Vehicle and unmanned aerial vehicle cooperative distribution path planning method considering heterogeneous fleet

Abstract

The invention discloses a vehicle and unmanned aerial vehicle low-altitude collaborative distribution path planning method considering a heterogeneous fleet. The method comprises the following steps: firstly, acquiring low-altitude logistics distribution demand data, distribution combination data composed of a distribution vehicle and a heterogeneous unmanned aerial vehicle, and distribution combination data composed of a vehicle and a heterogeneous unmanned aerial vehicle for low-altitude logistics distribution, and initially realizing distribution resource allocation; establishing a vehicle-heterogeneous unmanned aerial vehicle team low-altitude logistics cooperative distribution path optimization model by taking the shortest total distribution time and low total distribution cost as targets, improving the distribution timeliness of all demands of customers and optimizing the cost, and optimizing the vehicle and unmanned aerial vehicle low-altitude logistics cooperative distribution path by adopting an improved adaptive neighborhood search algorithm, so as to improve the distribution timeliness of all demands of the customers. And an overall distribution scheme of the terminal logistics demand is obtained. The method can be compatible with cooperative transportation of different types of unmanned aerial vehicles, so that corresponding unmanned aerial vehicle fleet combinations are configured for different transportation requirements, the transportation cost is saved, and meanwhile, the method has good vehicle-vehicle cooperative path planning solving efficiency.

Description

Technical Field

[0001] The present invention relates to the field of path planning and logistics distribution, and in particular to a vehicle-heterogeneous UAV collaborative distribution path planning method considering heterogeneous fleets. Background Art

[0002] With the rapid development of e-commerce, the contradiction between the limited logistics carrying capacity of the current short-distance freight system and the growing demand for logistics distribution has become increasingly significant. Although the traditional vehicle-based transportation mode has a large carrying capacity, it is costly and inefficient, and is also restricted by terrain and transportation facilities. At present, drone technology is gradually being applied to logistics distribution to achieve vehicle-drone collaborative distribution, so as to improve the timeliness of distribution, reduce costs, and also improve customer satisfaction. Therefore, the present invention proposes a multi-vehicle multi-drone collaborative distribution path planning method considering heterogeneous fleets, aiming to achieve the goal of shortest total distribution time and lower total cost by pairing multiple vehicles and multiple heterogeneous drones on the basis of the existing logistics transportation network. Compared with other existing air-to-ground logistics distribution methods based on vehicle-vehicle collaboration, the model constructed by this method is compatible with collaborative transportation of different types of drones, so as to configure corresponding drone fleet combinations for different transportation needs, further save transportation costs, and have good vehicle-vehicle collaborative path planning solution efficiency and path implementation feasibility. Summary of the invention

[0003] The present invention provides a multi-vehicle and multi-UAV collaborative delivery path planning method considering heterogeneous fleets to achieve the goals of minimizing total delivery time and lowering total cost.

[0004] To achieve the above goals, it mainly includes two stages and four main steps:

[0005] The method for low-altitude collaborative delivery of vehicles and drones considering heterogeneous fleets is characterized by comprising the following steps:

[0006] Phase 1: Vehicle-heterogeneous UAV transportation resource allocation for basic transportation needs

[0007] S1. Obtain the location and demand data of the warehouse of the low-altitude logistics distribution center and the customer demand nodes, as well as the distribution consortium data consisting of vehicles and heterogeneous drones used for low-altitude logistics distribution;

[0008] S2, based on the location and demand data of customer demand nodes and the data of the distribution consortium consisting of vehicles and heterogeneous drones used for low-altitude logistics distribution, realize preliminary distribution resource allocation according to the location distribution and demand of customer demand nodes;

[0009] Phase 2: Vehicle-UAV transport path planning based on detailed logistics environment

[0010] S3. To achieve the highest delivery timeliness for all customer needs, that is, the shortest total delivery time and the lowest total cost, a vehicle-heterogeneous drone fleet low-altitude logistics collaborative delivery path optimization model is established;

[0011] S4. The vehicle-heterogeneous UAV fleet low-altitude logistics collaborative distribution path optimization model is solved by using an improved adaptive neighborhood search algorithm.

[0012] As a further improvement of the present invention, the acquisition of the data of the distribution consortium consisting of low-altitude logistics distribution vehicles and heterogeneous drones in step S1 includes:

[0013] S101. Multiple drones are equipped on each vehicle for coordinated delivery, where different types of drones have different performance parameters of load capacity and battery life;

[0014] S102. Obtain the vehicle's delivery data and the drone's delivery data; the vehicle's delivery data includes: the time required for the vehicle to travel from delivery point m to delivery point n, and the time required for the vehicle to serve delivery point m; the drone's delivery data includes: the time required for a type G drone to travel from delivery point m to delivery point n, and the time required for a type G drone to serve delivery point m.

[0015] As a further improvement of the present invention, the acquisition of the distribution consortium data consisting of vehicles and heterogeneous drones for low-altitude logistics distribution in step S1 includes: clustering the location information of customer demand nodes to preliminarily determine the demand points served by each vehicle; according to the logistics distribution needs of the demand points, combined with the performance parameters of the drones, determine the type and quantity of drones loaded on each vehicle to achieve preliminary distribution resource allocation.

[0016] As a further improvement of the present invention, the steps for constructing the vehicle-heterogeneous drone fleet low-altitude logistics collaborative distribution path optimization model in step S3 are:

[0017] S401. According to the preliminary distribution resource allocation results, each drone departs from the central warehouse with the vehicle, takes off from any distribution point, completes the distribution task of another customer demand node, and then merges with the vehicle to other customer demand nodes; after completing all distributions, the drone and the vehicle return to the central warehouse together, with the shortest overall distribution time and the lowest total cost; the objective function of the vehicle-heterogeneous drone collaborative path problem model is the total distribution time mint and the total cost minr.

[0018] As a further improvement of the present invention, the constraint conditions of the vehicle-heterogeneous drone fleet low-altitude logistics collaborative distribution path optimization model include: customer service attribute constraints, vehicle path constraints and drone path constraints;

[0019] The constraints of the customer service attributes include:

[0020] use Formula, defining decision variables: To judge customers Whether it is in the itinerary k∈K of a drone of type g and delivered by the drone; Judging customers Whether the delivery is made by the vehicle numbered z, using two decision variables and 1 to ensure that the set of customers served by the vehicle or drone is only served once;

[0021] The path constraints of the vehicle include:

[0022] The vehicle is constrained to depart from the distribution center once and return to the distribution center once after delivery; each customer point is constrained to be served by the vehicle at most once; the constraints of the sub-loop of the vehicle's driving path are eliminated; there is only one vehicle to deliver customer point i, and the attribute of customer point i is that it can only be delivered by the vehicle, each customer point i is served once and the time the vehicle leaves is greater than the sum of the vehicle's arrival time and service time, ensuring that the vehicle leaves after completing the service requirements of the customer point; the time when the drone returns to the vehicle is earlier than the time when the vehicle leaves, ensuring the constraints of the coordinated operation between the vehicle and the drone;

[0023] The drone path constraints include:

[0024] Determine whether itinerary k∈K is flown by a type g drone along the arc (m,n); determine whether customer point i is delivered by a type g drone in itinerary k, and ensure that the type g drone is only launched or recovered once at a certain node; in a certain itinerary of a type g drone, the package weight set of all customer points does not exceed the maximum load capacity of this type of drone; in a certain itinerary of a type g drone, the remaining power of the drone can complete the entire itinerary and return to the vehicle; in a certain flight path of a type g drone, at most one of the drone's landing point and take-off point is in the set that the vehicle can reach / leave; in the flight path of a type g drone, a path can only be served by a drone once.

[0025] As a further improvement of the present invention, the formulas of the objective functions of the total delivery time mint and the total cost minr of the vehicle-heterogeneous drone fleet low-altitude logistics collaborative delivery path optimization model are as follows:

[0026]

[0027] Among them, i, j, b, f are the numbers of customer gathering points, C is the set of customer gathering points; U is the set of drones, G is the set of drone types; T is the set of vehicles; is the time required for vehicle numbered z to travel from the warehouse to node b; is the time required for vehicle numbered z to travel from node f to the warehouse; is the time required for vehicle numbered z to travel from node i to node j; is the time required for a type g drone to travel from node i to node j; The time required for vehicle number z to serve point i; The time required to serve point i for a drone of type g; r T is the fuel consumption per unit distance of the vehicle; v Tz is the average speed of vehicle numbered z; u g is the number of drones of type g; r Ug is the cost of a single drone of type g; B is the fixed cost of a single vehicle.

[0028] As a further improvement of the present invention, the improved adaptive neighborhood search algorithm is used to solve the step S4, including: solving the vehicle-heterogeneous drone fleet low-altitude logistics collaborative distribution path optimization model through a two-stage improved adaptive neighborhood search algorithm, the first stage generates an initial solution through taboo search, and the second stage obtains the local optimal vehicle and drone low-altitude logistics collaborative distribution path through the genetic simulated annealing algorithm judgment criterion.

[0029] As a further improvement of the present invention, the first stage generates an initial solution through taboo search, including:

[0030] S601, preprocessing relevant parameters of the vehicle-heterogeneous drone fleet low-altitude logistics collaborative distribution path optimization model according to the data set;

[0031] S602, adding all customer demand nodes to the unvisited node set;

[0032] S603, determine whether the customer demand node has been traversed, if so, execute S606, otherwise execute the next step;

[0033] S604, selecting a suitable unvisited customer demand node, and selecting a vehicle or drone for delivery based on the node attributes, including the currently selected node, the distance to other nodes, and whether it can be accessed by a drone;

[0034] S605. According to the selection result of S604, if the drone service condition is met, the customer demand node is added to the drone service path; if not, it is included in the vehicle service path; the node is marked as visited, and the process returns to S603;

[0035] S606: End the preliminary processing and generate an initial solution.

[0036] As a further improvement of the present invention, the second stage obtains the local optimal path for the coordinated delivery of low-altitude logistics between vehicles and drones by using the genetic simulated annealing algorithm judgment criteria, including:

[0037] S701, using ElimOperato neighborhood search algorithm to select the best preliminary solution;

[0038] S702a, randomly selecting a drone delivery solution on a path from the initial solution as solution p1;

[0039] S702b, delete plan p1, and reselect the drone type on the delivery path plan;

[0040] S702c, re-rank the customers on path p1 according to the three elements of cargo demand, service time window, and distance, and re-select the type of drone for the customers on this path according to the ranking of the customers; calculate the time and cost of the path to obtain the current solution, and determine whether to use the new solution to update the initial solution and the optimal solution according to the effect of the new solution;

[0041] S702d, repeat steps 702a to 702c until the iteration termination condition is reached.

[0042] The beneficial effects of the present invention are: a joint air-ground logistics transportation model based on vehicle-machine collaboration and a corresponding rapid solution method are proposed. A collaborative distribution model is constructed by evaluating the transportation capacity and performance of various types of drones and ground vehicles and evaluating logistics needs, and then a domain search algorithm is designed to quickly solve and obtain a logistics distribution plan for vehicle-machine collaboration. Compared with other existing air-ground logistics distribution methods based on machine-vehicle collaboration, the model constructed by this method is compatible with collaborative transportation of drones of different models, so that corresponding drone fleet combinations can be configured for different transportation needs, further saving transportation costs. At the same time, it has good efficiency in solving the solution of machine-vehicle collaborative path planning and the feasibility of path implementation, which is expected to make up for the current high time cost and economic cost of terminal logistics transportation in the "last mile". BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a flow chart of the present invention.

[0044] Figure 2 The diagram shows the vehicle-drone collaborative delivery model.

[0045] Figure 3 Flowchart for improving the adaptive neighborhood search algorithm to solve collaborative delivery routes. DETAILED DESCRIPTION

[0046] The present invention will be further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention.

[0047] As shown in the figure, the present invention provides a vehicle-drone low-altitude collaborative delivery path planning method considering a heterogeneous fleet, comprising the following steps:

[0048] Phase 1: Vehicle-heterogeneous UAV transportation resource allocation for basic transportation needs

[0049] S1. Obtain the location and demand data of the warehouse of the low-altitude logistics distribution center and the customer demand nodes, as well as the distribution consortium data consisting of vehicles and heterogeneous drones used for low-altitude logistics distribution;

[0050] S2, based on the location and demand data of customer demand nodes and the data of the distribution consortium consisting of vehicles and heterogeneous drones used for low-altitude logistics distribution, realize preliminary distribution resource allocation according to the location distribution and demand of customer demand nodes;

[0051] Phase 2: Vehicle-UAV transport path planning based on detailed logistics environment

[0052] S3. To achieve the highest delivery timeliness for all customer needs, that is, the shortest total delivery time and the lowest total cost, a vehicle-heterogeneous drone fleet low-altitude logistics collaborative delivery path optimization model is established;

[0053] S4. The vehicle-heterogeneous UAV fleet low-altitude logistics collaborative distribution path optimization model is solved by using an improved adaptive neighborhood search algorithm.

[0054] As a further improvement of the present invention, the acquisition of the data of the distribution consortium consisting of low-altitude logistics distribution vehicles and heterogeneous drones in step S1 includes:

[0055] S101. Multiple drones are equipped on each vehicle for coordinated delivery, where different types of drones have different performance parameters of load capacity and battery life;

[0056] S102. Obtain the vehicle's delivery data and the drone's delivery data; the vehicle's delivery data includes: the time required for the vehicle to travel from delivery point m to delivery point n, and the time required for the vehicle to serve delivery point m; the drone's delivery data includes: the time required for a type G drone to travel from delivery point m to delivery point n, and the time required for a type G drone to serve delivery point m.

[0057] In this embodiment, the acquisition of the distribution consortium data consisting of vehicles and heterogeneous drones for low-altitude logistics distribution in step S1 includes: clustering the location information of customer demand nodes to preliminarily determine the demand points served by each vehicle; according to the logistics distribution needs of the demand points, combined with the performance parameters of the drones, determine the type and quantity of drones loaded on each vehicle to achieve preliminary distribution resource allocation.

[0058] In this embodiment, the steps for constructing the vehicle-heterogeneous drone fleet low-altitude logistics collaborative distribution path optimization model in step S3 are:

[0059] S401. According to the preliminary distribution resource allocation results, each drone departs from the central warehouse with the vehicle, takes off from any distribution point, completes the distribution task of another customer demand node, and then merges with the vehicle to other customer demand nodes; after completing all distributions, the drone and the vehicle return to the central warehouse together, with the shortest overall distribution time and the lowest total cost; the objective function of the vehicle-heterogeneous drone collaborative path problem model is the total distribution time mint and the total cost minr.

[0060] In this embodiment, the constraints of the vehicle-heterogeneous drone fleet low-altitude logistics collaborative distribution path optimization model include: customer service attribute constraints, vehicle path constraints, and drone path constraints;

[0061] The constraints of the customer service attributes include:

[0062] use Formula, defining decision variables: To judge customers Whether it is in the itinerary k∈K of a drone of type g and delivered by the drone; Judging customers Whether the delivery is made by the vehicle numbered z, using two decision variables and 1 to ensure that the set of customers served by the vehicle or drone is only served once;

[0063] The path constraints of the vehicle include:

[0064] The vehicle is constrained to depart from the distribution center once and return to the distribution center once after delivery; each customer point is constrained to be served by the vehicle at most once; the constraints of the sub-loop of the vehicle's driving path are eliminated; there is only one vehicle to deliver customer point i, and the attribute of customer point i is that it can only be delivered by the vehicle, each customer point i is served once and the time the vehicle leaves is greater than the sum of the vehicle's arrival time and service time, ensuring that the vehicle leaves after completing the service requirements of the customer point; the time when the drone returns to the vehicle is earlier than the time when the vehicle leaves, ensuring the constraints of the coordinated operation between the vehicle and the drone;

[0065] The drone path constraints include:

[0066] Determine whether itinerary k∈K is flown by a type g drone along the arc (m,n); determine whether customer point i is delivered by a type g drone in itinerary k, and ensure that the type g drone is only launched or recovered once at a certain node; in a certain itinerary of a type g drone, the package weight set of all customer points does not exceed the maximum load capacity of this type of drone; in a certain itinerary of a type g drone, the remaining power of the drone can complete the entire itinerary and return to the vehicle; in a certain flight path of a type g drone, at most one of the drone's landing point and take-off point is in the set that the vehicle can reach / leave; in the flight path of a type g drone, a path can only be served by a drone once.

[0067] In this embodiment, the formulas of the objective functions of the vehicle-heterogeneous drone fleet low-altitude logistics collaborative delivery path optimization model, total delivery time mint and total cost minr, are as follows:

[0068]

[0069] Among them, i, j, b, f are the numbers of customer gathering points, C is the set of customer gathering points; U is the set of drones, G is the set of drone types; T is the set of vehicles; is the time required for vehicle numbered z to travel from the warehouse to node b; is the time required for vehicle numbered z to travel from node f to the warehouse; is the time required for vehicle numbered z to travel from node i to node j; is the time required for a type g drone to travel from node i to node j; The time required for vehicle number z to serve point i; The time required for a drone of type g to serve point i; r T is the fuel consumption per unit distance of the vehicle; v Tz is the average speed of vehicle numbered z; u g is the number of drones of type g; r Ug is the cost of a single drone of type g; B is the fixed cost of a single vehicle.

[0070] In this embodiment, the improved adaptive neighborhood search algorithm is used to solve step S4, including: solving the vehicle-heterogeneous drone fleet low-altitude logistics collaborative distribution path optimization model through a two-stage improved adaptive neighborhood search algorithm, the first stage generates an initial solution through taboo search, and the second stage obtains the local optimal vehicle and drone low-altitude logistics collaborative distribution path through the genetic simulated annealing algorithm judgment criterion.

[0071] In this embodiment, the first stage generates an initial solution through taboo search, including:

[0072] S601. Preprocess the relevant parameters of the vehicle and UAV path models according to the data set and format them into a format acceptable to the algorithm input;

[0073] S602, based on the network structure of available paths on the ground and in the air, add all customer demand nodes to the set of unvisited nodes and the set of directed arcs between the nodes;

[0074] S603, based on the determined vehicle and heterogeneous UAV fleet configuration information, generate an initial feasible solution as an initial path for vehicle-heterogeneous UAV collaborative delivery;

[0075] S604. Configure the machine-road collaborative path planning cost function: minimize the transportation cost of the vehicle-heterogeneous drone fleet, which mainly includes time cost and energy consumption cost.

[0076] S605, initialization algorithm parameters are as follows:

[0077] Initial temperature: The temperature parameter controls the probability of accepting a poor solution and affects the algorithm's ability to explore in the search space. The choice of initial temperature affects the algorithm's convergence speed and search quality.

[0078] Cooling rate: The cooling rate determines how the temperature gradually decreases as the number of iterations increases. A higher cooling rate will cause the temperature to decrease rapidly, but may fall into a local optimal solution early; a lower cooling rate will cause the temperature to decrease slowly, increasing the possibility of a global search.

[0079] Stopping condition: You can set a maximum number of iterations or define a satisfactory solution quality as a condition for stopping the search to ensure that the algorithm converges to a solution within a limited time.

[0080] S606: Determine whether the customer demand node has been traversed. If yes, execute S609; otherwise, execute the next step.

[0081] S607. Select a suitable unvisited customer demand node. According to the node attributes, such as the currently selected node, the distance to other nodes, whether it can be visited by drones, etc., select a vehicle or drone for delivery and calculate the transportation cost function:

[0082] Z=mint+minr

[0083]

[0084] S608, select the current path node and perform random modification, select the surrounding adjacent nodes to replace the selected node, and according to the calculation result of S607, if it meets the drone service conditions, then decide to add the customer demand node to the drone service path based on the Metropolis criterion, as shown in the following formula; if it does not meet the conditions, then add it to the vehicle service path; mark the node as visited, and return to S606;

[0085] A=e -ΔC / T

[0086] Among them, A is the probability of the node being accepted, ΔC is the difference between the cost function value of the new solution and the cost function value of the current solution, and T is the current temperature. As the temperature T decreases, the probability of accepting the difference solution gradually decreases, and the algorithm tends to accept a better solution, realizing the adaptive acceptance threshold adjustment of the domain search algorithm.

[0087] S609: End the preliminary processing and generate an initial solution.

[0088] In this embodiment, the second stage obtains the local optimal path for low-altitude logistics coordinated delivery of vehicles and drones through the genetic simulated annealing algorithm judgment criteria, including:

[0089] S701, using ElimOperato neighborhood search algorithm to select the best preliminary solution;

[0090] S702a, randomly selecting a drone delivery solution on a path from the initial solution as solution p1;

[0091] S702b, delete plan p1, and reselect the drone type on the delivery path plan;

[0092] S702c, re-rank the customers on path p1 according to the three elements of cargo demand, service time window, and distance, and re-select the type of drone for the customers on this path according to the ranking of the customers; calculate the time and cost of the path to obtain the current solution, and determine whether to use the new solution to update the initial solution and the optimal solution according to the effect of the new solution;

[0093] S702d, repeat steps 702a to 702c until the iteration termination condition is reached.

[0094] The technical means disclosed in the scheme of the present invention are not limited to the technical means disclosed in the above-mentioned implementation mode, but also include technical schemes composed of any combination of the above-mentioned technical features. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also regarded as the protection scope of the present invention.

Claims

1. A low-altitude collaborative delivery path planning method for vehicles and drones considering heterogeneous fleets, characterized by: The following steps are involved: Phase 1: Vehicle-heterogeneous UAV transportation resource allocation for basic transportation needs S1. Obtain the location and demand data of the warehouse of the low-altitude logistics distribution center and the customer demand nodes, as well as the distribution consortium data consisting of vehicles and heterogeneous drones used for low-altitude logistics distribution; S2, based on the location and demand data of customer demand nodes and the data of the distribution consortium consisting of vehicles and heterogeneous drones used for low-altitude logistics distribution, realize preliminary distribution resource allocation according to the location distribution and demand of customer demand nodes; Phase 2: Vehicle-UAV transport path planning based on detailed logistics environment S3. To achieve the highest delivery timeliness for all customer needs, that is, the shortest total delivery time and the lowest total cost, a vehicle-heterogeneous drone fleet low-altitude logistics collaborative delivery path optimization model is established; S4. The vehicle-heterogeneous UAV fleet low-altitude logistics collaborative distribution path optimization model is solved by using an improved adaptive neighborhood search algorithm.

2. The method for low-altitude collaborative delivery of vehicles and drones considering heterogeneous fleets as claimed in claim 1 is characterized in that: The acquisition of the data of the distribution consortium consisting of low-altitude logistics distribution vehicles and heterogeneous drones in step S1 includes: S101. Multiple drones are equipped on each vehicle for coordinated delivery, where different types of drones have different performance parameters of load capacity and battery life; S102. Obtain the vehicle's delivery data and the drone's delivery data; the vehicle's delivery data includes: the time required for the vehicle to travel from delivery point m to delivery point n, and the time required for the vehicle to serve delivery point m; the drone's delivery data includes: the time required for a type G drone to travel from delivery point m to delivery point n, and the time required for a type G drone to serve delivery point m.

3. The method for low-altitude collaborative delivery of vehicles and drones considering heterogeneous fleets as claimed in claim 1 is characterized in that: The acquisition of the data of the distribution consortium consisting of vehicles and heterogeneous drones for low-altitude logistics distribution in step S1 includes: clustering the location information of customer demand nodes to preliminarily determine the demand points served by each vehicle; according to the logistics distribution needs of the demand points, combined with the performance parameters of the drones, determine the type and quantity of drones loaded on each vehicle to achieve preliminary distribution resource allocation.

4. The method for low-altitude collaborative delivery of vehicles and drones considering heterogeneous fleets as claimed in claim 1 is characterized in that: The steps for constructing the vehicle-heterogeneous drone fleet low-altitude logistics collaborative distribution path optimization model in step S3 are: S401. According to the preliminary distribution resource allocation results, each drone departs from the central warehouse with the vehicle, takes off from any distribution point, completes the distribution task of another customer demand node, and then merges with the vehicle to other customer demand nodes; after completing all distributions, the drone and the vehicle return to the central warehouse together, with the shortest overall distribution time and the lowest total cost; the objective function of the vehicle-heterogeneous drone collaborative path problem model is the total distribution time mint and the total cost minr.

5. The method for low-altitude collaborative delivery of vehicles and drones considering heterogeneous fleets as claimed in claim 4 is characterized in that: The constraints of the vehicle-heterogeneous drone fleet low-altitude logistics collaborative distribution path optimization model include: customer service attribute constraints, vehicle path constraints and drone path constraints; The constraints of the customer service attributes include: use Formula, defining decision variables: To judge customers Whether it is in the itinerary k∈K of a drone of type g and delivered by the drone; Judging customers Whether the delivery is made by the vehicle numbered z, using two decision variables and 1 to ensure that the set of customers served by the vehicle or drone is only served once; The path constraints of the vehicle include: The vehicle is constrained to depart from the distribution center once and return to the distribution center once after delivery; each customer point is constrained to be served by the vehicle at most once; the constraints of the sub-loop of the vehicle's driving path are eliminated; there is only one vehicle to deliver customer point i, and the attribute of customer point i is that it can only be delivered by the vehicle, each customer point i is served once and the time the vehicle leaves is greater than the sum of the vehicle's arrival time and service time, ensuring that the vehicle leaves after completing the service requirements of the customer point; the time when the drone returns to the vehicle is earlier than the time when the vehicle leaves, ensuring the constraints of the coordinated operation between the vehicle and the drone; The drone path constraints include: Determine whether itinerary k∈K is flown by a type g drone along the arc (m,n); determine whether customer point i is delivered by a type g drone in itinerary k, and ensure that the type g drone is only launched or recovered once at a certain node; in a certain itinerary of a type g drone, the package weight set of all customer points does not exceed the maximum load capacity of this type of drone; in a certain itinerary of a type g drone, the remaining power of the drone can complete the entire itinerary and return to the vehicle; in a certain flight path of a type g drone, at most one of the drone's landing point and take-off point is in the set that the vehicle can reach / leave; in the flight path of a type g drone, a path can only be served by a drone once.

6. The method for low-altitude collaborative delivery of vehicles and drones considering heterogeneous fleets as claimed in claim 4 is characterized in that: The objective function total delivery time mint and total cost minr of the vehicle-heterogeneous drone fleet low-altitude logistics collaborative delivery path optimization model are as follows: Among them, i, j, b, f are the numbers of customer gathering points, C is the set of customer gathering points; U is the set of drones, G is the set of drone types; T is the set of vehicles; is the time required for vehicle numbered z to travel from the warehouse to node b; is the time required for vehicle numbered z to travel from node f to the warehouse; is the time required for vehicle numbered z to travel from node i to node j; is the time required for a type g drone to travel from node i to node j; The time required for vehicle number z to serve point i; The time required to serve point i for a drone of type g; r T is the fuel consumption per unit distance of the vehicle; v Tz is the average speed of vehicle numbered z; u g is the number of drones of type g; r Ug is the cost of a single drone of type g; B is the fixed cost of a single vehicle.

7. The method for low-altitude collaborative delivery of vehicles and drones considering heterogeneous fleets as claimed in claim 1 is characterized in that: In step S4, an improved adaptive neighborhood search algorithm is used for solution, including: solving the vehicle-heterogeneous drone fleet low-altitude logistics collaborative distribution path optimization model through a two-stage improved adaptive neighborhood search algorithm, in which the initial solution is generated through taboo search in the first stage, and the local optimal path of vehicle and drone low-altitude logistics collaborative distribution is obtained through the genetic simulated annealing algorithm judgment criterion in the second stage.

8. The method for low-altitude collaborative delivery of vehicles and drones considering heterogeneous fleets as claimed in claim 7 is characterized in that: The first stage generates an initial solution through tabu search, including: S601, preprocessing relevant parameters of the vehicle-heterogeneous drone fleet low-altitude logistics collaborative distribution path optimization model according to the data set; S602, adding all customer demand nodes to the unvisited node set; S603, determine whether the customer demand node has been traversed, if so, execute S606, otherwise execute the next step; S604, selecting a suitable unvisited customer demand node, and selecting a vehicle or drone for delivery based on the node attributes, including the currently selected node, the distance to other nodes, and whether it can be accessed by a drone; S605. According to the selection result of S604, if the drone service condition is met, the customer demand node is added to the drone service path; if not, it is included in the vehicle service path; the node is marked as visited, and the process returns to S603; S606: End the preliminary processing and generate an initial solution.

9. The method for low-altitude collaborative delivery of vehicles and drones considering heterogeneous fleets as claimed in claim 8, characterized in that: The second stage obtains the local optimal path for low-altitude logistics coordinated delivery of vehicles and drones through the genetic simulated annealing algorithm judgment criteria, including: S701, using ElimOperato neighborhood search algorithm to select the best preliminary solution; S702a, randomly selecting a drone delivery solution on a path from the initial solution as solution p1; S702b, delete plan p1, and reselect the drone type on the delivery path plan; S702c, re-rank the customers on path p1 according to the three elements of cargo demand, service time window, and distance, and re-select the type of drone for the customers on this path according to the ranking of the customers; calculate the time and cost of the path to obtain the current solution, and determine whether to use the new solution to update the initial solution and the optimal solution according to the effect of the new solution; S702d, repeat steps 702a to 702c until the iteration termination condition is reached.

Images