Unmanned aerial vehicle path planning method and device, and storage medium

Abstract

This invention relates to a method, apparatus, and storage medium for unmanned aerial vehicle (UAV) path planning. The method includes: Step S1: acquiring information about the target UAV and the target unmanned vehicle to be accessed, and determining the lower and upper bounds of the revisit time for the target UAV; Step S2: under the constraints of a speed range and revisit time, determining the nearest and farthest positions that the target UAV can reach by moving at a constant speed along a known straight line; Step S3: determining the target UAV's position s. i The nearest communication neighborhood and target unmanned vehicle i Does a common neighborhood exist between the farthest communication neighborhoods? If yes, proceed to step S4; otherwise, proceed to step S5. Step S4: Control the drone to traverse this common neighborhood to achieve target unmanned vehicle s i Revisit; Step S5: Construct a search channel based on the nearest and farthest locations, and control the drone to pass through this search channel to achieve the target unmanned vehicle s i A revisit. Compared with the prior art, the present invention has advantages such as reduced power consumption.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) path planning, and in particular to a UAV path planning method, apparatus, and storage medium. Background Technology

[0002] Heterogeneous multi-robot systems fully leverage the complementary functional and structural characteristics of different robot types to form cross-domain collaborative systems, expanding into richer application scenarios and fulfilling more complex task requirements. Air-ground collaborative unmanned systems, as an important branch of heterogeneous multi-robot systems, utilize unmanned aerial vehicles (UAVs) and unmanned vehicles (UAVs) to collaboratively execute tasks. UAVs offer high maneuverability and a wide field of view, enabling them to quickly capture dynamic targets, but their endurance and payload capacity are limited. While UAVs have a smaller perception range and slower movement, they can carry more mission payloads. Therefore, effectively combining the performance advantages of both types of robots can play a crucial role in complex tasks such as reconnaissance and surveillance, and disaster relief.

[0003] The messenger mechanism is a typical implementation of air-to-ground collaboration. This mechanism assumes that the task radius of unmanned vehicles (UAVs) is larger than their own communication radius, thus preventing them from communicating with each other. To achieve collaborative operation among UAVs, the UAV acts as a messenger, transmitting the status of other UAVs and relevant task execution information between them. Each UAV dynamically adjusts its decision-making regarding task execution after receiving the information from the UAV. The existence of the messenger mechanism not only enables task collaboration between UAVs and UAVs but also introduces constraints on their behavior. Specifically, the UAV's path planning is limited by the position and movement of the UAVs, while the UAVs' task planning depends on the task information provided by the UAVs and the information from other UAVs.

[0004] In the messenger mechanism, the UAV needs to periodically traverse each unmanned vehicle (UAV). This process involves determining the order in which the UAV traverses each UAV and planning the traversal path for each UAV. Considering the uncertainty of the environment or the movement of the UAV itself, the UAV cannot accurately predict the location of the UAV. To ensure that the UAV can successfully revisit each UAV, a maximum revisit time constraint is introduced. The revisit time affects the size of the intersection of the communication neighborhood determined by the UAV within two consecutive visit time periods to each UAV. Therefore, the candidate range of the UAV's communication neighborhood visit location in the decision variables depends on the revisit time. The above process is solved based on a rolling time-domain optimization method. First, the order in which the UAV visits each UAV is determined through heuristic rules. Then, the boundary of the UAV's revisit time for each UAV is determined. Finally, the appropriate reliable access strategy is selected and the path is planned based on whether a common neighborhood exists. Therefore, it requires a large amount of computing power. Summary of the Invention

[0005] The purpose of this invention is to provide a method, device, and storage medium for unmanned vehicle path planning. Under the condition that the direction of movement of the unmanned vehicle is known and the speed is uncertain, the invention innovatively introduces a maximum revisit time constraint and ensures that the messenger UAV can revisit each unmanned vehicle by proposing a reliable access strategy.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A method for unmanned aerial vehicle (UAV) path planning includes:

[0008] Step S1: Obtain the target unmanned vehicle s to be accessed i The current position and speed range, as well as the current position and heading angle of the drone, are used to determine the target drone vehicle s i The lower bound of the revisit time Δt min and the upper bound of the revisit time Δt max ;

[0009] Step S2: Within the speed range and the lower bound of the revisit time Δt min Upper bound of revisit time Δt max Under constraints, determine the target unmanned vehicle s i The closest position that can be reached by moving at a constant speed along a known straight line. and farthest position

[0010] Step S3: Determine the target unmanned vehicle s i Located in the nearest position Communication neighborhood and target unmanned vehicles i Located at the farthest position Between the communication neighbors, is there a common neighbor? If yes, then proceed to step S4; otherwise, proceed to step S5.

[0011] Step S4: Control the drone to traverse this public neighborhood to achieve target unmanned vehicle s i A follow-up visit;

[0012] Step S5: Based on the nearest location and farthest position Construct a search channel, and control the drone to pass through this search channel to achieve target unmanned vehicle s i A follow-up visit.

[0013] Step S1 includes:

[0014] Step S11: Obtain the initial position and heading angle of the UAV;

[0015] Step S12: Based on the current locations of all unvisited vehicles in this cycle, and in conjunction with the current location of the drone, determine the target unmanned vehicle that needs to be visited;

[0016] Step S13: Based on the target unmanned vehicle s that needs to be accessed i The current position and speed range, as well as the current position and heading angle of the drone, determine the target unmanned vehicle s i The lower bound of the revisit time Δt min and the upper bound of the revisit time Δt max .

[0017] Step S12 specifically includes:

[0018] Step S121: Calculate the geometric center point based on the current positions of all unmanned vehicles;

[0019] Step S122: Calculate the angle of each unvisited unmanned vehicle relative to the drone based on the obtained geometric center point;

[0020] Step S123: Sort the unvisited autonomous vehicles from smallest to largest according to their corresponding angles, and select the autonomous vehicle with the smallest angle as the target autonomous vehicle to be visited.

[0021] Step S13 specifically includes:

[0022] Step S131: Determine the time interval t between the last time the drone visited the target vehicle and the time when it completed a revisit to the previous vehicle within the current cycle. AB ;

[0023] Step S132: Calculate the target unmanned vehicle's speed within its speed range at time interval t. AB The farthest distance within and closest distance and the communication neighborhood corresponding to the farthest distance. and the communication neighborhood corresponding to the nearest distance

[0024] Step S133: Based on the vector Direction determination of UAVs for communication neighborhood The boundary range during access is defined, and N uniform samplings are performed within this boundary range to obtain N first sampling points;

[0025] Step S134: Assume the target unmanned vehicle starts from the nearest distance at the minimum speed. If it continues to move in a straight line, then the communication neighborhood will be affected. For each first sampling point on the boundary, calculate the time it takes for the drone to visit each first sampling point;

[0026] Step S135: Calculate the minimum access time of the UAV to all first sampling points.

[0027] Step S136: Based on the obtained minimum value Calculate the lower bound Δt of the revisit time of the drone to the target unmanned vehicle. min ;

[0028] Step S137: Based on the vector Direction determination of UAVs in the communication neighborhood The boundary range during access is defined, and N uniform samplings are performed within this boundary range to obtain N second sampling points;

[0029] Step S138: Assume the target autonomous vehicle starts from the furthest distance at maximum speed. If it continues to move in a straight line, then the communication neighborhood will be affected. For each second sampling point on the boundary, calculate the time it takes for the drone to visit each second sampling point;

[0030] Step S139: Calculate the maximum access time of the UAV to all second sampling points.

[0031] Step S1310: Based on the obtained maximum value Calculate the upper bound Δt of the revisit time of the drone to the target unmanned vehicle. max .

[0032] Step S2 includes:

[0033] Step S21: Calculate the minimum speed for the target unmanned vehicle. Lower bound of revisit time Δt min The closest position traveled along the inner straight line.

[0034] Step S22: Calculate the target unmanned vehicle's maximum speed The upper bound of the revisit time Δt max The farthest position traveled along the inner straight line

[0035] Step S4 specifically includes:

[0036] Step S41: Calculate the communication neighborhood and communication neighborhood The coordinates of the two intersection points P b G P t G ;

[0037] Step S42: Determine the coordinates P of the two intersection points b GP t G If the values ​​are the same, then the common neighborhood is a point. The drone is controlled to pass through the coordinates of this intersection point to achieve the target drone s. i If a revisit is required, then proceed to step S43;

[0038] Step S43: Using the nearest position Centered on the autonomous vehicle The communication radius is taken as the radius, and the coordinates of the two intersection points P are... b G P t G Multiple third sampling points are obtained uniformly along the defined communication domain boundary;

[0039] Step S44: Calculate the Dubins path length of each third sampling point, and select the third sampling point with the smallest Dubins path length among all third sampling points as the first alternative access point;

[0040] Step S45: Using the farthest position Centered on the autonomous vehicle The communication radius is taken as the radius, and the coordinates of the two intersection points P are... b G P t G Multiple fourth sampling points are obtained by uniformly applying sampling methods along the defined communication domain boundary;

[0041] Step S46: Calculate the Dubins path length of each fourth sampling point, and select the fourth sampling point with the smallest Dubins path length among all fourth sampling points as the second alternative access point;

[0042] Step S47: Select the one with the smaller Dubins path length between the first and second alternative access points as the first final access point;

[0043] Step S48: Control the drone to pass through this first final access point to achieve the target unmanned vehicle s i A follow-up visit.

[0044] Step S5 specifically includes:

[0045] Step S51: Using the nearest position Starting point, farthest position The endpoint is the target unmanned vehicle. The search channel is constructed using twice the communication radius as the channel width.

[0046] Step S52: Based on the vector The direction to determine the drone's communication neighborhood The boundary range during access is used to perform multiple samplings within this boundary range, resulting in multiple fifth sampling points;

[0047] Step S54: Filter the fifth sampling point according to the pre-configured rules, calculate the Dubins path length of the filtered fifth sampling point, and take the fifth sampling point with the smallest Dubins path length as the third alternative access point;

[0048] Step S55: Based on the vector The direction to determine the drone's communication neighborhood The boundary range during access is used to perform multiple samplings within this boundary range, resulting in multiple sixth sampling points;

[0049] Step S56: Calculate the Dubins path length of all sixth sampling points, and select the sixth sampling point with the smallest Dubins path length as the fourth alternative access point;

[0050] Step S57: Select the one with the shorter Dubins path length between the third and fourth alternative access points as the second final access point, and control the drone to pass through this second final access point to achieve the target unmanned vehicle s i A follow-up visit.

[0051] The filtering process for the fifth sampling point in step S54 specifically includes:

[0052] Step S541: Calculate the position of the UAV at the time when the previous unmanned vehicle completed its visit. With its heading angle h min Path length to each fifth sampling point and access time

[0053] Step S542: For each fifth sampling point, determine the path taken by the UAV along the vector. The direction, starting from the fifth sampling point, is related to the communication field. The coordinates of the meeting point are used as the first meeting point of the fifth sampling point;

[0054] Step S543: For each fifth sampling point, calculate the path length and access time of the UAV from the fifth sampling point to the corresponding first meeting position, taking that fifth sampling point as the starting point.

[0055] Step S544: Access time With access time The sum is less than the upper bound of the revisit time Δt max The fifth sampling point is retained, while the other fifth sampling points are filtered out.

[0056] A drone path planning device includes a memory, a processor, and a program stored in the memory, characterized in that the processor executes the program to implement the method described above.

[0057] A storage medium having a program stored thereon, which, when executed, implements the method described above.

[0058] Compared with the prior art, the present invention has the following beneficial effects:

[0059] 1. Given that the direction of motion of the unmanned vehicle is known but its speed is uncertain, a maximum revisit time constraint is innovatively introduced, and a reliable access strategy is proposed to ensure that the messenger drone can revisit each unmanned vehicle.

[0060] 2. It can be expanded to an air-ground collaborative system with multiple messenger drones and multiple unmanned vehicles. Due to the real-time nature of drone decision-making, this path planning method can be applied to a variety of mission scenarios.

[0061] 3. Specific final access points are designed for both public and non-public neighborhoods, which can effectively improve the success rate of revisits and reduce the power consumption of revisits. Attached Figure Description

[0062] Figure 1 This is a schematic diagram of the main steps of the method of the present invention;

[0063] Figure 2 This is a flowchart of a messenger UAV path planning method based on a deterministic access strategy provided in an embodiment of the present invention;

[0064] Figure 3 This is a schematic diagram illustrating a reliable access strategy for unmanned vehicles by drones in the presence of a public neighborhood, as described in an embodiment of the present invention.

[0065] Figure 4 This is a schematic diagram illustrating a reliable access strategy for unmanned vehicles by drones in the absence of a public neighborhood, as described in this embodiment of the invention.

[0066] Figure 5 This is a schematic diagram of the path taken by the UAV to complete one periodic traversal of each unmanned vehicle in an embodiment of the present invention;

[0067] Figure 6 This is a schematic diagram illustrating the path taken by the UAV during two periodic traversals of each unmanned vehicle in an embodiment of the present invention. Detailed Implementation

[0068] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0069] The path planning problem for messenger UAVs in a multi-point dynamic assembly mission can be described as follows:

[0070] Taking disaster relief as an example, survivors are known to be distributed in different locations in space. An air-ground collaborative system consisting of a drone and multiple unmanned vehicles (UAVs) is responsible for the survivor search and rescue mission. The survivors' vital signs change dynamically. Each UAV has a certain rescue capability value. Based on the survivors' vital signs and the status of other UAVs, the system dynamically decides which locations to visit and rescue. The mission ends when all survivors are successfully rescued. Because the UAVs have a large operating range but limited communication range, they cannot communicate directly with each other. The drone acts as a messenger, periodically traversing the communication neighborhood of each UAV to transmit survivor status and information about other UAVs. In this mission, due to environmental or their own motion uncertainties, the messenger drone may not be able to accurately determine the location of the UAVs, thus failing to ensure successful revisiting. Given that the UAVs always travel in a straight line and their speed varies within a finite range, the messenger drone's goal is to plan the shortest path to ensure successful revisiting of all UAVs.

[0071] In this embodiment of the invention, the messenger drone is modeled as a Dubins car subject to curvature constraints, and its kinematic model is as follows:

[0072]

[0073] Among them, (x A ,y A θ represents the position coordinates of the UAV. A Indicates the orientation angle, v A The speed of the drone is represented by r. A U represents the minimum turning radius of the drone. A Indicates control input, (x A ,y A ,θ A () represents the state of the drone, simplified to (P) A ,h A ), where P A Represents coordinates, h A Indicates direction.

[0074] The model for the path planning problem of a messenger drone under uncertain motion conditions of an autonomous vehicle can be represented as:

[0075]

[0076]

[0077]

[0078]

[0079]

[0080]

[0081]

[0082] Where, N G S represents the number of driverless cars. * This indicates the order in which the drone accesses each unmanned vehicle, specifically: P * This indicates the locations visited by the drones for each unmanned vehicle. and These represent the initial position and initial orientation of the drone, respectively. This indicates that drones are related to unmanned vehicles. i Access location within the communication neighborhood. Drones vs. Unmanned Vehicles i+1 Access location within the communication neighborhood. Indicates in Driverless cars i Location coordinates, Indicates the communication radius of the drone. Let represent the communication radius of the i-th autonomous vehicle, and T represent the set of revisit times. This indicates that drones are related to unmanned vehicles. i The time for the follow-up visit, This indicates that drones are related to unmanned vehicles. i The set of candidate access locations, This indicates that drones are related to unmanned vehicles. i The actual revisit path length.

[0083] In the constraints of the above problem model, the first term represents the UAV motion model constraint, the second term represents the communication radius constraint, the third term represents the UAV speed constraint, the fourth term represents the revisit time constraint, the fifth term represents the revisit location constraint, and the sixth term represents the revisit distance constraint.

[0084] The messenger drone path planning method based on deterministic access strategy provided in this application is conceived as follows: When the speed of the unmanned vehicle is unknown but bounded, the position range of each unmanned vehicle can be determined based on the upper and lower bounds of the speed and the travel time. First, a heuristic method is used to determine the order in which the drone visits each unmanned vehicle. Then, based on the position range of each unmanned vehicle, it can be determined whether a common neighborhood is generated at the nearest and farthest positions of the unmanned vehicles. Different access strategies are selected based on whether a common neighborhood exists. Finally, based on the idea of ​​rolling optimization, the dynamic adjustment of the drone's access order and access position for unmanned vehicles is realized.

[0085] Based on the aforementioned path planning problem and principles, the messenger UAV path planning method based on a deterministic access strategy provided in this embodiment of the invention has the following process: Figure 2 As shown, this is mainly achieved by sequentially accessing each drone within each cycle. The core of this approach lies in the path planning for accessing the drones.

[0086] Specifically, a drone path planning method, such as Figure 1 As shown, it includes:

[0087] Step S1: Obtain the target unmanned vehicle s to be accessed i The current position and speed range, as well as the current position and heading angle of the drone, are used to determine the target drone vehicle s i The lower bound of the revisit time Δt min and the upper bound of the revisit time Δt max ,

[0088] In the first iteration, the initial position of the drone is obtained. and heading angle And the starting position P of each driverless vehicle i G i = 1, 2, ..., N G and the speed range of each autonomous vehicle

[0089] Within each cycle, step S1 specifically includes:

[0090] Step S11: Obtain the initial position and heading angle of the UAV;

[0091] Step S12: Based on the current locations of all unvisited autonomous vehicles in this cycle, and in conjunction with the current location of the drone, determine the target autonomous vehicle to be visited, specifically including:

[0092] Step S121: Calculate the geometric center point q based on the current positions of all autonomous vehicles.

[0093]

[0094] Where: q i (i = 1, 2, ..., N) G ) represents the current position of the i-th unmanned vehicle, and q0 represents the current position of the drone.

[0095] Step S122: Calculate the angle of each unvisited unmanned vehicle relative to the drone based on the obtained geometric center point;

[0096] Step S123: Sort the unvisited autonomous vehicles from smallest to largest according to their corresponding angles, and select the autonomous vehicle with the smallest angle as the target autonomous vehicle to be visited.

[0097] Step S13: Based on the target unmanned vehicle s that needs to be accessed i The current position and speed range, as well as the current position and heading angle of the drone, determine the target unmanned vehicle s i The lower bound of the revisit time Δt min and the upper bound of the revisit time Δt max Specifically, it includes:

[0098] Step S131: Determine the time interval t between the last time the drone visited the target vehicle and the time when it completed a revisit to the previous vehicle within the current cycle. AB ;

[0099] Step S132: Calculate the target unmanned vehicle's speed within its speed range at time interval t. AB The farthest distance within and closest distance and the communication neighborhood corresponding to the farthest distance. and the communication neighborhood corresponding to the nearest distance

[0100] Step S133: Based on the vector Direction determination of UAVs in the communication neighborhood The boundary range during access is defined. Within this boundary range, N uniform sampling operations are performed to obtain N first sampling points. Each sampling point can be represented as:

[0101]

[0102] in, and They represent The x and y coordinates, Indicates driverless car s i The communication radius is θ, where θ represents the sampling angle.

[0103] Step S134: Assume the target unmanned vehicle is moving at minimum speed From the closest distance If it continues to move in a straight line, then the communication neighborhood will be affected. Each first sampling point on the boundary Calculate the time for the drone to visit each first sampling point.

[0104] Step S135: Calculate the minimum access time of the UAV to all first sampling points.

[0105] Step S136: Based on the obtained minimum value Calculate the lower bound Δt of the revisit time of the drone to the target unmanned vehicle. min :

[0106]

[0107] Step S137: Based on the vector Direction determination of UAVs in the communication neighborhood The boundary range during access is defined. Within this boundary range, N uniform sampling operations are performed to obtain N second sampling points. Each second sampling point can be represented as:

[0108]

[0109] in, and They represent The horizontal and vertical coordinates.

[0110] Step S138: Assume the target driverless car is traveling at its maximum speed. From the farthest distance If it continues to move in a straight line, then the communication neighborhood will be affected. For each second sampling point on the boundary, calculate the time it takes for the drone to visit each second sampling point.

[0111] Step S139: Calculate the maximum access time of the UAV to all second sampling points.

[0112] Step S1310: Based on the obtained maximum value Calculate the upper bound Δt of the revisit time of the drone to the target unmanned vehicle. max :

[0113]

[0114] Step S2: Within the speed range and the lower bound of the revisit time Δt min Upper bound of revisit time Δt max Under the constraints, determine the target unmanned vehicle s i The closest position that can be reached by moving at a constant speed along a known straight line. and farthest position include:

[0115] Step S21: Calculate the target unmanned vehicle s i at minimum speed Lower bound of revisit time Δt min The closest position traveled along the inner straight line.

[0116]

[0117] Step S22: Calculate the target unmanned vehicle s i At maximum speed The upper bound of the revisit time Δt max The farthest position traveled along the inner straight line

[0118]

[0119] Step S3: Determine the target unmanned vehicle s i Located in the nearest position Communication neighborhood and target unmanned vehicles i Located at the farthest position Between the communication neighbors, is there a common neighbor? If yes, then proceed to step S4; otherwise, proceed to step S5.

[0120] Step S4: As Figure 3 As shown, controlling the drone to traverse this public neighborhood enables the targeting of the unmanned vehicle s. i The revisit specifically includes:

[0121] Step S41: Calculate the communication neighborhood and communication neighborhood The coordinates of the two intersection points P b G P t G ;

[0122] Step S42: Determine the coordinates P of the two intersection points b G P t G If the values ​​are the same, then the common neighborhood is a point. The drone is controlled to pass through the coordinates of this intersection point to achieve the target drone s. i If the visit is repeated, then the public neighborhood S is... and For the resulting intersection, proceed to step S43;

[0123] Step S43: Using the nearest position Centered on the autonomous vehicle The communication radius is taken as the radius, and the coordinates of the two intersection points P are... b G P t G M third sampling points are uniformly obtained on the defined communication domain boundary. It can be represented as:

[0124]

[0125] in, and They represent The x and y coordinates, φ M Indicated by P b G and P t G The sampling angle is determined by two points.

[0126] Step S44: Calculate the Dubins path length for each third sampling point:

[0127]

[0128] The first term indicates that the drone has moved from the previous unmanned vehicle. i-1 End of visit Departure, arrival at the sampling point The Dubins path length, the second term indicates the distance the drone traveled from the sampling point. Depart, arrive at the next driverless car to visit. i+1 Location Dubins path length at that time This indicates that the drone has left the sampling point. The heading angle at that time.

[0129] And select the Dubins path length from all third sampling points. The smallest third sampling point is selected as the first alternative access point.

[0130] Step S45: Using the farthest position Centered on the autonomous vehicle The communication radius is taken as the radius, and the coordinates of the two intersection points P are... b G P t G Multiple fourth sampling points are obtained by uniformly applying sampling methods along the defined communication domain boundary;

[0131] Step S46: Calculate the Dubins path length of each fourth sampling point in the same way as the third sampling point, and select the fourth sampling point with the smallest Dubins path length among all the fourth sampling points as the second alternative access point;

[0132] Step S47: Select the one with the smaller Dubins path length between the first and second alternative access points as the first final access point;

[0133] Step S48: Control the drone to pass through this first final access point to achieve the target unmanned vehicle s i A follow-up visit.

[0134] Step S5: As Figure 4 As shown, based on the nearest position and farthest position Construct a search channel, and control the drone to pass through this search channel to achieve target unmanned vehicle s i The revisit specifically includes:

[0135] Step S51: Using the nearest position Starting point, farthest position The endpoint is the target unmanned vehicle. The search channel is constructed using twice the communication radius as the channel width.

[0136] Step S52: Based on the vector The direction to determine the drone's communication neighborhood The boundary range during access is used to perform multiple samplings within this boundary range, resulting in multiple fifth sampling points;

[0137] Step S54: Filter the fifth sampling point according to the pre-configured rules, and calculate the Dubins path length of the filtered fifth sampling point.

[0138]

[0139] in: For the drone to be compared with the previous unmanned vehicle i-1 End of visit With an orientation angle h min to sampling point Path length, For drones As the starting point, along arrive Path length, Indicates drones Starting from the designated point, proceed to the next driverless vehicle to be visited. i+1 Location Dubins path length at that time

[0140] The fifth sampling point with the shortest Dubins path length was selected as the third alternative access point.

[0141] The process of filtering the fifth sampling point specifically includes:

[0142] Step S541: Calculate the position of the UAV at the time when the previous unmanned vehicle completed its visit. With its heading angle h min Path length to each fifth sampling point and access time

[0143] Step S542: For each fifth sampling point, determine the path taken by the UAV along the vector. The direction, starting from the fifth sampling point, is related to the communication field. The coordinates of the meeting point are used as the first meeting point of the fifth sampling point;

[0144] Step S543: For each fifth sampling point, calculate the path length and access time of the UAV from the fifth sampling point to the corresponding first meeting position, taking that fifth sampling point as the starting point.

[0145] Step S544: Access time With access time The sum is less than the upper bound of the revisit time Δt max The fifth sampling point is retained, while the other fifth sampling points are filtered out.

[0146] Step S55: Based on the vector The direction to determine the drone's communication neighborhood The boundary range during access is used to perform multiple samplings within this boundary range, resulting in multiple sixth sampling points;

[0147] Step S56: Calculate the Dubins path length for all sixth sampling points.

[0148]

[0149] in: For the drone to be compared with the previous unmanned vehicle i-1 End of visit With an orientation angle h max To the sixth sampling point Path length, For drones As the starting point, along arrive The path length, where Indicates that the drone is from arrive Search for driverless cars i The angle of orientation at that time Indicates drones Starting point, to reach the next driverless car to visit. i+1 Location Dubins path length at that time

[0150] The sixth sampling point with the shortest Dubins path length was selected as the fourth alternative access point.

[0151] Step S57: Select the one with the shorter Dubins path length between the third and fourth alternative access points as the second final access point, and control the drone to pass through this second final access point to achieve the target unmanned vehicle s i A follow-up visit.

[0152] Then, the revisit of all autonomous vehicles within this cycle is completed through iteration, and this process is repeated multiple times until the preset maximum number of visit cycles is completed.

[0153] To verify the effectiveness of this invention, a simulation experiment was conducted to assess the path planning effect. The experiment used a system consisting of four unmanned vehicles (UAVs) and one drone. The initial position of the drone was (0,0), with an orientation angle of θ. The initial position of UAV 1 was (60,-10), with a speed range of [0.2 m / s, 0.7 m / s]. The initial position of UAV 2 was (-70,-20), with a speed range of [0.2 m / s, 1.5 m / s]. The initial position of UAV 3 was (40,60), with a speed range of [0.4 m / s, 1.5 m / s]. The initial position of UAV 4 was (-60,60), with a speed range of [0.5 m / s, 1.5 m / s]. The communication radius of each UAV was 5 m, and the communication radius of the drone was 20 m. The turning radius of the drone was 5 m, and the speed of the drone was 15 m / s. The paths taken by the drone during one and two periodic traversals of the UAVs are shown below. Figure 5 and Figure 6 As shown.

[0154] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for unmanned aerial vehicle (UAV) path planning, characterized in that, include: Step S1: Obtain the target driverless car to be accessed The current position and speed range, as well as the current position and heading angle of the drone, are used to determine the target drone vehicle. The next time of revisit and the upper limit of revisit time ; Step S2: Within the speed range and the lower boundary of revisit time Upper boundary of revisit time Under the constraints, determine the target unmanned vehicle The closest position that can be reached by moving at a constant speed along a known straight line. and farthest position ; Step S3: Identify the target driverless vehicle Located in the nearest position Communication neighborhood and target unmanned vehicles Located at the farthest position Between the communication neighbors, is there a common neighbor? If yes, then proceed to step S4; otherwise, proceed to step S5. Step S4: Control the drone to traverse this public neighborhood to achieve target unmanned vehicle A follow-up visit; Step S5: Based on the nearest location and farthest position Construct a search channel and control the drone to pass through this search channel to achieve target unmanned vehicle. A follow-up visit; Step S4 specifically includes: Step S41: Calculate the communication neighborhood and communication neighborhood Coordinates of the two intersecting points , ; Step S42: Determine the coordinates of the two intersection points , If the values ​​are the same, then the common neighborhood is a point. The drone is controlled to pass through the coordinates of this intersection point to achieve target drone operation. If a revisit is required, then proceed to step S43; Step S43: Using the nearest position Centered on the autonomous vehicle The communication radius is taken as the radius, and the coordinates of the two intersection points are taken as the radius. , Multiple third sampling points are obtained uniformly along the defined communication domain boundary; Step S44: Calculate the Dubins path length for each third sampling point: in, The Dubins path length for the third sampling point. This indicates that the drone is from the previous unmanned vehicle. End of visit Departure, arrival at the third sampling point The length of the Dubins path. This indicates that the drone started from the third sampling point. Depart, and arrive at the next driverless car to visit. Location Dubins path length at that time This indicates that the drone has left the sampling point. Heading angle at time And select the third sampling point with the smallest Dubins path length among all the third sampling points as the first alternative access point; Step S45: Using the farthest position Centered on the autonomous vehicle The communication radius is taken as the radius, and the coordinates of the two intersection points are taken as the radius. , Multiple fourth sampling points are obtained by uniformly applying sampling methods along the defined communication domain boundary; Step S46: Calculate the Dubins path length of each fourth sampling point, and select the fourth sampling point with the smallest Dubins path length among all fourth sampling points as the second alternative access point; Step S47: Select the one with the smaller Dubins path length between the first and second alternative access points as the first final access point; Step S48: Control the drone to pass through this first final access point to achieve target unmanned vehicle A follow-up visit.

2. The UAV path planning method according to claim 1, characterized in that, Step S1 includes: Step S11: Obtain the initial position and heading angle of the UAV; Step S12: Based on the current locations of all unvisited vehicles in this cycle, and in conjunction with the current location of the drone, determine the target unmanned vehicle that needs to be visited; Step S13: Based on the target unmanned vehicle that needs to be accessed The current position and speed range, as well as the current position and heading angle of the drone, determine the target unmanned vehicle. The next time of revisit and the upper limit of revisit time .

3. The UAV path planning method according to claim 2, characterized in that, Step S12 specifically includes: Step S121: Calculate the geometric center point based on the current positions of all unmanned vehicles; Step S122: Calculate the angle of each unvisited unmanned vehicle relative to the drone based on the obtained geometric center point; Step S123: Sort the unvisited autonomous vehicles from smallest to largest according to their corresponding angles, and select the autonomous vehicle with the smallest angle as the target autonomous vehicle to be visited.

4. The UAV path planning method according to claim 2, characterized in that, Step S13 specifically includes: Step S131: Determine the time interval between the last time the drone visited the target vehicle and the time when it completed a revisit to the previous vehicle within the current cycle. ; Step S132: Calculate the target unmanned vehicle's speed within its speed range at time intervals. The farthest distance within and closest distance and the communication neighborhood corresponding to the farthest distance. and the communication neighborhood corresponding to the nearest distance ; Step S133: Based on the vector Direction determination of UAVs for communication neighborhood Access is performed within the boundary range, within which... N Obtained by uniform sampling N The first sampling point; Step S134: Assume the target unmanned vehicle starts from the nearest distance at the minimum speed. If it continues to move in a straight line, then the communication neighborhood will be affected. For each first sampling point on the boundary, calculate the time it takes for the drone to visit each first sampling point; Step S135: Calculate the minimum access time of the UAV to all first sampling points. ; Step S136: Based on the obtained minimum value Calculate the lower bound of the revisit time of the drone to the target unmanned vehicle. ; Step S137: Based on the vector Direction determination of UAVs for communication neighborhood Access is performed within the boundary range, within which... N Obtained by uniform sampling N A second sampling point; Step S138: Assume the target autonomous vehicle starts from the furthest distance at maximum speed. If it continues to move in a straight line, then the communication neighborhood will be affected. For each second sampling point on the boundary, calculate the time it takes for the drone to visit each second sampling point; Step S139: Calculate the maximum access time of the UAV to all second sampling points. ; Step S1310: Based on the obtained maximum value Calculate the upper bound of the revisit time of the drone to the target unmanned vehicle. .

5. The UAV path planning method according to claim 1, characterized in that, Step S2 includes: Step S21: Calculate the minimum speed for the target unmanned vehicle. Lower boundary of revisit time The closest position traveled along the inner straight line. ; Step S22: Calculate the target unmanned vehicle's maximum speed Upper boundary of revisit time The farthest position traveled along the inner straight line .

6. The UAV path planning method according to claim 1, characterized in that, Step S5 specifically includes: Step S51: Using the nearest position Starting point, farthest position The endpoint is the target unmanned vehicle. The search channel is constructed using twice the communication radius as the channel width. Step S52: Based on the vector The direction to determine the drone's communication neighborhood The boundary range during access is used to perform multiple samplings within this boundary range, resulting in multiple fifth sampling points; Step S54: Filter the fifth sampling point according to the pre-configured rules, and calculate the Dubins path length of the filtered fifth sampling point. : in: For the drone to go from the previous unmanned vehicle End of visit With the direction angle to sampling point Path length, For drones As the starting point, along arrive Path length, Indicates drones Starting from the next driverless car to visit. Location Dubins path length at that time The fifth sampling point with the shortest Dubins path length was selected as the third alternative access point. Step S55: Based on the vector The direction to determine the drone's communication neighborhood The boundary range during access is used to perform multiple samplings within this boundary range, resulting in multiple sixth sampling points; Step S56: Calculate the Dubins path length for all sixth sampling points. : in: For the drone to go from the previous unmanned vehicle End of visit With the direction angle To the sixth sampling point Path length, For drones As the starting point, along arrive The path length, where Indicates that the drone is from arrive Search for driverless cars The angle of orientation at that time Indicates drones Starting point, to the next driverless car to visit. Location Dubins path length at that time The sixth sampling point with the shortest Dubins path length was selected as the fourth alternative access point. Step S57: Select the one with the shorter Dubins path length between the third and fourth alternative access points as the second final access point, and control the drone to pass through this second final access point to achieve the target unmanned vehicle. A follow-up visit.

7. The UAV path planning method according to claim 6, characterized in that, The filtering process for the fifth sampling point in step S54 specifically includes: Step S541: Calculate the position of the UAV at the time when the previous unmanned vehicle completed its visit. With its heading angle Path length to each fifth sampling point and access time ; Step S542: For each fifth sampling point, determine the path taken by the UAV along the vector. The direction, starting from the fifth sampling point, is related to the communication field. The coordinates of the meeting point are used as the first meeting point of the fifth sampling point; Step S543: For each fifth sampling point, calculate the path length and access time of the UAV from the fifth sampling point to the corresponding first meeting position, taking that fifth sampling point as the starting point. ; Step S544: Access time With access time The sum is less than the upper bound of the revisit time. The fifth sampling point is retained, while the other fifth sampling points are filtered out.

8. A drone path planning device, comprising a memory, a processor, and a program stored in the memory, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1-7.

9. A storage medium having a program stored thereon, characterized in that, When the program is executed, it implements the method as described in any one of claims 1-7.

Images