A decision-making method for collaborative delivery between vehicles and drones based on multi-dimensional cost fusion

Abstract

The present invention proposes a vehicle and drone collaborative delivery decision method based on multidimensional cost fusion, the method comprising: obtaining customer point information, road information, vehicle information and drone information; establishing a multidimensional cost objective function, and determining the weights of different dimensional costs; constructing three delivery mode models based on multidimensional cost fusion; using a branch and bound algorithm to solve the above three delivery mode models based on multidimensional cost fusion; according to the solution results, the optimal path planning scheme, that is, the optimal delivery mode, is determined, and the vehicle and drone collaborative delivery decision based on multidimensional cost fusion is completed. The present invention solves the problem of difficult delivery mode decision-making in the actual delivery process, and facilitates users to understand the potential costs and benefits of different delivery modes more comprehensively and accurately, so as to make more scientific decisions in actual delivery.

Description

Technical Field

[0001] The present invention relates to a collaborative distribution decision method, in particular to a vehicle and unmanned aerial vehicle collaborative distribution decision method based on multi-dimensional cost fusion. Background Art

[0002] Due to the rapid development of e-commerce logistics and the continuous growth of distribution demand, terminal distribution efficiency and flexibility have become key points. The traditional distribution model faces a series of challenges such as high cost, low efficiency and limited timeliness.

[0003] Terminal delivery in the field of logistics and transportation is usually completed by vehicles and delivery personnel, but the logistics and distribution industry is facing huge challenges: on the one hand, high labor costs have caused a sharp increase in terminal delivery costs; on the other hand, the pressure brought by high population density on road traffic in residential areas has also led to longer delivery time, resulting in a decrease in delivery efficiency.

[0004] Therefore, improving the efficiency of terminal delivery and reducing costs have become urgent issues to be solved in the industry. In order to improve the efficiency of terminal logistics delivery, the vehicle-UAV collaborative delivery system has gradually become a research hotspot in recent years: the vehicle-UAV collaborative delivery system realizes collaborative work in the delivery process by organically combining vehicles and drones; vehicles are responsible for long-distance transportation and carrying drones, while drones undertake short-distance, small-batch terminal delivery tasks. This system can not only overcome the high labor costs in traditional delivery, but also flexibly respond to problems such as traffic congestion, thereby improving the timeliness and efficiency of delivery.

[0005] Although there are currently a variety of delivery modes, including full vehicle, vehicle + single drone, and vehicle + multiple drones; in actual applications, making decisions on different delivery modes is a complex and difficult task. There are mutual constraints and differences in advantages and disadvantages among various modes. In actual operations, choosing a suitable delivery mode requires comprehensive consideration of multiple factors such as cost, timeliness, and feasibility. Summary of the invention

[0006] Purpose of the invention: The technical problem to be solved by the present invention is to provide a vehicle and drone collaborative delivery decision-making method based on multi-dimensional cost fusion in response to the shortcomings of the existing technology.

[0007] In order to solve the above technical problems, the present invention discloses a vehicle and drone collaborative delivery decision method based on multi-dimensional cost fusion, the method comprising:

[0008] Step S1, obtaining customer point information, road information, vehicle information and drone information;

[0009] Step S2: Establish a multi-dimensional cost objective function and determine the weights of costs in different dimensions;

[0010] Step S3: Based on the customer point information, road information, vehicle information and drone information, combined with the delivery point location, vehicle speed and drone range, as well as the weights of different dimensions determined in step S2, three delivery mode models based on multi-dimensional cost fusion are constructed;

[0011] Step S4, using the branch and bound algorithm to solve the above three distribution mode models based on multi-dimensional cost fusion;

[0012] Step S5: Based on the solution of step S4, the optimal path planning solution is determined, that is, the optimal delivery mode, and the vehicle and drone collaborative delivery decision based on multi-dimensional cost fusion is completed.

[0013] Furthermore, the three distribution mode models based on multi-dimensional cost fusion described in step S3 specifically include:

[0014] Full vehicle distribution model based on multi-dimensional cost fusion, vehicle and single drone collaborative distribution model based on multi-dimensional cost fusion, and vehicle and multiple drone collaborative distribution model based on multi-dimensional cost fusion.

[0015] Furthermore, the establishment of the multi-dimensional cost objective function described in step S2 specifically includes:

[0016] The multidimensional cost objective function at least includes: transportation cost and time cost, and the time cost and transportation cost are respectively denoted as f1(x) and f2(x); wherein the transportation cost includes: vehicle transportation cost and drone transportation cost.

[0017] Furthermore, the step S2 of determining the weights of the costs of different dimensions specifically includes:

[0018] Using weights α and β, a comprehensive objective function Z is constructed, which is the weighted sum of all individual objective functions, as follows:

[0019] Z=α·f1(x)+β·f2(x)

[0020] Unify the measurement units of time cost and transportation cost and standardize them as follows:

[0021] Assuming the time value per hour is b3, the time cost is expressed as:

[0022] Time cost = b3·T c+1

[0023] This gives us the final comprehensive cost function:

[0024] Z=α·f1(x)+β·b3·T c+1

[0025] Among them, T c+1 The time it takes for a transport vehicle to return to the distribution center after completing all delivery tasks.

[0026] Furthermore, the full vehicle distribution model based on multi-dimensional cost fusion described in step S3 specifically includes:

[0027] Objective function and constraints of the full vehicle delivery model based on multi-dimensional cost fusion, where:

[0028] The objective function of the full vehicle distribution model based on multi-dimensional cost fusion is expressed as follows:

[0029]

[0030] Among them, minZ means the minimum comprehensive cost, i and j are the transportation node numbers; α is the importance weight of the logistics enterprise to the transportation cost, β is the importance weight of the logistics enterprise to the timeliness; b1 is the unit transportation cost of the vehicle, b3 is the time value cost; N + is the set of transportation nodes that the vehicle may visit; x ij is the decision variable, i.e., the vehicle's driving path from transport node i to transport node j; d ij is the driving distance of the vehicle from transport node i to transport node j; T c+1 It is the time it takes for a transport vehicle to return to the distribution center after completing all delivery tasks;

[0031] The sum of the importance weights of transportation cost and transportation time for logistics companies is 1, that is:

[0032] α+β=1

[0033] The constraints include:

[0034] a1, a vehicle enters and exits any transport node at most once, expressed as follows:

[0035]

[0036]

[0037] Among them, C is the set of all transportation nodes;

[0038] a2, vehicle elimination sub-loop constraint, is expressed as follows:

[0039]

[0040] Among them, μ i is the visit order of transport node i, n is the total number of transport nodes, and N0 is the set of nodes that the vehicle may leave;

[0041] a3, the time constraint of the vehicle is expressed as follows:

[0042]

[0043] l i ≥t i +s i ·m i -M·(1-x ip )

[0044] Among them, t i is the time when the vehicle arrives at transport node i, l i is the time when the vehicle leaves the transport node i, v t is the average speed of the vehicle, s i is the service time of trucks and drones at each transportation node, m i The transport node i is served by the drone, and M is an infinite positive integer.

[0045] Furthermore, the vehicle and single drone collaborative delivery model based on multi-dimensional cost fusion described in step S3 specifically includes:

[0046] Objective function and constraints of the vehicle-single-UAV collaborative delivery model based on multi-dimensional cost fusion, where:

[0047] The objective function of the vehicle and single UAV collaborative delivery model based on multi-dimensional cost fusion is expressed as follows:

[0048]

[0049] Among them, b2 is the unit transportation cost of the drone, y ij is the decision variable, i.e., the flight path of the UAV from transport node i to transport node j; d′ ij is the flight distance of the drone from transport node i to transport node j; Tc+1 is the time it takes for the transport tools, i.e., vehicles and drones, to return to the distribution center after completing all delivery tasks;

[0050] The sum of the importance weights of transportation cost and transportation time for logistics companies is 1, that is:

[0051] α+β=1

[0052] The constraints include:

[0053] b1, there is only one drone, and it enters and exits any transport node at most once, which is represented as follows:

[0054]

[0055]

[0056] b2, vehicle elimination sub-loop constraint, is expressed as follows:

[0057]

[0058] b3, the traffic flow of vehicles and drones in and out of the transportation node is conserved, that is, each transportation node can only be served by a vehicle or drone once, which is expressed as follows:

[0059]

[0060] Among them, n i The transport node i is served by the drone;

[0061] b4, the vehicle must depart from and return to the same distribution center once, expressed as follows:

[0062]

[0063]

[0064] b5, the UAV’s cruising range constraint, that is, the UAV must complete the delivery task within its predetermined cruising range, expressed as follows:

[0065]

[0066] Among them, ε is the range of the UAV;

[0067] b6, the load limit of the drone, is used to indicate the maximum load capacity of the drone when performing a delivery mission, as follows:

[0068]

[0069] Among them, w i is the cargo weight at transport node i, and W is the load limit of the UAV;

[0070] b7, the vehicle arrival time constraint, is used to ensure that after the vehicle departs from any transport node, the time it arrives at the subsequent transport node is no earlier than the time it leaves the previous transport node plus the travel time between the two transport nodes, expressed as follows:

[0071]

[0072] b8, the departure time constraint of the vehicle, is used to ensure that the departure time of the vehicle from the transport node i is no earlier than the time it arrives at the node plus the total time required by the vehicle at the node, such as the service time and the additional time required by the driver to launch or recover the drone at the point, expressed as follows:

[0073]

[0074] Among them, t l The time required for the vehicle driver to release the drone, t r the time required to recover the drone for the vehicle driver;

[0075] b9, the vehicle's departure time constraint, is used to ensure that if transport node i is a drone recovery point, the vehicle's departure time from transport node i shall not be earlier than the time when the drone leaves the transport node k after completing the delivery plus the time required for the driver to release or recover the drone. This ensures that the truck will depart after completing the drone recovery operation, as shown below:

[0076] l i ≥l k ′+ / v d ·n k ·m i +t r ·n k ·m i -M·(1-y ki )

[0077] Among them, l k ′ is the moment when the UAV leaves the transport node i, v d is the average flight speed of the UAV;

[0078] b10, the arrival time constraint of the drone, is used to ensure that the drone arrives at any subsequent transport node no earlier than the time it leaves the previous node plus the flight time between the two nodes, expressed as follows:

[0079] t i ′≥l′ k +d′ ki / v d -M·(1-y ki )

[0080] Among them, t i ′ is the time when the UAV arrives at the transport node i;

[0081] b11, the departure time constraint of the drone, is used to ensure that the departure time of the drone from any transportation node is no earlier than the sum of the arrival time at the node and the service time required by the drone at the node;

[0082] l i ′≥(t i ′+s i )·n i +t i ·m i +t r ·y ki ·m i +t l·y ij ·m i -M·(1-y ij )

[0083] b12, the visit order constraint, is used to ensure that the vehicle does not repeatedly pass through the visited transport nodes, which is expressed as follows:

[0084]

[0085]

[0086] Among them, P ij is the access order between transport node i and transport node j;

[0087] b13, if the drone is launched from transport node i and recovered at transport node k, and the drone is launched from transport node l in the next step; this constraint will ensure that the time when the vehicle launches the drone at transport node i is earlier than the time when the drone returns to transport node k, that is, when a single drone departs from transport node i to serve customer point j, the vehicle will not launch the drone at the next visited transport node l before the drone returns to node k; if the time when the vehicle visits transport node l is earlier than the time when the vehicle visits transport node i, this constraint will not take effect, as shown below:

[0088]

[0089] b14, the type and range of the relevant variables, are expressed as follows:

[0090]

[0091]

[0092]

[0093] t0=0, l0=0, t0′=0, l0′=0

[0094] Among them, t0, l0, t0′ and l0′ are the zeroed times of truck arrival, truck departure, drone arrival and drone departure respectively.

[0095] Furthermore, the vehicle and multi-UAV collaborative delivery model based on multi-dimensional cost fusion described in step S3 specifically includes:

[0096] The objective function of the vehicle and multi-UAV collaborative delivery model based on multi-dimensional cost fusion is expressed as follows:

[0097]

[0098] The sum of the importance weights of transportation cost and transportation time for logistics companies is 1, that is:

[0099] α+β=1

[0100] The constraints include the constraints of the vehicle and single UAV collaborative delivery model based on multi-dimensional cost fusion, and include the following constraints:

[0101] c1, the vehicle does not travel between any two UAV nodes, which is represented as follows:

[0102]

[0103] c2, the drone does not travel between any two drone nodes, which is represented as follows:

[0104]

[0105] c3, the drone enters and exits any transport node at most once, which is expressed as follows:

[0106]

[0107]

[0108] c4, the number of drones launched at any launch point does not exceed k, expressed as follows:

[0109]

[0110] Among them, k is the number of unmanned aerial vehicles;

[0111] c5, the total number of drones does not exceed k, expressed as follows:

[0112]

[0113]

[0114] Furthermore, the step S4 uses the branch and bound algorithm to solve the three distribution mode models based on multi-dimensional cost fusion, specifically including:

[0115] Define decision variables: define relevant decision variables and objective functions that reflect multi-dimensional cost fusion;

[0116] Constraint integration: add constraints to different distribution mode models and incorporate the constraints into the above three distribution mode models;

[0117] Branching process: Create branches of the decision tree and branch according to the preset strategy, where each branch represents a possible solution or path choice;

[0118] Bounding operation: Calculate the upper and lower bounds for each branch to screen out potential optimal or suboptimal solutions. The upper bound is the limit of the best solution that the branch can reach, and the lower bound is the current estimate of the global optimal solution.

[0119] Pruning strategy: If the lower bound of a subproblem is worse than the known global upper bound or has no feasible solution, the subproblem branch is pruned;

[0120] Optimize the search process: Repeat the branch, bound, and prune process in the framework of the branch and bound method to find the optimal or near-optimal solution;

[0121] Extract solutions: When all nodes are explored or pruned, the best feasible solution found is the global optimal solution, that is, the potential optimal solutions under the three delivery modes are obtained respectively.

[0122] Furthermore, the decision in step S5 obtains the optimal path planning solution, which specifically includes:

[0123] After obtaining the potential optimization solutions under the three distribution modes in step 4, conduct a comprehensive evaluation and comparison of the optimal solutions based on the solution results of the three distribution modes in the same case, determine the decision criteria based on the minimization of total cost, and based on the decision criteria, select the optimal solution and the most suitable distribution strategy among the three distribution models.

[0124] Furthermore, the weights α and β described in step S2 are set to 0.7 and 0.3.

[0125] Beneficial effects:

[0126] The present invention achieves multi-faceted optimization in the field of logistics and distribution by introducing a vehicle-drone collaborative distribution model decision-making method that integrates multi-dimensional costs. First, the invention improves the existing collaborative distribution model of vehicles and drones, and more flexibly realizes efficient path design in urban environments, reducing time waste and energy consumption in the distribution process; secondly, the introduction of a multi-dimensional cost comprehensive model enables the system to not only focus on transportation costs, but also take into account multiple dimensions such as time value costs, thereby achieving cost minimization in multiple dimensions; finally, the method and system provided by the present invention have higher applicability, providing an innovative and efficient solution for modern logistics and distribution. BRIEF DESCRIPTION OF THE DRAWINGS

[0127] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments, and the above and / or other advantages of the present invention will become more clear.

[0128] Figure 1 The present invention is a flowchart of a decision-making method provided according to an embodiment of the present application.

[0129] Figure 2A schematic diagram of three delivery modes provided according to an embodiment of the present application.

[0130] Figure 3 A schematic diagram of vehicle-single-drone collaborative delivery provided according to an embodiment of the present application.

[0131] Figure 4 A schematic diagram of vehicle-multi-drone collaborative delivery provided according to an embodiment of the present application.

[0132] Figure 5 The figure is a comparison chart of the results of three delivery modes under different numbers of customers obtained according to an embodiment of the present application. DETAILED DESCRIPTION

[0133] The embodiment of the present application provides a vehicle-drone collaborative delivery mode decision method that considers multi-dimensional cost fusion, which innovatively solves the problem of insufficient multi-dimensional cost considerations and difficult delivery mode decision-making in the field of terminal delivery based on background technology. The present invention covers a variety of delivery modes such as full vehicle, vehicle + single drone, and vehicle + multiple drones, and quantitatively evaluates the pros and cons of each delivery mode, fully considering the comprehensive impact of different delivery modes in terms of time cost and money cost, and making decisions based on the multi-dimensional cost optimal principle, so as to facilitate a more comprehensive and accurate understanding of the potential costs and benefits of different delivery modes, so as to make more scientific decisions in actual delivery.

[0134] The technical solution in the embodiment of the present application is to solve the above technical problems, and the overall idea is as follows:

[0135] Currently, the field of terminal distribution generally has problems such as high cost and poor timeliness, and the traditional distribution model is unable to meet the needs of modern logistics. Traditional distribution models often cannot effectively solve these problems because they fail to fully consider multidimensional costs and cannot make scientific and reasonable decisions in complex distribution environments. Although existing studies have provided some distribution models, they have failed to provide clear guidance for distribution decisions in different scenarios based on the fusion of multidimensional costs, making it difficult to determine the optimal distribution strategy in practical applications. Therefore, this application aims to make up for the shortcomings of existing research and propose an innovative vehicle-UAV collaborative distribution model decision-making method that comprehensively considers time costs and money costs, and solves the decision-making problems in terminal distribution in an intelligent way.

[0136] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods:

[0137] The embodiment of the present invention provides a vehicle-UAV collaborative delivery mode decision method considering multi-dimensional cost fusion, such as Figure 1 As shown, the method comprises the following steps:

[0138] S1, obtain customer point information, road information, vehicle information and drone information;

[0139] S2. Establish a multi-dimensional cost objective function and determine the weights of different dimensions;

[0140] S3. Based on customer point information, road information, vehicle information and drone information, combined with factors such as the location of the delivery point, vehicle speed and drone range, a full vehicle delivery model, a vehicle-single drone delivery model and a vehicle-multi-drone delivery model are constructed that consider multi-dimensional cost fusion. The delivery schematic diagrams of the three delivery modes are shown in the figure. Figure 2 As shown;

[0141] S4, using the branch and bound algorithm to solve three distribution models considering multi-dimensional cost fusion;

[0142] S5, making a decision to obtain the optimal path planning solution under the distribution mode;

[0143] The embodiment of the present invention constructs an intelligent vehicle-drone collaborative delivery mode decision method by fully integrating time cost and money cost. This method not only solves the problem of the single dimension of the traditional mode, but also can flexibly select different delivery modes, thereby improving the terminal delivery efficiency while reducing the overall cost.

[0144] The implementation process of the embodiment of the present invention is described in detail below:

[0145] In step S1, customer point information, road information, vehicle information and drone information are obtained. The specific implementation process is as follows:

[0146] The computer obtains customer point information, road information, vehicle information, and drone information;

[0147] Customer point information includes customer point demand, service duration and coordinates;

[0148] Road information includes the road network distance between nodes and the flight distance between nodes;

[0149] Vehicle information includes vehicle unit transportation cost and vehicle travel speed;

[0150] Drone information includes the unit transportation cost of the drone, the time required to launch / recover the drone, the flight speed of the drone, and the endurance of the drone;

[0151] In step S2, a multi-dimensional cost objective function is established and the weights of different dimensions are determined. The specific implementation process is as follows:

[0152] S201. Clarify the cost items: clarify the dimensions of each cost item in each distribution mode, and take the minimization of vehicle transportation cost, drone transportation cost and time cost as the objective function;

[0153] S202. Comprehensively consider the importance of different cost items and set reasonable weights for each cost dimension;

[0154] S203. Standardize the measurement units of different cost items and normalize them into monetary units to facilitate comprehensive comparison;

[0155] The objective function of the vehicle-UAV collaborative delivery model considering multi-dimensional cost fusion is expressed by formula (1):

[0156]

[0157] Among them, i and j are node numbers; α is the importance weight of logistics companies to transportation costs, and β is the importance weight of logistics companies to timeliness; b1 is the unit transportation cost of the vehicle, b2 is the unit transportation cost of the drone, and b3 is the time value cost; N+ is the set of nodes that the vehicle may visit; xij is the decision variable, that is, the driving path of the vehicle from node i to node j, and yij is the decision variable, that is, the flight path of the drone from i to node j; dij is the time required for the vehicle to travel from node i to node j, and dij' is the time required for the drone to travel from node i to node j; Tc+1 is the time it takes for the vehicle and drone to return to the distribution center after completing all delivery tasks.

[0158] The sum of the importance weights of transportation cost and transportation time for logistics companies is 1, which is calculated by the following formula:

[0159] α+β=1 (2)

[0160] In step S3, based on customer point information, road information, vehicle information and drone information, combined with factors such as the location of the delivery point, the vehicle's driving speed and the drone's range, a full vehicle delivery model, a vehicle-single drone delivery model and a vehicle-multi-drone delivery model considering multi-dimensional cost integration are constructed. The specific implementation process is as follows:

[0161] S301. The constraint conditions of the full vehicle distribution model considering multi-dimensional cost fusion are expressed by formulas (3) to (7):

[0162]

[0163]

[0164]

[0165]

[0166] l i ≥t i +s i ·m i -M·(1-x ip ) (7)

[0167] in:

[0168] Formulas (3) to (4) indicate that a vehicle can enter and exit any customer node at most once;

[0169] Formula (5) represents the vehicle elimination sub-loop constraint;

[0170] Formulas (6) to (7) represent the time constraints of the vehicle;

[0171] l, i and j are customer node numbers, C is the set of all customer points; N0 is the set of nodes that the vehicle may leave, N+ is the set of nodes that the vehicle may visit; α is the importance weight of the logistics company to the transportation cost, β is the importance weight of the logistics company to the timeliness; b1 is the unit transportation cost of the vehicle, b3 is the time value cost; dij is the driving distance of the vehicle from node i to node j; vt is the average driving speed of the vehicle; μi is the visit order of node i; si is the service time of the vehicle at each customer point. ti is the time when the vehicle arrives at node i, li is the time when the vehicle leaves node i; xij is the decision variable, which represents the driving path of the vehicle from node i to node j.

[0172] S302: The constraints of the vehicle-single-UAV collaborative delivery model considering multi-dimensional cost fusion are expressed by formulas (8) to (27). The delivery schematic diagram is as follows: Figure 3 As shown:

[0173]

[0174]

[0175]

[0176]

[0177]

[0178]

[0179]

[0180]

[0181]

[0182] li ≥t i +s i ·m i +t r ·(∑ k y ki )+t l ·(∑ j y ij )-M·(1-x ip ) (17)

[0183] l i ≥l k ′+d′ ki / v′·n k ·m i +t r ·n k ·m i -M·(1-y ki ) (18)

[0184] t i ′≥l′ k +d′ ki / v′-M·(1-y ki ) (19)

[0185] l i ′≥(t i ′+s i )·n i +t i ·m i +t r ·y ki ·m i +t l ·y ij ·m i -M·(1-y ij ) (20)

[0186]

[0187]

[0188]

[0189]

[0190]

[0191]

[0192] T0=0,T0'=0,t0=0,t0'=0 (27)

[0193] in:

[0194] Formulas (8) to (9) indicate that there is only one drone and it enters and exits any customer node at most once;

[0195] Formula (10) represents the vehicle elimination sub-loop constraint;

[0196] Formula (11) indicates that each customer point can only be served once by a vehicle or drone;

[0197] Formulas (12) to (13) indicate that the vehicle must depart from and return to the distribution center once;

[0198] Formula (14) represents the range constraint of the UAV;

[0199] Formula (15) represents the load constraint of the UAV;

[0200] Formulas (16) to (18) represent the time constraints of the vehicle, where formula (16) ensures that the time when the vehicle arrives at the subsequent node is no earlier than the time when it departs from the previous node plus the travel time between the two nodes, formula (17) ensures that the time when the vehicle departs from the node is no earlier than the time when it arrives at the node plus the service time of the vehicle at the node and the time when the drone is released or recovered at the node, and formula (18) ensures that if a node is a drone merging node, the time when the vehicle departs from the node is no earlier than the time when the drone arrives at the node plus the time when the vehicle releases / recovers the drone, that is, the vehicle leaves the node only after releasing / recovering the drone;

[0201] Formulas (19) to (20) represent the time constraints of the drone. Formula (19) ensures that the time when the drone arrives at the subsequent node is no earlier than the time when it departs from the previous node plus the travel time between the two nodes. Formula (20) ensures that the time when the drone departs from a node is no earlier than the sum of the time when it arrives at the node and the service time of the drone at the node.

[0202] Formulas (21) to (22) represent the access order constraints of pij, ensuring that the vehicle will not repeatedly pass through the visited customer nodes;

[0203] Formula (23) means that if the drone is launched from node i and recovered at node k, and the drone is launched from the next node l (pil = 1), this constraint will ensure that the time when the vehicle launches the drone at node i is earlier than the time when the drone returns to node k, that is, when a single drone departs from point i to serve customer point j, the vehicle will not launch the drone at the next visited node l before the drone returns to node k; if the time when the vehicle visits point l is earlier than the time when it visits point i, then this formula will not be binding;

[0204] Formulas (24) to (27) represent the types and ranges of relevant variables;

[0205] 1. i and j are customer node numbers, C is the set of all customer points; N is the set of all nodes, N0 is the set of nodes that the vehicle may leave, and N+ is the set of nodes that the vehicle may visit; α is the importance weight of logistics companies to transportation costs, and β is the importance weight of logistics companies to timeliness; b1 is the unit transportation cost of the vehicle, b2 is the unit transportation cost of the drone, and b3 is the time value cost; dij is the driving distance of the vehicle from node i to node j; dij' is the flight distance of the drone from node i to node j; vt is the average driving speed of the vehicle, and vd is the average flight speed of the drone; tl is the time required for the vehicle driver to release the drone, and tr is the time required to recover the drone; si is the service time of the vehicle and the drone at each customer point. ti is the time when the vehicle arrives at node i, and li is the time when the vehicle leaves node i; ti' is the time when the drone arrives at node i, and li' is the time when the drone leaves node i; mi is a decision variable, indicating that customer node i is served by the vehicle, and ni is a decision variable, indicating that customer node i is served by the drone; xij is a decision variable, indicating the driving path of the vehicle from node i to node j, and yij is a decision variable, indicating the flight path of the drone from i to node j; Pij is a decision variable, indicating the visit order between customer nodes i and customer j; μi is the visit order of node i; ε is the range of the drone; wi is the cargo weight of customer point i, and W is the load limit of the drone; M is an infinite positive integer.

[0206] S303. Since the relevant constraints are similar to the vehicle-single-UAV collaborative delivery model considering multi-dimensional cost fusion, only the additional formula is explained here. The constraints of the vehicle-multi-UAV collaborative delivery model considering multi-dimensional cost fusion are expressed by formulas (28) to (34). The distribution diagram is shown in Figure 4 As shown:

[0207]

[0208]

[0209]

[0210]

[0211]

[0212]

[0213]

[0214] in:

[0215] Formula (28) indicates that the vehicle does not travel between two UAV nodes;

[0216] Formula (29) indicates that the UAV does not travel between two UAV nodes;

[0217] Formulas (30) to (31) indicate that a drone can enter and exit any customer node at most once;

[0218] Formula (32) indicates that the number of drones launched at any launch point does not exceed k;

[0219] Formulas (33) to (34) indicate that the total number of drones does not exceed k;

[0220] k is the total number of drone flights, and the other parameters have been explained in the decision-making method of vehicle-single drone collaborative delivery mode considering multi-dimensional cost fusion, and will not be repeated here.

[0221] In step S4, three distribution models considering multi-dimensional cost fusion are efficiently solved with the help of branch and bound algorithm;

[0222] In step S5, the optimal route planning scheme under the delivery mode is obtained through comparison and decision making. The comparison of the results of the three delivery modes when the number of customers is 6 and 10 is shown in the figure below. Figure 5 shown.

[0223] In a specific implementation, the present application provides a computer storage medium and a corresponding data processing unit, wherein the computer storage medium can store a computer program, and when the computer program is executed by the data processing unit, the invention content of the vehicle and drone collaborative delivery decision method based on multi-dimensional cost fusion provided by the present invention and some or all steps in each embodiment can be run. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0224] Those skilled in the art can clearly understand that the technical solutions in the embodiments of the present invention can be implemented by means of computer programs and their corresponding general hardware platforms. Based on such an understanding, the technical solutions in the embodiments of the present invention can be essentially or partly contributed to the prior art in the form of computer programs, i.e., software products, which can be stored in a storage medium and include several instructions for enabling a device including a data processing unit (which can be a personal computer, a server, a single-chip microcomputer, an MCU or a network device, etc.) to execute the methods described in various embodiments of the present invention or certain parts of the embodiments.

[0225] The present invention provides a thought and method for a vehicle and drone collaborative delivery decision method based on multi-dimensional cost fusion. There are many methods and ways to implement the technical solution. The above is only a preferred implementation of the present invention. It should be pointed out that for ordinary technicians in this technical field, without departing from the principle of the present invention, several improvements and modifications can be made, and these improvements and modifications should also be regarded as the protection scope of the present invention. All components not specified in this embodiment can be implemented by existing technologies.

Claims

1. A vehicle and drone collaborative delivery decision method based on multi-dimensional cost fusion, characterized in that: The method comprises: Step S1, obtaining customer point information, road information, vehicle information and drone information; Step S2: Establish a multi-dimensional cost objective function and determine the weights of costs in different dimensions; Step S3: Based on the customer point information, road information, vehicle information and drone information, combined with the constraint conditions and the weights of different dimensions determined in step S2, three distribution mode models based on multi-dimensional cost fusion are constructed; Step S4, using the branch and bound algorithm to solve the above three distribution mode models based on multi-dimensional cost fusion; Step S5: Based on the solution of step S4, the optimal path planning scheme is determined, that is, the optimal delivery mode, and the vehicle and drone collaborative delivery decision based on multi-dimensional cost fusion is completed; The three distribution mode models based on multi-dimensional cost fusion described in step S3 specifically include: A full vehicle delivery model based on multi-dimensional cost fusion, a vehicle and single drone collaborative delivery model based on multi-dimensional cost fusion, and a vehicle and multiple drone collaborative delivery model based on multi-dimensional cost fusion; The full vehicle distribution model based on multi-dimensional cost fusion described in step S3 specifically includes: Objective function and constraints of the full vehicle delivery model based on multi-dimensional cost fusion, where: The objective function of the full vehicle distribution model based on multi-dimensional cost fusion is expressed as follows: Among them, minZ means the minimum comprehensive cost, i and j are the transportation node numbers; α is the importance weight of the logistics enterprise to the transportation cost, β is the importance weight of the logistics enterprise to the timeliness; b1 is the unit transportation cost of the vehicle, b3 is the time value cost; N + is the set of transportation nodes that the vehicle may visit; x ij is the decision variable, i.e., the vehicle's driving path from transport node i to transport node j; d ij is the driving distance of the vehicle from transport node i to transport node j; T c+1 It is the time it takes for a transport vehicle to return to the distribution center after completing all delivery tasks; The sum of the importance weights of transportation cost and transportation time for logistics companies is 1, that is: α+β＝1 The constraints include: a1, a vehicle enters and exits any transport node at most once, expressed as follows: Among them, C is the set of all transportation nodes; a2, vehicle elimination sub-loop constraint, is expressed as follows: Among them, μi is the visit order of transport node i, n is the number of total transport nodes, and N0 is the set of nodes that the vehicle may leave; a3, the time constraint of the vehicle is expressed as follows: l i ≥t i +s i ·m i -M·(1-x ip ) Among them, t i is the time when the vehicle arrives at transport node i, l i is the time when the vehicle leaves the transport node i, v t is the average speed of the vehicle, s i is the service time of trucks and drones at each transportation node, m i The transport node i is served by the drone, and M is an infinite positive integer; The vehicle and single drone collaborative delivery model based on multi-dimensional cost fusion described in step S3 specifically includes: Objective function and constraints of the vehicle-single-UAV collaborative delivery model based on multi-dimensional cost fusion, where: The objective function of the vehicle and single UAV collaborative delivery model based on multi-dimensional cost fusion is expressed as follows: Among them, b2 is the unit transportation cost of the drone, y ij is the decision variable, i.e., the flight path of the UAV from transport node i to transport node j; d′ ij is the flight distance of the drone from transport node i to transport node j; Tc+1 is the time it takes for the transport tools, i.e., vehicles and drones, to return to the distribution center after completing all delivery tasks; The sum of the importance weights of transportation cost and transportation time for logistics companies is 1, that is: α+β＝1 The constraints include: b1, there is only one drone, and it enters and exits any transport node at most once, which is represented as follows: b2, vehicle elimination sub-loop constraint, is expressed as follows: b3, the traffic flow of vehicles and drones in and out of the transportation node is conserved, that is, each transportation node can only be served by a vehicle or drone once, which is expressed as follows: Among them, n i The transport node i is served by the drone; b4, the vehicle must depart from and return to the same distribution center once, expressed as follows: b5, the UAV’s cruising range constraint, that is, the UAV must complete the delivery task within its predetermined cruising range, expressed as follows: Among them, ε is the range of the UAV; b6, the load limit of the drone, is used to indicate the maximum load capacity of the drone when performing a delivery mission, as follows: Among them, w i is the cargo weight at transport node i, and W is the load limit of the UAV; b7, the vehicle arrival time constraint, is used to ensure that after the vehicle departs from any transport node, the time it arrives at the subsequent transport node is no earlier than the time it leaves the previous transport node plus the travel time between the two transport nodes, expressed as follows: b8, the vehicle’s departure time constraint, is used to ensure that the vehicle’s departure time from transport node i is no earlier than the time it arrives at the node plus the total time the vehicle needs to be at the node, including: service time, and the additional time the driver needs to release or recover the drone at the node, expressed as follows: Among them, t l The time required for the vehicle driver to release the drone, t r the time required to recover the drone for the vehicle driver; b9, the departure time constraint of the vehicle, is used to ensure that if the transport node i is the drone recovery point, the time when the vehicle leaves the transport node i shall not be earlier than the time when the drone leaves the transport node k after completing the delivery plus the time required for the driver to release or recover the drone. It is used to ensure that the truck departs after completing the drone recovery operation, which is expressed as follows: l i ≥l k ′+ / v d ·n k ·m i +t r ·n k ·m i -M·(1-y ki ) Among them, l k ′ is the moment when the UAV leaves the transport node i, v d is the average flight speed of the UAV; b10, the arrival time constraint of the drone, is used to ensure that the drone arrives at any subsequent transport node no earlier than the time it leaves the previous node plus the flight time between the two nodes, expressed as follows: t i ′≥l′ k +d′ ki / v d -M·(1-y ki ) Among them, t i ′ is the time when the UAV arrives at the transport node i; b11, the departure time constraint of the drone, is used to ensure that the departure time of the drone from any transportation node is no earlier than the sum of the arrival time at the node and the service time required by the drone at the node; l i ′≥(t i ′+s i )·n i +t i ·m i +t r ·y ki ·m i +t l ·y ij ·m i -M·(1-y ij ) b12, the visit order constraint, is used to ensure that the vehicle does not repeatedly pass through the visited transport nodes, which is expressed as follows: Among them, P ij is the access order between transport node i and transport node j; b13, if the drone is launched from transport node i and recovered at transport node k, and the drone is launched from transport node l in the next step; this constraint will ensure that the time when the vehicle launches the drone at transport node i is earlier than the time when the drone returns to transport node k, that is, when a single drone departs from transport node i to serve customer point j, the vehicle will not launch the drone at the next visited transport node l before the drone returns to node k; if the time when the vehicle visits transport node l is earlier than the time when the vehicle visits transport node i, this constraint will not take effect, as shown below: b14, the type and range of the relevant variables, are expressed as follows: t0=0, l0=0, t0′=0, l0′=0 Among them, t0, l0, t0′ and l0′ are the times of truck arrival, truck departure, drone arrival and drone departure respectively; The vehicle and multi-UAV collaborative delivery model based on multi-dimensional cost fusion described in step S3 specifically includes: The objective function of the vehicle and multi-UAV collaborative delivery model based on multi-dimensional cost fusion is expressed as follows: The sum of the importance weights of transportation cost and transportation time for logistics companies is 1, that is: α+β＝1 The constraints include the constraints of the vehicle and single UAV collaborative delivery model based on multi-dimensional cost fusion, and include the following constraints: c1, the vehicle does not travel between any two UAV nodes, which is represented as follows: c2, the drone does not travel between any two drone nodes, which is represented as follows: c3, the drone enters and exits any transport node at most once, which is expressed as follows: c4, the number of drones launched at any launch point does not exceed k, expressed as follows: Among them, k is the number of unmanned aerial vehicles; c5, the total number of drones does not exceed k, expressed as follows:

2. The vehicle and drone collaborative delivery decision method based on multi-dimensional cost fusion as claimed in claim 1, characterized in that: The establishment of the multi-dimensional cost objective function described in step S2 specifically includes: The multidimensional cost objective function at least includes: transportation cost and time cost, and the time cost and transportation cost are respectively denoted as f1(x) and f2(x); wherein the transportation cost includes: vehicle transportation cost and drone transportation cost.

3. The vehicle and drone collaborative delivery decision method based on multi-dimensional cost fusion as claimed in claim 2, characterized in that: The step S2 of determining the weights of the costs of different dimensions specifically includes: Using weights α and β, a comprehensive objective function Z is constructed, which is the weighted sum of all individual objective functions, as follows: Z=α·f1(x)+β·f2(x) Unify the measurement units of time cost and transportation cost and standardize them as follows: Assuming the time value per hour is b3, the time cost is expressed as: Time cost = b3·T c+1 This gives us the final comprehensive cost function: Z=α·f1(x)+β·b3·T c+1 Among them, T c+1 The time it takes for a transport vehicle to return to the distribution center after completing all delivery tasks.

4. The vehicle and drone collaborative delivery decision method based on multi-dimensional cost fusion as claimed in claim 3, characterized in that: The step S4 uses the branch and bound algorithm to solve the three distribution mode models based on multi-dimensional cost fusion, specifically including: Define decision variables: define relevant decision variables and objective functions that reflect multi-dimensional cost fusion; Constraint integration: add constraints to different distribution mode models and incorporate the constraints into the above three distribution mode models; Branching process: Create branches of the decision tree and branch according to the preset strategy, where each branch represents a possible solution or path choice; Bounding operation: Calculate the upper and lower bounds for each branch to screen out potential optimal or suboptimal solutions. The upper bound is the limit of the best solution that the branch can reach, and the lower bound is the current estimate of the global optimal solution. Pruning strategy: If the lower bound of a subproblem is worse than the known global upper bound or has no feasible solution, the subproblem branch is pruned; Optimize the search process: Repeat the branch, bound, and prune process in the framework of the branch and bound method to find the optimal or near-optimal solution; Extract solutions: When all nodes are explored or pruned, the best feasible solution found is the global optimal solution, that is, the potential optimal solutions under the three delivery modes are obtained respectively.

5. The vehicle and drone collaborative delivery decision method based on multi-dimensional cost fusion as claimed in claim 4, characterized in that: The decision in step S5 obtains the optimal path planning solution, which specifically includes: After obtaining the potential optimization solutions under the three distribution modes in step 4, conduct a comprehensive evaluation and comparison of the optimal solutions based on the solution results of the three distribution modes in the same case, determine the decision criteria based on the minimization of total cost, and based on the decision criteria, select the optimal solution and the most suitable distribution strategy among the three distribution models.

6. The vehicle and drone collaborative delivery decision method based on multi-dimensional cost fusion as claimed in claim 5, characterized in that: The weights α and β described in step S2 are set to 0.7 and 0.3.

Images