A method for generating a global delivery scheme using multiple unmanned aerial vehicles (UAVs) and a delivery system thereof.

Abstract

This invention relates to the field of unmanned aerial vehicle (UAV) delivery technology, specifically disclosing a method and system for generating global delivery plans for multiple UAVs. The generation of each UAV delivery plan includes: generating candidate flight paths for the UAVs based on a preset path planning strategy, meteorological data at the start time, and a defined flight area; determining the corresponding estimated flight time interval; then filtering future meteorological data based on the estimated flight time interval; and validating the existing candidate flight paths. If invalid, a new candidate flight path is generated based on the invalid interval of the candidate flight path, and its validity is verified again. This invention achieves safe, efficient, and optimal multi-objective mission path planning. The planned effective flight paths avoid all hazardous weather areas, ensuring the safe navigation of all UAVs on different routes.

Description

Technical Field

[0001] This invention relates to the field of drone delivery technology, and in particular to a method for generating a global delivery scheme for multiple drones and a delivery system. Background Technology

[0002] With the booming development and gradual maturation of the low-altitude economy, it is projected that by the mid-21st century (around 2050), unmanned transport aircraft, represented by propulsion-based configurations, will achieve large-scale mass production and be widely promoted and applied in logistics transportation, material distribution, and emergency support. These aircraft, with their advantages of simple structure, high payload capacity, and relatively low operating costs, are expected to become an important carrier of the low-altitude logistics network. However, the operational safety of unmanned transport aircraft faces severe challenges under complex and variable weather conditions. Especially in severe weather environments such as strong convection, thunderstorms, strong winds, and freezing conditions, small unmanned aircraft, due to their small size, low flight altitude, and limited wind resistance, are highly susceptible to safety accidents such as attitude loss, power failure, and even mid-air disintegration, leading to significant economic losses and safety risks.

[0003] Currently, to improve the navigation safety of unmanned aerial vehicles (UAVs) in complex weather environments, the industry generally attempts to pre-plan flight routes using high-precision weather forecast data provided by meteorological departments. By integrating meteorological elements such as wind speed, precipitation, cloud cover, and lightning, potentially hazardous weather areas can be identified to some extent, providing a reference for route adjustments and thus avoiding high-risk airspace. However, existing technologies lack a navigation decision-making algorithm that can deeply integrate weather forecast data and dynamically respond to weather changes. Most route planning still relies primarily on static path planning, failing to effectively incorporate the dynamic updates and evolution trends of meteorological information, making it difficult to fundamentally guarantee navigation safety. Summary of the Invention

[0004] To address the technical problems in the prior art, this invention provides a method for generating a global operation and delivery scheme for multiple unmanned aerial vehicles (UAVs) and a delivery system.

[0005] In a first aspect, the present invention provides a method for generating a drone delivery operation plan, the method comprising:

[0006] Obtain the starting point and ending point coordinates of the drone delivery operation, and determine the limited flight area of ​​the drone based on the starting point and ending point coordinates;

[0007] Acquire weather forecast meteorological data for a limited flight area; the weather forecast meteorological data includes meteorological data corresponding to multiple future times from the start time;

[0008] Based on the preset path planning strategy, meteorological data at the start time, and the defined flight area, candidate flight paths for the UAV are generated, and the expected flight time intervals corresponding to the candidate flight paths are calculated.

[0009] Meteorological data for future relevant times are filtered based on the predicted flight time interval, and the validity of existing candidate flight paths is verified based on the meteorological data for future relevant times.

[0010] If valid, the corresponding candidate flight path will be set as a valid flight path.

[0011] If invalid, the invalid interval of the candidate flight path is identified, and the candidate flight path is regenerated based on the meteorological data at the time corresponding to the invalid interval and the path planning strategy. The expected flight time interval corresponding to the candidate flight path is then calculated.

[0012] The validity of the candidate flight paths is verified again until a valid flight path is obtained or the search for a valid flight path fails.

[0013] A delivery plan for the drone is generated based on the effective flight path.

[0014] Furthermore, the method also includes:

[0015] Determine whether the expected flight time interval corresponding to the candidate flight path exceeds the preset drone operating time interval;

[0016] If the limit is exceeded, the candidate flight path will be set as an invalid flight path;

[0017] If the limit is not exceeded, the validity of the candidate flight path is verified.

[0018] Furthermore, in the acquisition of weather forecast meteorological data for the limited flight area, the weather forecast meteorological data is meteorological data for the next several days, with each day's time being a preset drone operating time interval;

[0019] The effective flight path includes the effective flight path for each day within a consecutive set number of days.

[0020] Furthermore, the path planning strategy includes:

[0021] The defined flight area is divided into grids; the size of the grid is calculated based on a preset UAV flight speed and a preset unit time.

[0022] Based on the meteorological data and preset danger thresholds, dynamic obstacle markings are applied to the grid at corresponding times;

[0023] The flight path is generated at the corresponding time according to the preset path planning algorithm and the marked dynamic obstacle grid;

[0024] The path planning algorithm includes:

[0025] Calculate the straight-line distance between the starting point coordinates and the ending point coordinates.

[0026] The normalized distance of the grid nodes is calculated based on the straight-line distance distance;

[0027] Divide the normalized distance values ​​into numerical ranges and set corresponding dynamic weighting factors;

[0028] Calculate the dynamic weighting heuristic function h based on the dynamic weighting factor and the straight-line distance;

[0029] The dynamically weighted heuristic function h is used as the estimated cost of the A* algorithm, and the A* algorithm is executed for path search.

[0030] Furthermore, in the path planning algorithm, before executing the A* algorithm for path search, the following steps are also included:

[0031] Obtain all grid cells along the straight line containing the starting point coordinates and the ending point coordinates;

[0032] Determine in turn whether the above grids are dynamic obstacle grids;

[0033] If so, then execute the A* algorithm to perform path search;

[0034] If not, the straight-line path from the starting point coordinates to the ending point coordinates will be used as the candidate flight path.

[0035] Furthermore, the method also includes:

[0036] The effective flight path is visualized by combining the meteorological data and the coordinates of the starting point and the ending point.

[0037] Furthermore, based on the effective flight path, a drone operation and delivery plan is generated, including:

[0038] Calculate the estimated flight time for each valid flight path, and select the valid flight path with the shortest time as the final flight path.

[0039] Secondly, the present invention provides a method for generating a global multi-UAV delivery operation plan, the method comprising:

[0040] Based on the above method, several valid flight paths are generated for each UAV; the starting point coordinates and ending point coordinates of each UAV's valid flight path are derived from at least one preset starting point geographic coordinates and multiple destination geographic coordinates.

[0041] Calculate the estimated flight time interval for each valid flight path;

[0042] The combined analysis of several effective flight paths corresponding to multiple drones is performed, and the total time length T1 used to complete the delivery task is calculated in combination with the expected flight time interval; wherein, the total time length T1 is calculated by the difference between the end time point and the start time point, the start time point is the start time of the effective flight path corresponding to the earliest departing drone, and the end time point is the arrival time of the effective flight path corresponding to the latest arriving drone.

[0043] A multi-drone delivery solution is generated based on the combination of effective flight paths with the shortest total time length T1.

[0044] Furthermore, the method includes:

[0045] Based on the above method, several valid flight paths for several future days are generated for multiple drones; the starting point coordinates and ending point coordinates of the valid flight path for each drone are derived from at least one preset starting point geographical coordinates and multiple destination geographical coordinates.

[0046] Calculate the estimated flight time interval for each valid flight path;

[0047] By combining several valid flight paths for several future days corresponding to multiple drones, a combination of valid flight paths for different date spans is generated with the aim of reaching all destination coordinates.

[0048] Calculate the date span, total time length T1, and total flight duration for different valid flight path combinations;

[0049] The optimal combination of effective flight paths is selected based on the total flight time, total time length T1, and date span, and a corresponding multi-drone delivery operation plan is generated.

[0050] Thirdly, the present invention provides a multi-drone global operation and delivery system, the system comprising a control center and multiple drones for operation and delivery; the control center is used to generate a multi-drone operation and delivery plan according to the above method, and the multiple drones are used to perform delivery operations according to the operation and delivery plan.

[0051] The UAV delivery scheme generation method of the present invention generates candidate flight paths for the UAV based on a preset path planning strategy and meteorological data at the start time, limited flight area, and other information. It then determines the corresponding expected flight time interval, filters future relevant meteorological data based on the expected flight time interval, and verifies the validity of existing candidate flight paths using this data. If valid, the corresponding candidate flight path is set as a valid flight path; if invalid, the invalid interval of the candidate flight path is identified, and a new candidate flight path is generated based on the meteorological data at the corresponding time of the invalid interval. The validity of the candidate flight path is then verified again until a valid flight path is obtained or the search for a valid flight path fails. Finally, a UAV delivery scheme is generated based on the valid flight path. This embodiment of the invention verifies the validity of the path after path generation by combining time-series meteorological data, considering changes in meteorological conditions at different times, and achieving dynamic obstacle avoidance, thereby obtaining a practically feasible and effective flight path. This solves the defect of existing technologies that do not deeply integrate meteorological forecast data during path planning and cannot dynamically respond to meteorological changes, fundamentally ensuring the reliability of the planned path and guaranteeing the navigation safety of the UAV during operation.

[0052] The multi-UAV global operation delivery scheme generation method and delivery system of the present invention realizes effective flight path planning for multi-UAV and multi-destination delivery operations, achieves the goal of safe, efficient and optimal multi-objective task path planning, and the planned effective flight path avoids all dangerous weather areas, ensuring the navigation safety of all UAVs on different routes. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1 This is a flowchart illustrating the steps of a method for generating a drone delivery solution according to an embodiment of the present invention.

[0055] Figure 2 This is a flowchart illustrating the steps of a method for generating a global multi-UAV delivery solution according to an embodiment of the present invention.

[0056] Figure 3 This is a flowchart illustrating the steps of another method for generating a multi-UAV global operation and delivery scheme according to an embodiment of the present invention.

[0057] Figure 4A visual schematic diagram (I) of an effective flight path generated for an embodiment of the present invention;

[0058] Figure 5 A visual schematic diagram (II) of the effective flight path generated for an embodiment of the present invention. Detailed Implementation

[0059] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0060] The present invention provides a method for generating a drone delivery operation scheme. First, it should be noted that the "drone" referred to in the present invention refers to an unmanned aerial vehicle, including the "unmanned transport aircraft" mentioned in the background art. The design scheme of the present invention is applicable to unmanned aerial vehicles of different sizes and different transport capabilities, and its application scenarios are not limited by the specific technical problems mentioned in the background art.

[0061] Specifically, such as Figure 1 As shown, the method of this embodiment of the invention includes:

[0062] Step S10: Obtain the starting point coordinates and ending point coordinates of the drone delivery operation, and determine the limited flight area of ​​the drone based on the starting point coordinates and ending point coordinates.

[0063] The defined flight area in this step is the scope for subsequent path planning. The starting and ending coordinates should be at least within the boundary of the defined flight area, or within it. The specific scope of the defined flight area can be set according to regional characteristics; this embodiment does not impose any restrictions.

[0064] Step S20: Obtain weather forecast meteorological data for the limited flight area.

[0065] In this step, the weather forecast meteorological data includes meteorological data corresponding to multiple future times from the start time. The method of this invention implements the corresponding functions by running a computer program on a processor. Therefore, to facilitate information processing in subsequent steps, a dataset can be generated by combining the information obtained in steps S10 and S20. The dataset includes not only date information but also meteorological data within the defined flight area at different times. The meteorological data should at least include wind speed and rainfall at different geographical coordinates. The time length of "from the start time to multiple future times" in this step can be set according to the actual situation. Taking a day as an example, assuming meteorological data is obtained from 3 AM to 9 PM, the meteorological data is not available at any given time but corresponds to different times, usually at 1-hour intervals. From 3 AM to 9 PM, a total of 18 hours, there will be 19 times, each time corresponding to a set of meteorological data. The storage structure of the dataset is not limited in this embodiment. Its content includes the start and end points, the numbers of relevant locations within the defined flight area, the corresponding geographical coordinates (x-value, y-value), and the date, forecast time, and corresponding wind speed and rainfall, which are then used to generate and store the dataset.

[0066] Step S30: Based on the preset path planning strategy, meteorological data at the start time, and the limited flight area, generate candidate flight paths for the UAV and calculate the expected flight time intervals corresponding to the candidate flight paths.

[0067] In this step, the scope of path planning is a limited flight area, in which the coordinates of the starting point and the ending point are known. A candidate flight path is planned first by combining the meteorological data at the starting time. Since this candidate flight path does not incorporate meteorological data at future times, its feasibility is uncertain and needs further verification. Therefore, this step is used to calculate the expected flight time interval of this candidate flight path.

[0068] As an example, suppose a drone departs from city 0 and flies to city 1 at 8:00 AM. A candidate flight path is generated based on the meteorological data at 8:00 AM. If it takes 3 hours and 30 minutes to complete the candidate flight path, the corresponding estimated flight time interval is 8:00 AM to 11:30 AM. Therefore, the drone should be between city 0 and city 1 at 10:00 AM or 11:00 AM. The flight path at that time should also avoid meteorological obstacles to ensure the safety of the drone flight. Therefore, the next step should be carried out to verify the validity of the candidate flight path. The estimated flight time interval calculated in this step is used as the time interval benchmark for subsequent validity verification.

[0069] Step S40: Filter meteorological data for future relevant times based on the expected flight time interval, and verify the validity of existing candidate flight paths based on the meteorological data for future relevant times.

[0070] If valid, proceed to step S50: set the corresponding candidate flight path as a valid flight path;

[0071] If invalid, proceed to step S60: identify the invalid intervals of the candidate flight path, regenerate the candidate flight path based on the meteorological data at the time corresponding to the invalid interval and the path planning strategy, and calculate the expected flight time interval corresponding to the candidate flight path;

[0072] If the flight duration corresponding to the candidate flight path does not exceed 1 hour (here, 1 hour refers to the interval of meteorological data), we assume that the meteorological data within 1 hour remains unchanged, that is, the meteorological obstacle points are fixed, so the candidate flight path is valid. However, if the flight duration corresponding to the candidate flight path exceeds 1 hour, the flight path at future relevant times needs to be re-evaluated, because the meteorological obstacle points at the two times may change. Based on the previous example, the relevant times in the future should include 9:00, 10:00, and 11:00. Therefore, it is necessary to verify whether there are any weather obstacles in the flight path from 9:00 to 10:00 in the meteorological data at 9:00. If there are, it means the verified candidate flight path is unusable; if not, it means the flight path from 9:00 to 10:00 is usable. Then, continue verifying whether there are any weather obstacles in the flight path from 10:00 to 11:00 in the meteorological data at 10:00. If there are, it means the verified candidate flight path is unusable; if not, it means the flight path from 10:00 to 11:00 is usable. Then, continue verifying whether there are any weather obstacles in the flight path from 11:00 to 11:30 in the meteorological data at 11:00. If there are, it means the verified candidate flight path is unusable; if not, it means the entire candidate flight path is a valid flight path.

[0073] If a candidate flight path is unavailable, i.e., it is verified as invalid, then a new candidate flight path is generated based on the meteorological data at the corresponding time of the invalid interval and the path planning strategy. For example, if there is a meteorological obstacle in the flight path from 9 o'clock to 10 o'clock in the meteorological data at 9 o'clock, then a new route from the new starting point to the destination should be planned with 9 o'clock as the current time and the position reached at 9 o'clock as the starting point, as a new candidate flight path. Similarly, the corresponding expected flight time interval is calculated, and step S40 is executed again to verify the validity of the candidate flight path until a valid flight path is obtained or the search for a valid flight path fails. At the same time, the valid flight path is saved.

[0074] Step S70: Generate a drone delivery plan based on the effective flight path.

[0075] The delivery plan can be sent to the drone, which will take off on time and complete the delivery operation according to the plan and the corresponding flight path. This embodiment does not limit the specific content of the delivery plan; those skilled in the art can set it up according to the drone's command execution and backend management.

[0076] The drone delivery scheme generation method of this invention generates candidate flight paths for the drone based on a preset path planning strategy and information such as meteorological data at the start time and limited flight area. It then determines the corresponding expected flight time interval, filters future relevant meteorological data based on the expected flight time interval, and verifies the validity of existing candidate flight paths using this data. If valid, the corresponding candidate flight path is set as a valid flight path; if invalid, the invalid interval of the candidate flight path is identified, and a new candidate flight path is generated based on the meteorological data at the corresponding time of the invalid interval. The validity of the candidate flight path is then verified again until a valid flight path is obtained or the search for a valid flight path fails. Finally, a drone delivery scheme is generated based on the valid flight path. This invention verifies the validity of the path by combining time-series meteorological data after path generation, considering changes in meteorological conditions at different times, and achieving dynamic obstacle avoidance. This results in a practically feasible and effective flight path, solving the deficiency of existing technologies that fail to deeply integrate meteorological forecast data during path planning and thus cannot dynamically respond to meteorological changes. This fundamentally ensures the reliability of the planned path and guarantees the navigation safety of the drone during operation.

[0077] Specifically, based on the above embodiments, the method further includes:

[0078] Determine whether the expected flight time interval corresponding to the candidate flight path exceeds the preset drone operating time interval;

[0079] If the limit is exceeded, the candidate flight path will be set as an invalid flight path;

[0080] If the limit is not exceeded, the validity of the candidate flight path is verified.

[0081] Assuming the maximum flight time of the drone is 18 hours, and the drone's operating time is set from 03:00 to 21:00 daily, then for safety reasons, the drone must arrive at its destination before 21:00. Therefore, this embodiment determines whether the expected flight time interval corresponding to a candidate flight path exceeds the preset drone operating time interval, thus filtering the candidate flight paths based on the time interval dimension. For example, if the expected flight time corresponding to a candidate flight path is 19 hours, then this candidate flight path should be determined as an invalid flight path. Similarly, if the expected flight time corresponding to a candidate flight path is 12 hours, and the corresponding expected flight time interval is 10:00-22:00, then this candidate flight path should also be determined as an invalid flight path.

[0082] The design of this embodiment ensures that the planned effective flight path can ensure that the drone reaches its destination within the drone's operating time range, thus ensuring the timeliness of the delivery operation.

[0083] More specifically, based on the above embodiments, the weather forecast meteorological data for the limited flight area is obtained as meteorological data for the next several days, with each day's time being a preset drone working time interval; the effective flight path includes the effective flight path for each day within a continuously set number of days.

[0084] Since this embodiment generates a valid flight path, the steps in the aforementioned embodiment include verifying the validity of the path. Taking the acquisition of meteorological data for the next 5 days as an example, including meteorological data for days 0 to 4 and time periods from 3:00 to 20:00 (a total of 18 hours), when the computer program performs the functions of the above steps, it iterates through all date spans (0 to 4 days). For each span, it iterates through all possible start dates (1 to 5 - max_date_span), where max_date_span is the date span (days). For each start date and the date range determined by the span, a valid path is planned for the target city (which needs to be verified across hours / days).

[0085] In this embodiment, the meteorological data is increased from one day to several future days. This allows for the planning of effective flight paths for multiple future days by traversing the timeframes of those future days. This facilitates the implementation of global operational delivery planning between multiple locations and multiple drones in subsequent embodiments.

[0086] Specifically, based on the above embodiments, the path planning strategy includes:

[0087] First, the restricted flight area is divided into grids;

[0088] Secondly, based on meteorological data and preset danger thresholds, dynamic obstacle markings are applied to the grid at corresponding times;

[0089] Finally, the flight path at the corresponding moment is generated based on the preset path planning algorithm and the marked dynamic obstacle grid.

[0090] In this embodiment, the size of the grid is calculated based on the preset drone flight speed and preset unit time. This embodiment of the invention does not limit the size of the grid. As an example, the drone's flight speed is a fixed value under all weather conditions and remains basically unchanged. Therefore, the size of each grid is set according to the distance the drone travels in 2 minutes.

[0091] The process of dynamically marking obstacles on the grid at corresponding times is based on preset danger thresholds. For example, if the two weather factors affecting drone crashes are wind speed and rainfall, then areas with wind speeds greater than or equal to 15 (this value is a standard value, only used as an example, and its unit is not limited) or rainfall greater than or equal to 4 (this value is a standard value, only used as an example, and its unit is not limited) can be defined as danger zones. Obstacle marking is performed according to the meteorological data at different times. As long as a grid contains a danger zone that meets the danger threshold, the grid is marked as an obstacle grid and should be bypassed during path planning. Through this step, the dataset of the aforementioned embodiment is updated to obtain a new dataset containing danger markers such as high_wind, high_rainfall, and both.

[0092] The path planning algorithm in this embodiment of the invention includes:

[0093] Step (1): Calculate the straight-line distance between the starting point and the ending point based on their coordinates. The straight-line distance is calculated using the Euclidean distance method, i.e., distance = sqrt((x_a - x_b)^2 + (y_a - y_b)^2), where (x_a, y_a) and (x_b, y_b) are the coordinates of the starting point and the ending point, respectively.

[0094] Step (2): Calculate the normalized distance of the grid nodes based on the straight-line distance.

[0095] The normalized distance is calculated as: normalized_distance = 1.0 - (current_distance / max_distance), where current_distance is the distance between the grid node and the endpoint coordinates, and max_distance is the value of the straight-line distance.

[0096] Step (3): Divide the normalized distance values ​​into numerical ranges and set the corresponding dynamic weighting factors.

[0097] The calculation of dynamic weighting factors involves the values ​​of relevant parameters and coefficients, as shown in the example:

[0098] When the normalized distance < 0.3, the dynamic weighting factor is calculated as: dynamic_factor = 0.3 + 2.0 * normalized_distance; when the normalized distance ≥ 0.3, the dynamic weighting factor is calculated as: dynamic_factor = 0.9 + 1.6 * normalized_distance. In this example, 0.3, 2.0, 0.9, and 1.6 are all adjustable parameters or coefficients and can be set according to path planning requirements.

[0099] Step (4): Calculate the dynamic weighting heuristic function h based on the dynamic weighting factor and the straight-line distance.

[0100] The dynamic weighted heuristic function is: h(a,b) = dynamic_factor * distance.

[0101] Step (5): Use the dynamically weighted heuristic function h as the estimated cost of the A* algorithm and execute the A* algorithm for path search.

[0102] The above examples demonstrate that when the distance to the destination coordinates is relatively far (normalized_distance < 0.3), the weight is smaller (0.3~2.0), favoring exploration during path planning; conversely, when the distance to the destination coordinates is relatively close, the weight is larger (2.0~2.5), favoring utilization during path planning. This embodiment designs a dynamic weighting heuristic function that dynamically adjusts the weighting factor based on the distance to the target, balancing exploration and utilization.

[0103] As an example, taking the starting point (0,0) and the target (10,10) as an example, max_distance=sqrt(10^2+10^2)≈14.14;

[0104] When the grid node is located at (2,2):

[0105] The distance between the grid node and the endpoint coordinates: current_distance = sqrt(8^2 + 8^2) ≈ 11.31;

[0106] The normalized distance is calculated as: normalized_distance = 1 - 11.31 / 14.14 ≈ 0.2 (< 0.3);

[0107] Therefore, the dynamic weighting factor is calculated as: dynamic_factor = 0.3 + 2.0 * 0.2 = 0.7.

[0108] The dynamic weighted heuristic function h = 0.7 * 14.14 ≈ 9.9;

[0109] When the grid node is located at (8,8):

[0110] The distance between the grid node and the endpoint coordinates: current_distance = sqrt(2^2 + 2^2) ≈ 2.83;

[0111] The normalized distance is calculated as: normalized_distance = 1 - 2.83 / 14.14 ≈ 0.8 (≥ 0.3);

[0112] Therefore, the dynamic weighting factor is calculated as: dynamic_factor = 0.9 + 1.6 * 0.8 = 2.18.

[0113] The dynamic weighted heuristic function h = 2.18 * 14.14 ≈ 30.83.

[0114] This shows that the weight of a grid node increases steadily as it gets closer to the endpoint.

[0115] More preferably, in the embodiments of the above path planning algorithm, before executing the A* algorithm for path search, the following steps are also included:

[0116] Step (a): Obtain all grid cells along the line containing the starting and ending coordinates;

[0117] Step (b): Determine whether the above grids are dynamic obstacle grids in sequence;

[0118] If so, it means there is an obstacle on the straight path from the starting point coordinates to the ending point coordinates, and the A* algorithm needs to be executed to search for the path.

[0119] If not, it means there are no obstacles on the straight path from the starting point coordinates to the ending point coordinates, and the straight path from the starting point coordinates to the ending point coordinates can be used as a candidate flight path.

[0120] This embodiment can combine the Bresenham algorithm to verify whether an unobstructed straight path exists between the starting point and the ending point before executing the A* algorithm to search for a path. This allows for a rapid check of straight paths, improving the search efficiency of the A* algorithm and reducing redundant calculations. The specific calculation process is not detailed in this embodiment; those skilled in the art can design corresponding software programs to implement the functions of this embodiment based on existing computational theories. Figure 4 The diagram shown is a schematic of the straight path obtained through this embodiment.

[0121] Specifically, based on the methods in the above embodiments, this embodiment also includes: visualizing the effective flight path in combination with meteorological data and the coordinates of the starting point and the ending point.

[0122] The visualization includes a generated path visualization image and related text result files. The visualization image contains information such as obstacles, start point, end point, and path, making it easier for staff to understand the path planning situation more intuitively.

[0123] Specifically, based on the aforementioned embodiments, a delivery plan for the drone is generated according to the effective flight paths, including: calculating the estimated flight time for each effective flight path, and using the effective flight path with the shortest time as the final flight path. The shortest path is planned under safe conditions to obtain the optimal delivery plan.

[0124] Furthermore, based on the above embodiments, the present invention also provides a method for generating a multi-UAV global operation and delivery scheme, specifically as follows: Figure 2 As shown, the method includes:

[0125] Step S101: Generate several valid flight paths corresponding to multiple drones according to the aforementioned method. In this step, the number of drones is increased compared to the method in the previous embodiment, and the starting point coordinates and ending point coordinates of the valid flight path of each drone are derived from at least one preset starting point geographical coordinates and multiple destination geographical coordinates. That is to say, the embodiment of the present invention plans flight routes for multiple drones with at least one starting point and multiple destinations. The process of generating a valid flight path between a certain starting point and destination is the same as the aforementioned process of generating a valid path for a single drone, so the specific generation process will not be described here.

[0126] This step can be viewed as generating valid flight paths by iterating through all possible times as departure times for all combinations of starting points and destinations. These valid flight paths have been verified against weather obstacles and are potentially usable for flight. Whether a path is selected as a flight path depends on further screening in subsequent steps.

[0127] Step S102: Calculate the estimated flight time interval for each valid flight path.

[0128] Step S103: Perform a combined analysis of several effective flight paths corresponding to multiple drones, and calculate the total time T1 required to complete the delivery task by combining the expected flight time interval.

[0129] The total time length T1 is calculated by subtracting the start time from the end time. The start time is the beginning time of the effective flight path corresponding to the earliest departing drone, and the end time is the arrival time of the effective flight path corresponding to the latest arriving drone.

[0130] Step S104: Generate a multi-drone operation and delivery plan based on the combination of effective flight paths with the shortest total time length T1.

[0131] Referring to the example of the aforementioned embodiment, assuming the maximum flight time of the drone is 18 hours (the drone's working time interval is 03:00-21:00), this embodiment configures 10 drones; city number 0 is designated as the departure city, i.e., the geographical coordinates of city 0 are the starting point geographical coordinates, which are the starting coordinates in the above steps; cities numbered 1 to 10 are designated as destination cities, i.e., the geographical coordinates of cities 1 to 10 are the destination geographical coordinates, for a total of 10 destination coordinates. In order to realize the delivery operation from city 0 to destination cities 1 to 10, this embodiment performs a combination analysis of several effective flight paths corresponding to multiple drones. After calculating the expected flight time interval corresponding to each effective flight path, the total time length T1 of feasible combinations is calculated according to step S103. When there are multiple feasible combinations, the effective flight path combination with the shortest total time length T1 value is selected to generate a multi-drone operation delivery plan.

[0132] Another scenario exists in this embodiment: if some destinations are reachable within a day (i.e., effective flight paths are available), while others are inaccessible due to weather conditions (i.e., effective flight paths are unavailable), such as... Figure 5 As shown, at a certain moment, the red area is a high-risk area, unsuitable for drone flight. Delivery is possible between cities 0 and 1, 2, 3, 4, 9, and 10, but not between cities 0 and 5, 6, 7, and 8. Therefore, city 0 can be designated as the departure city, and cities 1-10 can be designated as either departure or destination cities. Drones can depart from city 0, reach a city, and then use that city as their departure point to reach another city (provided delivery conditions are met). Therefore, in the analysis of effective flight path combinations, city 0 is not the sole departure city; rather, combinations are made based on the goal of completing the delivery task. Thus, cities 1-10 may also be possible departure cities. This combination process includes verifying the feasibility of different combinations, eliminating combinations with time conflicts, and finding the effective flight path combination with the shortest total time T1 to generate a multi-drone delivery operation plan.

[0133] Furthermore, to achieve the goal of completing deliveries in all cities, "cross-date" completion can also be considered. Specifically, this invention provides another embodiment of a multi-drone global operation delivery scheme generation method, as detailed below. Figure 3 As shown, the method includes:

[0134] Step S201: Generate several valid flight paths for several future days corresponding to multiple drones according to the method of the aforementioned embodiment. In this step, not only is the number of drones increased compared to the previous embodiment, but the number of "days" for path planning is also increased to solve the problem that it is impossible to complete all delivery tasks on the same day. Similarly, the starting point coordinates and ending point coordinates of the valid flight path of each drone are derived from at least one preset starting point geographical coordinates and multiple destination geographical coordinates.

[0135] Step S202: Calculate the estimated flight time interval for each valid flight path.

[0136] Step S203: Combine several valid flight paths for several future days corresponding to multiple drones to generate a combination of valid flight paths for different date spans with the aim of reaching all destination coordinates.

[0137] Step S204: Calculate the date span, total time length T1, and total flight duration for different valid flight path combinations;

[0138] Step S205: Select the optimal combination of effective flight paths according to the total flight time, total time length T1 and date span, and generate a multi-drone operation and delivery plan accordingly.

[0139] Taking the aforementioned example, for instance, obtaining meteorological data for the next 5 days, including 18 hours of meteorological data from days 0 to 4 and from 3:00 to 20:00, we traverse different date spans and start times to verify cross-time and cross-date path combinations, aiming to generate effective flight path combinations for different date spans by finding those that can reach all destination coordinates.

[0140] This embodiment implements path combination attempts across time periods and dates, calculating the date span, total time length T1, and total flight time for different effective flight path combinations. Specifically, the date span is represented by the following function: Minimize the total flight span time Total_Span_Time = (date_span * 24 * 60) + total_flight_time, where date_span is the date span (days), and total_flight_time is the total flight time of all drones (minutes). For step S205, which filters for the optimal effective flight path combination based on total flight time, total time length T1, and date span, when multiple effective flight path combinations exist, first, the combinations with the shortest total flight time are selected; then, the combinations with the smallest date span are selected from these; finally, the combination with the smallest total flight span time is chosen as the optimal solution.

[0141] To facilitate understanding of this embodiment, as an example, assume there are 3 target cities (e.g., cities numbered 1, 2, and 3), and all drones depart from city 0, with dates ranging from day 1 to day 5:

[0142] Traverse the date span: For example, when the span = 1 day (starting day 1, ending day 2);

[0143] Path collection and selection: On day 1 and day 2, find effective flight paths starting from city 0 for cities 1, 2 and 3 respectively (considering dynamic weather obstacles). During this process, record the effective flight path with the shortest flight time for each city on these two days.

[0144] Calculate the total flight span time: Assuming the flight times for the three paths are T1, T2, and T3 minutes respectively, then Total_Span_Time = (1 * 24 * 60) + (T1 + T2 + T3);

[0145] Solution comparison: Compared with solutions with span = 0 days (all tasks are completed on the same day) and span = 2 days, the solution with the smallest Total_Span_Time was ultimately selected as the optimal one.

[0146] This embodiment plans the optimal route combination for multiple destinations to minimize the total flight time, and prioritizes smaller date spans. For the same date span, it selects the option with the shortest total flight time, thereby choosing the optimal delivery solution.

[0147] Furthermore, the present invention also provides an embodiment of a multi-drone global operation and delivery system, the system comprising a control center and multiple drones for operation and delivery; the control center is used to generate an operation and delivery plan for multiple drones according to the method of the foregoing embodiment, and the multiple drones are used to perform delivery operations according to the operation and delivery plan.

[0148] The multi-UAV global operation delivery scheme generation method and delivery system of the present invention realizes effective flight path planning for multi-UAV and multi-destination delivery operations, achieves the goal of safe, efficient and optimal multi-objective task path planning, and the planned effective flight path avoids all dangerous weather areas, ensuring the navigation safety of all UAVs on different routes.

[0149] The present invention has been further described above with reference to specific embodiments. However, it should be understood that the specific description herein should not be construed as limiting the nature and scope of the present invention. Various modifications made to the above embodiments by those skilled in the art after reading this specification are all within the scope of protection of the present invention.

Claims

1. A method for generating a drone delivery plan, characterized in that, The method includes: Obtain the starting point and ending point coordinates of the drone delivery operation, and determine the limited flight area of ​​the drone based on the starting point and ending point coordinates; Acquire weather forecast meteorological data for a limited flight area; the weather forecast meteorological data includes meteorological data corresponding to multiple future times from the start time; Based on the preset path planning strategy, meteorological data at the start time, and the defined flight area, candidate flight paths for the UAV are generated, and the expected flight time intervals corresponding to the candidate flight paths are calculated. Meteorological data for future relevant times are filtered based on the predicted flight time interval, and the validity of existing candidate flight paths is verified based on the meteorological data for future relevant times. If valid, the corresponding candidate flight path will be set as a valid flight path. If invalid, the invalid interval of the candidate flight path is identified, and the candidate flight path is regenerated based on the meteorological data at the time corresponding to the invalid interval and the path planning strategy. The expected flight time interval corresponding to the candidate flight path is then calculated. The validity of the candidate flight paths is verified again until a valid flight path is obtained or the search for a valid flight path fails. A delivery plan for the drone is generated based on the effective flight path.

2. The method for generating a drone delivery plan as described in claim 1, characterized in that, The method further includes: Determine whether the expected flight time interval corresponding to the candidate flight path exceeds the preset drone operating time interval; If the limit is exceeded, the candidate flight path will be set as an invalid flight path; If the limit is not exceeded, the validity of the candidate flight path is verified.

3. The method for generating a drone delivery plan as described in claim 2, characterized in that, The weather forecast meteorological data for the limited flight area refers to meteorological data for the next several days, with each day's time being a preset operating time interval for the UAV. The effective flight path includes the effective flight path for each day within a consecutive set number of days.

4. The method for generating a drone delivery plan as described in claim 1, characterized in that, The path planning strategy includes: The defined flight area is divided into grids; the size of the grid is calculated based on a preset UAV flight speed and a preset unit time. Based on the meteorological data and preset danger thresholds, dynamic obstacle markings are applied to the grid at corresponding times; The flight path is generated at the corresponding time according to the preset path planning algorithm and the marked dynamic obstacle grid; The path planning algorithm includes: Calculate the straight-line distance between the starting point coordinates and the ending point coordinates. The normalized distance of the grid nodes is calculated based on the straight-line distance distance; Divide the normalized distance values ​​into numerical ranges and set corresponding dynamic weighting factors; Calculate the dynamic weighting heuristic function h based on the dynamic weighting factor and the straight-line distance; The dynamically weighted heuristic function h is used as the estimated cost of the A* algorithm, and the A* algorithm is executed for path search.

5. The method for generating a drone delivery plan as described in claim 4, characterized in that, The path planning algorithm, before executing the A* algorithm for path search, also includes: Obtain all grid cells along the straight line containing the starting point coordinates and the ending point coordinates; Determine in turn whether the above grids are dynamic obstacle grids; If so, then execute the A* algorithm to perform path search; If not, the straight-line path from the starting point coordinates to the ending point coordinates will be used as the candidate flight path.

6. The method for generating a drone delivery plan as described in claim 1, characterized in that, The method further includes: The effective flight path is visualized by combining the meteorological data and the coordinates of the starting point and the ending point.

7. The method for generating a drone delivery plan as described in claim 1, characterized in that, Based on the effective flight path, a drone delivery plan is generated, including: Calculate the estimated flight time for each valid flight path, and select the valid flight path with the shortest time as the final flight path.

8. A method for generating a global delivery scheme using multiple unmanned aerial vehicles (UAVs), characterized in that, The method includes: The method according to any one of claims 1 to 7 generates several valid flight paths corresponding to multiple UAVs; the starting point coordinates and ending point coordinates of the valid flight path of each UAV are derived from at least one preset starting point geographical coordinates and multiple destination geographical coordinates. Calculate the estimated flight time interval for each valid flight path; The combined analysis of several effective flight paths corresponding to multiple drones is performed, and the total time length T1 used to complete the delivery task is calculated in combination with the expected flight time interval; wherein, the total time length T1 is calculated by the difference between the end time point and the start time point, the start time point is the start time of the effective flight path corresponding to the earliest departing drone, and the end time point is the arrival time of the effective flight path corresponding to the latest arriving drone. A multi-drone delivery solution is generated based on the combination of effective flight paths with the shortest total time length T1.

9. The method for generating a multi-UAV global operation and delivery scheme as described in claim 8, characterized in that, The method includes: The method according to any one of claims 1 to 7 generates several valid flight paths for several future days corresponding to multiple drones; the starting point coordinates and ending point coordinates of the valid flight path of each drone are derived from at least one preset starting point geographical coordinates and multiple destination geographical coordinates. Calculate the estimated flight time interval for each valid flight path; By combining several valid flight paths for several future days corresponding to multiple drones, a combination of valid flight paths for different date spans is generated with the aim of reaching all destination coordinates. Calculate the date span, total time length T1, and total flight duration for different valid flight path combinations; The optimal combination of effective flight paths is selected based on the total flight time, total time length T1, and date span, and a corresponding multi-drone delivery operation plan is generated.

10. A multi-UAV global operation and delivery system, characterized in that, The system includes a control center and multiple drones for delivery operations; the control center is used to generate a multi-drone delivery operation plan according to the method of any one of claims 8-9, and the multiple drones are used to perform delivery operations according to the delivery operation plan.

Images