Freight unmanned aerial vehicle distribution station site selection method based on greedy overlapped circle algorithm

Abstract

The invention discloses a freight unmanned aerial vehicle distribution station site selection method based on a greedy overlapping circle algorithm, and aims to solve the problem of'last kilometer 'distribution of urban regional cargo transportation. The method comprises the following steps: firstly, determining a freight unmanned aerial vehicle model and a delivery service range according to geographic attributes of a target delivery area, the number and value attributes of delivery packages and the like; then, in combination with a GIS geographic information system, buffer area analysis is carried out on the site selection area of the unmanned aerial vehicle distribution site, and site candidate areas are determined; secondly, by taking minimization of the number of unmanned aerial vehicle distribution stations and the comprehensive distribution distance as optimization targets, constructing an unmanned aerial vehicle distribution station site selection optimization model; and finally, designing a greedy overlapping circle algorithm to solve the unmanned aerial vehicle distribution station site selection optimization model, and obtaining the minimum number and accurate positions of the unmanned aerial vehicle distribution stations. According to the technical scheme, cargo distribution delay caused by ground traffic congestion can be relieved, and the overall efficiency and customer satisfaction of tail-end cargo distribution in urban areas are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of facility location, and in particular to a freight unmanned aerial vehicle distribution site location method based on a greedy overlapping circle algorithm. BACKGROUND

[0002] There is a deep integration and mutual promotion symbiotic relationship between urban e-commerce and freight transportation. The rapid development of e-commerce not only promotes the prosperity of the logistics industry, but also lays a solid foundation for the sustainable growth of e-commerce. In the field of freight transportation, with the continuous exploration and innovation of technology application by many large logistics enterprises, unmanned aerial vehicle distribution technology has gradually matured, and its application scenarios have become increasingly widespread, and have gradually expanded from remote rural areas to urban core areas. The method of the present application makes full use of the significant advantages of unmanned aerial vehicles in transportation efficiency, cost control, and adaptation to small batch and high frequency transportation, and aims to effectively solve the problem of "last mile" distribution in urban area freight transportation, thereby further improving the overall efficiency of logistics services.

[0003] Freight unmanned aerial vehicle distribution site location optimization belongs to the facility location problem (LP), and the unmanned aerial vehicle distribution mode relying on logistics vehicles is suitable for small-scale distribution. For large-scale distribution, it is necessary to consider setting up unmanned aerial vehicle distribution sites to complete the distribution of users around the unmanned aerial vehicle distribution site, and the most important part is to reasonably select the location of the unmanned aerial vehicle distribution site, which is close to the facility location problem. The current research on this problem can be mainly divided into the maximum coverage problem, the set covering problem, the P-median problem and the fixed cost facility location problem. In the construction of unmanned aerial vehicle distribution network, the goal is to determine the optimal distribution site layout, which not only ensures the minimum number of site construction to reduce costs, but also takes into account the shortest comprehensive distribution path between each site and demand point to improve efficiency. In the face of unknown construction cost of distribution site and uncertainty of distribution demand, through reasonable planning, the least number of unmanned aerial vehicle distribution sites is established to maximize the initial investment. In view of the uncertainty of demand, the key factors directly related to distribution are accurately captured, which can effectively optimize the distribution strategy and minimize the overall distribution mileage, thereby significantly reducing operating costs. SUMMARY

[0004] The present application discloses a freight unmanned aerial vehicle distribution site location method based on a greedy overlapping circle algorithm, which aims to alleviate the delay of freight distribution caused by ground traffic congestion, improve the level of urban area end logistics service, and improve the distribution efficiency.

[0005] To achieve the above object, the technical scheme provided by the present application is:

[0006] A freight unmanned aerial vehicle distribution site location method based on a greedy overlapping circle algorithm, characterized by comprising the following steps:

[0007] Step 1: According to the geographical attributes of the target distribution area, the number and value attributes of the distribution parcels, etc., determine the model of the freight unmanned aerial vehicle and the distribution service range;

[0008] Step 2: In combination with the GIS geographic information system, perform buffer analysis on the unmanned aerial vehicle distribution site location area to determine the candidate area for the unmanned aerial vehicle distribution site;

[0009] Step 3: Taking the minimization of the number of unmanned aerial vehicle distribution sites and the comprehensive distribution distance as the optimization goal, construct an unmanned aerial vehicle distribution site location optimization model;

[0010] Step 4: Design a greedy overlapping circle algorithm to solve the unmanned aerial vehicle distribution site location optimization model to obtain the minimum number and accurate position of the unmanned aerial vehicle distribution site.

[0011] To optimize the above technical solution, the specific measures / limitations taken also include:

[0012] In step 1, first determine the geographical boundary range of the target distribution area, obtain the geographical information of the demand points in the target distribution area through the terminal, that is, the latitude and longitude coordinates of the distribution service range of the courier station. Then, according to the historical parcel data of the target distribution area, including the number, weight, value, and vulnerability of the distribution parcels, consider the related technical parameters such as the endurance mileage, flight speed, and load capacity of the unmanned aerial vehicle, determine the model of the freight unmanned aerial vehicle, and further determine the unmanned aerial vehicle distribution service range.

[0013] In step 2, according to the requirements of the site clearance conditions and the ground environment for the site location design of the unmanned aerial vehicle vertical take-off and landing field, in combination with the GIS geographic information system, use the geographic model analysis method to perform buffer analysis on the courier stations in the target distribution area, set a reasonable buffer distance for the demand points within the geographical boundary range determined in step 1 to generate a multi-ring buffer zone, and exclude limiting factors within the buffer zone that do not meet the requirements of the unmanned aerial vehicle vertical take-off and landing, such as airspace restriction zones, military sites, rivers, and mountains, to determine the candidate area for the freight unmanned aerial vehicle distribution site.

[0014] In step 3, for the candidate area of the freight unmanned aerial vehicle distribution site, consider the distribution of demand points in the target area and the unmanned aerial vehicle distribution service range, and under the premise of providing unmanned aerial vehicle distribution services for all demand points in the distribution area, establish an unmanned aerial vehicle distribution site location optimization model with the goal of minimizing the number of distribution sites and the comprehensive distribution distance:

[0015] Considering the construction cost of the unmanned aerial vehicle distribution site, the smaller the number of unmanned aerial vehicle distribution sites, the lower the construction cost, and the formula is as follows:

[0016] f1 = min(k)

[0017] Wherein, k is the number of unmanned aerial vehicle distribution station;

[0018] After determining the minimum number of unmanned aerial vehicle distribution station, the location of unmanned aerial vehicle distribution station needs to be further clarified, the distance between unmanned aerial vehicle distribution station and demand point should be as close as possible, and the objective function is as follows:

[0019]

[0020] Wherein, L ji is the unmanned aerial vehicle distribution distance from the jth unmanned aerial vehicle distribution station to the ith demand point; μ ji is a binary variable, which indicates whether the ith demand point is within the service range of the jth unmanned aerial vehicle distribution station;

[0021] Each demand point is located within the service range of at least one unmanned aerial vehicle distribution station, and the constraint condition is:

[0022]

[0023] The calculation formula of unmanned aerial vehicle distribution station to demand point is:

[0024]

[0025] The constraint condition for determining the demand point set within the service area of each unmanned aerial vehicle distribution station is:

[0026]

[0027] Wherein (a i , b i ) represents the longitude and latitude of the ith demand point, (x j , y j ) represents the longitude and latitude of the jth unmanned aerial vehicle distribution station, G j represents the set of demand points within the jth unmanned aerial vehicle distribution station, and r uav represents the endurance mileage of cargo unmanned aerial vehicle.

[0028] In step 4, a greedy overlapping circle algorithm is designed to solve the site optimization model described in step 3, and the accurate position of unmanned aerial vehicle distribution station is obtained, wherein the design process of the greedy overlapping circle algorithm is as follows:

[0029] Step 4.1: According to the distance value of all demand points from the origin, they are stored in the total demand point set Q in order from small to large;

[0030] Step 4.2: Let i = 1;

[0031] Step 4.3: Find the smallest number of demand points in Q, and set the demand point as the i-th center point O i ;

[0032] Step 4.4: Find the demand points within a distance of O i not greater than 2R, that is, find all demand points covered by a circle with a radius of 2R, which can be named as the i-th characteristic circle;

[0033] Step 4.5: Add these demand points to the candidate set, and name the candidate set as set S i , and remove them from the total demand point set Q;

[0034] Step 4.6: Determine the position of one or more UAV distribution stations within the i-th characteristic circle;

[0035] Step 4.7: Let i = i + 1, repeat steps 4.3 to 4.5 until the algorithm ends.

[0036] After the processing of step 4.5, it is still not possible to guarantee that all demand points in set S i can be covered by the service range of the UAV distribution station. Therefore, further processing of step 4.6 is required, which essentially finds the minimum enclosing circle of all demand points in set S i . The specific operation of step 4.6 is as follows:

[0037] Step 4.6.1: Make a circle with the i-th center node O i as the center and R as the radius, obtaining the i-th center circle;

[0038] Step 4.6.2: Let j = 1;

[0039] Step 4.6.3: Choose a point on the outline of the i-th center circle as the center to make a circle, obtaining a candidate circle; move the candidate circle around the center of the center circle for one revolution to determine the position of the candidate circle that covers the most demand points in the candidate set S i , which is taken as the covering circle ij ; remove the demand points within the covering circle from the candidate set S i and add them to the characteristic set U ij , respectively;

[0040] Step 4.6.4: Determine the specific method of one or more UAV distribution stations as follows:

[0041] ① If there is only one demand point in the characteristic set U ij , take the demand point as the UAV distribution station C ij , and directly enter step 4.6.5;

[0042] If the characteristic set U ij contains only two demand points, the midpoint of the line connecting the two demand points is taken as the UAV distribution station C ij , and directly enters step 4.6.5.

[0043] If the characteristic set U ij contains more than or equal to three demand points, the two demand points with the largest distance in the characteristic set U ij are taken as the first screened demand point p1 and the second screened demand point p2. The midpoint of the line connecting the first screened demand point p1 and the second screened demand point p2 is taken as the candidate node O ij . The demand point with the largest distance from the candidate node O ij except the first screened demand point p1 and the second screened demand point p2 is taken as the third screened demand point p3. If the distance between the third screened demand point p3 and the candidate node O ij is greater than R, step 4 is entered; otherwise, the candidate node O ij can be directly taken as the UAV distribution station C ij , and directly enters step 4.6.5.

[0044] Step 4: The first screened demand point p1, the second screened demand point p2, and the third screened demand point p3 are taken as the three vertices of a characteristic triangle, and the circumcenter of the characteristic triangle is taken as the UAV distribution station C ij , and then step 4.6.5 is entered.

[0045] Step 4.6.5: If there are still demand points in the candidate set S i , 1 is increased, and steps 4.6.3 to 4.6.4 are repeatedly executed; otherwise, step 4.7 is directly entered. The flowchart of the steps of the entire greedy overlapping circle algorithm is shown in Figure 3 .

[0046] Compared with the prior art, the present application has the following advantages:

[0047] The present application aims at the deficiency in the existing research on the optimization of unmanned freight aircraft distribution site location, and proposes a location method based on greedy overlapping circle algorithm which comprehensively considers the minimization of the number of distribution sites and the comprehensive distribution distance, for solving the accurate position of unmanned aircraft distribution site. Compared with other precise algorithms and heuristic algorithms, the greedy overlapping circle algorithm designed by the present application makes use of the strong local optimization ability of the greedy algorithm, and gives the search direction as a whole, making up for the deficiency of the poor overall optimization ability of the greedy algorithm. The algorithm proposed not only gives the specific position of the unmanned aircraft distribution site, but also has the advantages of less number of unmanned aircraft distribution sites and low time complexity, and has better performance than the existing algorithm. At the same time, the technical scheme is aimed at the characteristics of urban areas, combines the advantages of logistics vehicles and unmanned aircraft distribution, and studies the location of unmanned freight aircraft distribution sites in urban areas, and gives a more perfect location scheme, which can improve the logistics service level in urban areas and realize large-scale unmanned aircraft distribution scene.

[0048] EXPLANATION OF DRAWINGS

[0049] Figure 1 is the area when the algorithm flow is first carried out to step 4.5.

[0050] Figure 2 is the several cases when the algorithm flow is carried out to step 4.6.4.

[0051] Figure 3 is the flow chart of the greedy overlapping circle algorithm.

[0052] Figure 4 is the flow chart of the method of the present application.

[0053] Figure 5 is the result map of the location of unmanned aircraft distribution sites. DETAILED DESCRIPTION

[0054] The above content of the present application will be further explained in the form of examples, but this should not be understood as limiting the scope of the above subject matter of the present application to the following examples, and any technology realized based on the above content of the present application belongs to the scope of the present application.

[0055] The present application proposes a location method of unmanned freight aircraft distribution sites based on greedy overlapping circle algorithm, and the flow chart is as shown in Figure 4 , which comprises the following steps:

[0056] (1) According to the geographical properties of the target distribution area, the number and value properties of the distribution packages, etc., determine the model of unmanned freight aircraft and the distribution service range.

[0057] The specific steps in step (1) include:

[0058] Firstly, the geographical boundary of a city is determined, and the longitude and latitude coordinates of the Cainiao stations in the city are obtained through the collection terminal. In order to reduce the distribution cost and improve the distribution efficiency, the following standards are selected to screen the suitable Cainiao stations as the distribution demand points of the cargo unmanned aerial vehicle: ① selecting the Cainiao stations near the communities with high population density as the demand points; and ② selecting the relatively centrally located Cainiao station in the densely distributed Cainiao station cluster (the straight line distance between each Cainiao station is less than one kilometer) as the demand point.

[0059] According to the standards, a total of 30 Cainiao stations are screened as the distribution demand points of the cargo unmanned aerial vehicle. The service range of the demand points basically covers the whole urban area of the city, and the geographical position is also adjacent to the communities with high population density. Compared with the distribution mode of taking all the branches of the Cainiao stations as the demand points, the screening method can greatly improve the distribution efficiency, reduce the distribution cost, reduce the collision risk of the cargo unmanned aerial vehicle when performing the flight task, and is beneficial to the planning of the flight path of the cargo unmanned aerial vehicle. According to the historical package data of the city, including the number, weight, value and damageability of the distribution packages, the related technical parameters such as the endurance mileage and flight speed of the unmanned aerial vehicle and the load capacity are considered, the quad-rotor cargo unmanned aerial vehicle H4 of the SF Express Company is selected to perform the distribution task, and it is determined that the service range of the unmanned aerial vehicle distribution site is 5 kilometers.

[0060] (2) Buffer analysis is performed on the site selection area of the unmanned aerial vehicle distribution site in combination with the GIS geographic information system to determine the site candidate area.

[0061] According to the requirements of the site selection design of the unmanned aerial vehicle vertical take-off and landing field for the site clearance condition and the ground environment, in combination with the GIS geographic information system, a geographic model analysis method is adopted, buffer analysis is performed on the Cainiao stations of the city, a multi-ring buffer area with a buffer distance of 500 meters is generated by taking the Cainiao station as the center by means of the ArcGIS10.2 software, the limiting factors not meeting the requirements of the unmanned aerial vehicle vertical take-off and landing in the buffer area are excluded, the unsuitable site selection areas including rivers, coal machine plants, steel plants and military divisions are removed, and the suitability candidate layer of the site selection area of the unmanned aerial vehicle distribution site is obtained.

[0062] (3) The unmanned aerial vehicle distribution site selection optimization model is constructed by taking the minimization of the number of unmanned aerial vehicle distribution sites and the comprehensive distribution distance as the optimization target.

[0063] In the step (3), for the candidate area of the cargo unmanned aerial vehicle distribution point site, the distribution site number and the comprehensive distribution distance are minimized as the target under the premise of providing the unmanned aerial vehicle distribution service for all the demand points in the distribution area, and the unmanned aerial vehicle distribution site selection optimization model is established:

[0064] Considering the construction cost of the UAV distribution station, the smaller the number of UAV distribution stations, the lower the construction cost, and the formula is as follows:

[0065] f1=min(k)

[0066] Wherein, k is the number of UAV distribution stations;

[0067] After determining the minimum number of UAV distribution stations, the location of the UAV distribution station needs to be further determined, and the distance between the UAV distribution station and the demand point should be as close as possible, and the objective function is as follows:

[0068]

[0069] Wherein, L ji is the UAV distribution distance from the jth UAV distribution station to the ith demand point; μ ji is a binary variable, indicating whether the ith demand point is within the service range of the jth UAV distribution station;

[0070] Each demand point is located within the service range of at least one UAV distribution station, and the constraint condition is:

[0071]

[0072] The calculation formula of the UAV distribution station to the demand point is:

[0073]

[0074] The constraint condition for determining the demand point set within the service area of each UAV distribution station is:

[0075]

[0076] Wherein (a i , b i ) represents the longitude and latitude of the ith demand point, (x j , y j ) represents the longitude and latitude of the jth UAV distribution station, G j represents the set of demand points within the jth UAV distribution station, and r uav represents the endurance mileage of the freight UAV.

[0077] (4) Design a greedy overlapping circle algorithm to solve the UAV distribution station site optimization model and obtain the minimum number and accurate location of the UAV distribution station.

[0078] The specific steps in step (4) include algorithm design and model solving, and the design process of the greedy overlapping circle algorithm is as follows:

[0079] Step 4.1: Store all demand points into the total demand point set Q in ascending order based on their distance from the coordinate origin;

[0080] Step 4.2: Let i = 1;

[0081] Step 4.3: Find the demand point with the smallest number in Q and set it as the i-th center point O i ;

[0082] Step 4.4: Find the distance O i The demand points no larger than 2R are equivalent to finding all the demand points covered by the radius of 2R. This covering circle can be named the i-th characteristic circle;

[0083] Step 4.5: Add these demand points to the candidate set and name the transition set S i , and removed from the total demand point set Q;

[0084] Step 4.6: Determine the location of one or more drone delivery sites within the i-th characteristic circle;

[0085] Step 4.7: Let i = i + 1, and repeat steps 4.3 to 4.5 until The algorithm ends.

[0086] The results after steps 4.4 and 4.5 are as follows Figure 1 After step 4.5, it is still impossible to guarantee that the set S i All demand points in can be covered by the service range of the drone delivery station. Therefore, further processing of step 4.6 is required. Step 4.6 is actually to find the set S i The specific operation of step 4.6 is as follows:

[0087] Step 4.6.1: Take the i-th central node O i Draw a circle with as the center and R as the radius to get the i-th center circle;

[0088] Step 4.6.2: Set j = 1;

[0089] Step 4.6.3: Draw a circle with any point as the center on the contour of the i-th central circle to obtain a candidate circle; move the candidate circle around the center of the central circle for one circle to determine the coverage candidate set S i The candidate circular location with the most demand points is used as the covering circle ⊙ ij ; Remove the demand points in the coverage circle from the candidate set S i Remove and add feature set U ij ;

[0090] Step 4.6.4: IfFigure 2 As shown, the specific method for determining one or more drone delivery sites is as follows:

[0091] ① If the feature set U ij If there is only one demand point in the area, then this demand point is taken as the drone delivery site C. ij , and go directly to step 4.6.5;

[0092] ② If the feature set U ij If there are only two demand points, the midpoint of the line connecting the two demand points is taken as the drone delivery site C. ij , and go directly to step 4.6.5;

[0093] ③If the feature set U ij If the number of internal demand points is greater than or equal to three, then the feature set U ij The two demand points with the farthest distance between them are recorded as the first screening demand point p1 and the second screening demand point p2. The midpoint of the line connecting the first screening demand point p1 and the second screening demand point p2 is taken as the candidate node O. ij Find the demand point in the set that is farthest from the candidate node except the first screening demand point p1 and the second screening demand point p2, and record it as the third screening demand point p3. ij If the distance between candidate nodes O is greater than R, then proceed to step ④; otherwise, the candidate node O ij Can be directly used as drone delivery station C ij , and go directly to step 4.6.5;

[0094] ④ Use the first screening demand point p1, the second screening demand point p2, and the third screening demand point p3 as the three vertices of the characteristic triangle to establish the characteristic triangle. The circumcenter of the characteristic triangle can be used as the drone delivery site C. ij , then proceed to step 4.6.5.

[0095] Step 4.6.5: If the candidate set S i If there are still demand points in , it will increase by 1 and repeat steps 4.6.3 to 4.6.4; otherwise, go directly to step 4.7. The flowchart of the entire greedy overlapping circle algorithm steps is as follows Figure 3 shown.

[0096] The specific steps of solving the model are as follows:

[0097] According to step 1, the present technical solution selects the four-rotor unmanned aerial vehicle H4 of the SF Express Company to perform the delivery task. The maximum range is 10 kilometers, the maximum load is 104 kilograms, and the cruising speed is 43.2 kilometers per hour. Because the coordinate system established by the Python platform is based on latitude and longitude as coordinate axes, it is necessary to convert the cruising range of the cargo unmanned aerial vehicle into latitude and longitude as input values to perform site selection of the delivery station. According to the inquiry, at 27.8° latitude, one degree of longitude is about 98.31 km, and one degree of latitude is about 111 km, and the average one degree of longitude and latitude is about 105 km. Therefore, the service radius of the unmanned aerial vehicle delivery station should be half of the cruising range of the delivery unmanned aerial vehicle, 5 kilometers, which is converted into latitude and longitude about 0.04 degrees. The site selection results obtained by inputting the latitude and longitude coordinates of each demand point and the delivery radius are shown in Table 1. Figure 5 The number of minimum delivery stations obtained is 5, and the specific positions are (112.93, 27.97), (112.92, 27.88), (112.97, 27.82), (113.06, 27.90), and (113.06, 27.80).

[0098] The above is only the preferred embodiment of the present application, and does not limit the present application in any form. Any skilled person in the art can make any simple modification, equivalent replacement and improvement of the above embodiment according to the technical essence of the present application without departing from the scope of the present application.

Claims

1. A method for selecting a delivery station for UAVs based on a greedy overlapping circle algorithm, characterized in that: The following steps are involved: Step 1: Determine the model of the cargo drone and the delivery service range based on the geographical attributes of the target delivery area, the number and value of the delivered packages; Step 2: Combined with the GIS geographic information system, a buffer zone analysis is conducted on the drone delivery station site selection area to determine the candidate areas for drone delivery stations; Step 3: Build a drone delivery site selection optimization model with the optimization goal of minimizing the number of drone delivery sites and the comprehensive delivery distance; Step 4: Design a greedy overlapping circle algorithm to solve the drone delivery station location optimization model to obtain the minimum number and precise location of drone delivery stations.

2. The method for selecting a delivery site for a UAV using a greedy overlapping circle algorithm according to claim 1, wherein: The specific process of step 1 is as follows: First, determine the geographic boundaries of the target delivery area, and obtain the geographic information of the demand points in the target delivery area through the collection terminal, that is, the latitude and longitude coordinates of the Cainiao station within the delivery service range; then, based on the historical package data of the target delivery area, including the number, weight, value, and fragile item attributes of the delivered packages, considering the relevant technical parameters of the drone's range, flight speed, and load capacity, determine the model of the cargo drone, and further determine the drone's delivery service range.

3. The method for selecting a delivery site for a UAV using a greedy overlapping circle algorithm according to claim 1, wherein: The specific process of step 2 is as follows: Based on the site clearance conditions and ground environment requirements for the UAV vertical take-off and landing site selection design, combined with the GIS geographic information system, a geographic model analysis method is used to conduct a buffer zone analysis of the Cainiao Express Station within the target delivery area. Based on the geographic boundary range determined in step 1, a 500-meter buffer distance is set for the demand point to generate a multi-ring buffer zone. Restrictive factors that do not meet the UAV vertical take-off and landing requirements within the buffer zone are eliminated to determine the candidate areas for the site selection of freight UAV delivery points.

4. The method for selecting a delivery site for a UAV using a greedy overlapping circle algorithm according to claim 1, wherein: In step 3, the specific model is as follows: For the candidate areas for the location of cargo drone delivery points, the construction cost of drone delivery stations is considered. The smaller the number of drone delivery stations, the lower the construction cost. The formula is as follows: f1=min(k) Where k is the number of drone delivery sites; After determining the minimum number of drone delivery stations, the locations of the drone delivery stations need to be further clarified. Considering the minimization of the comprehensive distance from the drone delivery station to the demand point, the objective function is as follows: Among them, L ji is the drone delivery distance from the jth drone delivery station to the i-th demand point; μ ji is a binary variable indicating whether the i-th demand point is within the delivery service range of the j-th drone delivery station; Each demand point is within the service range of at least one drone delivery station, with the following constraints: The calculation formula for the distance from the drone delivery station to the demand point is: The constraints for determining the set of demand points within the service area of ​​each drone delivery station are: Among them (a i ,b i ) represents the longitude and latitude of the i-th demand point, (x j ,y j ) represents the longitude and latitude of the jth drone delivery site, G j represents the set of demand points in the jth drone delivery station, r uav Indicates the cruising range of the cargo drone.

5. The method for selecting a delivery site for a UAV using a greedy overlapping circle algorithm according to claim 1, wherein: The greedy overlapping circle algorithm process in step 4 is designed as follows: Step 4.1: Store all demand points into the total demand point set Q in ascending order based on their distance from the coordinate origin; Step 4.2: Let i = 1; Step 4.3: Find the demand point with the smallest number in Q and set it as the i-th center point O i ; Step 4.4: Find the distance O i The demand points no larger than 2R are equivalent to finding all the demand points covered by the radius of 2R. This covering circle can be named the i-th characteristic circle; Step 4.5: Add these demand points to the candidate set and name the candidate set S i , and removed from the total demand point set Q; Step 4.6: Determine the location of one or more drone delivery sites within the i-th characteristic circle; Step 4.7: Let i = i + 1, and repeat steps 4.3 to 4.5 until The algorithm ends.

6. The method for selecting a delivery site for a UAV using a greedy overlapping circle algorithm according to claim 5, wherein: In step 4.6: After the processing of step 4.5, it is still impossible to guarantee that the set S i All demand points in can be covered by the service range of the drone delivery station; therefore, further processing of step 4.6 is required. Step 4.6 is actually to find the set S i The specific operation of step 4.6 is as follows: Step 4.6.1: Take the i-th central node O i Draw a circle with as the center and R as the radius to get the i-th center circle; Step 4.6.2: Set j = 1; Step 4.6.3: Draw a circle with any point as the center on the contour of the i-th central circle to obtain a candidate circle; move the candidate circle around the center of the central circle for one circle to determine the coverage candidate set S i The candidate circular location with the most demand points is used as the covering circle ⊙ ij ; Remove the demand points in the coverage circle from the candidate set S i Remove and add feature set U ij ; Step 4.6.4: Determine one or more drone delivery sites as follows: ① If the feature set U ij If there is only one demand point in the area, then this demand point is taken as the drone delivery site C. ij , and go directly to step 4.6.5; ② If the feature set U ij If there are only two demand points, the midpoint of the line connecting the two demand points is taken as the drone delivery site C. ij , and go directly to step 4.6.5; ③If the feature set U ij If the number of internal demand points is greater than or equal to three, then the feature set U ij The two demand points with the farthest distance between them are recorded as the first screening demand point p1 and the second screening demand point p2; the midpoint of the line connecting the first screening demand point p1 and the second screening demand point p2 is taken as the candidate node O. ij ; Find the demand point in the set that is farthest from the candidate node except the first screening demand point p1 and the second screening demand point p2, and record it as the third screening demand point p3; if the third screening demand point p3 is farthest from the candidate node O ij If the distance between candidate nodes O is greater than R, then proceed to step ④; otherwise, the candidate node O ij Can be directly used as drone delivery station C ij , and go directly to step 4.6.5; ④ Establish a characteristic triangle with the first screening demand point p1, the second screening demand point p2, and the third screening demand point p3 as the three vertices of the characteristic triangle; feature The circumcenter of the triangle is the drone delivery site C ij , then go to step 4.6.5; Step 4.6.5: If the candidate set S i If there are still demand points in , it will be increased by 1 and steps 4.6.3 to 4.6.4 will be repeated; otherwise, go directly to step 4.7.

Images