A method for optimizing site selection and flow distribution of urban low-altitude logistics terminal unmanned aerial vehicle landing points

Abstract

This invention discloses a method for optimizing the location of take-off and landing points and allocating traffic for urban low-altitude logistics terminal drones. Based on demand analysis, and with the goal of minimizing total operating costs and using demand coverage as a penalty, a model for optimizing the location of terminal drone take-off and landing points and allocating traffic is established. Considering various constraints such as coverage limitations, capacity limitations, traffic conservation, service radius limitations, and allocation relationship limitations in several special areas, a heuristic solution algorithm based on adaptive large neighborhood search (ALNS) is employed to determine the number and location of terminal take-off and landing points and the drone traffic allocated to each point. Decision-makers can choose specific solutions based on coverage fairness and economic preferences, thereby significantly reducing operating costs and improving the efficiency of terminal facility layout while meeting basic coverage requirements. This invention achieves refined layout of terminal take-off and landing points and dynamic traffic allocation across nodes, enabling urban areas to achieve a balance between cost and coverage while meeting timeliness requirements.

Description

Technical Field

[0001] This invention relates to the field of urban low-altitude logistics infrastructure planning and management strategies, and in particular to a method for optimizing the selection of take-off and landing points and allocating traffic for urban low-altitude logistics terminal drones. Background Technology

[0002] The low-altitude economy is a comprehensive economic form that uses various manned and unmanned aircraft as carriers, drives low-altitude flight activities, and promotes the integrated development of related fields. Currently, the low-altitude economy has entered a rapid development stage, with urban low-altitude logistics applications showing significant growth potential. Low-altitude logistics refers to the delivery of goods using drones and other aircraft at relatively low airspace altitudes. Compared to traditional logistics methods, it offers advantages such as high efficiency, speed, reduced labor costs, and expanded service coverage.

[0003] With the development of low-altitude logistics, higher requirements have been placed on the planning and construction of infrastructure. However, the site selection and construction of low-altitude infrastructure in most areas is still mainly based on pilot exploration, lacking scientific site selection and layout methods and overall planning schemes. This has led to increasingly prominent problems such as unreasonable site selection and layout of take-off and landing fields and redundant construction of facilities, which have restricted the high-quality development of urban low-altitude logistics.

[0004] Regarding demand analysis, existing research has limited specific needs analysis for urban low-altitude logistics, with most studies focusing on demand analysis within the urban air mobility (UAM) sector. Demand analysis is crucial for facility site selection, as it accurately identifies high-frequency demand areas for low-altitude logistics within urban areas, preventing resource waste caused by site selection deviating from actual needs. Therefore, the demand analysis methods previously used by scholars in the UAM field can be extended to the field of low-altitude logistics demand analysis, and adapted to the actual application scenarios and characteristics of low-altitude logistics.

[0005] In the area of ​​drone traffic allocation, existing research often has a singular objective. Whether it focuses solely on minimizing transportation costs, time, or distance, or treats traffic allocation as a static process after site selection optimization, it neglects the complexity of real-world constraints such as multi-level low-altitude infrastructure networks, capacity, service radius, and restricted / no-fly zones, as well as the trade-offs involved in the decision-making process. Furthermore, only a few studies explicitly incorporate service coverage and equity, failing to consider the differentiated needs of remote and critical areas. Therefore, the design of subsequent allocation strategy models needs to comprehensively balance multiple objectives such as operating costs, service timeliness, and demand coverage, establishing an integrated allocation mechanism that is co-optimized with facility site selection to avoid resource imbalances and efficiency declines caused by a single-minded approach.

[0006] Regarding site selection methods, existing research often has a singular objective. Whether it's a cost-oriented or time-oriented optimization model, both neglect the complexity of real-world scenarios and the trade-offs involved in the decision-making process. Furthermore, only a few studies use demand coverage as the objective function, while in practical infrastructure site selection scenarios, considering demand coverage is crucial for social equity. Therefore, future site selection optimization models need to comprehensively balance multiple objectives such as cost input and demand coverage to avoid decision-making biases caused by a single-minded focus.

[0007] In summary, by deeply analyzing the demand characteristics of urban areas and the core factors of concern in the selection of take-off and landing points for low-altitude logistics terminals, this paper proposes a method for optimizing the selection of take-off and landing points for UAVs in urban low-altitude logistics terminals and allocating traffic. How to minimize the total operating cost while taking into account the demand coverage rate as a penalty, and comprehensively consider various constraints to determine the optimal location, number, and UAV configuration scheme of terminal take-off and landing points, and to solve the problem, is a key technology that urgently needs to be studied in the current planning of urban low-altitude logistics infrastructure. Summary of the Invention

[0008] Objective: To address the shortcomings of existing technologies, this invention proposes a method for optimizing the location of take-off and landing points and allocating traffic for urban low-altitude logistics terminal drones. It constructs a multi-level joint location and traffic allocation model for terminal take-off and landing points and demand points, aiming to minimize the total daily operating cost while considering demand coverage. This addresses the problem of coarse-grained granularity in existing research and the inability of static location selection alone to alleviate regional service imbalances. The method employs multiple constraints, including consideration of remote and important areas, capacity limitations, and traffic conservation. It uses an adaptive large neighborhood search (ALNS) to iteratively solve these constraints through destruction and repair, achieving refined layout of terminal take-off and landing points and dynamic traffic allocation across nodes. When static location selection alone is insufficient to meet balanced service needs, a collaborative strategy of facility layout optimization and multi-level traffic redistribution is further proposed, enabling all areas within the city to achieve a comprehensive balance between cost and coverage while meeting timeliness requirements.

[0009] Technical solution: The present invention provides a method for optimizing the location of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs, comprising the following steps: Step 1) Determine the analysis object and facility system, and establish a three-level facility system and research object consisting of city-level low-altitude logistics hubs, regional UAV take-off and landing sites and terminal take-off and landing points. Use grids as the basic research unit and spatial carriers for subsequent demand and site selection analysis.

[0010] Step 2) Obtain the total annual volume of regular express delivery business in the urban area. Integrate urban population density data, POI data, and questionnaire survey data, and use a quantitative calculation method for urban low-altitude logistics demand to obtain the daily low-altitude logistics demand for each grid cell. .

[0011] Step 3) Construct spatial geographic information data and candidate terminal take-off and landing point set within the given study city area. Based on the daily low-altitude logistics demand obtained in Step 2), construct demand point set, remote area demand point set, and important area demand point set, and calculate the overall demand coverage rate, remote area demand coverage rate, and important area demand coverage rate.

[0012] Step 4): Based on the candidate terminal take-off and landing point set, demand point set, and serviceability relationship between take-off and landing points and demand points obtained in Step 3), a terminal UAV take-off and landing point location optimization and traffic allocation model is established using the objective function minE with the goal of minimizing the total operating cost, and the overall demand coverage rate is used as a penalty term.

[0013] (1).

[0014] The established objective function consists of three parts: the first part is the cost amortized for the construction of take-off and landing points; the second part is the transportation cost, which increases with longer transportation distances and higher traffic volumes; and the third part is a penalty term for unmet demand coverage. P .

[0015] in, For alternative take-off and landing sites Construction cost of a single standard capacity landing point; It is an integer variable representing the alternative take-off and landing points. The number of take-off and landing points to be built at the location; Indicates the construction cost-sharing period; For take-off and landing field and take-off and landing points The distance between them; For take-off and landing field Assigned to take-off and landing points Transportation flow; This refers to the cost per flight kilometer during the drone's flight range; This indicates the penalty for unmet requirements.

[0016] Penalties for unmet needs Prioritize optimizing total operating costs while ensuring basic demand coverage. The calculation formula is as follows: ;

[0017] in, This is the minimum demand coverage threshold; This represents the highest demand coverage threshold. Indicates the take-off and landing point Assigned to demand points Transportation flow; It is a demand point The demand for traffic; To ensure demand coverage The following are the penalty factors; To ensure demand coverage to The reward factor for the interval.

[0018] Step 5) In combination with the actual situation, set special area coverage limits, capacity limits, flow conservation, service radius limits, and allocation relationship limits to ensure the feasibility of the solution and determine the number and location of take-off and landing points and the flow of equipped drones.

[0019] Step 6) To combine the takeoff and landing site selection problem with the traffic allocation problem, a multi-level traffic allocation algorithm is adopted. This aims to minimize operating costs while satisfying the capacity constraints and demand coverage targets of the takeoff and landing sites. The traffic allocation process is as follows:

[0020] 61) Initialize the takeoff and landing point capacity and the takeoff and landing field capacity. Takeoff and landing point capacity: Remaining capacity at take-off and landing points Takeoff and landing field capacity: Remaining capacity of the take-off and landing field Demand points and required traffic: Traffic to be covered in demand areas: .

[0021] 62) Connect each takeoff and landing point k to the nearest takeoff and landing field j; if the current takeoff and landing field has remaining capacity... Then establish a connection. and update Otherwise, choose the next nearest takeoff and landing site with sufficient capacity.

[0022] 63) Filter the distance to the established take-off and landing point k ≤ coverage radius. From demand point i, we obtain the set of candidate demand points. ; Filter out demand points in remote areas and demand points in important areas, and establish a set of remote areas. and important regions .

[0023] 64) Calculate the traffic required to reach the coverage threshold in remote areas. Traffic required to reach coverage thresholds in key areas ,in This represents the minimum coverage threshold for remote areas. This represents the threshold for expected demand coverage in key regions.

[0024] 65) Calculate the priority score for each takeoff and landing point. , ;in For remote areas of takeoff and landing point k, the priority score is... Priority score for important areas of takeoff and landing point k; Let i be the distance between the demand point i and the take-off and landing point k. The distance between takeoff and landing point k and takeoff and landing field j; allocate traffic flow to remote areas and important areas.

[0025] 66) Calculate the remaining unallocated traffic and update the takeoff and landing point value scores. ,in ; then allocate traffic flow; and allocate take-off and landing points according to Sort in descending order, traverse the remaining demand points i covered by each start / stop point k, and sort by... Sort in descending order; allocate transport flights to each demand point i. Update the capacity of take-off and landing points and take-off and landing fields, as well as the traffic flow between take-off and landing fields and take-off and landing points, and update the remaining uncovered traffic flow at the demand points.

[0026] 67) Verify the constraints in step 5) to ensure that the sum of the flow rates allocated from the take-off and landing points to the demand points is equal to the flow rates allocated from the take-off and landing field to the take-off and landing points.

[0027] Step 7), considering the characteristics of the take-off and landing point selection problem and the flow allocation problem, based on the Adaptive Large Neighborhood Search (ALNS), a heuristic initial solution generation strategy, a multi-level flow allocation strategy, and a diversified destruction-repair strategy, along with the multi-level flow allocation algorithm from Step 6), are employed for optimal addressing to improve convergence speed and solution set quality. The process is as follows:

[0028] Step 71), initialize the set of destruction operators D and their corresponding weights. Initialize the set of repair operators R and their corresponding weights. Generate an initial feasible solution: Filter valid candidate take-off and landing points; greedily select take-off and landing points to meet the minimum demand coverage rate in remote areas and the expected demand coverage rate in important areas; randomly select take-off and landing points to meet the overall demand coverage rate, and obtain the current solution. For the current solution Perform traffic allocation; check termination conditions; if the demand coverage target is reached, complete the initial solution generation. Give Conversely, incremental supplementation is performed, and an initial solution is set. The current optimal solution ,Will Assign to ,Will Assign to .

[0029] Step 72) Enter the iterative search phase. In each iteration, a destruction operator is selected based on probability to remove some of the take-off and landing points in the current initial feasible solution. Then, a repair operator is selected to reconstruct the initial feasible solution.

[0030] Step 73) Apply the flow allocation algorithm to the current solution to obtain the flow allocation scheme.

[0031] Step 74): Determine the objective function value of the new solution, decide whether to accept the new solution based on the Metropolis criterion, and update the current solution and optimal solution records; based on weights... Select the destruction operator Applying the destruction operator: Select the repair operator Apply the repair operator: Traffic allocation is performed on candidate solutions: the traffic allocation algorithm is described in step 6; based on the Metropolis criterion, it is determined whether to accept a candidate solution. , , = ;if or And with probability Accept the new interpretation. , = Conversely, the number of iterations .

[0032] Step 75) Adjust the weights of the destruction and repair operators and reduce the system temperature, etc.

[0033] Step 76), check the termination condition: if the number of iterations n < If the condition is met, return to step 72; otherwise, return the optimal solution.

[0034] In step 2), the quantitative calculation method for urban low-altitude logistics demand is as follows:

[0035] Step 21) Obtain the total volume of regular express delivery business in the urban area. The total volume of express delivery services is the sum of domestic same-city express delivery volume, domestic intercity express delivery volume, and international express delivery volume.

[0036] Step 22) Spatially correlate POI data (vector points) with population density data (raster surface) in the raster, with healthcare services and companies being the two types of POIs with higher demand importance.

[0037] Take population density percentage Percentage of POIs The mean value is the grid heat. And calculate the average latitude and longitude of the POI in the grid as the location of the demand center.

[0038] Step 23): Allocate the total amount according to heat to form the annual grid demand, based on grid heat. The total volume of regular express delivery business in urban areas The data is allocated to each grid cell to obtain the annual express delivery volume for that grid cell. .

[0039] Step 24), The potential low-altitude logistics volume is obtained, and the low-altitude logistics volume is divided by the number of days in a year to obtain the daily low-altitude logistics demand for each grid. .in, For the proportion of high-time-delivery goods, Selecting candidates for low-altitude logistics The number of days in a year.

[0040] In step 2), the population density percentage Percentage of POIs The grid heat .

[0041] The constraints in step 5) are: ; ;

[0042] in, To improve coverage of needs in remote areas; This represents the minimum threshold for demand coverage in remote areas. A collection of demand points in remote areas; It is the transportation flow allocated from takeoff and landing point k to demand point i; This represents the demand flow at demand point i; To ensure coverage of demand in key regions; This represents the expected threshold for coverage in key areas. It is a collection of demand points in important regions; This indicates the number of takeoff and landing points to be constructed at candidate takeoff and landing point k; This represents the maximum total number of landing and takeoff points to be constructed. For the capacity of takeoff and landing field j, ; The number of drones at takeoff and landing site j; The number of times the drone operates per day; This refers to the capacity of a single standard takeoff and landing point; Indicates whether the takeoff and landing point k is served by the takeoff and landing field j; Is demand point i served by takeoff and landing point k? It is the distance between the takeoff and landing point k and the takeoff and landing field j; The service coverage radius of the take-off and landing field; Let i be the distance between the demand point i and the take-off and landing point k. It is the service coverage radius of the take-off and landing point; It is a positive number; A value of 1 indicates that the minimum transport volume is 1 flight. This represents the transport flow allocated from landing field j to landing point k. This is a collection of no-fly zone locations; This is a set of geographically restricted locations.

[0043] In step 5), constraint (1) represents the minimum demand coverage rate constraint for remote areas, setting a minimum threshold for demand coverage rate in remote areas; constraint (2) represents the expected coverage rate constraint for important areas, setting an expected coverage threshold; and constraint (3) represents the total number of take-off and landing points to be constructed.

[0044] The range of values ​​for the variables in step 5) is: , , , , .

[0045] Step 71), initialize the set of destruction operators D and their corresponding weights. Initialize the set of repair operators R and their corresponding weights. Generate an initial feasible solution: Filter valid candidate take-off and landing points, greedily select take-off and landing points to meet the minimum coverage requirements in remote areas and the expected coverage requirements in important areas; randomly select take-off and landing points to meet the overall coverage requirements, and obtain the current solution. For the current solution Perform traffic allocation; check termination conditions; if the demand coverage target is reached, complete the initial solution generation. Assign to Conversely, incremental supplementation is performed, and an initial solution is set. The current optimal solution ,Will Assign to ,Will Assign to .

[0046] Step 74): Determine the objective function value of the new solution, decide whether to accept the new solution based on the Metropolis criterion, and update the current solution and optimal solution records; based on weights... Select the destruction operator Applying the destruction operator: Select the repair operator Apply the repair operator: According to step 6), traffic allocation is performed on candidate solutions; based on the Metropolis criterion, it is determined whether to accept a candidate solution. , , = ,if or And with probability Accept the new interpretation. , = Conversely, the number of iterations .

[0047] In step 3), for the established take-off and landing points k, the filter is performed with a distance ≤ coverage radius. From demand point i, we obtain the set of candidate demand points. If the distance to the nearest takeoff and landing field is greater than a threshold, The demand points are used as demand points in remote areas to obtain a set of demand points in remote areas. Identify important POIs and ordinary POIs, sort the number of important POIs in each grid cell in descending order, and determine the demand points for important regions based on the sorting.

[0048] In step 67), if there is no flow allocation at the take-off and landing points, then set... And delete the related links, i.e. .

[0049] In step 3), the overall demand coverage rate is the ratio of the actual transport flow satisfied by all demand points within the study area to the actual transport flow required by all demand points. The remote area demand coverage rate is the ratio of the actual transport flow satisfied by remote demand points within the study area to the actual transport flow required by remote demand points. The important area demand coverage rate is the ratio of the actual transport flow satisfied by important demand points within the study area to the actual transport flow required by important demand points.

[0050] Working Principle: The multi-level traffic allocation algorithm used in this invention focuses on multi-stage priority allocation and dynamic capacity coordination. It aims to minimize operating costs while meeting the capacity constraints of takeoff and landing points and landing fields, as well as demand coverage targets. The process is as follows: Initialize the capacity of takeoff and landing points and landing fields. Takeoff and landing point capacity: Remaining capacity Takeoff and landing field capacity: Remaining capacity Demand points and required traffic: Traffic to be covered in demand areas: Establish connections between take-off and landing points and take-off and landing fields; filter currently coverable demand points; prioritize the allocation of traffic to remote and important areas; greedily allocate remaining traffic; verify and adjust constraints; check for reaching termination conditions and return the optimal solution.

[0051] In step 7), the set of destruction operators D and their corresponding weights are first initialized. Initialize the set of repair operators R and their corresponding weights. 72) In each iteration of the iterative search phase, a destruction operator is selected based on probability to remove some take-off and landing points in the current initial feasible solution, and then a repair operator is selected to reconstruct the initial feasible solution; 73) A flow allocation scheme is obtained by using a flow allocation algorithm on the current initial feasible solution; 74) The objective function value of the new solution is determined, and a decision is made on whether to accept the new solution based on the Metropolis criterion, and the records of the current solution and the optimal solution are updated; 75) The weights of the destruction operator and the repair operator are adjusted and the system temperature is reduced; 6) Termination condition check: if the number of iterations n < If the condition is met, return to step 72; otherwise, return to the optimal solution.

[0052] As a preferred embodiment of the present invention, in step 7), considering the complexity of the UAV take-off and landing point selection optimization and traffic allocation model constraints, as well as the problem that traditional algorithms are prone to getting trapped in local optima, the ALNS algorithm with large neighborhood search characteristics was selected, and four destruction operators and three repair operators in the heuristic initial solution generation strategy were adopted.

[0053] The random destruction operator among the four destruction operators breaks the structure of the current solution by randomly removing a certain proportion of take-off and landing points, thus giving the algorithm the opportunity to explore other possible regions in the solution space, increasing the diversity of solutions and avoiding the algorithm from getting trapped in local optima.

[0054] The cost-benefit ratio-based destruction operator assesses the probability of removal for each take-off and landing point based on its cost-benefit ratio. It employs a roulette wheel strategy to select take-off and landing points to be removed, which increases randomness while ensuring that take-off and landing points with high construction costs and low coverage demand are removed first.

[0055] The long-distance, low-return sabotage operator assesses the probability of removal based on the distance-to-return ratio of each take-off and landing point. It employs a roulette-style sabotage strategy to select take-off and landing points to be removed, which increases randomness while ensuring priority removal of take-off and landing points with longer transport distances and lower coverage demand.

[0056] The facility utilization-based destruction operator is based on the principle of facility utilization. It assesses the probability of removal based on the facility utilization of each take-off and landing point and uses a roulette wheel destruction strategy to select take-off and landing points to be removed. This increases randomness while ensuring that take-off and landing points with low facility utilization are removed first.

[0057] Among the three repair operators, the urgent demand coverage repair operator, after removing the takeoff and landing points using the destruction operator, has a demand coverage rate below the urgent demand coverage rate threshold. At that time, a greedy algorithm is used to select high-coverage take-off and landing points to fill coverage gaps up to the minimum coverage threshold. This is to prevent incurring significant penalties and costs that could undermine the feasibility of a solution.

[0058] The facility utilization-based repair operator selectively adds high-utilization take-off and landing points, enabling more demand to be covered with fewer take-off and landing points, thereby reducing the total number of take-off and landing points and lowering construction cost allocation.

[0059] The cost-optimized repair operator selectively adds takeoff and landing points with high cost-effectiveness, enabling the system to meet basic demand coverage at the lowest cost, thereby reducing construction and transportation costs.

[0060] The solution obtained significantly reduces operating costs while meeting demand coverage and basic constraints. This method provides a high-quality solution for terminal take-off and landing site selection.

[0061] Beneficial effects: Compared with the prior art, the present invention has the following advantages: (1) This invention presents an optimization method for the selection of take-off and landing points for urban low-altitude logistics terminals using unmanned aerial vehicles (UAVs). This method addresses the issues of UAV take-off and landing point selection and traffic allocation. Based on meeting the minimum demand coverage threshold and aiming to minimize total operating costs, a take-off and landing point selection optimization and traffic allocation model is established, taking into account multiple constraints. Considering the characteristics of the take-off and landing point selection optimization and traffic allocation model and the problem itself, a heuristic solution algorithm is designed based on the ALNS (Adaptive Large Neighborhood Search) framework. Running the ALNS algorithm, the optimal urban low-altitude logistics terminal take-off and landing point selection scheme and UAV traffic allocation scheme are solved under complex constraints. This significantly reduces operating costs while meeting demand coverage and basic constraints.

[0062] (2) This invention simultaneously verifies the feasibility of the take-off and landing point location optimization and traffic allocation model and algorithm, making up for the shortcomings of existing methods that are too coarse-grained and difficult to balance demand coverage and total operating cost. Attached Figure Description

[0063] Figure 1 This is a flowchart of the method for optimizing the selection of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs according to the present invention.

[0064] Figure 2 This is a schematic diagram of the urban low-altitude logistics take-off and landing facility system constructed by the present invention.

[0065] Figure 3 This is a schematic diagram of the take-off and landing site selection research area constructed by this invention.

[0066] Figure 4 This is a diagram illustrating the method for establishing and solving the terminal take-off and landing point location model of the present invention.

[0067] Figure 5 This is a flowchart of the ALNS algorithm used in this invention.

[0068] Figure 6 This is the result of the take-off and landing point selection and traffic allocation of the present invention.

[0069] Figure 7 This invention relates to partition requirement coverage analysis.

[0070] Figure 8 This is a diagram showing the service coverage results specific to this invention. Detailed Implementation

[0071] like Figure 1 As shown, the method for optimizing the location of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs of the present invention includes the following steps: Step 1) Establish a three-tiered infrastructure system consisting of city-level low-altitude logistics hubs, regional-level drone take-off and landing points, and terminal take-off and landing points. Figure 2 The drones depart from city-level low-altitude logistics hubs, transit through regional drone take-off and landing fields, and then land precisely at various take-off and landing points, achieving efficient and convenient urban logistics services.

[0072] City-level low-altitude logistics hubs are primary sites in the urban low-altitude logistics take-off and landing facility system, undertaking the important tasks of centralized storage, sorting, and distribution of goods. They are usually located in the largest cargo airport in the city, sub-centers or satellite cities, large logistics parks, or multi-level transportation hubs. Area-level drone take-off and landing sites are key nodes connecting hubs and take-off and landing points, providing stable operational support for drones. They are mostly located in relatively open areas with fewer airspace restrictions on the outskirts or inside the city, such as abandoned industrial parks and large parking lots. Terminal drone take-off and landing points are the most numerous and widely distributed in all corners of the city, including densely populated areas such as residential communities, commercial complexes, and office buildings.

[0073] Step 2) Integrate urban population density data, POI data and questionnaire survey data to establish a quantitative calculation method for urban low-altitude logistics demand.

[0074] First, in step 21), obtain the total volume of regular express delivery business in the urban area. The express delivery volume is the sum of domestic same-city express delivery volume, domestic intercity express delivery volume, and international express delivery volume. The total volume of regular express delivery business for each district can be obtained from the district's government website or statistical bulletin. .

[0075] Step 22), calculate grid heat. And the central location. Spatially correlate POI data with population density data (grid surface) within the grid. Healthcare services and businesses are considered high-importance POIs; the number of high-importance POIs in each grid is calculated, with higher numbers indicating higher grid importance. The population density value of the spatially correlated demand points (grids) is then calculated. and the number of POIs Calculate their respective proportions of the total regional volume. and ,Right now , Then calculate. and The average value is used to obtain the grid heat. Calculate the average longitude and average latitude of all POIs in each analysis grid, and use this as the required center location for that analysis grid.

[0076] Step 23), allocate regular express delivery volume. Distribute the total express delivery volume to each grid cell according to grid heat, and obtain the annual express delivery volume for each grid cell. .

[0077] Step 24), calculate the demand for low-altitude logistics. First, determine the proportion of time-sensitive deliveries made by urban residents in their daily express delivery business, and their intention to choose low-altitude logistics in special scenarios. The average of the proportion of time-sensitive deliveries and the intention to choose low-altitude logistics will be used as the proportion of time-sensitive goods. and low-altitude logistics selection intention . Will( As for high-speed express delivery volume, this represents potential low-altitude demand. The subsequent optimization model uses daily transport volume, dividing the low-altitude demand by the number of days in a year to obtain the daily low-altitude logistics demand for each grid cell. ,Right now .

[0078] For all established take-off and landing points k, the selection distance is less than or equal to the coverage radius. From demand point i, we obtain the set of candidate demand points. If the distance to the nearest takeoff and landing field is greater than a threshold, The demand points are used as demand points in remote areas to obtain a set of demand points in remote areas. In this embodiment, healthcare services and companies are considered as high-importance POIs, and the remaining POIs are considered as ordinary POIs. The number of important POIs in each grid is calculated, sorted in descending order, and the top 10% of the grids are taken as important regional demand points.

[0079] To minimize operating costs while meeting basic demand coverage, three methods for calculating demand coverage are defined: overall demand coverage is the ratio of the actual transport volume met by all demand points within the study area to the actual transport volume required by all demand points; demand coverage in remote areas and demand coverage in important areas are calculated using the same method.

[0080] Design a 20km x 20km square research area, such as... Figure 3 The area was divided into 400 grids, spaced 1km apart, with a total of 400 demand points distributed across them. Each demand point was assumed to be located at the center of a grid. This square area was further divided into four zones: upper left, upper right, lower left, and lower right. No-fly zones and geographically restricted zones were established. After removing candidate points from the no-fly and geographically restricted zones, the four zones had 94, 62, 100, and 100 valid demand points, respectively. Two area-level take-off and landing fields were set up, equipped with 148 and 166 drones, respectively.

[0081] Step 4: Establish a model for optimizing the location of take-off and landing points for terminal UAVs and allocating traffic. The modeling approach is as follows: Figure 4 As shown.

[0082] The optimization of terminal take-off and landing site selection aims to scientifically and rationally determine the location and number of take-off and landing sites, and to formulate a drone traffic allocation plan to minimize operating costs while meeting basic coverage needs. Unlike the selection of area-level take-off and landing sites, which focuses more on demand coverage and construction costs, terminal take-off and landing site selection prioritizes operating costs. On the one hand, as a key node connecting end-point demand, the location of terminal take-off and landing sites directly determines the transportation distance and traffic flow when drones perform delivery tasks, thus directly affecting transportation costs. On the other hand, although the construction scale of terminal take-off and landing sites is relatively small, the number of sites is large, and the construction-allocated cost accounts for a significant proportion of the total daily operating cost. These two factors are interconnected. Lower transportation costs usually mean transportation over shorter distances, requiring the establishment of more take-off and landing sites, leading to an increase in construction-allocated costs. Conversely, lower construction-allocated costs usually mean establishing fewer take-off and landing sites or choosing locations with lower construction costs. The former may result in the inability to complete the scheduled transportation tasks, while the latter may lead to increased delivery distances and higher transportation costs. Therefore, the selection of terminal take-off and landing sites needs to comprehensively consider construction and transportation costs to minimize overall operating costs. A balance between these two factors can be found by rationally planning the location and number of take-off and landing sites and the drone traffic allocation scheme.

[0083] Therefore, the optimization model for take-off and landing site selection involves both the take-off and landing site selection problem and the traffic allocation problem. These two problems are interdependent. The former mainly determines the construction location and number of take-off and landing sites, thus determining the construction cost allocation; the latter determines the transportation cost by determining the allocation of UAV transport traffic. Together, they constitute the operating cost. A penalty term, representing the overall demand coverage rate, is incorporated into the objective function to avoid mandatory requirements that could lead to infeasible solutions during algorithm iteration. Constraints include minimum demand coverage rate constraints in remote areas, expected demand coverage rate constraints in important areas, capacity limits, traffic conservation, and coverage radius constraints.

[0084] The objective function consists of three parts: the first part is the cost of construction and allocation of take-off and landing points; the second part is the transportation cost, which increases with longer transportation distances and higher traffic volumes; and the third part is a penalty term P for unmet demand coverage.

[0085] ;

[0086] in, For alternative take-off and landing sites Construction cost of a single standard capacity landing point; It is an integer variable representing the alternative take-off and landing points. The number of take-off and landing points to be built at the location; Indicates the construction cost-sharing period; For take-off and landing field and take-off and landing points The distance between them; For take-off and landing field Assigned to take-off and landing points Transportation flow; This refers to the cost per flight kilometer during the drone's flight range; This indicates the penalty for unmet requirements.

[0087] The penalty term P for unmet demand coverage causes the terminal UAV take-off and landing point optimization and traffic allocation model to prioritize cost optimization while meeting the basic demand coverage. The calculation formula is as follows:

[0088] ;

[0089] in, This is the minimum demand coverage threshold; This represents the highest demand coverage threshold. Indicates the take-off and landing point Assigned to demand points Transportation flow; It is a demand point The demand for traffic; To ensure demand coverage The following are the penalty factors; To ensure demand coverage to The reward factor for the interval.

[0090] In this embodiment of the invention, the parameters of the terminal UAV take-off and landing point selection optimization and traffic allocation model are shown in Table 1:

[0091] Table 1 Model Parameter Settings ; Step 5) The terminal UAV take-off and landing point selection optimization and traffic allocation model needs to consider multiple constraints such as coverage limitations, capacity limitations, traffic conservation, service radius limitations, and allocation relationship limitations in special areas to ensure the feasibility of the solution: ; ;

[0092] In the above constraint formula, the specific meanings of the symbols are as follows:

[0093] in, To improve coverage of needs in remote areas; This represents the minimum threshold for demand coverage in remote areas. A collection of demand points in remote areas; It is the transportation flow allocated from takeoff and landing point k to demand point i; This represents the demand flow at demand point i; To ensure coverage of demand in key regions; This represents the expected threshold for coverage in key areas. It is a collection of demand points in important regions; This indicates the number of takeoff and landing points to be constructed at candidate takeoff and landing point k; This represents the maximum total number of landing and takeoff points to be constructed. For the capacity of takeoff and landing field j, ; The number of drones at takeoff and landing site j; The number of times the drone operates per day; This refers to the capacity of a single standard takeoff and landing point; Indicates whether the takeoff and landing point k is served by the takeoff and landing field j; Is demand point i served by takeoff and landing point k? It is the distance between the takeoff and landing point k and the takeoff and landing field j; The service coverage radius of the take-off and landing field; Let i be the distance between the demand point i and the take-off and landing point k. It is the service coverage radius of the take-off and landing point; It is a positive number; A value of 1 indicates that the minimum transport volume is 1 flight. This represents the transport flow allocated from landing field j to landing point k. This is a collection of no-fly zone locations; This is a set of geographically restricted locations.

[0094] Furthermore, the constraints in the formula are explained in turn: Specifically, constraint (1) represents the minimum demand coverage rate constraint for remote areas, setting a minimum threshold for demand coverage rate in remote areas to prevent these areas from being overlooked during site selection; constraint (2) represents the expected coverage rate constraint for important areas, setting an expected coverage rate threshold to guide the terminal UAV take-off and landing point site selection optimization and traffic allocation model to tilt resources toward important areas during the optimization process to meet the needs of important areas (such as hospitals and other emergency material demand points); constraint (3) is the total number of take-off and landing point construction constraints, the total number of take-off and landing point constructions shall not exceed the specified maximum number; constraints (4) and (5) are the capacity constraints for take-off and landing fields and take-off and landing points, the traffic allocated by each take-off and landing field and point shall not exceed its own capacity. Constraint (6) is a capacity constraint for take-off and landing fields and points; the flow allocated by each take-off and landing field and point cannot exceed its own capacity. Constraint (7) represents a flow conservation constraint; for any take-off and landing point, the transport flow allocated by the take-off and landing field to that point should be equal to the sum of the transport flows allocated by that point to the surrounding demand points. Constraint (8) is a constraint between take-off and landing points and take-off and landing fields; each take-off and landing point can be served by at most one take-off and landing field. Constraint (9) represents a connection constraint between take-off and landing points and demand points; if a take-off and landing point is established at location k, it can provide services to the surrounding demand points, and vice versa. Constraints (11) and (12) represent... The coverage radius constraint means that the take-off and landing point must be located within the effective service coverage radius of the take-off and landing field. Only demand points within the coverage radius of the take-off and landing point are considered to be effectively covered. Constraints (13)-(16) represent flow constraints. Only when take-off and landing point k is served by take-off and landing field j and demand point i is served by take-off and landing point k will flow (at least 1 flight) be allocated to take-off and landing point k and demand point i. Constraints (17)-(20) are no-fly zones and geographically restricted areas. Take-off and landing points cannot be built, and demand within them cannot be covered. Constraints (21)-(24) describe the range of values ​​for the variables.

[0095] Step 6) Incorporate a traffic allocation strategy into the ALNS framework to achieve coordinated optimization of site selection and traffic allocation. This traffic allocation strategy centers on multi-stage priority allocation and dynamic capacity coordination. It prioritizes traffic demand in remote and important areas through tiered coverage, minimizes transportation costs based on a cost-benefit scoring model, and ensures rational resource utilization through dynamic capacity tracking. The process is as follows:

[0096] 61) Screening of existing take-off and landing points ( >0), to obtain the set Clear its existing allocation relationship (The allocation relationship between take-off and landing points and demand points) (Flow between take-off and landing points and demand points) and (Flow rate between the landing field and the landing point). Initialize the landing point and landing field capacity. Landing point capacity: Remaining capacity Takeoff and landing field capacity: Remaining capacity Demand points and required traffic: Traffic to be covered in demand areas: .

[0097] 62) For each takeoff and landing point k, connect to the nearest takeoff and landing field j. Check the remaining capacity of the current takeoff and landing field; if... Then establish a connection. and update If capacity is insufficient, select the next nearest takeoff and landing site with sufficient capacity.

[0098] 63) For all existing take-off and landing points k, the screening distance shall not exceed the coverage radius. From demand point i, we obtain the set of candidate demand points. Select the demand points in remote areas and establish a set of remote areas. The key regional demand points are then selected to create a set of key regions. .

[0099] 64) Calculate the traffic required to reach coverage thresholds in remote and important areas. and , , ,in It is the minimum coverage threshold for remote areas. It is the threshold for expected demand coverage in important regions.

[0100] 65) Calculate the priority score for each takeoff and landing point. , .in This represents the priority score for remote areas at takeoff and landing point k, primarily considering the demand in remote areas that this takeoff and landing point can cover. The priority score for important areas is determined by the area that the take-off and landing point k can cover, which mainly considers the demand in important areas that the take-off and landing point can cover and the distance between the take-off and landing point and its respective take-off and landing field. Let i be the distance between the demand point i and the take-off and landing point k. The distance between takeoff and landing point k and takeoff and landing field j; finally, the flow of traffic in remote and important areas is allocated.

[0101] 66) Greedy allocation of remaining traffic; first calculate the remaining traffic to be allocated; second, update the takeoff and landing point value scores. ,in Then, traffic flow is allocated. The take-off and landing points are arranged according to... Sort in descending order, process each start and stop point k in turn, traverse the remaining demand points i covered by k, and sort by... Sort in descending order. Assign transport flights to each demand point i. Update the capacity of take-off and landing points and take-off and landing fields, as well as the traffic flow between take-off and landing fields and take-off and landing points. Update the remaining uncovered traffic flow at the demand points to avoid over-allocation.

[0102] 67) Verify the constraints to ensure and Logical consistency is ensured. The sum of the traffic allocated to demand points from each takeoff and landing point equals the traffic allocated to that takeoff and landing point by the landing yard. If a takeoff and landing point ultimately receives no traffic allocation, construction is cancelled, and settings are updated accordingly. and delete the related links ( ).

[0103] Step 7) In view of the characteristics of the take-off and landing point selection problem and the traffic allocation problem, a heuristic solution algorithm is designed based on the adaptive large neighborhood search (ALNS) framework to make it more suitable for the collaborative optimization problem of the present invention.

[0104] Multi-level traffic allocation strategy: A traffic allocation strategy is added to the ALNS framework to achieve coordinated optimization of site selection and traffic allocation. This traffic allocation strategy is based on multi-stage priority allocation and dynamic capacity coordination. It prioritizes traffic demand in remote and important areas through hierarchical coverage, minimizes transportation costs based on a cost-benefit scoring model, and ensures rational resource utilization through dynamic capacity tracking.

[0105] Heuristic initial solution generation strategy: The initial solution generation combines a greedy method and a random generation strategy. The greedy strategy quickly improves the coverage rate of special areas by calculating the value score of take-off and landing points. The random generation strategy supplements the random selection of take-off and landing points after the special coverage requirements are met, thereby enhancing the diversity of the solution space. The combination of the two achieves a balance between efficiency and diversity, enabling the algorithm to obtain a high-quality initial solution from the beginning.

[0106] Diverse destruction-repair strategies: Based on the terminal UAV take-off and landing point optimization and traffic allocation model, four complementary destruction operators and three repair operators were designed, forming 12 combinations. A weighting mechanism was used to enable the ALNS algorithm to automatically focus on efficient operators to deal with different problem stages, enhance the algorithm's exploration ability, and avoid getting trapped in local optima.

[0107] The destruction operators are as follows: 1) Random Disruption Operator: By randomly removing a certain proportion of take-off and landing points, the structure of the current solution is broken, thus giving the algorithm the opportunity to explore other possible regions in the solution space, increasing the diversity of solutions and avoiding the algorithm from getting trapped in local optima.

[0108] 2) Cost-effectiveness-based destruction operator: The probability of removal is evaluated based on the cost-effectiveness of each take-off and landing point. A roulette wheel strategy is used to select take-off and landing points to be removed, which increases randomness while ensuring that take-off and landing points with high construction costs and low coverage demand are removed first.

[0109] 3) Long-distance, low-return sabotage operator: The probability of removal is evaluated based on the distance-to-return ratio of each take-off and landing point. A roulette-style sabotage strategy is used to select take-off and landing points to be removed. This increases randomness while ensuring that take-off and landing points with longer transportation distances and lower coverage demand are removed first.

[0110] 4) Damage operator based on facility utilization: Based on the principle of facility utilization, the probability of removal is assessed according to the facility utilization of each take-off and landing point. A roulette wheel damage strategy is used to select the take-off and landing points to be removed, which increases randomness while ensuring that take-off and landing points with low facility utilization are removed first.

[0111] The repair operators are as follows: 1) Emergency Demand Coverage Repair Operator: After removing take-off and landing points using the destruction operator, if the demand coverage rate is lower than the emergency demand coverage rate threshold, a greedy algorithm is used to select take-off and landing points with high coverage to fill the coverage gap to the minimum coverage rate threshold, thus preventing large penalty costs and compromising the feasibility of the solution.

[0112] 2) Repair operator based on facility utilization: By selectively adding take-off and landing points with high utilization, more demand can be covered with fewer take-off and landing points, thereby reducing the total number of take-off and landing points and reducing construction cost allocation.

[0113] 3) Cost-optimized repair operator: This operator selectively adds take-off and landing points with high cost-effectiveness ratios, enabling the system to meet basic demand coverage at the lowest cost, thereby reducing construction and transportation costs.

[0114] Step 7) employs the heuristic solution algorithm proposed in Step 6, based on the Adaptive Large Neighborhood Search (ALNS) framework. Through heuristic initial solution generation strategies, multi-level traffic allocation strategies, and diverse destruction and repair strategies, it explores the optimal addressing scheme under complex constraints. The algorithm flowchart is shown below. Figure 5 .

[0115] 1) First, initialize the parameters and generate an initial feasible solution, setting the current initial solution as the optimal solution; 2) Then, enter the iterative search phase, selecting a destruction operator based on probability in each iteration to remove some take-off and landing points in the current solution, and then selecting a repair operator to reconstruct the solution; 3) Apply a flow allocation algorithm to the current solution to obtain a flow allocation scheme; 4) Evaluate the objective function value of the new solution, decide whether to accept the new solution based on the Metropolis criterion, and update the records of the current solution and the optimal solution; 5) Adjust the operator weights and reduce the system temperature, etc.; 6) Check the termination condition and return the optimal solution.

[0116] The ALNS algorithm is used to solve the problem, and the algorithm parameters are set as shown in Table 2: Table 2 ALNS Algorithm Parameter Settings

[0117] The results of the take-off and landing point location selection and flow allocation obtained from the solution are as follows: Figure 6 As shown, the objective function value is 92,083.09 yuan, of which the construction cost is 27,701.88 yuan / day and the transportation cost is 64,381.21 yuan / day. A total of 46 take-off and landing sites were constructed, with a demand coverage rate of 62.65%. Take-off and landing site number 166 was allocated 1,125 transport flights, with a flow allocation rate of 95.02%, while take-off and landing site number 250 was allocated 1,237 transport flights, with a flow allocation rate of 93.15%.

[0118] Figure 7 In the study area, the coverage rate of zone 1 was 53.50%, zone 2 was 64.41%, zone 3 was 68.30%, and zone 4 was 64.09%. It can be seen that the service coverage rate among different zones is relatively balanced and the distribution of take-off and landing points is relatively even, ensuring the fairness of service while meeting the needs of most areas in the urban area.

[0119] Figure 8 This example demonstrates coverage examples in remote and key areas. Of the 356 valid demand points in this case study, 35 were key demand points and 76 were remote demand points. The demand satisfaction rate for key demand points reached 78.69%, and the demand satisfaction rate for remote areas reached 23.79%, both meeting the pre-set coverage standards.

[0120] This method provides cost-effective take-off and landing site selection and traffic allocation schemes for urban low-altitude logistics systems, while meeting basic demand coverage requirements, satisfying pre-set constraints, giving priority to remote and important areas, and providing high-quality solutions.

Claims

1. A method for optimizing the location of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs, characterized in that, Includes the following steps: 1) Establish a three-tiered facility system and research object consisting of city-level low-altitude logistics hubs, regional UAV take-off and landing sites, and terminal take-off and landing points, with grids as the research unit; 2) Obtain the total annual express delivery volume in the urban area, and use a quantitative calculation method for urban low-altitude logistics demand to determine the daily low-altitude logistics demand for each grid cell. ; 3) Construct a set of spatial geographic information data and candidate terminal take-off and landing points, and construct a set of candidate demand points, a set of demand points in remote areas, and a set of demand points in important areas based on the daily low-altitude logistics demand of each grid. 4) With the goal of minimizing total operating costs, a terminal UAV take-off and landing point optimization and traffic allocation model is established using the objective function minE; ; in, For alternative take-off and landing sites Construction cost of a single standard capacity landing point; It is an integer variable representing the alternative take-off and landing points. The number of take-off and landing points to be built at the location; Indicates the construction cost-sharing period; For take-off and landing field and take-off and landing points The distance between them; For take-off and landing field Assigned to take-off and landing points Transportation flow; This refers to the cost per flight kilometer during the drone's flight range; This indicates the penalty for unmet requirements. ； in, This is the minimum demand coverage threshold; This represents the highest demand coverage threshold. Indicates the take-off and landing point Assigned to demand points Transportation flow; It is a demand point The demand for traffic; To ensure demand coverage The following are the penalty factors; To ensure demand coverage to The reward factor for the interval; 5) Set constraints such as coverage limits, capacity limits, traffic conservation, service radius limits, and allocation relationship limits for special areas, and determine the number and location of take-off and landing points and the traffic of equipped drones; 6) A multi-level traffic allocation algorithm is used for traffic allocation. The process is as follows: 61) Initialize the takeoff and landing point capacity and the takeoff and landing field capacity. Takeoff and landing point capacity: Remaining capacity at take-off and landing points Takeoff and landing field capacity: Remaining capacity of the take-off and landing field Demand points and required traffic: Traffic to be covered in demand areas: ; 62) Connect each takeoff and landing point k to the nearest takeoff and landing field j; if the current takeoff and landing field has remaining capacity... Then establish a connection. and update Otherwise, choose the next nearest takeoff and landing field with sufficient capacity. 63) Filter the distance to the established take-off and landing point k ≤ coverage radius. From demand point i, we obtain the set of candidate demand points. ; Filter out demand points in remote areas and demand points in important areas, and establish a set of remote areas. and important regions ; 64) Calculate the traffic required to reach the coverage threshold in remote areas. Traffic required to reach coverage thresholds in key areas ,in This represents the minimum coverage threshold for remote areas. The threshold for expected demand coverage in key areas; 65) Calculate the priority score for each takeoff and landing point. , ;in For remote areas of takeoff and landing point k, the priority score is... Priority score for important areas of takeoff and landing point k; Let i be the distance between the demand point i and the take-off and landing point k. The distance between takeoff and landing point k and takeoff and landing field j; allocate traffic flow between remote areas and important areas; 66) Calculate the remaining unallocated traffic and update the takeoff and landing point value scores. ,in ; then allocate traffic flow; and allocate take-off and landing points according to Sort in descending order, traverse the remaining demand points i covered by each start / stop point k, and sort by... Sort in descending order; allocate transport flights to each demand point i. Update the capacity of take-off and landing points and take-off and landing fields, as well as the traffic flow between take-off and landing fields and take-off and landing points; update the remaining uncovered traffic flow at the demand points. 67) Verify the constraints in step 5) to ensure that the sum of the flow rates allocated from the take-off and landing points to the demand points is equal to the flow rates allocated from the take-off and landing field to the take-off and landing points; 7) Optimal addressing is achieved using a heuristic initial solution generation strategy, a multi-level traffic allocation strategy, and diverse destruction and repair strategies. The process is as follows: 71) First, initialize the set of destruction operators D and their corresponding weights. Initialize the set of repair operators R and their corresponding weights. 72) In each iteration of the iterative search phase, a destruction operator is selected based on probability to remove some take-off and landing points in the current initial feasible solution, and then a repair operator is selected to reconstruct the initial feasible solution; 73) A flow allocation scheme is obtained by using a flow allocation algorithm on the current initial feasible solution; 74) The objective function value of the new solution is determined, and a decision is made on whether to accept the new solution based on the Metropolis criterion, and the records of the current solution and the optimal solution are updated; 75) The weights of the destruction operator and the repair operator are adjusted and the system temperature is reduced; 6) Termination condition check: if the number of iterations n < If the condition is met, return to step 72; otherwise, return the optimal solution.

2. The method for optimizing the location of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs according to claim 1, characterized in that: Step 2) is as follows: 21) Obtain the total volume of express delivery business in the urban area. Total volume of express delivery business in urban areas This represents the sum of intra-city express delivery volume, inter-city express delivery volume, and international express delivery volume. 22) Spatially correlate POI data with population density data in a raster and take the population density percentage. Percentage of POIs The mean value is the grid heat. The average longitude and average latitude of all POIs in the calculated raster are used as the required center location of the raster. 23), based on grid heat The total volume of urban express delivery services The data is allocated to each grid cell to obtain the annual express delivery volume for each cell. ; 24), by formula The daily low-altitude logistics demand for each grid cell is calculated. ;in, For the proportion of high-time-delivery goods, Selecting candidates for low-altitude logistics The number of days in a year.

3. The method for optimizing the location of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs according to claim 2, characterized in that: In step 2), the population density percentage Percentage of POIs The grid heat .

4. The method for optimizing the location of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs according to claim 1, characterized in that: The constraints in step 5) are: ； ； in, To improve coverage of needs in remote areas; This represents the minimum threshold for demand coverage in remote areas. A collection of demand points in remote areas; It is the transportation flow allocated from takeoff and landing point k to demand point i; This represents the demand flow at demand point i; To ensure coverage of demand in key regions; This represents the expected threshold for coverage in key areas. It is a collection of demand points in important regions; This indicates the number of takeoff and landing points to be constructed at candidate takeoff and landing point k; This represents the maximum total number of landing and takeoff points to be constructed. For the capacity of takeoff and landing field j, ; The number of drones at takeoff and landing site j; The number of times the drone operates per day; This refers to the capacity of a single standard takeoff and landing point; Indicates whether the takeoff and landing point k is served by the takeoff and landing field j; Is demand point i served by takeoff and landing point k? It is the distance between the takeoff and landing point k and the takeoff and landing field j; The service coverage radius of the take-off and landing field; Let i be the distance between the demand point i and the take-off and landing point k. It is the service coverage radius of the take-off and landing point; It is a positive number; A value of 1 indicates that the minimum transport volume is 1 flight. This represents the transport flow allocated from landing field j to landing point k. This is a collection of no-fly zone locations; This is a set of geographically restricted locations.

5. The method for optimizing the location of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs according to claim 4, characterized in that: In step 5), constraint (1) represents the minimum demand coverage rate constraint for remote areas, setting a minimum threshold for demand coverage rate in remote areas; constraint (2) represents the expected coverage rate constraint for important areas, setting an expected coverage threshold; and constraint (3) represents the total number of take-off and landing points to be constructed.

6. The method for optimizing the location of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs according to claim 4, characterized in that: The range of values ​​for the variables in step 5) is: , , , , .

7. The method for optimizing the location of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs according to claim 1, characterized in that: Step 71), initialize the set of destruction operators D and their corresponding weights. Initialize the set of repair operators R and their corresponding weights. Generate an initial feasible solution: Filter valid candidate take-off and landing points, greedily select take-off and landing points to meet the minimum coverage requirements in remote areas and the expected coverage requirements in important areas; randomly select take-off and landing points to meet the overall coverage requirements, and obtain the current solution. ; Allocate traffic to the current solution; Check the termination condition; if the demand coverage target is met, complete the initial solution generation. Assign to Conversely, perform incremental supplementation and set the initial solution. This is the current optimal solution.

8. The method for optimizing the location of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs according to claim 1, characterized in that: Step 74): Determine the objective function value of the new solution, decide whether to accept the new solution based on the Metropolis criterion, and update the current solution and optimal solution records; based on weights... Select the destruction operator Applying the destruction operator: Select the repair operator Apply the repair operator: According to step 6), traffic allocation is performed on candidate solutions; based on the Metropolis criterion, it is determined whether to accept a candidate solution. , , = ,if or And with probability Accept the new interpretation. , = Conversely, the number of iterations .

9. The method for optimizing the location of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs according to claim 1, characterized in that: In step 3), for the established take-off and landing points k, the filter is performed with a distance ≤ coverage radius. From demand point i, we obtain the set of candidate demand points. If the distance to the nearest takeoff and landing field is greater than a threshold, The demand points are used as demand points in remote areas to obtain a set of demand points in remote areas. The number of important POIs in each calculated raster is sorted in descending order to determine the demand points in important regions.

10. The method for optimizing the location of take-off and landing points and allocating traffic flow for urban low-altitude logistics terminal UAVs according to claim 1, characterized in that: In step 67), if there is no flow allocation at the take-off and landing points, then set... And delete the related links, i.e. .

Images