A logistics unmanned aerial vehicle route layering method based on main heading extraction

Abstract

The application discloses a logistics unmanned aerial vehicle route layering method based on main direction extraction, and comprises the following steps: (1) performing cluster analysis on a logistics unmanned aerial vehicle transport route according to logistics data, and identifying a transport trunk; (2) extracting a main direction and a secondary direction by analyzing the direction of the transport trunk, and performing height layering on a flight airspace; (3) matching different directions with different height layers based on the height layering of the airspace, matching the trunk contained in different directions with corresponding height layers, and finally matching the routes contained in different trunks with corresponding height layers. The main direction extracted by the application can guide the orderly and safe layering of each route, and provides effective theoretical support for reasonably managing unmanned aerial vehicle operation and reducing unmanned aerial vehicle collision risks.

Description

Technical Field

[0001] This invention relates to the fields of logistics drone management and data mining, specifically to a method for hierarchical analysis of logistics drone routes based on the extraction of primary headings. Background Technology

[0002] With the gradual opening of low-altitude airspace and the rapid development of the internet, drones have come into our view and are increasingly being used in the logistics field, becoming an effective means of solving the "last mile" problem. More and more domestic and foreign logistics companies are beginning to try using small drones for express delivery. Drone logistics is already one of the emerging fields of drone application, and logistics drones are gradually becoming a popular way to deliver parcels. Countries around the world are carrying out technical research on logistics drone delivery.

[0003] The rapid development of the drone industry has led to an increasing density of drone flights in low-altitude airspace, bringing a series of safety hazards, such as uncontrolled crashes and mid-air collisions. These incidents severely impact the operations of logistics drone companies and threaten public safety. Currently, the management methods for logistics drones are not mature enough, and relevant laws and regulations are incomplete, making risk management a major challenge. Logistics drone flight missions consist of several phases: takeoff, vertical climb, level flight, vertical descent, and landing. The main collision risk occurs during level flight, where logistics drones maintain a certain altitude. Therefore, dividing the low-altitude airspace where logistics drones operate into different altitude layers is an effective way to reduce the risk of mid-air collisions and improve operational safety. Summary of the Invention

[0004] Purpose of the invention: In view of the lack of low-altitude altitude layering technology, this invention proposes a method for layering logistics drone flight routes based on the extraction of main headings. By mining and analyzing the operational data of logistics drones, it provides an effective basis for the reasonable division of logistics drone flight route altitude layers.

[0005] Technical solution: A method for hierarchical routing of logistics drones based on main heading extraction, comprising the following steps:

[0006] (1) Cluster analysis of logistics drone transportation routes based on logistics data to identify the main transportation routes;

[0007] (2) By analyzing the direction of the main transport flow, the primary and secondary routes are extracted, and the flight airspace is divided into altitude layers.

[0008] (3) Based on the airspace height layering, different headings are matched to different height layers, the main streams contained in different headings are matched to the corresponding height layers, and finally the routes contained in different main streams are matched to the corresponding height layers.

[0009] Furthermore, in step (1), each logistics delivery route is designed as a logistics drone transportation route X, and all routes are clustered into several logistics drone transportation trunk lines Y.

[0010] X = {x1, x2, ..., x} i}

[0011] Y = {y1, y2, ..., y} j}

[0012] Where X is the set of all routes; Y is the set of all main rivers.

[0013] Furthermore, the cluster analysis in step (1) includes: KNN method, K-means method, density clustering method, hierarchical clustering method and spectral clustering method.

[0014] Furthermore, in step (2), the main stream y formed by clustering the s routes... j Its direction angle α j The calculation method is as follows

[0015]

[0016] Where β i For route x i The direction angle, ε is the angle in any direction;

[0017] According to each main stream y j Direction α j And number of routes q j Extract the main flight path of logistics drone transportation. Heading angle of the main course The calculation methods include unweighted calculation methods and weighted calculation methods; the heading angle of the main heading. The unweighted calculation method is

[0018]

[0019] Heading angle of the main course The weighted calculation method is as follows

[0020]

[0021] According to the main course Determine the secondary course And define the opposite direction of the primary and secondary headings as the primary heading counter-heading. and secondary heading counter-heading

[0022] To minimize transport conflicts between primary and secondary routes, the following constraints must be met:

[0023]

[0024]

[0025] That is, the difference between the secondary course and the primary course is not less than θ, where θ is the minimum course difference and the range of θ is 45° to 135°.

[0026] Furthermore, in step (2), the flight altitude range of the logistics drone is restricted according to the airspace management regulations, and the airspace in which the logistics drone can fly is divided into altitude layers, with each altitude layer being equally spaced.

[0027] The flight altitude limit for logistics drones is:

[0028] H min ≤h i ≤H max

[0029] In the formula h i For the drone to complete the flight path x i Flight altitude during operation, H min H is the minimum flight altitude for the drone. max This is the maximum flight altitude of the drone.

[0030] Furthermore, step (3) includes the following steps:

[0031] 3.1 Based on airspace altitude layering, different altitude layers are assigned to different headings;

[0032] To reduce relative conflicts during drone operation and ensure orderly drone flight, the extracted different headings are assigned to different altitude layers.

[0033]

[0034] in Let H1, H2, H3...H be the altitude layers extracted for the i-th heading. n For each height level;

[0035] 3.2 Match the main currents contained in different headings with corresponding altitude stratifications;

[0036] Each course is formed by clusters of several main streams, and different main streams within the same course are matched with the same altitude layer.

[0037]

[0038] In the formula The height layer of the main current belonging to the i-th heading;

[0039] 3.3 Match the routes included in different main rivers with corresponding altitude layers;

[0040] Each main stream is formed by clusters of several routes. Different routes clustered into the same main stream are matched with the same altitude layer.

[0041]

[0042] In the formula The height layer is the route belonging to the i-th main stream.

[0043] Beneficial Effects: This invention uses logistics data to cluster and analyze the transportation routes of logistics drones, identifying the main transportation flows. By analyzing the direction of these flows, it extracts primary and secondary flight paths, hierarchically stratifies the flight airspace, and matches different routes with corresponding altitude layers. This provides a feasible method for enterprises to conduct drone transportation in an orderly manner. This invention can extract the primary flight paths of logistics drone transportation, rationally allocate different altitude layers to each route while maintaining vertical spacing, reduce the collision risk of logistics drones during transportation, and provide reliable theoretical support for using data mining to help enterprises manage efficiently. Attached Figure Description

[0044] Figure 1 This is a flowchart of the logistics drone route layering method based on main heading extraction according to the present invention. Detailed Implementation

[0045] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0046] Figure 1 The diagram shows a flowchart of the present invention for assigning altitude layers to each UAV transport route. The present invention is based on the hierarchical layering of logistics UAV routes extracted from the main heading and mainly includes the following steps:

[0047] Step (1): Based on logistics data, cluster analysis is performed on the logistics drone transportation routes to identify the main transportation flow. This paper uses the line segment from the starting point to the destination of the logistics drone trip as the transportation route. Based on the logistics data, the route number, departure time, departure longitude, departure latitude, arrival time, arrival longitude, and arrival latitude can be obtained. The departure longitude and departure latitude constitute the route starting point, and the arrival latitude and arrival longitude constitute the route ending point. Let the research start time be t0, the research period be T, and the research time period be (t0, t0+T). Each logistics delivery path within the research time period is designed as a logistics drone transportation route X, and then the routes are clustered into the logistics drone transportation main flow Y. The cluster analysis includes KNN, K-means, density clustering, hierarchical clustering, and spectral clustering.

[0048] X = {x1, x2, ..., x} i}

[0049] Y = {y1, y2, ..., y} j}

[0050] Where X is the set of all routes; Y is the set of all main rivers.

[0051] Step (2) By analyzing the direction of the main transport flow, the primary and secondary headings are extracted, and the airspace is stratified by altitude.

[0052] For the main stream y formed by clustering s routes j Its direction angle α j The calculation method is as follows

[0053]

[0054] To minimize the average deviation angle of the route from its main current, where β i For route x i The direction angle, ε is the angle in any direction.

[0055] Then, based on each main stream y j Direction α j And number of routes q j Extract the main flight path of logistics drone transportation. Heading angle of the main course The calculation methods include unweighted calculation methods and weighted calculation methods; the heading angle of the main heading. The unweighted calculation method is

[0056]

[0057] Heading angle of the main course The weighted calculation method is as follows

[0058]

[0059] According to the main course Determine the secondary course The primary heading is the direction with the highest drone transport volume, and the secondary heading is the direction with the second highest transport volume. The opposite direction between the primary and secondary headings is defined as the primary heading counter-heading. and secondary heading counter-heading

[0060] To reduce the reverse collision between the UAV's primary and secondary headings, the following constraints must be satisfied:

[0061]

[0062]

[0063] That is, the difference between the secondary heading and the primary heading angle is not less than θ, where θ is the minimum heading difference and the range of θ is 45° to 135°.

[0064] According to airspace management regulations, the flight altitude range of logistics drones is restricted, and the airspace in which logistics drones can fly is divided into altitude layers, with each altitude layer being equally spaced.

[0065] The flight altitude limit for logistics drones is:

[0066] H min ≤h i ≤H max

[0067] In the formula h i For the drone to complete the flight path x i Flight altitude during operation, H min H is the minimum flight altitude for the drone. max This represents the maximum flight altitude for the drone. In actual transport, to avoid obstacles, the minimum flight altitude is set at 60m; however, since regulations restrict drones to flight below 120m, the maximum flight altitude is set at 120m. Therefore, the drone's altitude is limited to 60m to 120m. Within this 60m flight airspace, the airspace can be divided into several altitude levels based on the set altitude of each level.

[0068] Step (3) Based on the airspace altitude layering, different headings are matched to different altitude layers, the main streams contained in different headings are matched to the corresponding altitude layers, and finally the routes contained in different main streams are matched to the corresponding altitude layers.

[0069] The specific steps include:

[0070] (31) Based on the airspace altitude layering, different altitude layers are assigned to different headings;

[0071] To reduce relative conflicts during drone operation and ensure orderly drone flight, the extracted different headings are assigned to different altitude layers.

[0072]

[0073] in Let H1, H2, H3...H be the altitude layers extracted for the i-th heading. n For each height level.

[0074] Select the main course Secondary course and the opposite direction of the main course There are three headings, each assigned a different altitude level. Since the main heading has the highest drone traffic, to minimize flight conflicts between other headings and the main heading, the main heading... Assigned to the highest altitude layer; to reduce head-on collisions between drones, the main heading will be reversed. Assigned to the lowest altitude layer, which reduces the risk of collision with drones taking off and landing in the primary heading; secondary heading This is then assigned to the altitude layer between the main heading altitude layer and the main heading-opposite heading altitude layer. This divides the 60-meter airspace into three altitude layers, each with an altitude range of 20 meters.

[0075] Of course, if the flow rate in the secondary course reaches a certain threshold, the following conditions are met:

[0076] f s ≥λ·f p

[0077] Then a secondary heading and reverse heading need to be selected, where f s For secondary heading flow; f p λ represents the flow rate along the main course; λ is the flow coefficient, λ∈(0,1).

[0078] The specific flow coefficient is set by the enterprise based on actual logistics data statistics. For example, if λ = 0.8, and the flow on the secondary route reaches 80% of the flow on the primary route, then it is necessary to select a secondary route in reverse. Furthermore, the airspace was further refined by adding a new heading-opposite heading altitude layer in the intermediate altitude layer to reduce flight conflicts between UAVs. This resulted in the airspace being divided into four altitude layers, each with an altitude range of 15 meters.

[0079] (32) Match the main streams contained in different headings with corresponding height layers;

[0080] Each course is formed by clusters of several main streams, and different main streams within the same course are matched with the same altitude layer.

[0081]

[0082] In the formula It represents the height layer of the main current belonging to the i-th heading.

[0083] (33) Match the routes contained in different main streams with corresponding altitude layers;

[0084] Each main stream is formed by clusters of several routes. Different routes clustered into the same main stream are matched with the same altitude layer.

[0085]

[0086] In the formula The height layer is the route belonging to the i-th main stream.

[0087] The technical solution of the present invention will be further illustrated below with specific embodiments.

[0088] The research period T is one day (i.e., 24 hours). The case study object is a logistics drone, and 10 pieces of logistics data are acquired. These are designed into 10 flight routes, X = {x1, x2, ..., x...} 10 The relevant route information is shown in Table 1.

[0089] Table 1. Flight route information for the study period, August 20, 2022

[0090]

[0091]

[0092] 1. Based on the logistics data in the example, perform cluster analysis on the logistics drone transportation routes to identify the main routes.

[0093] Using the commonly used K-means clustering method, every two routes are clustered into one main flow. The above 10 routes are clustered together to form 5 transport main flows, with main flows Y = {y1, y2, ..., y5}. The corresponding routes are shown in Table 2.

[0094] Table 2. Correspondence between each main river and each route

[0095] <![CDATA[Main stream number y i > <![CDATA[Flight Route Number x i > 1 2，5 2 1，6 3 4，9 4 7，8 5 3，10

[0096] The routes corresponding to y1 are x2 and x5; the routes corresponding to y2 are x1 and x6; the routes corresponding to y3 are x4 and x9; the routes corresponding to y4 are x7 and x8; ​​and the routes corresponding to y5 are x3 and x9. 10 .

[0097] 2. By analyzing the directions of the five main transport routes, the KNN flow clustering method is used to extract the primary and secondary headings, and the airspace is stratified by altitude.

[0098] The method for calculating the direction angle α1 of y1 is as follows:

[0099]

[0100] Where β i Let x2 and x5 be the heading angles of the routes, and ε be the angle in any direction. The heading angles of other main currents can be calculated similarly.

[0101] Further analysis of the directions of the five main streams reveals the primary headings. In this embodiment, no weight is applied to the number of routes on the main streams; only the sum of the deviation angles of the main streams from their respective headings is minimized. The calculation method is as follows

[0102]

[0103] Calculate the main heading It is 100°. Based on the main heading. Determine the secondary course The primary heading is the direction with the highest drone transport volume, and the secondary heading is the direction with the second highest transport volume. The opposite directions of the primary and secondary headings are defined as the primary heading reverse headings. and secondary heading counter-heading To reduce adverse collisions between UAVs on the primary and secondary routes, a certain course difference is established between the primary and secondary routes to divert flight paths, satisfying the following constraints:

[0104]

[0105]

[0106] That is, the difference between the secondary course and the primary course angle is not less than θ, where θ is the minimum course difference; in this case, θ is taken as 60°. The secondary course can be calculated based on the route information and constraints. The angle is 210°, therefore it can be determined that the main heading is in the opposite direction. 280°, secondary heading counter-heading It is 30°.

[0107] In this case, the main course Corresponding to the three main streams y1, y2, and y3, the secondary headings This corresponds to the two main streams, Y4 and Y5.

[0108] According to airspace management regulations, the flight altitude range of logistics drones is restricted, and the airspace in which logistics drones can fly is divided into altitude layers with equal intervals between each layer.

[0109] The flight altitude limit for logistics drones is:

[0110] H min ≤h i ≤H max

[0111] In the formula h i For the drone to complete the flight path x i Flight altitude during operation, H min H is the minimum flight altitude for the drone. max This represents the maximum flight altitude for the drone. In actual transport, to avoid obstacles, the minimum flight altitude is set at 60m; however, since regulations restrict drones to flight below 120m, the maximum flight altitude is set at 120m. Therefore, the drone's altitude is limited to 60m to 120m. Within this 60m flight airspace, the airspace can be divided into several altitude levels based on the set altitude of each level.

[0112] 3. Based on airspace altitude layering, different headings are matched to different altitude layers, then the main streams contained in different headings are matched to the corresponding altitude layers, and finally the routes contained in different main streams are matched to the corresponding altitude layers.

[0113] (31) Based on the airspace altitude layering, different altitude layers are assigned to different headings;

[0114] To reduce relative conflicts during drone operation and ensure orderly drone flight, the extracted different headings are assigned to different altitude layers.

[0115]

[0116] in Let H1, H2, H3...H be the altitude layers extracted for the i-th heading. n For each height level.

[0117] Select the main course Secondary course and the opposite direction of the main course There are three headings, each assigned a different altitude level. Since the main heading has the highest drone traffic, to minimize flight conflicts between other headings and the main heading, the main heading... Assigned to the highest altitude layer; to reduce head-on collisions between drones, the main heading will be reversed. Assigned to the lowest altitude layer, which reduces the risk of collision with drones taking off and landing in the primary heading; secondary heading This is then assigned to the altitude layer between the main heading altitude layer and the main heading-opposite heading altitude layer. This divides the 60-meter airspace into three altitude layers, each with an altitude range of 20 meters.

[0118] Of course, if the flow in the secondary heading reaches a certain threshold, in this case, to reach 80% of the flow in the primary heading, it is necessary to select an additional secondary heading in the opposite direction. Furthermore, the airspace was further refined by adding a new heading-opposite heading altitude layer in the intermediate altitude layer to reduce flight conflicts between UAVs. This resulted in the airspace being divided into four altitude layers, each with an altitude range of 15 meters.

[0119] In this case, there are 6 primary routes and 4 secondary routes. The traffic volume in the secondary routes is less than 80% of that in the primary routes, so there is no need to select additional secondary routes with opposite traffic flow. Therefore, the airspace from 60m to 120m is divided into 3 altitude layers, with the primary route... Assigned to the first altitude level, ranging from 100m to 120m; secondary heading Assign a second altitude level, ranging from 80m to 100m; main heading reverse course. A third height level is allocated, with a height range of 60m to 80m.

[0120]

[0121] (32) Match the main streams contained in different headings with corresponding height layers;

[0122] Each course is formed by clusters of several main streams, and different main streams within the same course are matched with the same altitude layer.

[0123]

[0124] In the formula It represents the height layer of the main current belonging to the i-th heading.

[0125] In this example, the main streams y1, y2, and y3 correspond to the primary course. Assigned to the primary heading altitude layers, with the main current's Y4 and Y5 corresponding to the secondary headings. Assigned to a secondary heading altitude level.

[0126]

[0127] (33) Match the routes contained in different main streams with corresponding altitude layers;

[0128] Each main stream is formed by clusters of several routes. Different routes clustered into the same main stream are matched with the same altitude layer.

[0129]

[0130] In the formula The height layer is the route belonging to the i-th main stream.

[0131] In this case, two air routes correspond to one main current. Therefore, each main current altitude layer contains the altitude layers of the two air routes. Each air route altitude layer has a corresponding main current altitude layer, and each main current altitude layer has a corresponding heading altitude layer. Each heading altitude layer is allocated a pre-defined airspace altitude layer. This process is repeated layer by layer from the air routes to the designated altitude layers. The altitude layers for each air route are shown in Table 3.

[0132] Table 3 Flight Line Height Levels

[0133] <![CDATA[Flight Route Number x i > <![CDATA[Height layer H i > 1 1 2 1 3 2 4 1 5 1 6 1 7 2 8 2 9 1 10 2

[0134] Big data provides the ability to continuously observe large samples of urban logistics, which is particularly important for the supervision and control of complex adaptive systems. By mining and analyzing logistics drone transportation data and stratifying flight routes, an effective management method is provided for enterprises.

[0135] The above description is merely a further detailed explanation of the present invention through a specific embodiment and should not be construed as limiting the invention to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the inventive concept, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A method for hierarchical classification of logistics drone routes based on primary heading extraction, characterized in that, The method includes the following steps: (1) Cluster analysis of logistics drone transportation routes based on logistics data to identify the main transportation routes; (2) By analyzing the direction of the main transport flow, the primary and secondary routes are extracted, and the airspace is divided into altitude layers. (3) Based on the airspace height layering, different headings are matched with different height layers, then the main streams contained in different headings are matched with the corresponding height layers, and finally the routes contained in different main streams are matched with the corresponding height layers. The main stream formed by clustering the s routes in step (2) Its direction angle The calculation method is as follows in For the route The direction angle, Angles in any direction; According to each main stream direction and number of routes Extract the main flight path of logistics drone transportation. ; Heading angle of the main course The calculation methods include unweighted calculation methods and weighted calculation methods; the heading angle of the main heading. The unweighted calculation method is Heading angle of the main course The weighted calculation method is as follows According to the main course Determine the secondary course And define the opposite direction of the primary and secondary headings as the primary heading counter-heading. and secondary heading counter-heading ; In step (2), to reduce transport conflicts between the primary and secondary routes, the following constraints should be met: That is, the difference between the secondary course and the primary course is not less than , To minimize the heading difference, The range is 45° to 135°; Step (3) includes the following steps: 3.1 Based on airspace altitude stratification, different altitude layers are assigned to different headings; in For the extracted first Each heading's altitude level, , , ... For each height level; 3.2 Match the main currents contained in different headings with corresponding altitude stratification; Each course is formed by clusters of several main streams, and different main streams within the same course are matched with the same altitude layer. In the formula For belonging to the first The altitude level of the main current in each course; 3.3 Match the routes contained in different main rivers with corresponding altitude layers; Each main stream is formed by clusters of several routes. Different routes clustered into the same main stream are matched with the same altitude layer. In the formula For belonging to the first The altitude layer of the main stream's route.

2. The method for hierarchical routing of logistics drones based on main heading extraction according to claim 1, characterized in that, In step (1), each logistics delivery route is designed as a logistics drone transportation route X, and then all routes are clustered into several logistics drone transportation trunk lines Y. in, This is the collection of all routes; It is the set of all main streams.

3. The method for hierarchical routing of logistics drones based on main heading extraction according to claim 1, characterized in that, The cluster analysis in step (1) includes: KNN method, K-means method, density clustering method, hierarchical clustering method and spectral clustering method.

4. The method for hierarchical routing of logistics drones based on main heading extraction according to claim 1, characterized in that, In step (2), the flight altitude range of the logistics drone is restricted according to the airspace management regulations, and the airspace in which the logistics drone can fly is divided into altitude layers, with each altitude layer being equally spaced. The flight altitude limit for the logistics drone is as follows: In the formula For drones to complete their flight paths Flight altitude during operation Minimum flight altitude for drones This is the maximum flight altitude of the drone.

Images