Express transportation network design method, device and equipment

Abstract

The application relates to the technical field of operations research, in particular to an express delivery network design method, device and equipment, which comprises the following steps: determining a decision variable according to express package delivery demand; identifying the number of vehicles on each feasible vehicle path, the number of packages on each feasible package path and the distribution mechanism of vehicle capacity on each sub-path of the decision variable; constructing a vehicle space-time network according to the number of vehicles on each feasible vehicle path, constructing a package space-time network according to the number of packages on each feasible package path, establishing a double space-time network of express delivery according to the vehicle space-time network, the package space-time network, the distribution mechanism of vehicle capacity on each sub-path, the transfer limit and the transfer cost, determining a drivable path set of vehicles and packages, and generating an express delivery network design model to meet the delivery demand. Therefore, the problems that the transportation cost and efficiency are difficult to balance, the network design is complex, and the data processing capacity is insufficient in the prior art are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of operations research, in particular to a method, device and equipment for designing express delivery network. BACKGROUND

[0002] With the rapid development of e-commerce, the express delivery industry is booming at an unprecedented speed, not only deeply changing people's shopping way, but also greatly promoting the innovation and transformation of the logistics industry. However, behind this rapid growth, there are also double challenges of transportation cost control and efficiency improvement, which put higher requirements on express companies.

[0003] In the face of huge transportation volume and complex transportation network, transportation cost has become an important issue that express companies cannot avoid. Traditionally, express companies rely on freight integration systems to reduce costs, and realize the scale effect of batch transportation by combining small goods. However, the indispensable transfer link in this process has become a big problem for cost control. Transfer not only requires additional labor and equipment investment, but also increases time cost, making the whole transportation chain more complex and difficult to manage.

[0004] In order to cope with this challenge, the transportation network design problem is explored to seek more refined models and algorithms to optimize the transportation process, improve transportation efficiency and reduce transportation cost. However, the traditional model is not competent in dealing with the complex and variable actual situation of Chinese express industry. These models are often based on simplified assumptions, such as ignoring transfer cost or limiting vehicles to make direct transportation between two points, which cannot accurately reflect the complexity and dynamics in the actual transportation process. SUMMARY

[0005] The present application provides a method, device and equipment for designing express delivery network to solve the problems of difficult balance between transportation cost and efficiency, complex network design and insufficient data processing capacity in the prior art.

[0006] The first aspect of this application provides a method for designing an express delivery transportation network, comprising the following steps: determining decision variables based on the demand for express parcel transportation, and obtaining transshipment restrictions and transshipment costs; identifying the number of vehicles on each feasible vehicle route, the number of parcels on each feasible parcel route, and the allocation mechanism of vehicle capacity on each sub-route for the decision variables; constructing a vehicle spatiotemporal network based on the number of vehicles on each feasible vehicle route, constructing a parcel spatiotemporal network based on the number of parcels on each feasible parcel route, and establishing a dual spatiotemporal network for express delivery transportation based on the vehicle spatiotemporal network, the parcel spatiotemporal network, the allocation mechanism of vehicle capacity on each sub-route, transshipment restrictions, and transshipment costs; determining the set of drivable routes for vehicles and parcels based on the dual spatiotemporal network, generating an express delivery transportation network design model based on the set of drivable routes, and using the express delivery transportation network design model to meet the demand for express parcel transportation.

[0007] Optionally, a vehicle spatiotemporal network is constructed based on the number of vehicles on each feasible vehicle route, including: enumerating all feasible vehicle routes, wherein the number of transfer centers along the route is less than a preset value, and for the same transfer center sequence, different departure times correspond to different vehicle routes; enumerating all sub-paths of each vehicle route to form a sub-path set, and constructing a final sub-path set based on the sub-path sets corresponding to all vehicle routes, wherein a sub-path is an edge connecting two spatiotemporal nodes of transfer centers on the route, and all packages on the edge are loaded from the starting point of the sub-path until they are unloaded at the ending point of the sub-path; and constructing a vehicle spatiotemporal network based on the final sub-path set.

[0008] Optionally, a package spatiotemporal network is constructed based on the number of packages on each feasible package path, including: obtaining a set of spatiotemporal nodes of the transfer center; a set of all sub-paths and waiting edges on each feasible package path; and constructing the package spatiotemporal network based on the set of spatiotemporal nodes of the transfer center, the set of all sub-paths and waiting edges.

[0009] Optionally, the transshipment restrictions are as follows: parcels cannot switch sub-routes within transshipment centers where parcel transshipment operations are not possible; transshipment costs are calculated based on the number of sub-routes for each feasible parcel route; and vehicle capacity is allocated on each sub-route as follows: vehicle route p has x p There are 10 vehicles, and the corresponding path contains n. p An edge, denoted by (i, j), represents a sub-path connecting the i-th and j-th transfer centers. The variable... This represents the proportion of capacity allocated to subpath (i, j), where 1 ≤ m ≤ n on the m-th edge (m-1, m). p , containing subsets of subpaths Subpath subset The capacity constraint is:

[0010]

[0011] The capacity is obtained by summing the capacities of all vehicle paths assigned to the sub-paths s = (i, j).

[0012]

[0013] If loading packages at the i′-th transit center on path p is prohibited, then 1≤i′≤n p -1 removes the variable. And the i′-th capacity constraint.

[0014] Alternatively, the express delivery network design model is as follows:

[0015]

[0016] in, This is the cost of operating the regular bus service. For the cost of operating the overtime bus, The cost of a feasible package route, x p Let p be the number of vehicles on the vehicle path. Indicates vehicle transportation network All feasible regular bus routes, This indicates all feasible routes for extra buses. Indicates parcel delivery network All feasible package paths in the list. Indicates x on the path p The proportion of capacity allocated to a vehicle on subpath s, w k Indicates the weight of the package. y represents the proportion of capacity allocated to subpath (i, j). kq Let N be the weight of the k-th demand along the package path q. f This represents the number of vehicles in the regular convoy f. Let n be the set of subpaths derived from vehicle path p. p Let p be the number of edges contained in the vehicle path.

[0017] Optionally, a courier transportation network design model is used to meet the needs of courier parcel transportation, including: selecting an initial subset of vehicle and parcel paths based on the courier transportation network design model; generating a restricted master problem based on the initial subset of vehicle and parcel paths; generating a pricing subproblem based on the linear relaxation solution of the restricted master problem; generating a test number based on the pricing subproblem; if the test number has a negative column, generating parcel path subproblems and vehicle path subproblems based on the simultaneous generation of rows and columns; expanding the path subset of the restricted master problem based on the solution results of the parcel path subproblems and vehicle path subproblems; otherwise, establishing a cutting plane subproblem based on strong inequalities and cut set inequalities; establishing a branch and bound tree; solving the cutting plane subproblem with linear relaxation at the tree nodes of the branch and bound tree; if there are non-integer vehicle path variables, branching the number of vehicles on the vehicle spatiotemporal network edges and the edge variables of the parcel spatiotemporal network; if there are integer vehicle path variables, the courier transportation network design is completed.

[0018] Alternatively, the strong inequality is:

[0019]

[0020] The cut set inequality is

[0021]

[0022] in, This indicates that all originating centers are in And the destination is The total weight of the package.

[0023] Alternatively, the branch constraint can take the form of:

[0024] ∑ p x p ≤z or ∑ p x p ≥z,

[0025] Where z is a natural number.

[0026] A second aspect of this application provides an express delivery network design apparatus, comprising: a determination module for determining decision variables based on express parcel transportation needs and obtaining transshipment restrictions and transshipment costs; an identification module for identifying the number of vehicles on each feasible vehicle path, the number of parcels on each feasible parcel path, and the allocation mechanism of vehicle capacity on each sub-path; a construction module for constructing a vehicle spatiotemporal network based on the number of vehicles on each feasible vehicle path, constructing a parcel spatiotemporal network based on the number of parcels on each feasible parcel path, and establishing a dual spatiotemporal network for express delivery based on the vehicle spatiotemporal network, the parcel spatiotemporal network, the allocation mechanism of vehicle capacity on each sub-path, transshipment restrictions, and transshipment costs; and a generation module for determining a set of drivable paths for vehicles and parcels based on the dual spatiotemporal network, generating an express delivery network design model based on the set of drivable paths, and using the express delivery network design model to meet the express parcel transportation needs.

[0027] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the express delivery network design method as described in the above embodiments.

[0028] Therefore, this application has the following beneficial effects:

[0029] This application's embodiments construct a fine-grained model that closely reflects actual transportation needs by precisely determining decision variables such as vehicle and parcel routes, combined with the allocation mechanism of vehicle and parcel quantities and vehicle capacity. This model not only optimizes vehicle and parcel scheduling, reducing waiting and idle time and improving transportation efficiency, but also reduces transportation costs by considering transshipment restrictions and costs, thereby enhancing enterprise profitability. Simultaneously, the dual-temporal network model demonstrates strong flexibility, capable of responding to various transportation demands and changes, ensuring the smooth completion of transportation tasks. Furthermore, it optimizes resource allocation, improves resource utilization, and directly enhances customer satisfaction by increasing transportation efficiency and reducing delays. Thus, it solves the technical problems of existing technologies, such as the difficulty in balancing transportation costs and efficiency, complex network design, and insufficient data processing capabilities.

[0030] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0031] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0032] Figure 1 This is a flowchart of the express delivery network design method provided according to an embodiment of this application;

[0033] Figure 2 This is a flowchart of the branch pricing segmentation algorithm provided according to an embodiment of this application;

[0034] Figure 3 This is an example diagram of a courier transportation network design device provided according to an embodiment of this application;

[0035] Figure 4 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0036] The embodiments of this application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0037] Transportation network design, as the cornerstone of the express delivery industry, focuses on the efficient planning of vehicle and parcel flow to achieve transportation goals in the shortest possible time and reduce overall costs. Faced with complex transit center networks and diverse parcel demands, express delivery companies need to meticulously calculate vehicle operating costs and parcel transfer fees for each route, striving for optimal resource allocation.

[0038] Given the significant proportion of transportation costs in the operations of express delivery companies, building and optimizing freight consolidation systems has become crucial for cost reduction. These systems achieve a dual improvement in transportation efficiency and frequency by consolidating multiple small parcels for centralized transport. However, the transshipment process—transferring packages from one vehicle to another—while promoting consolidation efficiency, also introduces additional costs, including labor, equipment, and time. Therefore, how to accurately model and manage transshipment in transportation network design has become a hot topic of common interest in both academia and industry.

[0039] During the modeling process, the spatial coupling between packages and vehicles, as well as the temporal synchronization between vehicles, constitute the core challenges, making the problem complex and difficult to solve. To simplify the process, some studies focus on specific scenarios where transshipment costs are negligible, vehicles only perform point-to-point direct transportation, or the model scale is small. While these are convenient for analysis, they are difficult to directly apply to the complex real-world environment of the express delivery industry.

[0040] Given the unique characteristics of express delivery—significant transshipment costs, vehicles flexibly traversing multiple hubs, and a massive number of vehicles and packages—traditional simplified models fall short. To address this, scholars have explored more advanced modeling methods, such as sub-path-based models. While these methods have proven effective in specific sectors like liner shipping, their limited predefined paths and lack of time-based considerations make them unsuitable for the complexity and dynamism of express delivery networks.

[0041] The following description, with reference to the accompanying drawings, outlines an express delivery network design method, apparatus, and equipment according to embodiments of this application. Addressing the problem mentioned in the background art of the inability to calculate transportation networks in practical use, this application provides an express delivery network design method. In this method, by accurately determining decision variables such as vehicle and parcel routes, and combining the number of vehicles and parcels with a vehicle capacity allocation mechanism, a fine-grained model closely reflecting actual transportation needs is constructed. This model not only optimizes vehicle and parcel scheduling, reducing waiting and idle time and improving transportation efficiency, but also reduces transportation costs by considering transshipment restrictions and costs, thereby enhancing the profitability of enterprises. Simultaneously, the dual-temporal-space network model exhibits strong flexibility, capable of responding to various transportation demands and changes, ensuring the smooth completion of transportation tasks. Furthermore, it optimizes resource allocation, improves resource utilization, and directly enhances customer satisfaction by increasing transportation efficiency and reducing delays. Thus, it solves the problems of difficulty in balancing transportation costs and efficiency, complex network design, and insufficient data processing capabilities in existing technologies.

[0042] Specifically, Figure 1 This is a flowchart illustrating a method for designing an express delivery network, as provided in an embodiment of this application.

[0043] like Figure 1 As shown, the express delivery network design method includes the following steps:

[0044] In step S101, decision variables are determined based on the express parcel transportation demand, and transfer restrictions and transfer costs are obtained.

[0045] Among them, express parcel transportation demand can be the parcel transportation tasks that express companies need to complete within a specific time period, including specific information such as the parcel's origin, destination, estimated delivery time, parcel weight or volume, etc. Decision variables can include vehicle routes, parcel allocation plans, transfer point selection, etc. Transfer restrictions can be the conditions or restrictions set on the transfer of parcels between different vehicles or transfer centers during transportation, and transfer costs can be the additional costs incurred by the parcel during transportation due to transfer operations.

[0046] It is understood that the embodiments of this application, by accurately determining decision variables, obtaining transshipment restrictions and transshipment costs, help to select lower-cost and more efficient routes and transshipment methods when planning transportation schemes.

[0047] In step S102, the number of vehicles on each feasible vehicle path, the number of parcels on each feasible parcel path, and the allocation mechanism of vehicle capacity on each sub-path are identified as decision variables.

[0048] It is understood that by identifying the number of vehicles on each feasible vehicle route, the embodiments of this application can ensure that the required number of vehicles is minimized while meeting transportation needs; determining the number of parcels on each feasible parcel route helps to achieve a balanced distribution of parcels and reduce transportation pressure caused by parcel concentration; and reasonably allocating vehicle capacity on each sub-route can ensure that vehicles can load as many parcels as possible without overloading, thereby reducing the number of transportation trips and transfer costs.

[0049] Specifically, the decision variables include the number of vehicles on each feasible vehicle route, the number of parcels on each feasible parcel route, and the allocation scheme of vehicle capacity on each sub-route. Assume there are L transfer centers in the express delivery network, and the decision period is a discrete T unit time period. There are K parcel transportation demands, each demand consisting of a quadruple. <o k d k , σ k w k > indicates that o k d k σ represents the origin and destination of the package, respectively. k This indicates the time when the package can begin shipping, and wk indicates the weight of the package.

[0050] Using directed graphs Represents the vehicle spatiotemporal network, in which A set of spatiotemporal nodes of the transit center. Let the set of edges between the transfer centers be denoted as . This is the set of waiting edges for adjacent time periods within the same transfer center. A feasible vehicle path p is a set of edges in the network that are connected end-to-end, and the decision variable x... p This represents the number of vehicles on path p. Similarly, based on the set of edges formed by the sub-paths of all feasible vehicle paths, combined with the set of spatiotemporal nodes of the transfer center, a spatiotemporal network for packages is constructed. Likewise, the feasible path q for the k-th demand can be defined. k Decision variable y kq ∈[0,1] represents the proportion of parcels transported via this route. The set of sub-paths of vehicle path p is defined as follows: Decision variables Indicates x on the path p The proportion of capacity that a vehicle is allocated to on sub-path s.

[0051] In step S103, a vehicle spatiotemporal network is constructed based on the number of vehicles on each feasible vehicle route, a parcel spatiotemporal network is constructed based on the number of parcels on each feasible parcel route, and a dual spatiotemporal network for express delivery is established based on the vehicle spatiotemporal network, the parcel spatiotemporal network, the allocation mechanism of vehicle capacity on each sub-route, transfer restrictions, and transfer costs.

[0052] Among them, the vehicle spatiotemporal network can display information such as the routes that a vehicle may take and the transfer centers it stops at in different time periods, while the parcel spatiotemporal network can show the entire transportation process of a parcel from its origin to its destination, including its stops and transfers at various transfer centers. The dual spatiotemporal network refers to a comprehensive model that considers both the vehicle spatiotemporal network and the parcel spatiotemporal network.

[0053] It is understood that the embodiments of this application, by constructing a spatiotemporal network of vehicles and parcels, can not only significantly improve transportation efficiency and reduce empty runs and waiting time by optimizing route selection and achieving precise matching of supply and demand, but also effectively reduce transportation costs by reducing unnecessary transfer links and optimizing vehicle loading rates to reduce transfer costs and vehicle empty runs.

[0054] In this embodiment of the application, a vehicle spatiotemporal network is constructed based on the number of vehicles on each feasible vehicle route, including: enumerating all feasible vehicle routes, wherein the number of transfer centers along the route is less than a preset value, and for the same transfer center sequence, different departure times correspond to different vehicle routes; enumerating all sub-paths of each vehicle route to form a sub-path set, and constructing a final sub-path set based on the sub-path sets corresponding to all vehicle routes, wherein a sub-path is an edge connecting two spatiotemporal nodes of transfer centers on the route, and all packages on the edge are loaded from the starting point of the sub-path until they are unloaded at the ending point of the sub-path; and constructing a vehicle spatiotemporal network based on the final sub-path set.

[0055] A transit center can be a transfer station for packages to be transported from one region to another.

[0056] It is understood that the embodiments of this application ensure comprehensive transportation planning and cost control by fully enumerating feasible vehicle routes and considering transshipment center constraints. Fine-grained route division and unified processing provide a solid foundation for route optimization and selection. The construction of the vehicle spatiotemporal network not only significantly improves transportation efficiency by reducing waiting and empty runs and enabling precise parcel delivery, but also enhances the flexibility and decision-making efficiency of transportation planning.

[0057] Specifically, in express delivery, packages can be transported by multiple vehicles in succession. The process of changing vehicles at intermediate transfer centers is called transfer. Directly modeling transfers means describing the correspondence between packages and vehicles on every edge, which introduces great complexity. By introducing sub-paths, packages are transported by the same vehicle on each sub-path, so switching sub-paths indicates a transfer has occurred. Using sub-paths to construct a spatiotemporal network of packages decouples packages and vehicles in the model, reducing the model's complexity.

[0058] Given a vehicle path p, a sub-path is an edge connecting two spatiotemporal nodes of a transfer center on the path. All packages on a sub-path board the vehicle at its starting point and disembark at its ending point. Therefore, a package switching sub-paths indicates a transfer has occurred. For path n... p The vehicle routes for each transfer center can be established in a total of n. p (n p +1) / 2 subpaths. The process of enumerating all subpaths can be divided into the following three steps:

[0059] a. In the vehicle spatiotemporal network, enumerate all feasible vehicle paths, where the number of transfer centers a vehicle passes through does not exceed 4. For the same sequence of transfer centers, different departure times correspond to different vehicle paths.

[0060] b. For each vehicle path p, enumerate all its sub-paths to form a set.

[0061] c. Collect the set of sub-paths corresponding to all vehicle paths to form the final set of sub-paths.

[0062] In this embodiment of the application, a package spatiotemporal network is constructed based on the number of packages on each feasible package path, including: obtaining a set of spatiotemporal nodes of the transfer center; a set of all sub-paths and waiting edges on each feasible package path; and constructing a package spatiotemporal network based on the set of spatiotemporal nodes of the transfer center, the set of all sub-paths and waiting edges.

[0063] Among them, the spatiotemporal node of the transfer center refers to the status of the transfer center at a specific point in time.

[0064] It is understood that by constructing a parcel spatiotemporal network, the embodiments of this application can comprehensively and in real time grasp the location, time and flow status of express parcels in the road network, thereby ensuring the efficiency and accuracy of the transportation process. This not only provides a solid foundation for route optimization and reduces transportation time and costs by using advanced algorithms, but also greatly improves management efficiency.

[0065] Specifically, using directed graphs Represents the spatiotemporal network of packages, in which A set of spatiotemporal nodes of the transit center. The set is a collection containing all subpaths and waiting edges. and Similar to the vehicle spatiotemporal network, the flow of packages and their corresponding paths are represented in the package spatiotemporal network, which is decoupled from the vehicle path representation in terms of network representation.

[0066] In this embodiment, the transshipment restriction is as follows: parcels cannot switch sub-paths within a transshipment center where parcel transshipment operations are not possible; the transshipment cost is calculated based on the number of sub-paths for each feasible parcel route; and the vehicle capacity allocation mechanism on each sub-path is as follows: vehicle route p has x p There are 10 vehicles, and the corresponding path contains n. p An edge, denoted by (i,j), represents a sub-path connecting the i-th and j-th transfer centers, with variables... This represents the proportion of capacity allocated to subpath (i,j), where 1≤m≤n on the m-th edge (m-1,m). p , containing subsets of subpaths Subpath subset The capacity constraint is:

[0067]

[0068] The capacity is obtained by summing the capacities of all vehicle paths assigned to the sub-paths s = (i,j).

[0069]

[0070] If loading packages at the i′-th transfer center on path p is prohibited, then 1≤i′≤n p-1 Then remove the variable. And the i′-th capacity constraint.

[0071] Specifically, because a package is transported by only one vehicle along a sub-path, there is no transshipment during this process. Each switch of sub-path within a package's route means a change to another vehicle, thus constituting one transshipment. In the express delivery network, some transfer centers cannot perform package transshipment operations; packages cannot switch sub-paths within these centers. This type of constraint is called a transshipment restriction. In the package spatiotemporal network, modeling the transshipment restriction is complete simply by setting package routes that do not meet the above constraints as infeasible routes. The number of transshipments during package transportation is equal to the number of sub-paths contained in that package route minus one. By counting the number of sub-paths for each feasible package route, the required transshipment cost can be calculated.

[0072] The dual-spatiotemporal network representation decouples the movement of vehicles and packages. However, since the capacity of a sub-path depends on the vehicle path, a capacity allocation mechanism is needed to determine the proportion of capacity allocated to each vehicle path's sub-paths. Assume there is a vehicle path p with a sub-path x... p There are n vehicles, and the path contains n vehicles. p An edge. Let (i,j) represent the sub-path connecting the i-th and j-th transfer centers. Variables This represents the proportion of capacity allocated to subpath (i,j). On the m-th edge (m-1,m), 1≤m≤n. pIt contains a subset of paths. The total capacity of these subpaths must equal x. p The capacity of the vehicle is such that the following capacity constraints apply:

[0073]

[0074] The capacity is obtained by summing the capacities of all vehicle paths assigned to the sub-paths s = (i,j).

[0075]

[0076] If loading packages at the i′-th transfer center on path p is prohibited, then 1≤i′≤n p-1 Then remove the variable. And the i′-th capacity constraint.

[0077] In step S104, the set of drivable paths for vehicles and parcels is determined based on the dual spatiotemporal network, and a courier transportation network design model is generated based on the set of drivable paths. The courier transportation network design model is then used to meet the transportation needs of courier parcels.

[0078] Among them, the set of drivable paths refers to the set of all possible paths that vehicles can travel to complete the parcel transportation task under the dual spatiotemporal network framework, based on factors such as traffic rules, vehicle performance, and parcel demand.

[0079] It is understood that the embodiments of this application, by constructing a dual spatiotemporal network, achieve a comprehensive consideration of time and space factors in the transportation environment, thereby accurately determining the set of drivable routes for vehicles and packages. This not only improves transportation efficiency and ensures the smooth progress of transportation tasks by reducing detours and waiting times, but also enhances real-time scheduling capabilities, enabling the system to flexibly respond to various emergencies. Simultaneously, the express delivery network design model generated based on the dual spatiotemporal network effectively reduces transportation costs, energy consumption, and emissions by optimizing resource allocation and route planning, thus promoting the development of green logistics.

[0080] Specifically, express delivery companies typically manage a fleet of scheduled vehicles, each with a specific origin and destination, and required to travel along a route from that origin to the destination. This indicates the assembly of the regular fleet of buses. To meet unpredictable package delivery needs, courier companies can also hire extra buses from third-party providers; these extra buses can travel along any route.

[0081] For each regular shift team Use sets Indicates vehicle transportation network All feasible regular bus routes. Each route All originate from the regular bus transfer center. f Departure, and final arrival at destination d f Similarly, using sets This represents all feasible routes for extra buses, which can depart from and arrive at any transfer center. (Using sets...) This represents all feasible vehicle routes.

[0082] For each package shipping requirement Use sets Indicates parcel delivery network All feasible package paths. Each path All originate from the initial transshipment center o of demand k. k Departure, and final arrival at destination d k .

[0083] In this embodiment of the application, the express delivery network design model is as follows:

[0084]

[0085] in, This is the cost of operating the regular bus service. For the cost of operating the overtime bus, The cost of a feasible package route, x p Let p be the number of vehicles on the vehicle path. Indicates vehicle transportation network All feasible regular bus routes, This indicates all feasible routes for extra buses. Indicates parcel delivery network All feasible package paths in the list. Indicates x on the path p The proportion of capacity allocated to a vehicle on subpath s, w k Indicates the weight of the package. y represents the proportion of capacity allocated to subpath (i,j). kq Let N be the weight of the k-th demand along the package path q. f This represents the number of vehicles in the regular convoy f. Let n be the set of subpaths derived from vehicle path p. p Let p be the number of edges contained in the vehicle path.

[0086] Specifically, in the vehicle spatiotemporal network, D a Let τ represent the distance traveled along edge a. a This indicates its travel time. In the package spatiotemporal network, τ s Let represent the travel time of subpath s, which is the sum of the travel times of all edges it contains. Assume the cost per unit distance for regular buses and extra buses is C.R and C E The unit weight parcel shipping cost is C. O The time penalty for shipping a unit weight package within a unit time period is C. T .

[0087] The cost of a feasible vehicle route is the sum of the travel costs of vehicles on each edge. For regular buses, the cost is... For overtime shuttle buses, the cost is... Similarly, the cost of a feasible parcel route is the sum of the parcel flow costs on each sub-route, including parcel transfer fees and time penalties.

[0088] Mathematical expression for a design model of an express delivery network with transshipment restrictions and transshipment costs.

[0089]

[0090] in, This is the cost of operating the regular bus service. For the cost of operating the overtime bus, The cost of a feasible package route, x p Let p be the number of vehicles on the vehicle path. Indicates vehicle transportation network All feasible regular bus routes, This indicates all feasible routes for extra buses. Indicates parcel delivery network All feasible package paths in the list. Indicates x on the path p The proportion of capacity allocated to a vehicle on subpath s, w k Indicates the weight of the package. y represents the proportion of capacity allocated to subpath (i,j). kq Let N be the weight of the k-th demand along the package path q. f This represents the number of vehicles in the regular convoy f. Let n be the set of subpaths derived from vehicle path p. p Let p be the number of edges contained in the vehicle path.

[0091] The optimization objective is to minimize the total cost, including vehicle operating costs, parcel transfer costs, and parcel delivery time penalties. The first constraint ensures that the number of scheduled vehicles used meets fleet requirements; the second constraint ensures that the result meets the transportation needs of all parcels; the third constraint ensures that the total weight of parcels on each sub-path does not exceed the sub-path's capacity; the fourth constraint is the capacity allocation constraint from vehicle path to sub-path; other constraints specify the variable types.

[0092] In this embodiment, the express delivery network design model is used to meet the needs of express parcel transportation. This includes: selecting an initial subset of vehicle and parcel paths based on the express delivery network design model; generating a restricted master problem based on the initial subset of vehicle and parcel paths; generating a pricing subproblem based on the linear relaxation solution of the restricted master problem; generating a test number based on the pricing subproblem; if the test number has a negative column, generating parcel path subproblems and vehicle path subproblems based on the simultaneous generation of rows and columns; expanding the path subset of the restricted master problem based on the solution results of the parcel path subproblems and vehicle path subproblems; otherwise, establishing a cutting plane subproblem based on strong inequalities and cut set inequalities; establishing a branch and bound tree; performing linear relaxation solution on the cutting plane subproblem at the tree nodes of the branch and bound tree; if there are non-integer vehicle path variables, branching the number of vehicles on the vehicle spatiotemporal network edges and the edge variables of the parcel spatiotemporal network; if there are integer vehicle path variables, the express delivery network design is completed.

[0093] It is understood that the embodiments of this application, by constructing a complex optimization framework and combining linear relaxation, pricing subproblems, cutting plane techniques, and branch and bound algorithms, achieve precise optimization of express delivery network design. This framework not only ensures that vehicles and parcels can be transported efficiently along optimal or suboptimal routes, effectively reducing transportation costs, but also significantly improves service quality and customer satisfaction.

[0094] In this embodiment of the application, the strong inequality is:

[0095]

[0096] The cut set inequality is

[0097]

[0098] in, This indicates that all originating centers are in And the destination is The total weight of the package.

[0099] In this embodiment of the application, the branch constraint takes the form of:

[0100] ∑ p x p ≤z or Σ p x p ≥z,

[0101] Where z is a natural number.

[0102] Specifically, a branch-and-price slicing algorithm is designed to address the express delivery network design problem with transit constraints and costs. First, an initial subset of vehicle and parcel routes is selected, forming a constrained master problem. Its linear relaxation is solved, and its dual value is used to construct a pricing subproblem. Then, the pricing subproblems for vehicle and parcel routes are solved separately to generate new routes with negative test numbers, expanding the route subset. For the parcel pricing subproblem, classic column generation is used; for the vehicle pricing subproblem, a method for simultaneous row and column generation is developed. Furthermore, a dynamic cutting plane generation module is designed to improve the lower bound of linear relaxation, a strong branching module is developed to select branch variables to accelerate the search for feasible solutions, and a heuristic module is developed to obtain the upper bound of the problem to aid in pruning, such as... Figure 2 As shown, it includes the following:

[0103] S1, Pricing Subproblem: Generate vehicle and package routes with negative test numbers; using γ f ,π k ,α s ≤0 and β pm Let represent the dual variables of the first four constraints in the express delivery network design model. In the linear relaxation problem, each column represents a vehicle route or a package route. Therefore, the pricing subproblem involves finding feasible columns with negative test numbers and can be decomposed into two subproblems: the package route subproblem and the vehicle route subproblem.

[0104] S2, Package Route Subproblem: Column Generation; For transportation demand k and its corresponding package route q, variable y kq The test number is:

[0105]

[0106] in, It is a positive value, while π k The sign is arbitrary. Therefore, only if the dual variable π k A path with a negative test number only exists when the test number is strictly positive. A feasible package path starts from the initial spatiotemporal node [o(k), σ(k)] of demand k and terminates at some spatiotemporal node of the destination transfer center d(k). The path with the minimum test number can be obtained directly using Dijkstra's shortest path algorithm. At any point during the execution of Dijkstra's algorithm, if the current cost exceeds π... k The algorithm can terminate early if the condition is met. The branch-and-price splitting algorithm solves the corresponding package path subproblem for each demand and adds the package path with the smallest negative test number to the path subset.

[0107] S3, Vehicle Path Subproblem: Row and Column Generation; For vehicle path... There are two variables: x p and (or The test numbers for these two variables are as follows:

[0108]

[0109] 0≤i<j≤n p

[0110] Where p is the regular bus route, When p is the route for the overtime bus The test number for vehicle path p can be defined from these two sets of test numbers:

[0111]

[0112] Subtracting the m-th and m+1-th constraints, we have:

[0113]

[0114] This means that nodes m = 1, 2, ..., n p If the inflow of -1 equals the outflow, then the minimization problem can be transformed into minimizing the flow from node 0 to node n. p The shortest path problem, with weight α ij ≤0.

[0115] For any subpath Use sets This represents all paths with the same start and end points as s, meaning that subpaths s can be derived from s. Derived from any path in the given path. The distance of the subpath s is defined as:

[0116]

[0117] in, To represent the optimal path, use Representing a path The number of edges contained.

[0118] The shortest path problem corresponding to the vehicle path subproblem is defined in the auxiliary network. The network includes a parcel network. The set of nodes and the set of edges. Each sub-path s in the array contains a dual value α. s Distance l s Optimal path and the number of transshipment centers n s For each regular convoy The cost of subpath s is equal to C. O ·l s +U·α s For overtime buses, the cost is C. R ·ls +U·α s This shortest path problem contains a resource constraint regarding the number of transit centers, i.e., the number of sub-paths n. s The sum of the values ​​cannot exceed 4.

[0119] Since edge weights in the shortest path problem can be negative, a labeling algorithm is designed to determine the shortest path for the overtime bus sub-problem. For the regular bus route sub-problem, an enumeration method is designed to calculate and compare the test number for each regular bus route. Assume the final path consists of m sub-paths {s1, s2, ..., sm}. m If the vehicle path is composed of}, then the vehicle path can be represented as The branch-and-price splitting algorithm solves the corresponding vehicle routing subproblem for each regular bus fleet and adds the path with the smallest negative test number to the path subset; for extra buses, it adds the vehicle route with the smallest negative test number to the path subset.

[0120] S4. Cutting plane: Improves the optimal value of the linear relaxation problem. The branch pricing cutting algorithm implements two types of cutting planes: strong inequality and cut set inequality.

[0121] Among them, the strong inequality states that for any demand k, the strong inequality holds true on the edge. A lower bound is applied to the number of vehicles:

[0122]

[0123] When the weight required for parcel delivery is much less than the capacity of the vehicles, the strong inequality ensures a greater number of vehicles, thus reducing the fractional solution.

[0124] Strong inequalities introduce additional dual variables that affect the calculation of the test statistic. Using δ... ks Let represent the dual variables associated with strong inequalities, and let the set [SI] = {(k,s)} represent the indices of existing strong inequalities. For each element (k,s) ∈ [SI], in the k-th wrapping path subproblem, -δ ks Adding this to the cost of edge s preserves the structure of the wrapper path subproblem without requiring modifications to the algorithm. For the vehicle path subproblem, for each edge... ∑ k:(k,s)∈[SI] w k ·δ ks By adding it to its cost, the structure of the vehicle routing subproblem can be preserved without modifying the algorithm.

[0125] Cut-set inequality: This applies to the set of transfer centers. Divide into non-empty subsets and its supplement Construct cut set In the vehicle spatiotemporal network, all originating transfer centers are... The final transfer center is located at The edges belong to the cut set use This indicates that all originating centers are in And the destination is The total weight of the package. The cut-set inequality ensures that... All sides have sufficient vehicle capacity to meet the needs from Flow direction Total required weight of the package:

[0126]

[0127] For parcel originating transfer centers Enumerate all its subsets to form a set. For each set of transit centers will be cut set The relevant cut-set inequalities are added to the RMP linear relaxation. To preserve the structure of the vehicle path subproblem after adding the cut-set inequalities, the optimal path is directly modified. The cost of subpath s. Using a set. Represents the index of all current cut-set inequalities, using Let represent the dual variable related to the cut-set inequality. The optimal path and sub-path costs are calculated by modifying the distance of edge 'a' in the vehicle spatiotemporal network.

[0128]

[0129] Wherein, the constant C for the regular bus problem is C0. R The answer to the overtime vehicle issue is C. E No further modifications to the algorithm are required.

[0130] Cutting plane generation process: Strong inequality generation is embedded in the column generation module. During the generation process, the branch pricing cutting algorithm checks the strong inequality corresponding to each pair (a,k), where... and All violated strong inequalities are added to the relaxation. If a violated strong inequality exists, the column generation module is re-executed. The pricing subproblem module and the cutting plane generation module are only stopped if there are no negative test path and no violated strong inequalities.

[0131] S5, Strong Branching Module: Accelerates the search for integer solutions; to obtain feasible solutions, a branch-and-bound tree is designed, and the pricing and cutting modules are embedded into the branch-and-bound framework. After finding the optimal solution with linear relaxation at the tree node, if there are non-integer vehicle path variables, branching is performed and two child nodes are created.

[0132] First, branch based on the number of vehicles on the spatial edges, where a spatial edge refers to all edges in the vehicle spatiotemporal network that share the same starting and ending transfer centers. Once all spatial edge variables are integers, branch based on the spatiotemporal edge variables. The branching constraint takes the form of:

[0133] ∑ p x p ≤z or ∑ p x p ≥z, This constraint is similar to the cut-set inequality, and therefore the structure of the pricing subproblem is maintained in the same way.

[0134] To further improve the lower bound, a strong branching module was designed. This module tests a set of non-integer variables and selects the variable with the best branching result for branching. In this invention, the 10 variables furthest from integers are selected as candidate variables. For each candidate variable, only the RMP relaxation containing the generated columns is solved to evaluate the lower bound of each child node. Using the weakest lower bound of the child node as a score, the strong branching module selects the variable with the highest score for branching. For tree search, this invention uses the best lower bound priority rule to select nodes in the first 20 nodes of the branch and bound tree, and uses the depth-first rule for the remaining cases.

[0135] S6. Heuristic Solution Finding: A heuristic module is designed to obtain an ideal integer solution as early as possible during the traversal. Based on this integer solution, the size of the branch-and-bound tree can be reduced and the search accelerated. The heuristic module temporarily removes all branch constraints and directly solves the RMP containing the currently generated paths. For express delivery network design problems of different sizes, different time limits can be set for the heuristic module to achieve optimal performance. The heuristic module is executed once after searching every 5 branch-and-bound tree nodes.

[0136] The express delivery network design method proposed in this application constructs a fine-grained model that closely reflects actual transportation needs by accurately determining decision variables such as vehicle and parcel routes and combining the number of vehicles and parcels with the vehicle capacity allocation mechanism. This model not only optimizes vehicle and parcel scheduling, reduces waiting and idle time, and improves transportation efficiency, but also reduces transportation costs and enhances enterprise profitability by considering transshipment restrictions and costs. Simultaneously, the dual-temporal network model demonstrates strong flexibility, capable of responding to various transportation demands and changes, ensuring the smooth completion of transportation tasks. Furthermore, it optimizes resource allocation, improves resource utilization, and directly enhances customer satisfaction by increasing transportation efficiency and reducing delays. Therefore, it solves the problems of balancing transportation costs and efficiency, complex network design, and insufficient data processing capabilities in existing technologies.

[0137] The following is a detailed explanation of the express delivery network design method through a specific embodiment:

[0138] 1) The problem is modeled using sub-paths and dual spatiotemporal networks, constructing a path-based mathematical model; a design model for an express delivery network with transshipment constraints and costs is established for this problem. The goal of the transportation network design problem is to determine the routes of vehicles and packages, given the demand for express parcel transportation, to minimize the total transportation cost, including vehicle travel costs, parcel transshipment costs, and parcel delivery time penalties.

[0139] 2) Design of a branch-and-price slicing algorithm: This paper designs a branch-and-price slicing algorithm to solve the express delivery network design problem with transit constraints and transit costs. The algorithm first selects an initial subset of vehicle and parcel routes to form a restricted master problem, solves for linear relaxation, and uses its dual value to construct a pricing subproblem. Then, it solves the pricing subproblems for vehicle and parcel routes separately to generate new routes with negative test numbers, expanding the route subset. For the parcel pricing subproblem, this invention uses classical column generation; for the vehicle pricing subproblem, a method for simultaneous row and column generation is developed. Furthermore, a dynamic cutting plane generation module is designed to improve the lower bound of linear relaxation, a strong branching module is developed to select branching variables to accelerate the search for feasible solutions, and a heuristic module is developed to obtain the upper bound of the problem to aid in pruning.

[0140] 3) Implementation of the branch-and-price splitting algorithm; written in JAVA programming language. During algorithm execution, the RMP linear relaxation problem belongs to linear programming, while the RMP problem solved in the heuristic module belongs to mixed-integer linear programming.

[0141] Specifically, the commercial solver Gurobi was used to solve the linear programming and integer programming models in the branch pricing algorithm. The proposed express delivery network design model with transfer restrictions and transfer costs was tested on real data from a certain express delivery company. The total transportation costs of vehicle and parcel scheduling decisions obtained by the new model and the traditional service network design model were compared. The transfer costs of the traditional service network design model were calculated based on the obtained decisions, as shown in Table 1 below.

[0142] Table 1 Sample Size of Companies in Multiple Regions

[0143]

[0144] Table 1 shows the sample size used in the experiment, with each sample corresponding to a regional company. Each regional company is responsible for sending parcels originating from a transit center within its region to transit centers nationwide. For each regional company, Table 1 lists its region, the number of transit centers in that region, the number of transit centers in that region without transit restrictions, the number of transit centers at the parcel's destination, the number of parcels requested, and the number of regular trucks in its fleet.

[0145] The loading capacity of both regular and extra buses is 11 tons; the cost per kilometer for regular and extra buses is 6 yuan and 7 yuan respectively; the parcel transfer fee is 0.2 yuan per kilogram; the parcel transportation time penalty is 0.2 yuan per kilogram per day, as shown in Table 2 below.

[0146] Table 2 Model Parameters

[0147]

[0148] Table 3 shows the simulation results for a 7-day sample in Jilin Province.

[0149] Table 3 Simulation results for a 7-day sample in Jilin region

[0150]

[0151] Table 3 lists the total transportation cost, parcel transfer cost, final error bound of the branch pricing algorithm for both the traditional model and this model, as well as the cost reduction ratio of the total transportation cost of this model compared to the traditional model. On average, this model achieves a cost reduction performance of 1.56% compared to the traditional model.

[0152] The simulation results for regional samples of different sizes are shown in Table 4 below.

[0153] Table 4 Cost Reduction Ratio in Different Regions

[0154]

[0155] The results are the average of the sample results over 7 days. Table 4 lists the total transportation cost, the final error bound of the branch pricing algorithm, and the cost reduction ratio of the total transportation cost of the traditional model compared to the traditional model for both the traditional model and the present model. On average, the present model achieves a cost reduction performance of more than 1% compared to the traditional model.

[0156] In summary, a deep modeling approach was developed for the express delivery network design problem using the sub-path concept and a dual spatiotemporal network framework. A significant feature of this model is its direct and explicit simulation of the transshipment stages of packages during transportation, thus achieving a more refined depiction of the transportation process. To efficiently solve this complex model, an advanced algorithm based on branch pricing was developed, demonstrating superior performance in both optimization efficiency and solution accuracy.

[0157] Compared to traditional service network design models, this model is not only more rigorous in theory but also demonstrates significant advantages in practice. It can more accurately estimate total transportation costs, closely reflecting the complex and ever-changing situations in actual operations and effectively avoiding cost estimation biases. More importantly, by optimizing transshipment strategies and route planning, it can significantly reduce overall transportation costs, bringing economic benefits to express delivery companies.

[0158] In specific applications of express delivery transportation in various regions, the constructed express delivery network design model, which incorporates transshipment constraints and costs, successfully minimized the total transportation cost. Furthermore, the vehicle routing solutions output by the model highly align with real-world operational needs, validating the model's effectiveness and practicality in guiding actual transportation decisions. These achievements not only highlight the advanced nature of the express delivery network design model that considers transshipment constraints and costs but also fully demonstrate the powerful capability of the branch-and-price splitting algorithm in solving such complex optimization problems.

[0159] Next, the express delivery network design device according to the embodiments of this application is described with reference to the accompanying drawings.

[0160] Figure 3 This is a block diagram of a courier transportation network design device according to an embodiment of this application.

[0161] like Figure 3 As shown, the express delivery network design device 10 includes: a determination module, an identification module 200, a construction module 300, and a generation module 400.

[0162] The system comprises the following modules: 100: Determines decision variables based on parcel transport demand and obtains transfer restrictions and costs; 200: Identifies the number of vehicles on each feasible vehicle route, the number of parcels on each feasible parcel route, and the allocation mechanism of vehicle capacity on each sub-route; 300: Constructs a vehicle spatiotemporal network based on the number of vehicles on each feasible vehicle route, constructs a parcel spatiotemporal network based on the number of parcels on each feasible parcel route, and establishes a dual spatiotemporal network for express transport based on the vehicle spatiotemporal network, parcel spatiotemporal network, vehicle capacity allocation mechanism on each sub-route, transfer restrictions, and transfer costs; 400: Determines the set of drivable routes for vehicles and parcels based on the dual spatiotemporal network, generates an express transport network design model based on the set of drivable routes, and uses the express transport network design model to meet parcel transport demands.

[0163] It should be noted that the foregoing explanation of the express delivery network design method embodiment also applies to the express delivery network design device of this embodiment, and will not be repeated here.

[0164] The express delivery network design device proposed in this application constructs a fine-grained model that closely reflects actual transportation needs by accurately determining decision variables such as vehicle and parcel routes and combining the number of vehicles and parcels with the vehicle capacity allocation mechanism. This model not only optimizes vehicle and parcel scheduling, reduces waiting and idle time, and improves transportation efficiency, but also reduces transportation costs and enhances enterprise profitability by considering transshipment restrictions and costs. Simultaneously, the dual-temporal network model demonstrates strong flexibility, capable of responding to various transportation demands and changes, ensuring the smooth completion of transportation tasks. Furthermore, it optimizes resource allocation, improves resource utilization, and directly enhances customer satisfaction by increasing transportation efficiency and reducing delays. Therefore, it solves the problems of difficulty in balancing transportation costs and efficiency, complex network design, and insufficient data processing capabilities in existing technologies.

[0165] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include:

[0166] The memory 401, the processor 402, and the computer program stored on the memory 401 and capable of running on the processor 402.

[0167] When processor 402 executes the program, it implements the express delivery network design method provided in the above embodiments.

[0168] Furthermore, electronic devices also include:

[0169] Communication interface 403 is used for communication between memory 401 and processor 402.

[0170] The memory 401 is used to store computer programs that can run on the processor 402.

[0171] The memory 401 may include high-speed RAM (Random Access Memory) memory, and may also include non-volatile memory, such as at least one disk storage.

[0172] If the memory 401, processor 402, and communication interface 403 are implemented independently, then the communication interface 403, memory 401, and processor 402 can be interconnected via a bus to complete communication between them. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0173] Optionally, in a specific implementation, if the memory 401, processor 402, and communication interface 403 are integrated on a single chip, then the memory 401, processor 402, and communication interface 403 can communicate with each other through an internal interface.

[0174] Processor 402 may be a CPU (Central Processing Unit), an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement embodiments of this application.

[0175] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0176] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0177] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0178] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any of the following techniques known in the art, or a combination thereof: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (FPGAs), field-programmable gate arrays (FPGAs), etc.

[0179] Those skilled in the art will understand that all or part of the steps of the methods implementing the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0180] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for designing an express delivery network, characterized in that, Includes the following steps: Decision variables are determined based on the demand for express parcel transportation, and transit restrictions and transit costs are obtained; The decision variables are identified as follows: the number of vehicles on each feasible vehicle path, the number of parcels on each feasible parcel path, and the allocation mechanism of vehicle capacity on each sub-path. The sub-path is an edge connecting two spatiotemporal nodes of transfer centers on the path. All parcels on the edge are loaded from the starting point of the sub-path until they are unloaded at the end point of the sub-path. A vehicle spatiotemporal network is constructed based on the number of vehicles on each feasible vehicle route, and a parcel spatiotemporal network is constructed based on the number of parcels on each feasible parcel route. A dual spatiotemporal network for express delivery is established based on the vehicle spatiotemporal network, the parcel spatiotemporal network, the vehicle capacity allocation mechanism on each sub-route, the transfer restrictions, and the transfer costs. The transfer restrictions are: parcels cannot switch sub-routes in transfer centers where parcel transfer operations are not possible; the transfer costs are calculated based on the number of sub-routes for each feasible parcel route; and the vehicle capacity allocation mechanism on each sub-route is as follows: vehicle route... Above A vehicle, and the corresponding path contains Edge, using Indicates the connection of the first and the Sub-paths of each transit center, variables Indicates assignment to a subpath The capacity ratio, in the first Strip edge superior, , containing subsets of subpaths Subpath subset The capacity constraint is: ； The capacity is obtained by summing the capacities of all vehicle paths assigned to the sub-path s=(i,j): ； in, For the path The capacity ratio of vehicles allocated to subpath s, where U is the vehicle's loading capacity; if driving on the path is prohibited... The Then remove the variable. and the One capacity limitation constraint; Based on the dual spatiotemporal network, a set of drivable paths for vehicles and parcels is determined. A courier transportation network design model is generated based on this set of drivable paths. This courier transportation network design model is then used to meet the transportation needs of the courier parcels. The courier transportation network design model is as follows: in, This is the cost of operating the regular bus service. For the cost of operating the overtime bus, Cost of a feasible package route. For vehicle routes The number of vehicles on board Indicates vehicle transportation network All feasible regular bus routes, This indicates all feasible routes for extra buses. Indicates parcel delivery network All feasible package paths in the list. Indicates the path Vehicles assigned to sub-paths The above is the capacity ratio, where U is the vehicle's loading capacity. Indicates the weight of the package. Indicates assignment to a subpath The capacity ratio Let be the weight percentage of the k-th requested package on package path q. For regular shift motorcade The number of vehicles in the middle, For vehicle routes The set of derived subpaths For vehicle routes The number of edges contained; The step of using the express delivery network design model to meet the express parcel transportation needs includes: selecting an initial subset of vehicle and parcel paths based on the express delivery network design model; generating a restricted master problem based on the initial subset of vehicle and parcel paths; generating a pricing subproblem based on the linear relaxation solution of the restricted master problem; generating a test number based on the pricing subproblem; if the test number has a negative column, generating a parcel path subproblem and a vehicle path subproblem based on the simultaneous generation of rows and columns; expanding the path subset of the restricted master problem based on the solution results of the parcel path subproblem and the vehicle path subproblem; otherwise, establishing a cutting plane subproblem based on strong inequalities and cut set inequalities; establishing a branch and bound tree; performing linear relaxation solution on the cutting plane subproblem at the tree nodes of the branch and bound tree; if there are non-integer vehicle path variables, branching the number of vehicles on the vehicle spatiotemporal network edge and the edge variables of the parcel spatiotemporal network; if there are integer vehicle path variables, the express delivery network design is completed.

2. The express delivery network design method according to claim 1, characterized in that, The construction of the vehicle spatiotemporal network based on the number of vehicles on each feasible vehicle path includes: Enumerate all feasible vehicle routes, where the number of transfer centers the vehicle passes through is less than a preset value. For the same sequence of transfer centers, different departure times correspond to different vehicle routes. Enumerate all sub-paths for each vehicle route to form a sub-path set, and then construct the final sub-path set based on the sub-path sets corresponding to all vehicle routes. The vehicle spatiotemporal network is constructed based on the final set of sub-paths.

3. The express delivery network design method according to claim 1, characterized in that, The step of constructing a package spatiotemporal network based on the number of packages on each feasible package path includes: Obtain the set of spatiotemporal nodes of the transit center; Based on the set of all sub-paths and waiting edges on each feasible package path; The package spatiotemporal network is constructed based on the set of spatiotemporal nodes of the transfer center, the set of all sub-paths, and the set of waiting edges.

4. The express delivery network design method according to claim 1, characterized in that, The strong inequality is: The cut set inequality is: in, This indicates that all originating centers are in And the destination is The total weight of the package.

5. The express delivery network design method according to claim 1, characterized in that, The form of branch constraint is: or , in, Let be a natural number.

6. A device for designing an express delivery network, characterized in that, The device is equipped with the express delivery network design method as described in any one of claims 1-5, including: The determination module is used to determine decision variables based on the transportation needs of express parcels, and to obtain transshipment restrictions and transshipment costs; The identification module is used to identify the number of vehicles on each feasible vehicle path, the number of parcels on each feasible parcel path, and the allocation mechanism of vehicle capacity on each sub-path for the decision variables. The construction module is used to construct a vehicle spatiotemporal network based on the number of vehicles on each feasible vehicle route, construct a parcel spatiotemporal network based on the number of parcels on each feasible parcel route, and establish a dual spatiotemporal network for express delivery based on the vehicle spatiotemporal network, the parcel spatiotemporal network, the vehicle capacity allocation mechanism on each sub-route, the transfer restrictions, and the transfer costs. The generation module is used to determine the set of drivable paths for vehicles and packages based on the dual spatiotemporal network, generate a courier transportation network design model based on the set of drivable paths, and use the courier transportation network design model to meet the courier package transportation requirements.

7. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement the express delivery network design method according to any one of claims 1-5.

Images