Subway transport capacity resource collaborative optimization configuration method for luggage preposed consignment

Abstract

The invention discloses a subway transport capacity resource collaborative optimization configuration method for luggage preposed consignment, and belongs to the technical field of luggage consignment optimization, and the method comprises the following steps: S1, obtaining a complete consignment demand and train transport capacity basic parameter set through a network platform; s2, a station set and a train set used for executing a user luggage consignment task are screened out with the running time of the subway train and the luggage storage facility allocation condition as screening conditions; s3, establishing an integer linear programming model with the goal of minimizing the comprehensive cost of integrating the compartment configuration cost, the residence time cost and the peak penalty cost; and S4, a branch pricing algorithm is adopted, and a luggage consignment scheme with the optimal comprehensive cost is output based on operation constraint conditions related to subway trains and user demands. By the adoption of the method, the contradiction between the requirement for light travel and the requirement for carrying large luggage to go back and forth in a transportation hub is solved, and the resource utilization rate of an existing subway system is increased.

Description

Technical Field

[0001] This invention relates to the field of baggage check-in optimization technology, and in particular to a method for collaborative optimization of subway capacity resources for pre-check-in of baggage. Background Technology

[0002] With the increasing development of urban rail transit networks, the contradiction between passengers' demand for "light travel" and the current situation of carrying large luggage to and from the city center and transportation hubs (airports and train stations) is becoming increasingly prominent.

[0003] Traditional luggage storage services require passengers to store and retrieve their luggage at the same location, which fails to fundamentally solve the problem of passengers carrying luggage throughout their journey. The key reason is that during off-peak hours, the large-scale subway network generally has redundant capacity that is not effectively reused, resulting in a waste of resources. This also means that the core need of passengers to travel light cannot be fundamentally met through the coordination of appropriate transportation resources, creating a mismatch between demand and resources. Summary of the Invention

[0004] The purpose of this invention is to provide a method for the coordinated optimization of subway capacity resources for pre-delivery of luggage, thereby solving the above-mentioned technical problems.

[0005] To achieve the above objectives, this invention provides a method for the collaborative optimization of subway capacity resources for pre-delivery baggage, comprising the following steps: S1. Based on baggage shipping demand and subway capacity information requirements, collect users’ baggage shipping demand data and obtain train timetables through the network platform to obtain a complete set of shipping demand and train capacity basic parameters. S2. Based on the baggage demand data collected in step S1, and using the subway train operating times and the availability of baggage storage facilities as filtering conditions, a set of stations and a set of trains that can be used to carry out baggage transport tasks for users are selected. S3. Based on the station set and train set from step S2, establish an integer linear programming model with the goal of minimizing the comprehensive cost of integrating carriage configuration cost, dwell time cost, and peak penalty cost. S4. Based on the integer linear programming model in step S3, the branch pricing algorithm is adopted, and based on the operational constraints related to subway trains and user demand, the baggage check-in scheme with the best overall cost is output.

[0006] Preferably, the user's baggage check-in request data in step S1 includes the weight of the user's baggage, the origin station information to be checked in, the time information of the baggage arriving at the origin station, and the departure time of the user's subsequent flight or train.

[0007] Preferably, in step S2, before screening the stations and trains, the usage is assumed to be limited, including: limiting the research object to a single direction of a single subway line containing a railway station or an airport station; limiting the user's luggage delivery task to not deviating from the task route or being cancelled in the actual operation process.

[0008] Preferably, the operation constraints of the branch-and-price algorithm in step S4 include: the delivery time limit constraint that the luggage must be delivered to the target location station before the user departs, the loading and unloading operation time constraint of the luggage within the limited stop time at the station, the physical loading capacity constraint of the carriage, and the minimum load rate constraint set to ensure resource utilization efficiency.

[0009] Preferably, the specific steps of the algorithm in step S4 include: S41, based on the delivery demand in step S1 and the train capacity basic parameter set, input the branch-and-price algorithm as input data; S42, taking the search tree as the solving framework, judging whether the search tree is empty or the upper bound and the lower bound corresponding to the search tree are equal; if yes, output the optimal luggage delivery scheme and the optimal target value, and the algorithm terminates; if no, execute step S43; S43, select a node to be processed from the search tree; S44, based on the node to be processed selected in step S43 and the integer linear programming model in step S3, construct a restrictive linear master problem RLMP containing operation constraints, and solve the operation constraints in the RLMP by using a high-performance mathematical programming solver Gurobi to obtain corresponding dual variables; S45, pass the dual variables obtained by solving the RLMP in step S44 to the luggage carriage configuration sub-problem, and solve the luggage carriage configuration sub-problem by using the label method; S46, judge whether the reduced cost obtained by solving the luggage carriage configuration sub-problem is less than 0; if yes, add the corresponding train to the RLMP in step S44 and return to step S44; if no, execute step S47; S47, execute the branching operation on the result of Reducedcost>0 obtained in step S46, and judge whether the current solution is greater than the upper bound; if yes, execute step S49; if no, execute step S48; S48, judge whether the solution obtained in step S47 is a fractional solution; if yes, update the lower bound, branch and generate a sub-node, and execute step S49; if no, update the upper bound and execute step S49; S49, delete the corresponding node in the search tree, and return to step S2.

[0010] Preferably, the search tree in step S42 is empty, all to-be-solved sub-problems have been traversed, the upper bound is the actual optimal target value, and the lower bound is the theoretical optimal target value lower limit. The upper bound = the lower bound surface actual solution target value is equal to the theoretical optimal target value.

[0011] Preferably, the dual variable in step S44 is the value coefficient corresponding to the operation constraint condition, so as to quantify the optimal target value of the main problem under the operation constraint condition.

[0012] Therefore, the subway capacity resource collaborative optimization configuration method for the above-mentioned pre-transport luggage has the beneficial effects of: 1. Through the collaborative optimization of subway capacity resources and pre-transport luggage demand, the realistic contradiction between passenger light travel demand and carrying heavy luggage to and from the transportation hub is effectively resolved. The subway peak period redundant capacity is used to fundamentally solve the passenger travel pain point and greatly improve the travel convenience. The resource utilization efficiency of the existing subway system is significantly improved, a new logistics service scene and revenue model are developed for the subway operator, and the win-win of passenger travel experience improvement and subway operation benefit improvement is realized.

[0013] 2. The integer linear programming model integrates the carriage configuration cost, the detention time cost, the peak punishment cost and other multi-dimensional optimization targets, and integrates the subway operation core constraint condition to construct a logical and quantifiable collaborative optimization decision tool. The tool provides a systematic analysis framework and scientific decision basis for the precise matching of pre-transport luggage demand and subway capacity resources, realizes the global optimal decision orientation under multi-objective and multi-constraint, effectively avoids resource waste or demand mismatch caused by scattered decision, lays a quantitative analysis foundation for efficient solving of the optimal transport scheme, and ensures that the decision process has scientificity and operability.

[0014] 3. The branch and price algorithm uses the search tree framework, the collaborative solving logic of the restrictive linear main problem and the sub-problem, and the efficient adaptation of the Gurobi solver and the label method. The algorithm has significant solving efficiency and can quickly traverse the multi-constrained and high-dimensional decision space, greatly shortens the solving time of large-scale problems, meets the demand for rapid decision-making in actual operation, has outstanding solving accuracy, quantifies the constraint value of the dual variable, locks the integer optimal solution through the branch operation, ensures that the output result is consistent with the theoretical optimal target value, effectively avoids the cost waste or resource mismatch caused by the approximate solution, provides strong adaptive and efficient solving support for the integer linear programming model, and guarantees the scientificity and operability of the collaborative optimization scheme in a large-scale scene.

[0015] The technical solutions of the present application will be further described in detail below with reference to the drawings and embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A flow chart of a subway transport capacity resource collaborative optimization configuration method for pre-transport of luggage is provided. DETAILED DESCRIPTION

[0017] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the embodiments of the present application are further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the embodiments of the present application and not to limit the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of the present application. Examples of the embodiments are shown in the drawings, wherein the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout.

[0018] It should be noted that the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or server comprising a series of steps or units does not have to be limited to only those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to the process, method, product or device.

[0019] The embodiments of the present application are described in detail below with reference to the drawings.

[0020] The service coverage of luggage transport in the prior art is limited to a few sites and does not form a large-scale collaborative network. The problem of traditional storage requiring access in the same place has not been completely solved. Most explorations have failed to systematically integrate subway redundant transport capacity and passenger luggage transport demand, and the full-process optimization adaptation of pre-transport of luggage and subway transport capacity resources has not been achieved.

[0021] The present application is designed based on the above analysis, and the subway transport capacity resource collaborative optimization configuration method for pre-transport of luggage is shown in FIG. 1. Figure 1 The subway transport capacity resource collaborative optimization configuration method for pre-transport of luggage comprises the following steps: S1, based on luggage transport demand and subway transport capacity basic information demand, collecting user luggage transport demand data through a network platform and obtaining train schedules to obtain a complete set of transport demand and train transport capacity basic parameters; The user luggage transport demand data in step S1 includes the weight of the user's luggage, the starting station information of the luggage to be transported, the time information of the luggage arriving at the starting station, and the departure time of the subsequent flight or train to be taken by the user.

[0022] By comprehensively collecting user luggage weight, check-in starting station, luggage arrival starting station time, and subsequent flight / train departure time, etc. core demand data, the precise description of the check-in demand is realized, and key data support is provided for subsequent screening of suitable stations and trains, and construction of an integer linear programming model under multiple constraints, i.e. luggage weight can match the loading capacity constraint of the car, the starting station and arrival time can dock the train operation time to realize precise matching of transport capacity, and the subsequent travel departure time can guarantee the luggage delivery time limit requirement, which ensures that the check-in demand and subway transport capacity resources are more targeted in matching, effectively avoiding configuration deviation caused by missing key demand information, such as overloading and delivery delay, etc. It lays a precise and reliable demand data foundation for whole-process collaborative optimization, and then ensures that the subsequent optimization scheme meets the actual travel demands of users and adapts to the actual subway operation.

[0023] S2, based on the luggage demand data information collected in step S1, and screening based on the running time of the subway train and the equipped luggage storage facilities to screen out a set of stations and a set of trains for performing the luggage task of the check-in user; In step S2, before screening the stations and trains, the usage is assumed to be limited, including: limiting the research object to a single direction of a single subway operating line containing a railway station or an airport station; limiting the user's luggage check-in task to not deviating from the task route or being canceled in the actual operation process. This effectively reduces the occupation of invalid transport capacity resources, improves the utilization efficiency of core resources, and reduces the complexity of subsequent optimization configuration by clearly defining the operation scenario, avoiding the problem of scheme disconnection caused by ambiguous scenarios or resource mismatch, ensuring that the selected resource set is highly consistent with the actual operation logic and user check-in demand, and laying a solid foundation for resource adaptation for subsequent multi-dimensional cost modeling and efficient solution.

[0024] S3, based on the set of stations and trains in step S2, an integer linear programming model is established with the minimum comprehensive cost of integrating car configuration cost + detention time cost + peak penalty cost as the goal; S4, based on the integer linear programming model in step S3, a branch and price algorithm is used, and based on the operation constraint conditions related to the subway train and user demand, the optimal luggage check-in scheme with the minimum comprehensive cost is output.

[0025] The specific steps of the algorithm in step S4 include: S41, based on the check-in demand and train capacity basic parameter set in step S1, as input data, input into the branch and price algorithm; S42, taking the search tree as a solving framework, judging whether the search tree is empty or the upper bound of the search tree is equal to the lower bound; if yes, outputting the optimal baggage handling scheme and the optimal target value, and terminating the algorithm; if no, executing step S43; The empty search tree in step S42 means that all sub-problems to be solved have been traversed, the upper bound is the actual optimal target value, the lower bound is the theoretical optimal target value, and the upper bound = the lower bound means that the actual solution target value is equal to the theoretical optimal target value.

[0026] S43, selecting a node to be processed from the search tree; S44, based on the node to be processed selected in step S43 and the integer linear programming model in step S3, constructing a restrictive linear master problem RLMP containing operation constraints, and using a high-performance mathematical programming solver Gurobi to solve the operation constraints in the RLMP to obtain corresponding dual variables; The dual variable in step S44 is the value coefficient of the corresponding operation constraint, which quantifies the optimal target value of the master problem.

[0027] Gurobi accurately obtains the dual variable corresponding to each operation constraint by solving the operation constraints in the RLMP, such as delivery time limit, loading and unloading time, loading capacity, and minimum loading rate. This variable quantifies the value coefficient of each operation constraint on the optimal target value of the master problem. On the one hand, it provides key input basis for the subsequent solving of the baggage compartment configuration sub-problem, enabling the label method to efficiently judge the optimization space of the sub-problem solution, i.e., whether the cost is less than 0, and then quickly filter out effective columns that can improve the optimality of the master problem and feedback to the RLMP. On the other hand, by quantifying the constraint value through the dual variable, the algorithm can accurately identify the core constraints that significantly affect the global optimization, avoiding ineffective search and redundant calculation, which not only improves the overall solving efficiency of the branch-and-price algorithm, but also ensures that the operation constraints are strictly and reasonably considered in the whole process of solving, providing key constraint quantization and solving support for the final output of the baggage handling scheme that meets the actual operation requirements and has optimal and feasible performance.

[0028] S45, passing the dual variable obtained by solving the RLMP in step S44 to the baggage compartment configuration sub-problem, and solving the baggage compartment configuration sub-problem by the label method; By passing the dual variables obtained from the RLMP solution to the baggage car configuration subproblem, the labeling method can efficiently handle multi-dimensional decisions such as baggage and car matching, transportation route planning, and operational constraints such as loading capacity and loading / unloading time in the subproblem. Ultimately, it solves for feasible baggage car configuration schemes, including specific car allocation, baggage loading arrangements, and cross-station transportation connection plans, while simultaneously calculating the reduced cost corresponding to each scheme. The dual variables provide precise optimization guidance for solving the subproblem, ensuring that the subproblem solution is highly aligned with the overall cost minimization objective of the main problem, avoiding the generation of invalid or deviating configuration schemes. The efficient solution characteristics of the labeling method significantly shorten the computation time of the subproblem. Furthermore, by reducing costs, it can be directly determined whether the configuration scheme can improve the current optimal solution of the main problem; that is, a reduced cost < 0 indicates a valid column. This significantly improves the overall solution efficiency of the branch pricing algorithm and ensures the feasibility and optimality of the subproblem solution, providing core support for the accurate output of the globally optimal baggage handling scheme.

[0029] S46. Determine whether the reduced cost obtained from solving the luggage compartment configuration subproblem is less than 0; if yes, add the corresponding column to RLMP in step S44 and return to step S44; if no, proceed to step S47. S47. Perform a branch operation on the result of Reducedcost > 0 obtained in step S46, and determine whether the current solution is greater than the upper bound; if yes, proceed to step S49; if no, proceed to step S48. By directly deleting the corresponding node where the current solution is greater than the upper bound, precise pruning of the search tree is achieved, effectively avoiding invalid searches and redundant calculations, significantly reducing the overall search space of the algorithm, and improving the solution efficiency of the branch pricing algorithm. Solutions less than or equal to the upper bound are guided into the subsequent fractional solution judgment process, laying a solid foundation for updating the upper and lower bounds of the algorithm and locking the integer optimal solution. This ensures that the algorithm always converges around the goal of minimizing the global comprehensive cost and avoids deviating from the optimization direction. At the same time, the constraints are gradually refined through branch operations, pushing the solution to approach the integer feasible solution from the fractional solution. This ensures that the final output baggage check-in plan not only meets the integer decision requirements of actual operation but also has global optimality, providing an efficient and accurate solution path for solving large-scale complex NP-hard problems.

[0030] S48. Determine whether the solution obtained in step S47 is a fractional solution; if yes, update the lower bound, branch and generate child nodes, and execute step S49; if no, update the upper bound and execute step S49. By determining whether the solution obtained in step S47 is a fractional solution, the feasibility type of the current solution is clarified. If it is a fractional solution, the lower bound of the algorithm is updated and a new child node is generated by branching. If it is an integer solution, the upper bound of the algorithm is updated. This process achieves dynamic optimization of the upper and lower bounds of the algorithm and iterative expansion of the search tree nodes. It accurately adapts to the actual operational needs of integer decisions such as carriage allocation and baggage loading in baggage handling schemes, avoiding the practical ineffectiveness of theoretical solutions due to their non-integer characteristics. The dynamic updating of the upper and lower bounds continuously narrows the search range of the optimal solution, accelerating the convergence speed of the algorithm towards the global optimal solution. At the same time, the generation of new child nodes by branching ensures the comprehensive exploration of potential better solutions. Updating the upper bound locks the currently found optimal feasible solution, effectively avoiding the omission of the optimal solution. This works in synergy with the pruning operation mentioned earlier, ensuring the accuracy of the algorithm's solution and further improving the efficiency of the branch pricing algorithm in handling large-scale NP-hard problems. This provides key iterative optimization support for the final output of a baggage handling scheme that is both practical and optimal.

[0031] S49. Delete the corresponding node in the search tree and return to step S2.

[0032] By deleting nodes in the search tree that have already been processed, such as those where the current solution is greater than the upper bound, after a fractional solution branch, or after an integer solution is confirmed, the search tree is dynamically simplified, and the process returns to step S2 to re-enter the search tree traversal and node processing loop. Deleting invalid or processed nodes significantly reduces the memory usage of the search tree, avoids repeated traversal of nodes that have no optimization value, further compresses the algorithm's search space, and improves the solution efficiency of subsequent iterations. It ensures that all nodes to be processed in the search tree can be traversed by the system, avoiding the omission of potential optimal solutions. This works in conjunction with the branching, pruning, and upper / lower bound updates mentioned earlier to drive the algorithm to continuously converge toward the global integer optimal solution. This not only ensures the integrity and systematic nature of the solution process but also further improves the efficiency of the branch pricing algorithm in handling large-scale NP-hard problems through the dynamic optimization of the search tree. It provides closed-loop iterative process support for the final output of a baggage check-in solution that meets actual operational requirements and has the best overall cost.

[0033] In summary, through a collaborative design involving precise data support, adaptive resource selection, multi-objective modeling, and efficient algorithm solutions, the minimum load factor was relaxed from 80% to 20%, significantly reducing the total baggage waiting time by 74.01%. Adding two baggage trains after off-peak hours reduced the total waiting time from 62,550 seconds to 54,922 seconds, effectively resolving the core conflict between passengers traveling light and those carrying large luggage to and from transportation hubs, while also fully activating redundant subway capacity during off-peak hours and avoiding resource waste. By comprehensively collecting key baggage handling needs and subway capacity parameters, and combining them with operational scenario constraints to achieve precise selection of station and train resources, a data and resource foundation closely aligned with practical operations was established for optimization. Finally, a systematic collaborative optimization was constructed based on an integer linear programming model integrating multi-dimensional costs. The decision-making framework, through a dedicated branch pricing algorithm, breaks through the efficiency and optimality bottlenecks of the traditional commercial solver Gurobi in large-scale NP-hard problems. Utilizing RLMP and iterative solutions to subproblems, quantified constraints of dual variables, and a closed-loop mechanism of branch pruning and dynamic updates of upper and lower bounds, it achieves efficient and accurate solutions in large-scale scenarios. This ensures the practical feasibility of baggage handling solutions under multiple constraints such as delivery time and loading capacity, while also achieving the global optimal goal of minimizing overall costs. Ultimately, it significantly improves passenger travel convenience and opens up new logistics services and revenue models for subway operators, achieving a win-win situation for improved user experience, increased subway resource utilization efficiency, and operational benefits. It possesses strong practical value for solving large-scale logistics scheduling problems in the real world.

[0034] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for collaborative optimization of subway capacity resources for pre-delivery of luggage, characterized in that: Includes the following steps: S1. Based on baggage shipping demand and subway capacity information requirements, collect users’ baggage shipping demand data and obtain train timetables through the network platform to obtain a complete set of shipping demand and train capacity basic parameters. S2. Based on the baggage demand data collected in step S1, and using the subway train operating times and the availability of baggage storage facilities as filtering conditions, a set of stations and a set of trains that can be used to carry out baggage transport tasks for users are selected. S3. Based on the station set and train set from step S2, establish an integer linear programming model with the goal of minimizing the comprehensive cost of integrating carriage configuration cost, dwell time cost, and peak penalty cost. S4. Based on the integer linear programming model in step S3, the branch pricing algorithm is adopted, and based on the operational constraints related to subway trains and user demand, the baggage check-in scheme with the best overall cost is output.

2. The method for collaborative optimization of subway capacity resources for pre-delivery baggage as described in claim 1, characterized in that: The user's baggage check-in request data in step S1 includes the weight of the user's baggage, the origin station information for the baggage to be checked in, the time information for the baggage to arrive at the origin station, and the departure time of the user's subsequent flight or train.

3. The method for collaborative optimization of subway capacity resources for pre-delivery of baggage as described in claim 2, characterized in that: In step S2, before screening stations and trains, assumptions about the usage need to be made, including: limiting the research object to a single direction of a single subway line that includes a railway station or airport station; and limiting that the user's baggage check-in task will not deviate from the task route or be canceled during actual operation.

4. The method for collaborative optimization of subway capacity resources for pre-delivery baggage as described in claim 3, characterized in that: The operational constraints of the branch pricing algorithm in step S4 include: the delivery time constraint that baggage must be delivered to the target station before the user departs, the loading and unloading time constraint of baggage within the limited stopping time at the station, the physical loading capacity constraint of the carriage, and the minimum full load rate constraint set to ensure resource utilization efficiency.

5. A method for collaborative optimization of subway capacity resources for pre-delivery of baggage, as described in claim 4, is characterized in that: The specific steps of the algorithm in step S4 include: S41. Based on the shipping demand and train capacity basic parameter set in step S1, the data is input into the branch pricing algorithm. S42. Using the search tree as the solution framework, determine whether the search tree is empty or whether the upper and lower bounds of the search tree are equal. If yes, output the optimal baggage check-in plan and the optimal target value, and the algorithm terminates. If no, proceed to step S43. S43. Select the node to be processed from the search tree; S44. Based on the nodes to be processed selected in step S43 and the integer linear programming model in step S3, construct a restricted linear master problem RLMP containing operational constraints, and use the high-performance mathematical programming solver Gurobi to solve the operational constraints in RLMP to obtain the corresponding dual variables. S45. Transfer the dual variables obtained from solving RLMP in step S44 to the baggage car configuration subproblem, and solve the baggage car configuration subproblem by the labeling method. S46. Determine whether the reduced cost obtained from solving the luggage compartment configuration subproblem is less than 0; if yes, add the corresponding column to RLMP in step S44 and return to step S44; if no, proceed to step S47. S47. Perform a branch operation on the result of Reducedcost > 0 obtained in step S46, and determine whether the current solution is greater than the upper bound; if yes, proceed to step S49; if no, proceed to step S48. S48. Determine whether the solution obtained in step S47 is a fractional solution; if yes, update the lower bound, branch and generate child nodes, and execute step S49; if no, update the upper bound and execute step S49. S49. Delete the corresponding node in the search tree and return to step S2.

6. The method for collaborative optimization of subway capacity resources for pre-delivery of baggage as described in claim 5, characterized in that: In step S42, an empty search tree indicates that all subproblems to be solved have been traversed. The upper bound is the actual optimal objective value, and the lower bound is the lower limit of the theoretical optimal objective value. The upper bound = the lower bound indicates that the actual solution objective value is equal to the theoretical optimal objective value.

7. A method for collaborative optimization of subway capacity resources for pre-delivery of baggage, as described in claim 6, is characterized in that: The dual variable in step S44 is the value coefficient of the corresponding operational constraint, in order to quantify the optimal objective value of the operational constraint for the main problem.

Images