Unmanned aerial vehicle group cooperative game task allocation and path planning method and system

Abstract

The invention provides an unmanned aerial vehicle group cooperative game task allocation and path planning method and system, and belongs to the technical field of unmanned aerial vehicle control. Unmanned aerial vehicle endurance time, search capability, release-recovery point and task point space distribution are integrated, and an efficiency and coverage oriented cooperative search task allocation model is constructed; based on a cooperative game theory, making a time-distance balanced unmanned aerial vehicle group dynamic target point selection strategy; and judging the return opportunity of the unmanned aerial vehicle, resolving the distance relationship between the non-searched task points and the target return path, judging whether the non-searched task points can be added in the return or not, maximally utilizing the endurance time, and generating the shortest return path of the unmanned aerial vehicle. According to the method, the problems of unreasonable task allocation, incomplete search coverage, dynamic change of task points and the like when an unmanned system cluster composed of unmanned aerial vehicles searches for multiple task points in an unstructured environment can be solved, and support is provided for research, development and application in the aspect of cooperative task execution of the unmanned system.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle control, and in particular to a method and system for allocating tasks and planning paths for a group of unmanned aerial vehicles in a cooperative game based on a cooperative game. Background Art

[0002] The technology of cooperative dynamic search and path planning of drone swarms is an important development direction in the field of modern unmanned systems, which covers multiple key technical fields such as autonomous control of drones, multi-agent systems, sensor technology, and efficient communication networks. The core of this technology is that multiple drones can work together in complex and dynamic environments to perform efficient search, monitoring and other complex tasks.

[0003] The collaborative dynamic search and path planning technology of drone swarms has a wide range of application scenarios and significant application potential. For example, in disaster emergency response: after earthquakes, floods, forest fires and other disasters, traditional search and rescue methods face challenges such as vast areas and poor access, which limits the efficiency of rescue. However, drone swarms can quickly cover a large area and accurately locate the victims through efficient search algorithms and autonomous flight capabilities, thereby improving search and rescue efficiency and the chance of survival. Secondly, in the field of environmental monitoring, such as wildlife protection and monitoring, forest resource surveys, and environmental pollution monitoring, drone swarms can monitor surface conditions, animal and plant activities, and environmental indicators such as air quality in real time by carrying various sensors and camera equipment, providing data support and early warning services, which helps to protect the ecological environment and natural resources. In addition, in the fields of security patrols and border monitoring: collaborative search technology can enable drones to monitor border areas in real time, monitor important facilities and densely populated areas, identify abnormal situations and respond in a timely manner, strengthen border security management, and improve the level of public safety prevention.

[0004] However, there are still many challenges to realize the widespread application of UAV swarm collaborative search and path planning technology. The adaptability, stability and efficiency of intelligent control algorithms are key issues. The existing collaborative algorithms still have many defects in complex search environments and dynamic and numerous task points. It is necessary to develop efficient path planning, task allocation and collaborative decision-making algorithms to achieve good collaboration between multiple UAVs. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for dynamic task allocation and path planning of a group of drones in a cooperative game based on cooperative game, so as to solve at least one technical problem existing in the above-mentioned background technology.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a method for dynamic task allocation and path planning of a group of drones in a cooperative game, comprising:

[0008] The collaborative search task allocation model for efficiency and coverage is constructed by integrating the UAV's flight time, search capability, release-recovery points, and spatial distribution of task points.

[0009] Based on the collaborative search task allocation model oriented towards efficiency and coverage, the shortest path method is used as the basis, and based on the cooperative game theory, a dynamic target point selection strategy for drone swarms with time-distance equilibrium is formulated, so that the drone swarm can dynamically select the next target point and form a search path through the cooperative game strategy starting from the release point;

[0010] Determine the return time of the drone, and solve the distance relationship between the unsearched mission points and the target return path, decide whether to add unsearched mission points in the return journey, maximize the use of flight time, and generate the shortest drone return path.

[0011] Furthermore, the integrated UAV flight time, search capability, release-recovery points and spatial distribution of task points are used to construct a collaborative search task allocation model for efficiency and coverage, including:

[0012] The purpose is to use multiple UAVs to collaboratively search for multiple task points that have been marked in the search space, and to build a UAV search capability model as follows:

[0013] S U =(x U ,y U ,h U ,V U ,T U ,k U )

[0014] where x U ,y U is the two-dimensional coordinate of the UAV at a certain horizontal flight height, h U is the flight altitude of the UAV, V U is the uniform flying speed of the UAV, T U is the maximum flight time of the UAV, k U Indicates the number of the drone in the drone group;

[0015] The simplified search task point model is constructed as follows:

[0016] S T =(x T ,y T ,k T )

[0017] In the formula, x T ,yT is the location coordinate of the search task, k T Number the task points to simplify the search for task points as point tasks;

[0018] The drone release-recovery point model is constructed as follows:

[0019] S re =(x re ,y re ,k re ,nu re )

[0020] In the formula, x re ,y re is the location coordinate of the drone release-recovery point, k re is the number of the release-recovery point, nu re The capacity of the release-reclaim point.

[0021] Furthermore, the cooperative search task target modeling of the drone swarm is as follows:

[0022]

[0023]

[0024] Among them, T U is the set of all points that the UAV can pass through, T U =(T re ∪T s );T re Set the release and recovery points for drones (t re ∈T re ), T s is the search task point set, is a Boolean variable. When the drone u is assigned to the starting point or mission point t ui Go to uj When executing the search task, this variable is 1. Indicates that the drone is assigned by t ui Go to uj The flight time required to perform search missions;

[0025] The UAV performs the search mission under the following constraints:

[0026]

[0027] When the drone swarm performs a search mission, it is necessary to meet the requirement that the drones work together to complete the search of as many mission points as possible within the drone endurance time, and only one member is allowed to search each mission point once. After completing the search mission, the drone returns to the release-recovery point, N T Indicates the number of mission points, UT To gather the drones, The number of drones deployed.

[0028] Furthermore, based on the shortest path method and cooperative game theory, a time-distance balanced UAV swarm dynamic target point selection strategy is formulated, so that the UAV swarm dynamically selects the next target point from the release point through the cooperative game strategy and forms a search path, including:

[0029] The drones take off from their respective release-recovery points, calculate the distances from all mission points to the drone's current position, select the mission point with the smallest distance to form the optimal target point sequence, and simultaneously select the suboptimal target point with the next smallest distance to form the suboptimal target point sequence;

[0030] Check whether there are repeated elements in the optimal target point sequence. If there are repeated elements, it means that multiple drones simultaneously choose the same mission point as the next target to go to. At this time, the first judgment function is formulated based on the time each drone completes the search mission and the cost of giving up the optimal target point. The drone with the smallest function value is regarded as the owner of the optimal target point. Other drones that have not won the optimal target point will select the mission point at its corresponding position in the suboptimal target point sequence as the next search target. If the function values ​​are the same, the optimal target point is randomly assigned to one of the competing drones, and other drones will select the mission point at its corresponding position in the suboptimal target point sequence to form the first stage target point sequence.

[0031] Furthermore, the first determination function is as follows:

[0032]

[0033] In the formula Indicates the position of the drone with sequence number i before making a decision The Euclidean distance difference between the optimal target point and the suboptimal target point, α, β are weight parameters, α+β=1, time ui Indicates the time it takes for UAV i to complete the current mission point.

[0034] The target sequence of the first stage is analyzed for the second time to see if there are still repeated elements. If there are still repeated elements, the number is m, then the drone U corresponding to the repeated elements (which do not appear in the optimal target point sequence) is calculated. second-m The geometric center of the location Center, the repeated task points in the first stage target sequence (and not appearing in the optimal target point sequence) and the unsearched task points are taken as candidate groups K second-m , calculate the distance from the task point in the candidate group to the geometric center, and sort these task points from small to large According to the member who completes the task first (timeui The smaller the value, the sooner the i-th drone completes the previous task) Priority selection rules:

[0035]

[0036] From the sorted K second-m First, intercept the first m mission points; assign these m mission points to the m drones in turn to form the target point sequence next_point for all drones performing the mission.

[0037] Furthermore, the determination of the return timing of the drone, and the calculation of the distance relationship between the unsearched mission points and the target return path, determine whether to add unsearched mission points during the return journey, maximize the use of the flight time, and generate the shortest return path for the drone, include:

[0038] Calculate the distance from the current position of each drone after completing the previous task The time to fly to the target location and from the target to the nearest release-recovery point is:

[0039]

[0040] This time is equal to the UAV flight time T U Compare, if time_jud ui <T U , the drone executes the action to go to the target point; if time_jud ui >T U , the drone gives up going to the target point and immediately calculates the return path.

[0041] If the drone needs to return to the release-recovery point immediately, calculate the current drone position and distance to the current The straight line formed by the nearest release-recovery point is used to calculate the distance from the remaining unassigned task points to the straight line, as well as the vector dot product operation between the task point and the drone position, the nearest release-recovery point and the drone position, and the task point with the smallest distance to the straight line is selected:

[0042]

[0043] The filtered As the next target point of the return drone, calculate time_jud ui And T U Compare:

[0044] If the obtained time_jud ui <T UThen add this point to the next_point sequence of the corresponding drone, and the drone moves to the next target point as a whole; if time_jud ui >T U , then it means that the drone cannot add other mission points on the way back. The drone returns to the nearest release-recovery point from the current position in a straight line, and the drone that returns early is removed from the drone set U that performs the mission. T , indicating that the drone has completed the mission and returned and will not participate in subsequent search missions.

[0045] When all drones return to the release-recovery point or all mission points have been searched, the drone search mission is completed. By connecting the mission points searched by each drone in chronological order, an intuitive drone search path can be formed.

[0046] In a second aspect, the present invention provides a drone group collaborative dynamic task allocation and path planning system, comprising:

[0047] A construction module is used to integrate the UAV's flight time, search capability, release-recovery points and spatial distribution of task points to build a collaborative search task allocation model for efficiency and coverage;

[0048] The dynamic selection module is used to formulate a dynamic target point selection strategy for drone swarms based on the efficiency and coverage-oriented collaborative search task allocation model based on the shortest path method and cooperative game theory, so that the drone swarm can dynamically select the next target point from the release point through the cooperative game strategy, calculate the time to reach the target point and return to the release-recovery point, and compare it with the endurance time to determine the return time of the drone. During the return journey, the distance relationship between the unsearched task points and the target return path is calculated to decide whether to add the task point to the return path;

[0049] The generation module generates the drone search and return paths according to the target point decision of each drone.

[0050] In a third aspect, the present invention provides a non-transitory computer-readable storage medium, which is used to store computer instructions. When the computer instructions are executed by a processor, the dynamic task allocation and path planning method for collaborative game of drone groups as described in the first aspect is implemented.

[0051] In a fourth aspect, the present invention provides a computer device comprising a memory and a processor, wherein the processor and the memory communicate with each other, the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the dynamic task allocation and path planning method for collaborative game of drone groups as described in the first aspect.

[0052] In a fifth aspect, the present invention provides an electronic device, comprising: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes instructions for implementing the dynamic task allocation and path planning method for collaborative game of drone groups as described in the first aspect.

[0053] The beneficial effects of the present invention are as follows: it can solve the problems of unreasonable task allocation, incomplete search coverage, dynamic changes in task points, etc. when an unmanned system cluster composed of drones searches for multiple task points in an unstructured environment, and provide support for the research and development and application of collaborative task execution of unmanned systems.

[0054] Additional advantages of the present invention will be more clearly given in the following description or learned through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0056] Figure 1 This is a detailed diagram of the implementation principle algorithm of the method for dynamic task allocation and path planning of a group of drones in a cooperative game based on cooperative game described in an embodiment of the present invention.

[0057] Figure 2 The present invention is a flowchart of a method for dynamic task allocation and path planning of a cooperative game of a group of drones based on cooperative game according to an embodiment of the present invention.

[0058] Figure 3 Schematic diagram of an application scenario of the method for dynamic task allocation and path planning of a group of drones in a cooperative game based on cooperative game according to an embodiment of the present invention.

[0059] Figure 4 Schematic diagram of the algorithm for allocating repeated task points in the drone search for suboptimal solutions according to an embodiment of the present invention.

[0060] Figure 5 A schematic diagram of an algorithm for planning the return path of a drone and determining whether to add a mission point according to an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The embodiments described below by the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

[0062] It should be understood by those skilled in the art that unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art to which this invention belongs.

[0063] It should also be understood that terms, such as those defined in commonly used dictionaries, should be understood to have a meaning consistent with that in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless as defined herein.

[0064] Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a", "an", "said" and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements and / or groups thereof.

[0065] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. Different embodiments or examples described in this specification and features of different embodiments or examples may be combined and combined by those skilled in the art without contradiction.

[0066] To facilitate understanding of the present invention, the present invention is further explained below with reference to specific embodiments in conjunction with the accompanying drawings, and the specific embodiments do not constitute a limitation on the embodiments of the present invention.

[0067] Those skilled in the art should understand that the drawings are merely schematic diagrams of embodiments, and the components in the drawings are not necessarily necessary for implementing the present invention.

[0068] Example 1

[0069] In this embodiment 1, a drone swarm collaborative cooperative game dynamic task allocation and path planning system is first provided, and the system includes: a construction module, which is used to comprehensively consider the drone's endurance time, search capability, release-recovery point and task point spatial distribution, and construct a collaborative search task allocation model for efficiency and coverage; a dynamic selection module, which is used to formulate a time-distance balanced drone swarm dynamic target point selection strategy based on the collaborative search task allocation model for efficiency and coverage, based on the shortest path method and cooperative game theory, so that the drone swarm dynamically selects the next target point from the release point through the cooperative game strategy, and determines the drone's return timing, solves the distance relationship between the unsearched task point and the target return path on the return journey, and decides whether to add the unsearched task point on the return journey to maximize the use of the endurance time; a generation module, which generates the drone search and return path according to the target point decision of each drone.

[0070] In this embodiment, the above system is used to implement a method for dynamic task allocation and path planning of a group of drones in a cooperative game, including:

[0071] A collaborative search task allocation model for efficiency and coverage is constructed by integrating factors such as drone flight time, search capability, release-recovery points, and spatial distribution of task points.

[0072] Based on the shortest path method and cooperative game theory, a time-distance equilibrium dynamic target point selection strategy for drone swarms is formulated, so that the drone swarm can dynamically select the next target point through a cooperative game strategy starting from the release point;

[0073] Determine the return timing of the drone, and solve the distance relationship between the unsearched mission points and the target return path, decide whether to add the unsearched mission points to the return path, and generate the drone search and return path according to the target point decision of each drone.

[0074] The above-mentioned integrated UAV flight time, search capability, release-recovery points and spatial distribution of task points are used to construct a collaborative search task allocation model for efficiency and coverage, including:

[0075] The purpose is to use multiple UAVs to collaboratively search for multiple task points that have been marked in the search space, and to build the UAV search capability model as follows:

[0076] S U =(x U ,y U ,h U ,V U ,T U ,k U ) (1)

[0077] where xU ,y U The UAV is at a certain horizontal height h U The two-dimensional coordinates of h U is the flight altitude of the UAV, V U is the uniform flying speed of the UAV, T U is the maximum flight time of the UAV, k U It indicates the number of the drone in the drone group. The members in the drone group in the present invention have exactly the same functions, search capabilities, endurance, appearance, etc. The number is the unique identifier of the drone.

[0078] The simplified search task point model is constructed as follows:

[0079] S T =(x T ,y T ,k T ) (2)

[0080] In the formula, x T ,y T is the location coordinate of the search task, simplifying the search task point to a point task. At this time, the time to perform the search task at this point is close to 0, k T The number of the task point.

[0081] The drone release-recovery point model is constructed as follows:

[0082] S re =(x re ,y re ,k re ,nu re ) (3)

[0083] In the formula, x re ,y re is the position coordinate of the release-recovery point of the drone. This patent is aimed at non-movable release-recovery points, that is, all release-recovery points are fixed position coordinates, simplifying the search task point to a point task. At this time, the time to perform the search task at this point is close to 0, k re is the number of the release-recovery point, nu re For the release-recovery point capacity, that is, the maximum number of drones that can be accommodated, it is stipulated that when a drone returns to a release-recovery point, it does not have to return to the starting position, but only needs to land at one of the release-recovery points that has not reached capacity.

[0084] The modeling of the UAV swarm collaborative search mission objective is as follows:

[0085]

[0086]

[0087] Formula (4) is the time objective function of the UAV task allocation, T U is the set of all points that the UAV can pass through, T U =(T re ∪T s );T re Set the release and recovery points for drones (t re ∈T re ), T s is the search task point set, is a Boolean variable. When the drone u is assigned to the starting point or mission point t ui Go to uj When executing the search task, this variable is 1. Indicates that the drone is assigned by t ui Go to uj The flight time required to perform the search task. Formula (5) is the objective function of the task allocation coverage of the drone. This patent takes into account that in actual search tasks, the search environment is large, the task points are numerous, and the number of drones is limited. Some task points may not be searched within the drone's flight time.

[0088] The UAV performs the search mission under the following constraints:

[0089]

[0090] When the drone swarm performs a search mission, it is necessary to meet the requirement that the drones work together to complete the search of as many mission points as possible within the drone endurance time, and only one member is allowed to search each mission point once. After completing the search mission, the drone returns to the release-recovery point, N T Indicates the number of mission points, U T To gather the drones, is the number of dispatched drones. Constraints 6, 7, and 8 in (6) collectively indicate that each mission point is visited at most once, and the path of each drone cannot have a back-and-forth situation. Constraint 9 indicates that the search path of each drone is continuous.

[0091] The shortest path method is based on cooperative game theory to formulate a time-distance balanced UAV swarm dynamic target point selection strategy, so that the UAV swarm dynamically selects the next target point from the release point through the cooperative game strategy and forms a search path, including:

[0092] In the first step, the drones take off from their respective release-recovery points, calculate the distances from all mission points to the current location of the drone, select the mission point with the smallest distance, and form the optimal target point sequence. At the same time, select the suboptimal target point with the smallest distance to form a suboptimal target point sequence in Indicates the mission point number selected by the UAV numbered n;

[0093] The second step is to check whether there are repeated elements in the optimal target point sequence, which means that multiple drones simultaneously choose the same task point as the next target to go to. At this time, a strategy (7) is formulated based on the time it takes for each drone to complete the search task and the cost of giving up the optimal target point. The drone with the smallest function value is selected as the optimal target point. Other drones that have not won the optimal target point will select the task point at its corresponding position in the suboptimal target point sequence as the next search target. If the function values ​​are the same, the optimal target point is randomly assigned to one of the competing drones, and the other drones will select the task point at its corresponding position in the suboptimal target point sequence to form the target point sequence of the first stage. Design the first judgment function as follows:

[0094]

[0095] In the formula Indicates the position of the drone with sequence number i before making a decision The Euclidean distance difference between the optimal target point and the suboptimal target point, α, β are weight parameters, α+β=1, time ui Indicates the time it takes for UAV i to complete the current mission point.

[0096] The third step is to sequence the first stage targets Perform a second analysis to see if there are still repeated elements. If there are still repeated elements, filter out the repeated elements (which do not appear in the optimal target point sequence) and the corresponding drones. The number of repeated elements and drones is m, then calculate the drone U corresponding to the repeated element. second-m The geometric center of the position Center, the repeated elements in the once sequence (and not in the min nex-point The task points that have not been searched are regarded as candidate groups. The distances from the task points in the candidate groups to the geometric center are calculated, and they are sorted from small to large. The first m task points K are intercepted from the first one. second-m , based on the rule that the member who completes the task first is given priority:

[0097]

[0098] Assign these m mission points to these m drones to form a target point sequence for all drones performing the mission.

[0099] The above-mentioned determination of the return timing of the drone, and solving the distance relationship between the unsearched mission points and the target return path, decides whether to add the unsearched mission points to the return path, and generates the drone search and return path according to the target point decision of each drone, including:

[0100] Step 1: Calculate the time it takes for each drone to fly from its current position to the target point and from the target point to the nearest release-recovery point after completing the previous mission, that is:

[0101]

[0102] where time_jud ui It represents the return judgment parameter of drone i, is the current mission point location of UAV i, is the next mission point corresponding to the drone in the next_point sequence. The release-recovery point closest to the next target task point. If time_jud ui <T U Then the drone numbered i will go to the target point; if time_jud ui >T U The drone immediately returned to the release-recovery point.

[0103] Step 2: If the drone needs to return to the release-recovery point immediately after the previous step, calculate the straight line formed by the current drone position and the nearest release-recovery point, calculate the distance from the remaining unassigned task points to the straight line, and the vector dot product of the two vectors formed by the task point and the drone position, the nearest release-recovery point and the drone position, and select the task point with the vector dot product greater than 0 and the smallest distance to the straight line:

[0104]

[0105] Use it as an alternative point to calculate formula (9), if the obtained time_jud ui <T U Then add this point to the next_point sequence of the corresponding drone, and the drone moves to the next target point as a whole. After arriving, the optimal target point sequence min is calculated repeatedly. nex-point , suboptimal target point sequence s-min nex-point , and follow the following steps to determine the target point sequence next_point and solve the return judgment parameter time_jud ui If time_jud ui >T U, then it means that the drone cannot add other mission points on the way back. The drone returns to the nearest release-recovery point from the current position in a straight line, and the drone that returns early is removed from the drone set U that performs the mission. T , indicating that the drone has completed the mission and returned and will not participate in subsequent search missions.

[0106] When all drones return to the release-recovery point or all mission points have been searched, the drone search mission is completed. By connecting the mission points searched by each drone in chronological order, an intuitive drone search path can be formed.

[0107] Example 2

[0108] The implementation principle and algorithm details of the method for dynamic task allocation and path planning of cooperative game based on cooperative game of drone groups provided in this embodiment are as follows: Figure 1 As shown in the figure, the main process includes the construction of the UAV swarm collaborative search task allocation model, the allocation of search target points based on cooperative game, and the judgment of the UAV return timing and path generation.

[0109] The processing flow of the method for dynamic task allocation and path planning of cooperative game of drone groups provided in this embodiment is as follows: Figure 2 As shown, the processing steps include the following:

[0110] Step S1: The above-mentioned comprehensive UAV flight time, search capability, release-recovery points and spatial distribution of task points are used to construct a collaborative search task allocation model for efficiency and coverage.

[0111] The schematic diagram of the application scenario of the method for dynamic task allocation and path planning of cooperative game based on UAV swarm provided in this embodiment is as follows: Figure 3 As shown. The purpose is to use multiple UAVs to collaboratively search for multiple task points that have been marked in the search space, and to build the UAV search capability model as follows:

[0112] S U =(x U ,y U ,h U ,V U ,T U ,k U ) (1)

[0113] where x U ,y U is the two-dimensional coordinate of the UAV on a certain horizontal plane, h U is the flight altitude of the UAV, V U is the uniform flying speed of the UAV, T U is the maximum flight time of the UAV, kU Indicates the number of the drone in the drone group. In this embodiment, the members of the drone group have exactly the same functions, search capabilities, endurance, appearance, etc. The number is the unique identifier of the drone. At the same time, it is stipulated that the drone maintains a uniform speed during the search mission, and the drone group takes off at the same time when performing the search mission.

[0114] The simplified search task point model is constructed as follows:

[0115] S T =(x T ,y T ,k T ) (2)

[0116] In formula (2), x T ,y T is the location coordinate of the search task, simplifying the search task point to a point task. At this time, the time to perform the search task at this point is close to 0, k T is the number of the task point. The number is the unique identifier of the task point. It can solve the problem of dynamic increase of task points, that is, the task points will increase over time, and the newly added task points will obtain a new unique corresponding number k T .

[0117] The drone release-recovery point model is constructed as follows:

[0118] S re =(x re ,y re ,k re ,nu re ) (3)

[0119] In the formula, x re ,y re is the position coordinate of the release-recovery point of the drone. For non-movable release-recovery points, that is, all release-recovery points are fixed position coordinates. The position of the release-recovery point is determined by the search environment and is generally at the boundary of the search environment. k re is the number of the release-recovery point, nu re For the release-recovery point capacity, that is, the maximum number of drones that can be accommodated, it is stipulated that when a drone returns to a release-recovery point, it does not have to return to the starting position, but only needs to land at one of the release-recovery points that has not reached capacity.

[0120] The modeling of the UAV swarm collaborative search mission objective is as follows:

[0121]

[0122]

[0123] Formula (4) is the time objective function of the UAV task allocation, T U is the set of all points that the UAV can pass through, T U =(T re ∪T s );T re Set the release and recovery points for drones (t re ∈T re ), T s is the search task point set, is a Boolean variable. When the drone u is assigned to the starting point or mission point t ui Go to uj When executing the search task, this variable is 1. Indicates that the drone is assigned by t ui Go to uj The flight time required to perform the search task. Formula (5) is the objective function of the UAV's task allocation coverage rate. Considering that in actual search tasks, the search environment is large, the task points are numerous, and the number of UAVs is limited, some task points may not be searched within the UAV's flight time.

[0124] The UAV performs the search mission under the following constraints:

[0125]

[0126] When the drone swarm performs a search mission, it is necessary to meet the requirement that the drones work together to complete the search of as many mission points as possible within the drone endurance time, and only one member is allowed to search each mission point once. After completing the search mission, the drone returns to the release-recovery point, N T Indicates the number of mission points, U T To gather the drones, is the number of dispatched drones. The first expression of formula (6) indicates that the number of dispatched drones does not exceed the total number of drones in the drone group; the second expression indicates that the time from takeoff to return of each drone must be within the endurance time; the third expression indicates that the total number of mission points searched by all drones does not exceed the total number of mission points; the fourth expression indicates that the number of drones leaving the release-recovery point must be equal to the number of drones returning to the release-recovery point; the fifth expression indicates that the number of drones recovered by each release-recovery point cannot exceed its maximum capacity; the sixth, seventh, and eighth expressions indicate that each mission point only needs to be visited once at most, and there cannot be a situation of returning back and forth; the ninth expression indicates that the search mission point of each drone has a unique entrance and exit, and there will be no path interruption; the tenth and eleventh expressions indicate that each drone performing a mission needs to take off and return to the release-recovery point.

[0127] Step S2: Based on the shortest path method and cooperative game theory, a time-distance balanced UAV swarm dynamic target point selection strategy is formulated, so that the UAV swarm dynamically selects the next target point from the release point through the cooperative game strategy and forms a search path, including:

[0128] In the first step, the drones take off from their respective release-recovery points at the same time, calculate the distances from all mission points to the current location of the drone, select the mission point with the smallest distance, and form the optimal target point sequence. For example, min nex-point (13,2,2,2), means there are 4 drones executing the task at this time, and the mission point numbered 2 is the closest to the positions of the last three drones. At the same time, the suboptimal target point with the second smallest distance is selected to form a suboptimal target point sequence For example, s-min nex-point (19,4,3,3), where Indicates the mission point number selected by the UAV numbered n;

[0129] The second step is to check whether there are repeated elements in the optimal target point sequence, which means that multiple drones simultaneously select the same mission point as the next target to go to. According to the above example, the mission point numbered 2 is the closest to the positions of the last three drones, that is, the mission point is contested by three drones. At this time, a strategy (7) is formulated based on the time it takes each drone to complete the search mission and the cost of giving up the optimal target point. The drone with the smallest function value is selected as the optimal target point. Other drones that have not won the optimal target point will select the mission point at its corresponding position in the suboptimal target point sequence as the next search target. If the function values ​​are the same, the optimal target point will be randomly assigned to one of the competing drones, and the other drones will select the mission point at its corresponding position in the suboptimal target point sequence, forming the first stage target point sequence. Design the first judgment function as follows:

[0130]

[0131] In formula (7) Indicates the position of the drone with sequence number i before making a decision The Euclidean distance difference between the optimal target point and the suboptimal target point, α, β are weight parameters, α+β=1, time ui It indicates the time taken by drone i to complete the current task point. Assuming that the second drone has the smallest judgment function, the task point numbered 2 is assigned to it. The third and fourth drones choose the corresponding (3,3) in the suboptimal solution, so once (13,2,3,3) is obtained.

[0132] The third step is to sequence the first stage targets Perform a second analysis to see if there are still duplicate elements. If there are still duplicate elements, the number is m, such as Figure 4 , then calculate the drone U corresponding to the repeated element second-m The geometric center Center of the location. For the previous example, there are still two repeated elements (3,3) in once. Then calculate the geometric center of the last two drones. Take the repeated elements in the once sequence (excluding the task points and corresponding drones determined by formula 7) and the unsearched task points as candidate groups. Calculate the distance from the task points in the candidate group to the geometric center, sort them from small to large, and intercept the first m task points K from the first one. second-m , based on the rule that the member who completes the task first is given priority:

[0133]

[0134] Assign these m mission points to these m drones to form a target point sequence for all drones performing the mission. According to the above example, the last two drones compete for the task point numbered 3. Assuming that the fourth drone completes the search task of the previous point first, the task point numbered 3 is assigned to it, because theoretically, this point is the shortest distance from the geometric center of the third and fourth drones. Then the task point that is second only to the task point numbered 3 from the geometric center is assigned to the third drone, assuming it is the task point numbered 9, and we can get next_point(13,2,9,3).

[0135] Step S3: The above-mentioned determination of the return timing of the drone, and the calculation of the distance relationship between the unsearched mission points and the target return path, determine whether to add unsearched mission points in the return journey, maximize the use of the flight time, and generate the drone search and return path according to the target point decision of each drone, including:

[0136] Step 1: Calculate the time it takes for each drone to fly from its current position to the target point and from the target point to the nearest release-recovery point after completing the previous mission, that is:

[0137]

[0138] where time_jud ui It represents the return judgment parameter of drone i, is the current mission point location of UAV i, is the next mission point corresponding to the drone in the next_point sequence. The release-recovery point closest to the next target task point. If time_jud ui <T UThen the drone numbered i will go to the target point; if time_jud ui >T U The drone immediately returned to the release-recovery point.

[0139] Step 2: If the drone needs to return to the release-recovery point immediately after the previous step, Figure 5 , then calculate the straight line formed by the current drone position and the nearest release-recovery point, calculate the distance from the remaining unassigned task points to the straight line, and the vector dot product operation of the two vectors formed by the task point and the drone position, the nearest release-recovery point and the drone position, and select the task point with the vector dot product result greater than 0 and the smallest distance to the straight line:

[0140]

[0141] When the vector dot multiplication result is greater than 0, it means that the drone's forward direction is consistent with the direction of the nearest release-recovery point, to avoid the drone flying in the opposite direction. The selected mission point is used as the target point. Calculation formula (9), if the obtained time_jud ui <T U Then add this point to the next_point sequence of the corresponding drone, and the drone moves to the next target point as a whole. After arriving, it repeatedly calculates the nearest target point, the suboptimal target point, and then determines the target point according to the subsequent cooperative competition rules. The cycle continues until the end condition is reached. If time_jud ui >T U , then it means that the drone cannot add other mission points on the way back. The drone returns to the nearest release-recovery point from the current position in a straight line, and the drone that returns early is removed from the drone set U that performs the mission. T , indicating that the drone has completed the mission and returned and will not participate in subsequent search missions.

[0142] The drone starts to calculate the next stage target point from takeoff, determines whether it needs to return, and whether it can add mission points on the way back. When all drones return to the release-recovery point or all mission points are searched, the drone search mission is completed and the cycle is stopped. Connecting the mission points searched by each drone in chronological order can form an intuitive drone search path.

[0143] In summary, the cooperative game-based UAV group collaborative cooperative game dynamic task allocation and path planning method provided by the embodiment of the present invention takes searching as many task points as possible in the shortest possible time as the core goal, simplifies the search task into a point task, constructs a group search model for release-recovery points, UAVs, search tasks and targets, calculates the optimal task point sequence and the suboptimal task point sequence closest to the current UAV position based on the shortest path, and uses the cooperative competition rule to obtain the final target task point sequence through two-stage allocation for the repeated task points in the optimal task points that are jointly competed by multiple UAVs, determines whether each UAV executes the target point to meet the endurance time constraint, plans the return path for the UAVs that do not meet the constraint, calculates whether the task point can be added to the path, and finally realizes dynamic UAV search task allocation and path planning. It can solve the problems of unreasonable task allocation, incomplete search coverage, dynamic changes in task points, etc. when the unmanned system cluster composed of UAVs searches for multiple task points in an unstructured environment, and provides support for the research and development and application of cooperative task execution of unmanned systems.

[0144] Example 3

[0145] This embodiment 3 provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the above-mentioned method for dynamic task allocation and path planning of a group of drones in a cooperative game is implemented. The method includes:

[0146] The collaborative search task allocation model for efficiency and coverage is constructed by integrating the UAV's flight time, search capability, release-recovery points, and spatial distribution of task points.

[0147] Based on the efficiency and coverage-oriented collaborative search task allocation model, the shortest path method is used as the basis, and based on cooperative game theory, a dynamic target point selection strategy for drone swarms with time-distance equilibrium is formulated, so that the drone swarm can dynamically select the next target point and form a search path through the cooperative game strategy starting from the release point;

[0148] Determine the return timing of the drone, and solve the distance relationship between the unsearched mission points and the target return path, decide whether to add unsearched mission points on the return journey, maximize the use of flight time, and generate the drone search and return path according to the target point decision of each drone.

[0149] Example 4

[0150] This embodiment 4 provides a computer device, including a memory and a processor, the processor and the memory communicate with each other, the memory stores program instructions executable by the processor, the processor calls the program instructions to execute the above-mentioned drone group collaborative game dynamic task allocation and path planning method, the method comprising:

[0151] The collaborative search task allocation model for efficiency and coverage is constructed by integrating the UAV's flight time, search capability, release-recovery points, and spatial distribution of task points.

[0152] Based on the collaborative search task allocation model oriented towards efficiency and coverage, the shortest path method is used as the basis, and based on the cooperative game theory, a dynamic target point selection strategy for drone swarms with time-distance equilibrium is formulated, so that the drone swarm can dynamically select the next target point and form a search path through the cooperative game strategy starting from the release point;

[0153] Determine the return timing of the drone, and solve the distance relationship between the unsearched mission points and the target return path, decide whether to add unsearched mission points on the return journey, maximize the use of flight time, and generate the drone search and return path according to the target point decision of each drone.

[0154] Example 5

[0155] This embodiment 5 provides an electronic device, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes instructions to implement the above-mentioned method for dynamic task allocation and path planning of drone group collaborative game, the method comprising:

[0156] The collaborative search task allocation model for efficiency and coverage is constructed by integrating the UAV's flight time, search capability, release-recovery points, and spatial distribution of task points.

[0157] Based on the collaborative search task allocation model oriented towards efficiency and coverage, the shortest path method is used as the basis, and based on the cooperative game theory, a dynamic target point selection strategy for drone swarms with time-distance equilibrium is formulated, so that the drone swarm can dynamically select the next target point and form a search path through the cooperative game strategy starting from the release point;

[0158] Determine the return timing of the drone, and solve the distance relationship between the unsearched mission points and the target return path, decide whether to add unsearched mission points on the return journey, maximize the use of flight time, and generate the drone search and return path according to the target point decision of each drone.

[0159] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0160] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0161] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0162] These computer program instructions can also be loaded onto a computer or other programmable data processing device, and a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0163] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative work on the basis of the technical solution disclosed in the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for task allocation and path planning of a group of drones in a cooperative game, characterized in that: include: The collaborative search task allocation model for efficiency and coverage is constructed by integrating the UAV's flight time, search capability, release-recovery points, and spatial distribution of task points. Based on the efficiency and coverage-oriented collaborative search task allocation model, the shortest path method is used as the basis, and based on cooperative game theory, a dynamic target point selection strategy for drone swarms with time-distance equilibrium is formulated, so that the drone swarm can dynamically select the next target point and form a search path through the cooperative game strategy starting from the release point; Determine the return timing of the drone, and solve the distance relationship between the unsearched mission points and the target return path, decide whether to add unsearched mission points on the return journey, maximize the use of flight time, and generate the drone search and return path according to the target point decision of each drone.

2. The method for task allocation and path planning of UAV swarm collaborative game according to claim 1 is characterized in that: The above-mentioned integrated UAV flight time, search capability, release-recovery point and task point spatial distribution are used to construct a collaborative search task allocation model, including: The purpose is to use multiple UAVs to collaboratively search for multiple task points that have been marked in the search space, and to build a UAV search capability model as follows: S U =(x U ,y U ,h U ,V U ,T U ,k U ) where x U ,y U is the two-dimensional coordinate of the UAV at a certain horizontal height, h U is the flight altitude of the UAV, V U is the uniform flying speed of the UAV, T U is the maximum flight time of the UAV, k U Indicates the number of the drone in the drone group; The simplified search task point model is constructed as follows: S T =(x T ,y T ,k T ) In the formula, x T ,y T is the location coordinate of the search task, simplifying the search task point to a point task; The drone release-recovery point model is constructed as follows: S re =(x re ,y re ,k re ,now re ) In the formula, x re ,y re is the location coordinate of the drone release-recovery point, k re is the number of the release-recovery point, nu re To release-reclaim point capacity.

3. The method for task allocation and path planning of UAV swarm collaborative game according to claim 1, characterized in that: Constructing a collaborative search task allocation model for efficiency and coverage: That is, the collaborative search task allocation goals for efficiency and coverage, and constructing a drone swarm collaborative search task target model as follows: Among them, T U is the set of all points that the UAV can pass through, T U =(T re ∪T s );T re Set the release and recovery points for drones (t re ∈T re ), T s is the set of search task points, is a Boolean variable. When the drone u is assigned to the starting point or mission point t ui Go to uj When executing the search task, this variable is 1. Indicates that the drone is assigned by t ui Go to uj The flight time required to perform search missions; The UAV performs the search mission under the following constraints: When the drone swarm performs a search mission, it needs to meet the following requirements: within the drone flight time, each drone starts from the release-recovery point, and the drones work together to complete the search of as many mission points as possible, and all mission points are allowed to be searched by at most one member once. The search path of each drone is continuous and uninterrupted, and it returns to the release-recovery point after completing the search mission. T Indicates the number of mission points, U T To gather the drones, The number of drones deployed.

4. The method for task allocation and path planning of UAV swarm collaborative game according to claim 1, characterized in that: Based on the shortest path method and cooperative game theory, a time-distance equilibrium UAV swarm dynamic target point selection strategy is formulated, so that the UAV swarm dynamically selects the next target point from the release point through the cooperative game strategy and forms a search path, including: The drones take off from their respective release-recovery points, calculate the distances from all mission points to the drone's current position, select the mission point with the smallest distance to form the optimal target point sequence, and simultaneously select the suboptimal target point with the next smallest distance to form the suboptimal target point sequence; Check whether there are repeated elements in the optimal target point sequence, which means that multiple drones simultaneously select the same mission point as the next target to go to. At this time, the first judgment function is formulated based on the time each drone completes the search mission and the cost of giving up the optimal target point. The drone with the smallest function value is regarded as the owner of the optimal target point. Other drones that have not won the optimal target point will select the mission point at its corresponding position in the suboptimal target point sequence as the next search target. If the function values ​​are the same, the optimal target point is randomly assigned to one of the competing drones, and other drones will select the mission point at its corresponding position in the suboptimal target point sequence to form the first stage target point sequence. The target sequence of the first stage is analyzed for the second time to see if there are still repeated elements. If there are still repeated elements (and they do not appear in the optimal target point sequence), the number is m, and the corresponding drones U of these repeated elements are calculated. second-m The geometric center Center of the location, the task points that are repeated in the first stage target sequence but do not appear in the optimal target point sequence and the task points that have not been searched are taken as candidate groups, the distances from the task points in the candidate group to the geometric center are calculated, and they are sorted from small to large. The rule of giving priority to members who complete the task first, that is, the second judgment function, is used to form the target point sequence of all drones performing the task.

5. The method for task allocation and path planning of UAV swarm collaborative game according to claim 4 is characterized in that: The first judgment function is as follows: In the formula Indicates the position of the drone with sequence number i before making a decision The Euclidean distance difference between the optimal target point and the suboptimal target point, α, β are weight parameters, α+β=1, time ui Indicates the time it takes UAV i to complete the current mission point; According to the rule that members who complete the task first are given priority, the second judgment function is as follows: From the sorted K second-m First, intercept the first m mission points; assign these m mission points to the m drones in turn to form the target point sequence next_point for all drones performing the mission.

6. The method for task allocation and path planning of UAV swarm collaborative game according to claim 1, characterized in that: The above-mentioned determination of the return timing of the drone, and the calculation of the distance relationship between the unsearched mission points and the target return path, determine whether to add unsearched mission points in the return journey, maximize the use of the endurance time, and generate the drone search and return path according to the target point decision of each drone, including: To determine the timing of the drone's return, calculate the time it takes for each drone to fly from its current location to the target location and from the target location to the nearest release-recovery point after completing its previous mission: where time_jud ui Used as the return judgment parameter of drone i, is the current mission point location of UAV i, is the next mission point corresponding to the drone in the next_point sequence. The release-recovery point closest to the next target task point. If time_jud ui <T U Then the drone numbered i will go to the target point; if time_jud ui >T U The drone immediately returned to the release-recovery point. The drone starts to calculate the next stage target point from takeoff, determines whether it needs to return, and whether it can add mission points on the way back. When all drones return to the release-recovery point or all mission points are searched, the drone search mission is completed and the cycle is stopped. Connecting the mission points searched by each drone in chronological order can form an intuitive drone search path.

7. The method for task allocation and path planning of UAV swarm collaborative game according to claim 6, characterized in that: Calculate the relationship between the unsearched task points and the target return path, and determine whether the unsearched task points can be added to the return path: that is, calculate the straight line formed by the current drone position and the nearest release-recovery point, calculate the distance from the remaining unassigned task points to the straight line, and the vector dot product operation of the two vectors formed by the task point and the drone position, the nearest release-recovery point and the drone position, and select the task point whose vector dot product result is greater than 0 and whose distance to the straight line is the smallest: When the vector dot multiplication result is greater than 0, it means that the drone's forward direction is consistent with the direction of the nearest release-recovery point, to avoid the drone flying in the opposite direction. The selected mission point is used as the target point. Calculate time and time_jud ui , if the obtained time_jud ui <T U Then this point is added to the next_point sequence of the corresponding drone, and the drone moves to the next target point as a whole. After arriving, it repeatedly calculates the nearest target point and the suboptimal target point, and then determines the target point according to the subsequent cooperative competition rules. The cycle continues until all drones return to the release-recovery point or all mission points are searched. If still time_jud ui >T U , then it means that the drone cannot add other mission points on the way back. The drone returns to the nearest release-recovery point from the current position in a straight line, and the drone that returns early is removed from the drone set U that performs the mission. T , indicating that the drone has completed the mission and returned and will not participate in subsequent search missions.

8. A UAV swarm collaborative game task allocation and path planning system, characterized in that: include: A construction module is used to integrate the UAV's flight time, search capability, release-recovery points and spatial distribution of task points to build a collaborative search task allocation model for efficiency and coverage; The dynamic selection module is used to formulate a time-distance balanced dynamic target point selection strategy for the drone swarm based on the collaborative search task allocation model oriented to efficiency and coverage, based on the shortest path method and cooperative game theory, so that the drone swarm can dynamically select the next target point from the release point through the cooperative game strategy and form a search path, determine the return time of the drone, and solve the distance relationship between the unsearched task point and the target return path, and determine whether the unsearched task point can be added to the return path; The generation module generates the drone search and return paths according to the target point decision of each drone.

9. A computer device, characterized in that: It includes a memory and a processor, the processor and the memory communicate with each other, the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the drone group collaborative game task allocation and path planning method as described in any one of claims 1-8.

10. An electronic device, characterized in that: include: A processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes instructions for implementing the method for task allocation and path planning of a group of drones in collaborative game as described in any one of claims 1 to 8.

Images