between any two arriving trains shall not be less than the minimum safe arrival interval. ,in, As a 0-1 auxiliary variable, it can flexibly adapt to the interval requirements under different departure sequences by cooperating with the large constant M.

[0057] Furthermore, the adoption of the rolling time-domain strategy specifically includes:

[0058] The global optimization problem is divided into rolling time-domain windows with fixed time lengths. Each window constitutes a local subproblem, covering passenger and freight trains, stations, and section resources within its scope. The focus is on optimizing train arrival and departure times and stopping decision variables within the window, and global coherence is ensured through boundary parameter propagation. For each window, an integer programming sub-model is constructed, guided by a weighted multi-objective function, integrating stopping, safety intervals, boundary connections, time windows, and transfer constraints. The Gurobi solver is called to obtain local optimal solutions. After solving, the train states across windows are extracted as inputs for the next window. The process is rolled forward and local solutions are temporarily stored. Finally, the solutions are concatenated to form the global solution.

[0059] Each window contains newly connected trains and cross-window trains continuing from the previous window. The final state parameters of cross-window trains from the previous window are extracted, including arrival and departure times. These parameters are embedded as hard constraints into the integer programming sub-model of the current window to ensure that the solution of each local window satisfies the current constraints and is connected with the previous window, thereby generating a high-quality initial feasible solution that satisfies all constraints.

[0060] Furthermore, the step of selecting a neighborhood operator from four types of neighborhood structures based on the roulette wheel method to perturb S includes:

[0061] Based on the principle of prioritizing passenger trains, four types of variable neighborhood structures are designed to achieve coordinated optimization of freight train efficiency and passenger train punctuality.

[0062] Time window adjustment for multiple freight trains: Randomly select n (n≥2) freight trains and coordinate the adjustment of their time windows to optimize overall resource utilization;

[0063] Adjustment of time window for single freight trains: For single freight trains, a new time window will be randomly selected and assigned.

[0064] Adjustment of time window for multiple passenger trains: Randomly select n (n≥2) passenger trains and adjust their departure time and stop schedule within the maximum allowable deviation range;

[0065] Adjustment of time window for single passenger train: Randomly select one passenger train and assign its time window to another suitable time window to resolve extreme conflicts;

[0066] The above four types of neighborhood structures are selected through a roulette wheel selection process, and their probability allocation is dynamically adjusted according to the priority of the optimization objective: 40% for multiple freight trains, 30% for a single freight train, 20% for multiple passenger trains, and 10% for a single passenger train.

[0067] The train time domain window is based on the spatiotemporal matching of the departure time window and the rolling time domain window. The set of windows that can be assigned to the train is determined by the intersection of intervals. The specific calculation method is as follows:

[0068] First, the operational planning period is discretized based on a fixed time length, forming a set of continuous and non-overlapping time domain windows. = Single window The time boundary is defined as: ,in, For window The start time, For window The end time, and = This ensures that the planning period is fully covered in terms of time.

[0069] Next, regarding existing passenger trains Its departure time window is determined by the existing scheduled departure time. Deviation from the maximum allowable amount Sure: For the newly added freight trains Its departure time window is from the originating station The permitted departure times are given directly: = .in For train The departure time window.

[0070] Therefore, the train Set of assignable time domain windows It is determined by the non-empty intersection of its departure time window and each time domain window; specifically, for any window If the following conditions are met: ,but Belongs to the train The set of assignable windows;

[0071] Finally, the train The set of assignable time domain windows is as follows: .

[0072] Furthermore, starting from the current solution S, the simulated annealing iteration process is initiated, specifically including:

[0073] The core parameters of the simulated annealing algorithm include: initial temperature, cooling coefficient, and iteration termination condition. In the simulated annealing algorithm, the Metropolis criterion is the core mechanism for achieving finite acceptance of suboptimal solutions. It balances solution space exploration and convergence through a probabilistic acceptance strategy. The core logic is as follows:

[0074] For the neighborhood solution generated by the variable neighborhood search If its objective function value If a neighboring solution is better, then the neighboring solution is directly accepted, and the current solution is updated accordingly. ;like If the neighborhood solution is worse than the current solution, then calculate the acceptance probability: = ;

[0075] in, For the current iteration temperature, generate a uniformly distributed interval. random numbers ,like If the solution is not optimal, then the non-optimal neighborhood solution is accepted with probability to maintain the diversity of solution space exploration; otherwise, it is rejected and the current better solution is retained. This criterion makes the algorithm more likely to escape local optima in the initial iteration and gradually converge to high-quality solutions in later iterations, ensuring global optimization capability.

[0076] A high-speed rail passenger and freight train timetable optimization system, the high-speed rail passenger and freight train timetable optimization system comprising:

[0077] Objectives and Decisions Module: Determine the optimization objectives, decision variables, objective functions, and constraints, and construct an optimization model for the high-speed railway passenger and freight train timetable;

[0078] Rolling temporal decomposition module: Based on the model and the freight train operation plan, the global problem is decomposed into temporally related local subproblems using a rolling temporal strategy. The linear programming solver is called to solve and concatenate the subproblems one by one to generate an initial feasible solution S0, which is then used as the initial current solution S.

[0079] Simulated annealing iteration module: Starting from the current solution S, the simulated annealing iteration process begins. In each iteration, a neighborhood operator is selected from four types of neighborhood structures using a roulette wheel approach to perturb S. For each time window affected by the perturbation, the linear programming solver is called again to obtain a local solution. and local solutions By splicing the solutions together, a candidate solution S′ is formed.

[0080] Acceptance and Update Module: The acceptance and update of solutions are controlled according to the Metropolis criterion: if S′ is better than S, then S is updated to S′; if it is worse than S, then it is accepted with a preset probability, otherwise it remains unchanged; at the same time, if the current solution is better than the historical best solution, then the best solution S* is updated.

[0081] Global solution optimization module: After multiple iterations, the algorithm continuously perturbs, reconstructs and optimizes the current solution, gradually improving the global solution, and finally outputs the optimized optimal passenger and freight train operation diagram S*.

[0082] A computer device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the high-speed rail passenger and freight train timetable optimization method described in any one of the above descriptions.

[0083] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the high-speed rail passenger and freight train timetable optimization method described above.

[0084] A computer program product includes a computer program that, when executed by a processor, implements the steps of the high-speed rail passenger and freight train timetable optimization method described above.

[0085] Compared with the prior art, the advantages of the present invention are as follows:

[0086] (1) This invention focuses on the core issue of coordinated optimization of high-speed railway passenger and freight train timetables, and innovates to address the gaps in existing research. Current high-speed railway timetables are centered on passenger transport, and existing research does not adequately consider the transfer and freight time windows of high-speed rail freight. This paper is the first to incorporate train transfer and freight time windows into the coordinated optimization framework of high-speed railway passenger and freight train timetables, clearly defining "passenger train priority" as the principle. Under the premise of ensuring the existing passenger transport order, it specifically studies the compilation of timetables for high-speed rail freight trains, filling the research gap in the coordinated optimization of high-speed rail freight trains and passenger trains.

[0087] (2) This invention constructs a multi-objective integer programming model. The innovation of the model lies in the integration of unique constraints of high-speed rail freight and the balance of multi-objective collaborative optimization. Based on the traditional timetable compilation constraints, the model adds train transfer and connection constraints and freight time window constraints, with the multi-objective goal of "minimizing the total travel time of freight trains, the total deviation time of existing passenger trains, and the delay time of freight delivery".

[0088] (3) This invention designs a hybrid optimization algorithm (RH-SA-VNS) that integrates rolling temporal decomposition strategy, simulated annealing global search, and variable neighborhood local optimization. This algorithm decomposes the global problem into temporally related local subproblems through rolling temporal decomposition, reducing the solution complexity; it uses simulated annealing to achieve global search, avoiding getting trapped in local optima; and it combines variable neighborhood search for local optimization, improving the quality of the solution. The algorithm innovatively designs four types of neighborhood structures and balances the optimization needs of passenger and freight trains through priority sorting, effectively improving the solution efficiency and stability of large-scale problems, providing a feasible technical path for optimizing high-speed rail passenger and freight train timetables. Attached Figure Description

[0089] Figure 1 An optimization diagram for the collaborative operation of existing passenger and freight trains;

[0090] Figure 2 A cargo transshipment time consideration diagram provided for embodiments of this application;

[0091] Figure 3 A diagram illustrating the cargo transportation time window provided in this application embodiment;

[0092] Figure 4 Neighborhood operator 1 provided in the embodiments of this application;

[0093] Figure 5 Neighborhood operator 2 provided in the embodiments of this application;

[0094] Figure 6 Neighborhood operator 3 is provided for embodiments of this application;

[0095] Figure 7Neighborhood operator 4 provided in the embodiments of this application;

[0096] Figure 8 This is a schematic diagram of the rolling time-domain generation strategy provided in an embodiment of this application;

[0097] Figure 9 The Hangzhou-Ningbo-Wenzhou high-speed railway map provided for embodiments of this application;

[0098] Figure 10 A diagram illustrating the algorithm search iteration process provided in an embodiment of this application;

[0099] Figure 11 This is an embodiment of the optimized operation diagram for passenger and freight trains provided in this application. Detailed Implementation

[0100] The specific implementation of the present invention is described below with reference to embodiments:

[0101] It should be noted that the structures, proportions, sizes, etc. shown in this specification are only used to complement the content disclosed in the specification for those skilled in the art to understand and read, and are not intended to limit the conditions under which the present invention can be implemented. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0102] Furthermore, the terms such as "upper," "lower," "left," "right," "middle," and "one" used in this specification are merely for clarity of description and are not intended to limit the scope of the invention. Any changes or adjustments to their relative relationships, without substantially altering the technical content, should also be considered within the scope of the invention.

[0103] Example 1:

[0104] With the gradual improvement of the high-speed rail network and the continuous growth of logistics demand, how to efficiently formulate freight train timetables while ensuring the priority operation of passenger trains has become a key issue in unlocking the potential of high-speed rail freight. This invention focuses on the coordinated optimization of high-speed rail passenger and freight train timetables, with the core being the optimization of existing passenger train timetables. Based on the operational schedule, additional freight trains will be rationally arranged. The operational plan, while also taking into account the cargo To meet the transportation requirements such as transit and connection needs and time window constraints, and to achieve efficient coordination of passenger and freight transportation.

[0105] Specifically, the research scenario is a high-speed railway, whose physical structure consists of a collection of stations. The line consists of existing passenger trains with pre-existing schedules and additional freight trains requiring different speed levels. The freight train operation plan is known, including the route, origin and destination stations, frequency, and stops. These freight trains use fixed formations and are not reorganized during transport. They primarily transport high-value goods such as fresh produce and express parcels. These goods may involve transshipment at a specific station. ( From a freight train ( Transferred to another freight train ( Continue transportation, such as Figure 2 As shown. It is worth noting that the cargo transshipment plan is given and can be obtained from the prior cargo transportation plan, and transshipment only occurs when cargo is loaded and unloaded at the originating and terminating stations of the freight train. Therefore, it is necessary to consider the train's transshipment connection time at the station, including the minimum time required for cargo loading and unloading, train stops, and other operations. and maximum Time range, determined by the indicator parameter Indicate whether there is a connection between the trains.

[0106] Meanwhile, different types of goods have specific time windows for transportation: fresh produce needs to be delivered within a short period to ensure quality, while express parcels must be delivered before the promised latest delivery time. Therefore, goods ( Total transportation time from the origin station to the destination station It must be controlled at its corresponding earliest allowed delivery time. With the latest permitted delivery time Within this range, exceeding this range will incur delay costs, such as... Figure 3 As shown.

[0107] Several constraints must be met during the timetable preparation process. First, the principle of passenger train priority must be upheld; the operation of newly added freight trains must not excessively disrupt the operation of existing passenger trains, and passenger train departure times must be considered. Deviations in interval travel time and stop time must be controlled within acceptable range. Inside. Secondly, the train In the interval The minimum requirements for operation are as follows: and maximum Running time constraints; train If at the station If a stop is made, the stop time must meet the minimum requirement. and maximum Stopping time constraints apply, and for freight trains, loading and unloading time must also be considered. Furthermore, at the same station... Trains departing or arriving must maintain a safe headway, including departure headway. and arrival interval The departure order of trains here is not fixed. Freight trains must depart from the originating station. Permissible time range to It must be within the scope of the train and maintain reasonable connections with relevant passenger or freight trains.

[0108] In summary, the objective of this invention is to achieve multi-objective synergistic optimization by optimizing the departure times, stop schedules, and interval travel times of freight trains, as well as the fine-tuning of passenger trains, while meeting the aforementioned capacity constraints and transportation requirements: firstly, minimizing the total travel time of freight trains to improve freight efficiency; secondly, minimizing the operational deviation of existing passenger trains to ensure passenger service quality; and thirdly, minimizing the total delay time in freight delivery to meet the timeliness requirements of freight transportation.

[0109] I. Model Building:

[0110] 1.1 Model Assumptions and Symbol Explanation:

[0111] The model of this invention makes the following assumptions:

[0112] (1) Allow trains to add stops based on the preset stopping plan in order to improve the flexibility and feasibility of timetable preparation.

[0113] (2) The freight train is a dedicated freight train with fixed formation, and its train operation plan is known.

[0114] (3) Passenger train operation plan is used as basic data input, and the existing passenger train operation plan is known information.

[0115] (4) The line's transport capacity is sufficient to support the mixed operation of existing passenger trains and newly added freight trains.

[0116] (5) The unit of goods shall not be split during transportation until it is delivered to the destination, so as to ensure the integrity of the goods and transportation efficiency.

[0117] The sets, parameters, variable symbols, and descriptions used in the constructed model are shown in Table 1-3.

[0118] Table 1 - Set Symbols and Explanations

[0119]

[0120] Table 2 - Parameter Symbols and Explanations

[0121]

[0122] Table 3 - Variable Symbols and Explanations

[0123]

[0124] 1.2 Objective Function:

[0125] This invention addresses the problem of coordinated optimization of passenger and freight train schedules on high-speed railways by constructing a multi-objective optimization framework and setting objective functions from three dimensions: freight efficiency, passenger stability, and freight timeliness.

[0126] ① The total travel time of the newly added freight trains is the shortest.

[0127] (1)

[0128] Equation (1) calculates the number of new freight trains arriving at the originating station. To the final stop The total travel time is obtained by summing the arrival and departure time differences, with the aim of minimizing this value. The core of this objective is to fully leverage the "high speed" advantage of high-speed rail freight, address the time-sensitive needs of high-value goods such as fresh produce and express delivery, improve transportation efficiency by reducing transit time, and simultaneously reduce the occupation of track resources by freight trains, creating more flexible scheduling space for mixed passenger and freight operations.

[0129] ② Existing buses deviate the least

[0130] (2)

[0131] Objective (2) covers three parts: departure time deviation, interval travel time deviation, and stop time deviation of existing passenger trains at the starting station. The overall deviation is measured by summing the absolute values ​​and the goal is to minimize it. This objective reflects the "passenger train priority" operation principle, which aims to ensure the punctuality and stability of passenger services, avoid significant interference to passenger travel plans caused by the scheduling of new freight trains, and balance the priority relationship between passenger and freight transportation.

[0132] ③ Minimize delivery delays

[0133] (3)

[0134] in: .

[0135] Objective (3) Based on the actual delivery time of the goods With time window boundaries, including the earliest allowed time and latest permitted time The relationship between the two factors is used to calculate and sum the delay penalties to minimize the total delay time. This objective aims to meet the timeliness requirements of freight transportation, ensure that goods are delivered within the stipulated time frame, reduce value loss or default risk caused by delays, and improve the service reliability of high-speed rail freight.

[0136] This invention employs a weighted summation method to integrate three sub-objectives into a single objective function. By assigning different weights, it balances the optimization requirements of freight efficiency, passenger stability, and cargo timeliness, as detailed below:

[0137] (4)

[0138] in, , , These are the weighting coefficients for the three sub-objectives. These weighting coefficients can be set according to actual operational needs. For example, during peak passenger periods, such as holidays, the weighting coefficients can be increased. Prioritize passenger travel; during periods of surge in freight demand, such as e-commerce promotions, rates may be appropriately increased. and To improve freight efficiency and reliability.

[0139] 1.3 Constraints:

[0140] (1) Constraints related to the time of arrival of goods

[0141] (5)

[0142] Constraint (5) represents the flow of goods. Delivery time From the station to its destination train The arrival times are determined by a weighted summation. As an indicator parameter, it represents the cargo flow specified in the cargo transportation plan. Whether by train Transportation. This constraint ensures that the delivery time of goods matches the actual arrival time of the transporting train, providing a basis for calculating delivery delays.

[0143] (2) Number of stops

[0144] (6)

[0145] Constraint (6) stipulates that trains must stop at pre-set stations, while allowing additional stops to improve the flexibility and feasibility of timetable preparation.

[0146] (3) Stop time constraints

[0147] (7)

[0148] (8)

[0149] Constraints (7) and (8) indicate the train If at the station If the train stops at a station, the stopping time must be within the minimum stopping time. and maximum stop time Between. The stopping time for freight trains includes the time for loading and unloading goods, while the stopping time for passenger trains must meet the basic time requirements for passenger boarding and alighting, constraining the parameters. The 0-1 characteristic ensures that the station does not stop ( = 0), the train's stop time is 0.

[0150] (4) Interval running time constraints

[0151] (9)

[0152] (10)

[0153] Constraints (9) and (10) limit the train From the station to The interval running time of +1 must be within the minimum interval running time. With the maximum interval running time Within the range. Among them, Additional time for train departure Additional time for stopping, by comparing with stop variables , The correlation between the two is used to characterize the impact of station stops on the total travel time of the section.

[0154] (5) Departure time constraints

[0155] (11)

[0156] (12)

[0157] Constraint (11) indicates that for existing passenger trains Its departure times at each station No earlier than the originally planned time And the latest deviation from the original plan shall not exceed the allowable threshold. To ensure the punctuality of passenger services. Constraint (12) indicates that for newly added freight trains Its starting station The departure time must be limited to the minimum allowable time. With maximum allowed time between.

[0158] (6) Transfer connection time constraints

[0159] (13)

[0160] (14)

[0161] Constraints (13) and (14) indicate that when a transfer connection exists, the connection time between the two trains must be within the minimum transfer time. With maximum transit time Between these, the constraints are automatically relaxed to ensure the smooth operation of cargo loading, unloading, and transfer; when there is no continuity, the constraints are automatically relaxed through a large constant M.

[0162] (7) Train interval constraints

[0163] (15)

[0164] (16)

[0165] (17)

[0166] Constraints (15)-(17) ensure the same station The departure interval between any two trains shall not be less than the minimum safe departure interval. The arrival interval between any two arriving trains shall not be less than the minimum safe arrival interval. .in, As a 0-1 auxiliary variable, it can flexibly adapt to the interval requirements under different departure sequences by cooperating with the large constant M.

[0167] II. Algorithm Design

[0168] 2.1 Overall Algorithm Idea:

[0169] Given that the high-speed railway passenger and freight train timetable optimization model constructed in this invention, which considers transfer connections and time windows, has NP-hard characteristics and is difficult to solve efficiently in polynomial time using exact algorithms, and that traditional heuristic algorithms are prone to getting stuck in local optima or having low solution efficiency when dealing with multi-objective collaboration, complex constraint coupling, and large-scale problems, this invention proposes a hybrid optimization algorithm (Rolling Horizon - Simulated Annealing - Variable Neighborhood Search, RH-SA-VNS) that integrates rolling time-domain decomposition strategy, simulated annealing global search, and variable neighborhood local optimization.

[0170] The RH-SA-VNS algorithm uses "decomposition-perturbation-reconstruction-optimization" as its core logic. First, it decomposes the global problem into temporally related local subproblems through rolling time domain and calls the linear programming solver to generate an initial feasible solution. Then, during the simulated annealing iteration, it selects neighborhood operators from four types of neighborhood structures (multiple freight train adjustments, single freight train adjustments, multiple passenger train adjustments, and single passenger train adjustments) to perturb the current solution and calls the linear programming solver again to solve the local solutions of each window to splice them to generate a new solution. Finally, it controls the acceptance and updating of solutions through the Metropolis criterion, balancing global exploration and local development while ensuring the feasibility of the solution.

[0171] This algorithm deeply integrates the windowing solution capability of rolling time domain with the global optimization characteristics of simulated annealing. In each iteration, it reconstructs the feasibility of perturbation solutions through local subproblem reconstruction, reducing the solution complexity of large-scale problems while ensuring the spatiotemporal coordination of passenger and freight train timetables. This effectively adapts to the unique constraints of high-speed rail freight under the "passenger train priority" principle. The pseudocode of the algorithm is as follows:

[0172] 2.2 Variable Neighborhood Structure Design

[0173] The rational design of the neighborhood structure is key to local optimization efficiency. Its core lies in achieving coordinated optimization of freight train operating efficiency and passenger train punctuality through differentiated operations. Given the constraint of "passenger train priority" in the high-speed rail passenger-freight mixed operation scenario, this invention designs four types of variable neighborhood structures: time-domain window adjustment neighborhood operators for multiple freight trains, single freight trains, multiple passenger trains, and single passenger trains. Targeted perturbation strategies are used to balance local optimization accuracy and global constraint satisfaction, as detailed below:

[0174] ① Neighborhood Operator 1: Time Domain Window Adjustment for Multiple Freight Trains

[0175] Randomly select n freight trains (n≥2) to form an adjustment group. Using train-time domain window assignment as the operation unit, coordinately perturb the time window allocation of trains within the group, such as... Figure 4 As shown, yellow represents selectable time domain windows, and green represents occupied time domain windows. Specifically, by randomly permuting the assignment relationships of trains in adjacent time windows, the overall optimization of freight trains' spatiotemporal resource utilization is achieved, reducing resource conflicts across windows. The design basis of this neighborhood operation is that the cluster adjustment of freight trains can quickly eliminate large-scale operational conflicts, and is particularly suitable for the synergistic optimization of improving freight efficiency and line resource utilization. Its disturbance intensity is adapted to the spatiotemporal distribution characteristics of freight transportation demand.

[0176] ② Neighborhood Operator 2: Time Domain Window Adjustment for Single Freight Trains

[0177] Taking a single freight train as the operational object, a new time window is randomly selected within its allowed departure time window for assignment, such as... Figure 5 As shown. The characteristic of this operation is to explore fine-grained optimization space through high-frequency local fine-tuning, which is particularly suitable for improving the operating efficiency of freight trains in local time periods. Its disturbance range is limited by the time window flexibility of freight trains, avoiding significant interference with the operation of passenger trains.

[0178] ③ Neighborhood Operator 3: Adjustment of Time Domain Window for Multiple Passenger Trains

[0179] Randomly select n (n≥2) passenger trains to form an adaptation group, and within its maximum permissible deviation... Within the scope, coordinate adjustments will be made to departure times and stop schedules, such as... Figure 6 As shown. The adjustment logic is: guided by the disturbance demand of freight trains, potential conflicts are mitigated through fine-tuning of passenger train timing, but the total deviation time must be strictly controlled, i.e., the objective function. The increase in passenger trains. The design of this area aims to balance the principle of "passenger priority" with the needs of freight optimization, freeing up reasonable operating space for freight trains through limited adaptation of passenger trains.

[0180] ④ Neighborhood Operator 4: Adjustment of Time Domain Window for Single Passenger Trains

[0181] As a fallback strategy, a freight train is randomly selected, its current assignment window is obtained, and it is randomly assigned to another time domain window, such as... Figure 7 As shown, this region, through minor adjustments to passenger trains, aims to resolve extreme conflicts while maximizing the passenger experience.

[0182] The four types of neighborhood structures are selected using a roulette wheel selection method, with their probability allocation dynamically adjusted based on the priority of the optimization objectives: adjustment for multiple freight trains (40%), adjustment for a single freight train (30%), adjustment for multiple passenger trains (20%), and independent adjustment for a single passenger train (10%). This design retains the local optimization capability of variable neighborhood search while strengthening the constraint logic of "freight efficiency optimization conforming to passenger train punctuality" through differentiated configuration of operation weights, thus ensuring a precise fit between the neighborhood structure and the model's multi-objective requirements.

[0183] The train time domain window is based on the spatiotemporal matching of the departure time window and the rolling time domain window. The set of windows that can be assigned to the train is determined by the intersection of intervals. The specific calculation method is as follows:

[0184] First, the operational planning period is discretized based on a fixed time length (e.g., 120 minutes) to form a continuous and non-overlapping set of time domain windows. = Single window The time boundary is defined as: ,in, For window The start time, For window The end time, and = This ensures that the planning period is fully covered in terms of time.

[0185] Next, regarding existing passenger trains Its departure time window is determined by the existing scheduled departure time. Deviation from the maximum allowable amount Sure: For the newly added freight trains Its departure time window is from the originating station The permitted departure times are given directly: = .in For train The departure time window.

[0186] Therefore, the train Set of assignable time domain windows It is determined by the non-empty intersection of its departure time window and each time domain window. Specifically, for any window... If the following conditions are met: ,but Belongs to the train A set of assignable windows.

[0187] Finally, the train The set of assignable time domain windows is as follows: .

[0188] 2.3 Obtaining Feasible Solutions Using a Rolling Time-Domain Strategy

[0189] In the RH-SA-VNS algorithm of this invention, it is necessary to ensure that the solutions obtained during the iteration process are feasible. This allows the algorithm to start from a solution space that satisfies all constraints, gradually improving the quality of the solutions through local optimization and global exploration. This avoids the algorithm's neighborhood search getting bogged down in conflict resolution of numerous infeasible solutions, which could lead to low iteration efficiency or even failure to converge. Since the constructed optimization model is an NP-hard problem with strong constraint coupling and a large search space, generating feasible solutions is difficult. Therefore, to efficiently generate feasible solutions that satisfy complex constraints, this invention proposes a feasible solution generation strategy based on the rolling time domain.

[0190] This strategy decomposes the global problem into temporally related local subproblems, which are then solved using precise algorithms such as branch and bound or solvers such as Gurobi, thereby progressively constructing the solution space and dynamically maintaining constraint satisfaction. Figure 8 As shown. The combination of the decomposition capability of the rolling time domain and the solution capability of Gurobi, etc., reduces the complexity of the problem while ensuring the quality of local and global solutions, ultimately generating feasible solutions that satisfy all constraints efficiently. The core idea of ​​this rolling time domain method is: to divide the global optimization problem into time-related rolling time domain windows with a fixed time length, each window constitutes a local subproblem, covering passenger and freight trains, stations and section resources within the scope, focusing on the optimization of variables such as train arrival and departure times and stopping decisions within the window, and ensuring global coherence through boundary parameter propagation; to construct an integer programming sub-model for each window, guided by a weighted multi-objective function, integrating constraints such as stopping, safety intervals, boundary connections, time windows and transfers, and calling the Gurobi solver to obtain local optimal solutions; after solving, the train status across windows is extracted ( Figure 8 The yellow line in the middle serves as the input for the next window, scrolling forward and temporarily storing local solutions, which are then concatenated into a global solution. Each window contains newly arrived trains and cross-window trains continuing from the previous window. The final state parameters of cross-window trains from the previous window are extracted, including arrival and departure times. These parameters are embedded as hard constraints into the integer programming sub-model of the current window, ensuring that the solution of each local window satisfies both the current constraints and is connected to the previous window, thereby generating a high-quality initial feasible solution that satisfies all constraints.

[0191] The pseudocode for the feasible solution generation method in the rolling time domain is as follows:

[0192] 2.4 Simulated Annealing Parameter Settings

[0193] The core parameters of the simulated annealing algorithm include the initial temperature, cooling coefficient, and iteration termination condition, which need to be configured appropriately to balance solution space exploration and convergence efficiency. Its working mechanism and design logic are as follows:

[0194] ① Initial temperature The initial "energy level," which serves as the starting point for algorithm iterations, determines the probability of accepting a non-optimal solution in the initial stage. A higher initial temperature can endow the algorithm with stronger global exploration capabilities, avoiding premature entrapment in local optima.

[0195] ②Cooling coefficient This is a key parameter controlling the rate of temperature decay, enabling dynamic switching of the search strategy. Its function is to allow the algorithm to retain a large perturbation space in the early stages of iteration for global exploration, and then gradually focus on the neighborhood of high-quality solutions for local fine-tuning as the temperature decreases. The setting of the cooling coefficient must follow the principle of "smooth transition," using exponential cooling to ensure a natural transition in the exploration process, avoiding both excessively rapid convergence leading to the loss of optimal solutions and preventing iterative redundancy from increasing computational costs.

[0196] ③ Iteration Termination Condition: A dual constraint mechanism is constructed to ensure both algorithm efficiency and solution quality. The temperature threshold, from a thermodynamic convergence perspective, defines the critical state where the algorithm "freezes." The number of consecutive times no better solution is found determines whether the solution tends to stabilize, based on actual search efficiency. The logic is set as follows: when the temperature drops to an extremely low level or the solution iteration stagnates, the calculation is terminated promptly, balancing optimization effectiveness and time cost.

[0197] In simulated annealing, the Metropolis criterion is the core mechanism for achieving "finite acceptance of suboptimal solutions." It balances solution space exploration and convergence through a probabilistic acceptance strategy. The core logic is as follows:

[0198] For the neighborhood solution generated by the variable neighborhood search If its objective function value If a neighboring solution is better, then the neighboring solution is directly accepted, and the current solution is updated accordingly. ;like If the neighborhood solution is worse than the current solution, then calculate the acceptance probability: = .

[0199] in, The current iteration temperature. Generate a uniformly distributed temperature within the interval. random numbers ,like If a solution in a non-optimal neighborhood is found, it is accepted with probability to maintain the diversity of the solution space exploration; otherwise, it is rejected, and the current better solution is retained. This criterion allows the algorithm to have a higher probability of escaping local optima in the initial iterations, and gradually converges to a high-quality solution in later iterations, ensuring global optimization capability.

[0200] Example 2:

[0201] Based on the same inventive concept, this application also provides a system for optimizing the high-speed rail passenger and freight train timetables as described above. The solution provided by this system is similar to the solution described in the above method. Therefore, the specific limitations of one or more embodiments of the high-speed rail passenger and freight train timetable optimization system provided below can be found in the limitations of the high-speed rail passenger and freight train timetable optimization method described above, and will not be repeated here.

[0202] In one exemplary embodiment, a high-speed rail passenger and freight train timetable optimization system is provided, comprising:

[0203] Objectives and Decisions Module: Determine the optimization objectives, decision variables, objective functions, and constraints, and construct an optimization model for the high-speed railway passenger and freight train timetable;

[0204] Rolling temporal decomposition module: Based on the model and the freight train operation plan, the global problem is decomposed into temporally related local subproblems using a rolling temporal strategy. The linear programming solver is called to solve and concatenate the subproblems one by one to generate an initial feasible solution S0, which is then used as the initial current solution S.

[0205] Simulated annealing iteration module: Starting from the current solution S, the simulated annealing iteration process begins. In each iteration, a neighborhood operator is selected from four types of neighborhood structures using a roulette wheel approach to perturb S. For each time window affected by the perturbation, the linear programming solver is called again to obtain a local solution. and local solutions By splicing the solutions together, a candidate solution S′ is formed.

[0206] Acceptance and Update Module: The acceptance and update of solutions are controlled according to the Metropolis criterion: if S′ is better than S, then S is updated to S′; if it is worse than S, then it is accepted with a preset probability, otherwise it remains unchanged; at the same time, if the current solution is better than the historical best solution, then the best solution S* is updated.

[0207] Global solution optimization module: After multiple iterations, the algorithm continuously perturbs, reconstructs and optimizes the current solution, gradually improving the global solution, and finally outputs the optimized optimal passenger and freight train operation diagram S*.

[0208] Example 3:

[0209] This embodiment provides an application scenario. To verify the effectiveness of the proposed high-speed railway passenger and freight train timetable collaborative optimization model and the RH-SA-VNS algorithm, the Hangzhou-Ningbo-Wenzhou high-speed railway, a typical busy trunk line in the Yangtze River Delta region, is selected as the research object. This line is an important core component of China's coastal high-speed railway network, undertaking the passenger and freight transport functions connecting the Yangtze River Delta economic zone with the southeast coastal region. The line is 425 kilometers long and has 17 stations along its route. Figure 9 As shown.

[0210] This invention defines the upstream direction as Hangzhou East → Ningbo → Wenzhou South, with station numbers increasing sequentially along this direction, i.e., the station set. ={1,2,…,17}, of which Hangzhou East (Station 1), Ningbo (Station 6), and Wenzhou South (Station 17) are the main originating and terminating stations, undertaking key functions such as train turnaround and freight transshipment. The line operation routes are set in three types according to the distribution of major stations: 1→6 (Hangzhou East-Ningbo), 6→17 (Ningbo-Wenzhou South), and 1→17 (Hangzhou East-Wenzhou South), covering the entire line and major sections.

[0211] The existing timetable includes 60 passenger trains, with planned arrival and departure times based on actual operational data. To ensure passenger service stability, the maximum permissible deviation from the initial station departure time for existing passenger trains is limited to 45 minutes. The planning period is 24 hours a day, from 0:00 to 24:00, to align with the daily operational time cycle of the railway. Twenty new dedicated freight trains will be added, covering three routes: 6 trains will operate along the entire line (1→17), 9 trains will operate along sections 1→6, and 5 trains will operate along sections 6→17, matching the spatial distribution of freight demand. The total freight demand is 200 items, covering high-value-added goods such as fresh produce and express parcels. The departure time windows for freight trains will be set according to freight transport demand. The connection time for freight trains at transfer stations is set between 30 and 90 minutes, including loading and unloading, train stops, and other operational time. The interval travel time will be set differently based on line conditions and train type, with a minimum departure interval. =5 minutes, minimum arrival interval =4 minutes, ensuring a safe distance between trains; the minimum stop time at stations along the route is set at 2 minutes, and the maximum stop time is 9 minutes; additional stop time. With additional time for startup All intervals are set to 2 minutes to correct for the impact of station stops on the running time of the section.

[0212] The model and algorithm established in this invention are used to solve the above example. Based on the problem characteristics and multiple pre-experiment debugging, the algorithm parameters are set as follows: Window length is set to 120 minutes, and the total number of windows is 12. The number of freight trains adjusted is 4, and the number of passenger trains adjusted is 2. Objective function weights. , and All values ​​were set to 1. The simulated annealing parameters were set to an initial temperature of 100°C, a cooling coefficient of 0.95, a termination temperature of 0.1°C, and a maximum of 200 iterations. All computational experiments were performed on a computer equipped with an Intel(R) Core(TM) i3-10105 CPU @ 3.70GHz and 32.00 GB of RAM, using Microsoft Windows 11 (64-bit) operating system. Our algorithm was written in Python 3.11 and solved within the time-domain window using Gurobi 10.0.3.

[0213] Finally, the algorithm solves the problem, and its search and iteration process is as follows: Figure 10 As shown in the figure, the convergence graph of the algorithm presents the trend of the objective function value with the number of iterations. In the first 50 iterations or so, due to the high temperature and strong neighborhood search perturbation, the objective function value drops sharply, with a significant decrease. Subsequently, the objective function value stabilizes at 2819, and the curve tends to be horizontal, indicating that the algorithm converges to a stable optimal solution, verifying the effectiveness of the simulated annealing and variable neighborhood search collaborative mechanism.

[0214] The statistical results of the basic example solution are shown in Table 4. As can be seen, the objective function value is 2819, with the total travel time for freight trains being 2181 minutes, indicating that the time-domain window adjustment and neighborhood search strategy effectively reduced freight train travel time. The total deviation time for passenger trains is only 401 minutes, demonstrating the algorithm's advantage in ensuring passenger train punctuality and meeting the "passenger train priority" operation requirements. The total delay time for freight delivery is 235 minutes, indicating that the punctuality of freight transportation is guaranteed and meets the actual needs of cargo owners. Overall, the various indicators synergistically confirm the effectiveness of the proposed algorithm in balancing passenger and freight transportation efficiency and punctuality, and the output solution is practically feasible for operation.

[0215] Table 4 - Statistics of Indicators for Operation Chart Optimization Scheme

[0216]

[0217] The optimized train timetable is as follows Figure 11 As shown in the figure, the blue lines represent passenger trains, the red lines represent freight trains, the horizontal axis represents time, and the vertical axis represents stations. The figure demonstrates that the trains are distributed in an orderly manner along the time axis, with relatively uniform departure intervals, reflecting the algorithm's effective fulfillment of train interval constraints. The output timeline exhibits clear spatiotemporal order and practical feasibility, showcasing the algorithm's effectiveness in coordinating passenger and freight train operations and optimizing resource allocation.

[0218] It is understood that this invention focuses on the collaborative optimization of high-speed railway passenger and freight train timetables, taking into account transfer connections and time windows. Addressing the issue that existing high-speed railway timetables are centered on passenger transport and insufficiently consider freight-related constraints, this invention incorporates train transfer connections and freight time windows into the collaborative optimization framework for high-speed railway passenger and freight train timetables, clearly defining the "passenger train priority" principle. This represents a breakthrough in the collaborative optimization of high-speed railway freight trains and passenger trains, providing a completely new approach to the compilation of high-speed railway freight timetables.

[0219] It is understood that the multi-objective integer programming model constructed in this invention integrates constraints unique to high-speed rail freight, such as train transfer and connection constraints and freight time window constraints. With the objective of "minimizing the total travel time of freight trains, the total deviation time of existing passenger trains, and the freight delivery delay time," it systematically characterizes the multi-dimensional coupling relationship between requirements and constraints in optimizing passenger and freight train timetables. Theoretical analysis confirms that the model has NP-hard characteristics, providing a clear basis for algorithm design in terms of complexity. The proposed RH-SA-VNS hybrid optimization algorithm integrates the windowed solution capability of rolling time domain, the global optimization characteristics of simulated annealing, and the local development capability of variable neighborhood search. Through the dynamic selection and perturbation of four types of neighborhood structures, it effectively balances the solution efficiency and solution quality for large-scale problems. Case verification based on the Hangzhou-Ningbo-Wenzhou high-speed railway shows that the total travel time of freight trains, the total deviation time of passenger trains, and the total freight delivery delay time are coordinated and controlled in the optimized scheme, with an objective function value of 2819, and all indicators meeting actual operational requirements.

[0220] In one exemplary embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0221] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0222] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0223] This is accomplished by instructing related hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes described in the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0224] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0225] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for optimizing the timetable of high-speed passenger and freight trains, characterized in that, The method for optimizing the high-speed rail passenger and freight train timetable includes: The optimization objective, decision variables, objective function and constraints are determined, and a high-speed railway passenger and freight train operation timetable optimization model is constructed. Based on the model and the freight train operation plan, the global problem is decomposed into time-related local subproblems using a rolling time-domain strategy. The linear programming solver is called to solve and splice the subproblems one by one to generate an initial feasible solution S0, which is then used as the initial current solution S. Starting from the current solution S, the simulated annealing iteration process begins. In each iteration, a neighborhood operator is selected from four types of neighborhood structures using a roulette wheel approach to perturb S. For each time window affected by the perturbation, the linear programming solver is called again to obtain a local solution. and local solutions By splicing the solutions together, a candidate solution S′ is formed. The acceptance and updating of solutions are controlled according to the Metropolis criterion: if S′ is better than S, then S is updated to S′; if it is worse than S, then it is accepted with a preset probability, otherwise it remains unchanged; at the same time, if the current solution is better than the historical best solution, then the best solution S* is updated. After multiple iterations, the algorithm continuously perturbs, reconstructs, and optimizes the current solution, gradually improving the global solution and finally outputting the optimized passenger and freight train operation diagram S*.

2. The method for optimizing high-speed rail passenger and freight train schedules considering transfers and time windows as described in claim 1, characterized in that, The determination of the objective function specifically includes: A multi-objective optimization framework is constructed, and objective functions are set from three dimensions: freight efficiency, passenger transport stability, and cargo timeliness. The total travel time for the newly added freight trains is the shortest: (1) Equation (1) calculates the number of new freight trains departing from the originating station. To the final stop The total travel time is obtained by summing the differences between arrival and departure times. This indicates that freight train i has arrived at its destination. Arrival time, This indicates that freight train i originates from the starting station. Departure time For the collection of new freight trains; The existing bus has the least deviation: (2) To assemble existing passenger trains, This is the set of stations that train i passes through; Let be the departure time of train i at station j; Let i be the planned departure time of the existing passenger train i at station j; Equation (2) includes three parts: the departure time deviation of the existing passenger train at the starting station, the running time deviation between stations, and the stopping time deviation. The overall deviation is measured by summing the absolute values ​​and the goal is to minimize it. Minimize delivery delays: (3) in: ; The delivery delay penalty function for goods g depends on the actual delivery time. With respect to the allowed time window boundaries, For the aggregation of cargo flow, The actual delivery time of the goods flow g; Equation (3) is based on the actual delivery time of the goods. With time window boundaries, including the earliest allowed time and latest permitted time Given the relationship, calculate the delay penalty and sum it; Finally, a weighted summation method is used to integrate the three sub-objectives into a single objective function. By assigning different weights, the optimization requirements of freight efficiency, passenger stability, and cargo timeliness are balanced, as detailed below: (4) in, , , These are the weighting coefficients for the three sub-objectives.

3. The method for optimizing high-speed rail passenger and freight train schedules considering transfers and time windows as described in claim 1, characterized in that, The determination of the constraints specifically includes: (1) Constraints related to the time of arrival of goods (5) Constraint (5) represents the flow of goods. Delivery time From the station to its destination train The arrival times are determined by a weighted summation. As an indicator parameter, it represents the cargo flow specified in the cargo transportation plan. Whether by train Transportation; this constraint ensures that the delivery time of goods is consistent with the actual arrival time of the transporting train, providing a basis for calculating delivery delays; (2) Number of stops (6) Constraint (6) stipulates that trains must stop at pre-set stations, while allowing additional stops to improve the flexibility and feasibility of timetable preparation; (3) Stop time constraints (7) (8) Constraints (7) and (8) indicate the train If at the station If the train stops at a station, the stopping time must be within the minimum stopping time. and maximum stop time Between these parameters, the stopping time for freight trains includes the time required for loading and unloading goods, while the stopping time for passenger trains must meet the basic time requirements for passenger boarding and alighting. The 0-1 characteristic ensures that the station does not stop ( = 0), the train's stop time is 0; (4) Interval running time constraints (9) (10) Constraints (9) and (10) limit the train From the station to The interval running time of +1 must be within the minimum interval running time. With the maximum interval running time Within the range; among which, Additional time for train departure Additional time for stopping, by comparing with stop variables , The correlation between stops and the total travel time of the section is characterized. (5) Departure time constraints (11) (12) Constraint (11) indicates that for existing passenger trains Its departure times at each station No earlier than the originally planned time And the latest deviation from the original plan shall not exceed the allowable threshold. To ensure the punctuality of passenger services; constraint (12) indicates that for newly added freight trains Its starting station The departure time must be limited to the minimum allowable time. With maximum allowed time between; (6) Transfer connection time constraints (13) (14) Constraints (13) and (14) indicate that when a transfer connection exists, the connection time between the two trains must be within the minimum transfer time. With maximum transit time Between these, to ensure the smooth operation of cargo loading, unloading, and transfer; when there is no continuity, the constraint is automatically relaxed by using a large constant M; (7) Train interval constraints (15) (16) (17) Constraints (15)-(17) ensure the same station The departure interval between any two trains shall not be less than the minimum safe departure interval. The arrival interval between any two arriving trains shall not be less than the minimum safe arrival interval. ,in, As a 0-1 auxiliary variable, it can flexibly adapt to the interval requirements under different departure sequences by cooperating with the large constant M.

4. The method for optimizing high-speed rail passenger and freight train schedules considering transfers and time windows as described in claim 1, characterized in that, The adoption of the rolling time-domain strategy specifically includes: The global optimization problem is divided into rolling time-domain windows with fixed time lengths. Each window constitutes a local subproblem, covering passenger and freight trains, stations, and section resources within its scope. The focus is on optimizing train arrival and departure times and stopping decision variables within the window, and global coherence is ensured through boundary parameter propagation. For each window, an integer programming sub-model is constructed, guided by a weighted multi-objective function, integrating stopping, safety intervals, boundary connections, time windows, and transfer constraints. The Gurobi solver is called to obtain local optimal solutions. After solving, the train states across windows are extracted as inputs for the next window. The process is rolled forward and local solutions are temporarily stored. Finally, the solutions are concatenated to form the global solution. Each window contains newly connected trains and cross-window trains continuing from the previous window. The final state parameters of cross-window trains from the previous window are extracted, including arrival and departure times. These parameters are embedded as hard constraints into the integer programming sub-model of the current window to ensure that the solution of each local window satisfies the current constraints and is connected with the previous window, thereby generating a high-quality initial feasible solution that satisfies all constraints.

5. The method for optimizing high-speed rail passenger and freight train schedules considering transfers and time windows according to claim 1, characterized in that, The step of selecting a neighborhood operator from four types of neighborhood structures based on a roulette wheel betting method and perturbing S includes: Based on the principle of prioritizing passenger trains, four types of variable neighborhood structures are designed to achieve coordinated optimization of freight train efficiency and passenger train punctuality. Time window adjustment for multiple freight trains: Randomly select n (n≥2) freight trains and coordinate the adjustment of their time windows to optimize overall resource utilization; Adjustment of time window for single freight trains: For single freight trains, a new time window will be randomly selected and assigned. Adjustment of time window for multiple passenger trains: Randomly select n (n≥2) passenger trains and adjust their departure time and stop schedule within the maximum allowable deviation range; Adjustment of time window for single passenger train: Randomly select one passenger train and assign its time window to another suitable time window to resolve extreme conflicts; The above four types of neighborhood structures are selected through a roulette wheel selection process, and their probability allocation is dynamically adjusted according to the priority of the optimization objective: 40% for multiple freight trains, 30% for a single freight train, 20% for multiple passenger trains, and 10% for a single passenger train. The train time domain window is based on the spatiotemporal matching of the departure time window and the rolling time domain window. The set of windows that can be assigned to the train is determined by the intersection of intervals. The specific calculation method is as follows: First, the operational planning period is discretized based on a fixed time length, forming a set of continuous and non-overlapping time domain windows. = Single window The time boundary is defined as: ,in, For window The start time, For window The end time, and = This ensures that the planning period is fully covered in terms of time. Next, regarding existing passenger trains Its departure time window is determined by the existing scheduled departure time. Deviation from the maximum allowable amount Sure: For the newly added freight trains Its departure time window is from the originating station The permitted departure times are given directly: = .in For train Departure time window; Therefore, the train Set of assignable time domain windows It is determined by the non-empty intersection of its departure time window and each time domain window; specifically, for any window If the following conditions are met: ,but Belongs to the train The set of assignable windows; Finally, the train The set of assignable time domain windows is as follows: .

6. The method for optimizing high-speed rail passenger and freight train schedules considering transfers and time windows according to claim 1, characterized in that, Starting from the current solution S, the simulated annealing iteration process is initiated, which specifically includes: The core parameters of the simulated annealing algorithm include: initial temperature, cooling coefficient, and iteration termination condition. In the simulated annealing algorithm, the Metropolis criterion is the core mechanism for achieving finite acceptance of suboptimal solutions. It balances solution space exploration and convergence through a probabilistic acceptance strategy. The core logic is as follows: For the neighborhood solution generated by the variable neighborhood search If its objective function value If a neighboring solution is better, then the neighboring solution is directly accepted, and the current solution is updated accordingly. ;like If the neighborhood solution is worse than the current solution, then calculate the acceptance probability: = ; in, For the current iteration temperature, generate a uniformly distributed interval. random numbers ,like If the solution is not optimal, then the non-optimal neighborhood solution is accepted with probability to maintain the diversity of solution space exploration; otherwise, it is rejected and the current better solution is retained. This criterion makes the algorithm more likely to escape local optima in the initial iteration and gradually converge to high-quality solutions in later iterations, ensuring global optimization capability.

7. A high-speed rail passenger and freight train timetable optimization system, characterized in that, The high-speed rail passenger and freight train timetable optimization system includes: Objectives and Decisions Module: Determine the optimization objectives, decision variables, objective functions, and constraints, and construct an optimization model for the high-speed railway passenger and freight train timetable; Rolling temporal decomposition module: Based on the model and the freight train operation plan, the global problem is decomposed into temporally related local subproblems using a rolling temporal strategy. The linear programming solver is called to solve and concatenate the subproblems one by one to generate an initial feasible solution S0, which is then used as the initial current solution S. Simulated annealing iteration module: Starting from the current solution S, the simulated annealing iteration process begins. In each iteration, a neighborhood operator is selected from four types of neighborhood structures using a roulette wheel approach to perturb S. For each time window affected by the perturbation, the linear programming solver is called again to obtain a local solution. and local solutions By splicing the solutions together, a candidate solution S′ is formed. Acceptance and Update Module: The acceptance and update of solutions are controlled according to the Metropolis criterion: if S′ is better than S, then S is updated to S′; if it is worse than S, then it is accepted with a preset probability, otherwise it remains unchanged; at the same time, if the current solution is better than the historical best solution, then the best solution S* is updated. Global solution optimization module: After multiple iterations, the algorithm continuously perturbs, reconstructs and optimizes the current solution, gradually improving the global solution, and finally outputs the optimized optimal passenger and freight train operation diagram S*.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the steps of the high-speed rail passenger and freight train timetable optimization method according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the high-speed rail passenger and freight train timetable optimization method as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the high-speed rail passenger and freight train timetable optimization method as described in any one of claims 1-6.

Images