Civil heterogeneous unmanned cluster path coordination method and system based on unmanned cooperation

Abstract

The invention discloses a civil heterogeneous unmanned cluster path coordination method and system based on unmanned-unmanned cooperation, and belongs to the field of civil unmanned aerial vehicle path coordination, and the method comprises the steps: obtaining an unmanned aerial vehicle cooperative flight task, carrying out the analysis of the unmanned aerial vehicle cooperative flight task, and building a subtask monomer; idle unmanned aerial vehicle self-inspection is executed, a feature set of idle unmanned aerial vehicles is established, feature matching is performed, and unmanned aerial vehicle-subtask mapping is established; establishing a search cluster and a following cluster; according to the pre-flight trajectory, executing real-time environmental data acquisition by using the search cluster; and pre-flight trajectory adjustment is carried out to complete path coordination of the unmanned aerial vehicle cooperative flight task. The technical problems that in the prior art, the environment adaptability of civil unmanned aerial vehicle cluster path planning is insufficient, and the heterogeneous cluster cooperation efficiency is not high are solved, and the technical effects that environment perception and trajectory optimization are carried out through cooperation of the search cluster and the follow-up cluster, and the environment adaptability and cooperation efficiency of the heterogeneous unmanned aerial vehicle cluster are improved are achieved.

Description

Technical Field

[0001] The present invention relates to the field of civil unmanned aerial vehicle path coordination, and in particular to a civil heterogeneous unmanned aerial vehicle cluster path coordination method and system based on human-machine collaboration. Background Art

[0002] With the rapid development of drone technology, the application of civil drones is becoming more and more extensive, such as agricultural plant protection, power inspection, disaster relief, etc. In order to complete complex tasks, multiple drones are usually required to work together to form a drone cluster. However, the existing cluster path planning methods are mostly aimed at homogeneous drones, and lack application to heterogeneous drone clusters. In practical applications, due to task requirements and performance limitations, heterogeneous drone collaboration is becoming more and more common, and the path planning of homogeneous drone clusters is applied to the path planning of heterogeneous drone clusters, which will lead to low efficiency of heterogeneous cluster collaboration; at the same time, when facing complex environments, such as complex and changeable terrain, electromagnetic interference, and numerous obstacles, it will make it difficult for the pre-planned path to meet the needs of real-time changes, and the environmental adaptability of heterogeneous drone clusters is insufficient. Summary of the invention

[0003] This application aims to solve the technical problems of insufficient environmental adaptability of civil UAV cluster path planning and low efficiency of heterogeneous cluster collaboration in the prior art by providing a civil heterogeneous unmanned cluster path coordination method and system based on unmanned collaboration.

[0004] In view of the above problems, the present application provides a civilian heterogeneous unmanned cluster path coordination method and system based on unmanned collaboration.

[0005] The first aspect disclosed in the present application provides a path coordination method for a civil heterogeneous unmanned cluster based on unmanned collaboration, the method comprising: obtaining a UAV collaborative flight mission, parsing the UAV collaborative flight mission, and establishing N subtask monomers; performing initial environment collection according to the UAV collaborative flight mission, establishing a basic environment library, performing self-check of idle UAVs, and establishing a feature set of idle UAVs, the feature set including performance characteristics and power characteristics, performing feature matching with the basic environment library, N subtask monomers and the feature set, completing task pairing of idle UAVs with N subtask monomers, and establishing N UAV-subtask mappings; classifying N UAVs according to the N UAV-subtask mappings, and establishing a search cluster and a following cluster; planning the flight trajectory of the search cluster and the pre-flight trajectory of the following cluster according to the corresponding subtasks, and performing real-time environment data collection with the search cluster; adjusting the pre-flight trajectory with the real-time environment data collection results and the subtask monomers of the following cluster, and updating the flight control parameters to complete the path coordination of the UAV collaborative flight mission.

[0006] Another aspect disclosed in the present application provides a civil heterogeneous unmanned cluster path coordination system based on unmanned collaboration, the system comprising: a flight mission acquisition module, used to acquire unmanned aerial vehicle collaborative flight missions, and analyze the unmanned aerial vehicle collaborative flight missions to establish N subtask monomers; a task mapping establishment module, used to perform initial environment collection according to the unmanned aerial vehicle collaborative flight missions, establish a basic environment library, perform idle unmanned aerial vehicle self-checks, and establish a feature set of idle unmanned aerial vehicles, the feature set including performance characteristics and power characteristics, perform feature matching with the basic environment library, N subtask monomers and feature sets, complete task pairing of idle unmanned aerial vehicles with N subtask monomers, and establish N unmanned aerial vehicle-subtask mappings; an unmanned aerial vehicle classification module, used to classify N unmanned aerial vehicles according to the N unmanned aerial vehicle-subtask mappings, establish a search cluster and a following cluster; an environmental data collection module, used to plan the flight trajectory of the search cluster and the pre-flight trajectory of the following cluster according to the corresponding subtasks, and perform real-time environmental data collection with the search cluster; a trajectory adjustment module, used to adjust the pre-flight trajectory with the real-time environmental data collection results and the subtask monomers of the following cluster, and update the flight control parameters to complete the path coordination of the unmanned aerial vehicle collaborative flight mission.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] By acquiring and analyzing the cooperative flight mission of UAVs, establishing N subtask monomers, and decomposing complex tasks into multiple subtasks, the foundation is laid for the task allocation of heterogeneous UAVs; initial environment collection is carried out according to the cooperative flight mission of UAVs, a basic environment library is established, self-check of idle UAVs is performed, and a feature set of idle UAVs is established. Feature matching is performed with the basic environment library, N subtask monomers and feature sets to complete the task pairing of idle UAVs with N subtask monomers, and N UAV-subtask mapping is established. Heterogeneous UAV features are obtained through self-checking and matched with environmental features and task requirements to achieve the optimal pairing of heterogeneous UAVs and subtasks; N UAVs are classified according to the N UAV-subtask mapping, search clusters and following clusters are established, heterogeneous UAVs are grouped, and search clusters and following clusters with different functions are constructed. Following cluster, prepare for collaborative execution of tasks; plan the flight trajectory of the search cluster and the pre-flight trajectory of the following cluster according to the corresponding sub-tasks, and use the search cluster to perform real-time environmental data collection to provide data support for path optimization of the following cluster, and realize division of labor and cooperation within the cluster; adjust the pre-flight trajectory based on the real-time environmental data collection results and the sub-task monomers of the following cluster, and update the flight control parameters to complete the path coordination of the UAV collaborative flight mission. According to the real-time environmental information fed back by the search cluster, the technical solution of dynamically optimizing the pre-planned path of the following cluster solves the technical problems of insufficient environmental adaptability of path planning of civil UAV clusters and low efficiency of heterogeneous cluster collaboration in the existing technology, and achieves the technical effect of improving the environmental adaptability and collaboration efficiency of heterogeneous UAV clusters through collaborative environmental perception and trajectory optimization by the search cluster and the following cluster.

[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A flow chart of a civilian heterogeneous unmanned cluster path coordination method based on human-assisted collaboration is provided for an embodiment of the present application;

[0011] Figure 2 A flow chart of real-time trajectory update in a civilian heterogeneous unmanned cluster path coordination method based on human-assisted collaboration is provided for an embodiment of the present application;

[0012] Figure 3 A structural schematic diagram of a civilian heterogeneous unmanned cluster path coordination system based on unmanned collaboration is provided for an embodiment of the present application.

[0013] Explanation of reference numerals: flight mission acquisition module 11, mission mapping establishment module 12, drone classification module 13, environmental data collection module 14, trajectory adjustment module 15. DETAILED DESCRIPTION

[0014] The overall idea of ​​the technical solution provided by this application is as follows:

[0015] The embodiments of the present application provide a path coordination method and system for a civil heterogeneous unmanned cluster based on collaboration between unmanned and unmanned vehicles. Through the steps of task analysis, UAV feature matching, cluster division of labor and collaboration, environmental perception and path optimization, a complete heterogeneous unmanned cluster path coordination mechanism is constructed to effectively improve the environmental adaptability and collaborative efficiency of civil heterogeneous unmanned clusters.

[0016] Specifically, first, the UAV collaborative flight mission is analyzed and divided into multiple subtasks to lay the foundation for subsequent task allocation. Secondly, through UAV self-inspection and feature matching, the characteristics and mission requirements of heterogeneous UAVs are fully considered to achieve the optimal pairing of UAVs and subtasks. Furthermore, according to the task type, heterogeneous UAVs are divided into search clusters and following clusters. The search cluster is responsible for real-time environmental perception, and the following cluster dynamically adjusts the pre-planned path based on the feedback of environmental information. The two work together to complete complex tasks.

[0017] In summary, this application makes full use of the characteristics of heterogeneous UAV clusters, realizes task allocation within the cluster through task analysis and feature matching; introduces search clusters and following clusters, adopts division of labor, cooperation, environmental perception and path optimization to improve the environmental adaptability of heterogeneous clusters as a whole; and through information interaction and collaboration within the cluster, effectively improves the efficiency of cluster task execution.

[0018] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically described below in conjunction with the drawings in the specification.

[0019] Embodiment 1

[0020] like Figure 1 As shown, the embodiment of the present application provides a civilian heterogeneous unmanned cluster path coordination method based on human-machine collaboration, the method comprising:

[0021] S1: Obtain the UAV collaborative flight mission, analyze the UAV collaborative flight mission, and establish N subtask monomers;

[0022] In an embodiment of the present application, the UAV cooperative flight mission is sent by the ground control station through a wireless communication link. The UAV cooperative flight mission adopts a structured task description format, such as XML, JSON, etc., which contains the key parameters and constraints of the task. After receiving the UAV cooperative flight mission, the key information described in the UAV cooperative flight mission, such as the task type, target area, time requirements, etc., is extracted through the task parser to construct a task parameter table. Among them, the task parser uses regular expressions and syntax analysis technology to parse the task. Then, according to the parsed task parameter table, combined with the task decomposition rule base, the overall UAV cooperative flight mission is divided to generate N subtask monomers. Among them, the task decomposition rule base predefines the decomposition mode and constraints of different types of tasks. For example, for the area detection task, it is divided into several sub-areas according to the size and shape of the area; for the target tracking task, it is divided into several tracking subtasks according to the number of targets and motion characteristics.

[0023] By conducting a structured analysis of the UAV collaborative flight mission and converting the overall mission into multiple executable subtasks, the foundation is laid for the subsequent UAV task allocation and path planning.

[0024] S2: Perform initial environment collection according to the UAV collaborative flight mission, establish a basic environment library, perform self-check of idle UAVs, establish a feature set of idle UAVs, the feature set includes performance features and power features, perform feature matching with the basic environment library, N subtask monomers and feature set, complete task pairing of idle UAVs with N subtask monomers, and establish N UAV-subtask mapping;

[0025] In an embodiment of the present application, according to the requirements of the UAV collaborative flight mission, initial environmental information collection is performed on the mission execution area through environmental perception equipment such as cameras, radars, sensors, etc., including topography, meteorological conditions, electromagnetic environment and other data, to obtain the environmental parameters in the mission execution area and store them in the basic environment library to provide necessary reference data for subsequent mission planning and path optimization.

[0026] At the same time, the currently idle drones are self-checked to comprehensively evaluate their status and capabilities. The self-check content includes the working status of key components such as the drone's power system, navigation equipment, communication module, mission payload, and the remaining battery power. Based on the self-check results, a feature set of idle drones is established, which includes the performance characteristics (such as speed, range, load, etc.) and power characteristics (such as battery capacity, remaining flight time, etc.) of each idle drone.

[0027] After obtaining the feature set of the basic environment library, N subtask monomers and idle drones, feature matching is performed to assign the most suitable drone to each subtask monomer. The goal of feature matching is to maximize the task execution efficiency and reliability of the drone cluster under the premise of meeting the requirements of drone collaborative flight missions. For example, the task type, target area, execution conditions and other attributes of each subtask monomer are analyzed, and the relevant environmental features such as terrain undulation, weather level, electromagnetic interference intensity, etc. are extracted from the basic environment library to characterize the subtask's requirements for drone capabilities; from the feature set of idle drones, the performance parameters such as speed, range, load, maneuverability, and battery capacity, remaining power, energy consumption rate, etc. of each drone are extracted to characterize the drone's task execution capability and endurance; the task parameters and drone parameters are matched so that each successfully matched drone satisfies the corresponding subtask monomer, and N drone-subtask mappings are obtained, so that each subtask monomer establishes a one-to-one correspondence with the most suitable idle drone.

[0028] By fully adapting the UAV collaborative flight missions to the UAV cluster capabilities, the matching of mission requirements and UAV heterogeneous characteristics is achieved based on environmental perception, the mission execution quality of a single UAV is improved, and the overall efficiency of the cluster can be maximized, providing support for path planning and collaborative optimization.

[0029] S3: Classify N drones according to the N drone-subtask mapping, and establish search clusters and follow clusters;

[0030] In an embodiment of the present application, after determining N drone-subtask mappings, the drones are classified and divided into roles to form a search cluster and a following cluster with clear division of labor, which is conducive to achieving coordination and complementarity within the cluster during the execution of the task.

[0031] First, the task attributes and environmental characteristics of each drone in the N drone-subtask mapping are analyzed, and the N drones are divided into several groups according to the similarity and complementarity of the tasks. The drones in each group have similar tasks and can form a closely coordinated team. Then, one or several drones are selected from each group as the search nodes of the group to form a search cluster. Among them, the selection of the search node takes into account sensor configuration, maneuverability, communication capability, etc.; sensors are used for environmental perception, such as high-resolution cameras, lidar, etc., to obtain detailed environmental information; maneuverability is used to flexibly adjust the track and speed to adapt to complex and changing environments; communication capability is used to share environmental information with other drones in real time and reliably. Then, the drones of non-search nodes are aggregated to form a following cluster. The drones in the following cluster form a follow-guide collaborative relationship with the search node according to the subtask mapping relationship. The drones in the following cluster obtain environmental perception results in real time through information interaction with the search node, and dynamically adjust their own tracks and behavior decisions according to environmental characteristics and task requirements.

[0032] The division of search clusters and following clusters is dynamic and flexible. The roles of drones can be switched dynamically according to the progress of mission execution and changes in the environment. In addition, the initial cluster division results are not static, but provide a starting point for cluster collaboration. During mission execution, the composition of the cluster and the roles of the nodes are adaptively adjusted according to the actual situation.

[0033] Through cluster division, the drones form a search and follow formation structure. The search cluster opens up routes and perceives the environment, while the following cluster follows closely to perform specific tasks. Through division of labor and cooperation, a closed-loop intelligence of environmental perception, path planning, and target disposal is formed at the cluster level, thereby improving the accuracy and efficiency of task execution.

[0034] S4: planning the flight trajectory of the search cluster and the pre-flight trajectory of the following cluster according to the corresponding subtasks, and performing real-time environmental data collection with the search cluster;

[0035] In an embodiment of the present application, after determining the search cluster and the following cluster, respective flight trajectories are planned for the search cluster and the following cluster, and the search cluster is used to perceive the environment in real time to provide information support for precise control of the following cluster.

[0036] The flight trajectory planning of the search cluster follows the principles of maximizing information gain, combining local obstacle avoidance with global optimization, and dynamic adjustment under resource constraints. In the planning process, the existing environmental prior information is fully utilized, and the trajectory is dynamically optimized and adjusted based on the real-time feedback of the search cluster. Among them, trajectory planning uses classic algorithms such as A* and RRT, or intelligent optimization technologies such as reinforcement learning.

[0037] The pre-flight trajectory planning of the following cluster is coordinated with the flight trajectory of the searching cluster. The appropriate route and flight parameters are selected according to the requirements of the subtask, and a certain amount of adjustment is reserved. The pre-flight trajectory provides a feasible and safe route basis for the UAVs in the following cluster, reducing the difficulty and range of real-time adjustments. During the execution process, the following cluster optimizes the pre-flight trajectory in real time based on the perception results of the searching cluster.

[0038] The search cluster uses sensors to obtain environmental information during flight, such as topography, obstacle locations, target status, etc., and collects real-time environmental data. The collected data is preprocessed and fused to form local environmental perception results, which are used as real-time environmental data collection results and shared with the following cluster through the cluster's internal communication network.

[0039] Through the perception actions of the search cluster and the pre-flight trajectory planning of the following cluster, the unification of environmental perception and task planning is achieved in time and space. The search cluster opens up information channels in front, and the following cluster adjusts its heading according to the prior trajectory and real-time perception results, forming close coordination within the cluster and effectively improving the efficiency and adaptability of task execution in complex environments.

[0040] S5: Pre-flight trajectory adjustment is performed based on the real-time environmental data collection results and the sub-task monomers of the following cluster, and the flight control parameters are updated to complete the path coordination of the UAV collaborative flight mission.

[0041] In an embodiment of the present application, after the search cluster obtains the real-time environmental data collection results, the pre-flight trajectory following the cluster is dynamically optimized, and the flight control parameters are adjusted accordingly, thereby achieving path coordination of the drone cluster in a dynamic environment.

[0042] When adjusting the pre-flight trajectory, model predictive control, dynamic programming, intelligent optimization and other methods are used to comprehensively consider environmental threats, mission requirements, resource constraints, cluster collaboration and other factors, and optimize the path under the premise of ensuring flight safety and mission completion to achieve pre-flight trajectory adjustment. At the same time, the flight control parameters such as speed, altitude, and heading are adjusted accordingly to execute the flight trajectory of the subtask monomers in the following cluster after the adjustment. Among them, trajectory adjustment and flight control parameter update are continuous closed-loop processes. The following cluster periodically optimizes decisions based on the latest environmental information provided by the search cluster, so that the drone cluster can autonomously coordinate in a dynamic unknown environment, flexibly respond to changes, and complete the established tasks. When all subtask monomers of the following cluster have completed their tasks, the path coordination of the drone collaborative flight mission is completed.

[0043] Through the steps of task analysis, drone matching, cluster formation, environmental perception, path optimization, etc., a complete drone cluster collaborative control process is constructed, which gives full play to the intelligent performance of drone clusters, such as environmental perception, task planning, dynamic decision-making, etc., and can improve the environmental adaptability, task flexibility and cluster robustness of drone clusters in complex dynamic environments, and provide technical support for drone collaborative control in complex mission environments. At the same time, the search cluster and the following cluster form an efficient and collaborative path optimization mechanism through clear division of labor, real-time perception, and dynamic decision-making during the entire flight process, which provides strong support for the autonomous collaboration of drone clusters in dynamic unknown environments, and effectively improves the environmental adaptability and task execution efficiency of the cluster.

[0044] Furthermore, the embodiments of the present application include:

[0045] The basic environment library, N subtask monomers and feature sets are input into the intelligent matching model to perform matching. The steps are as follows:

[0046] The task feature matching unit performs performance adaptation matching on N subtask monomers and feature sets to obtain traversal adaptation values;

[0047] The power matching unit performs power adaptation matching of N subtask monomers and feature sets to generate power elimination constraints;

[0048] Performing feature set adaptation evaluation under the basic environment library through the environment impact unit to generate an environment adaptation value, wherein the task feature matching unit, the power matching unit and the environment impact unit are all processing units of the intelligent matching model;

[0049] The traversal adaptation value, power elimination constraint and environmental adaptation value are normalized, and matching is completed based on the processing results.

[0050] In a feasible implementation, an intelligent matching model is pre-built to comprehensively consider the basic environment library, N subtask monomers and the feature set of idle drones, and evaluate the matching degree between drones and subtasks. The intelligent matching model includes three processing units, namely, a task feature matching unit, a power matching unit and an environmental impact unit. Among them, the task feature matching unit is responsible for analyzing the specific requirements of each subtask on the performance of the drone, such as aerial photography accuracy, flight speed, maneuverability and other indicators, and calculating the matching degree with the actual performance parameters of the drone to obtain the traversal adaptation value; the power matching unit focuses on the endurance requirements of the subtask and the power status of the drone, and by comparing the estimated energy consumption and battery capacity, it determines whether the drone meets the flight time required for task execution and generates power elimination constraints; the environmental impact unit introduces terrain, meteorology, electromagnetic and other environmental information in the basic environment library, evaluates the impact of environmental conditions on the performance of each drone when performing each task, and gives the environmental adaptation value. After obtaining the attribute set and feature set of the basic environment library, N subtask monomers, the attribute set and feature set of the basic environment library, N subtask monomers are input into the intelligent matching model execution to trigger the matching execution process.

[0051] First, the task feature matching unit extracts the task requirement features related to the performance of the UAV from the attribute set of N subtask monomers, such as the requirements for aerial photography resolution, flight speed, and maneuverability. Then, the task feature matching unit obtains the performance parameters of each idle UAV from the feature set, such as camera resolution, maximum flight speed, pitch angular velocity, etc. Next, the task feature matching unit matches and calculates the task requirement features with the performance parameters of the UAV by designing a matching evaluation function. Among them, the matching evaluation function can be constructed based on different methods such as weighted distance measurement, fuzzy logic reasoning, and machine learning to measure the degree to which the actual performance of the UAV meets the task requirements. Through the matching evaluation function, the performance matching degree of each idle UAV for each subtask is quantified into a numerical value, which is the adaptation value. The task feature matching unit traverses all UAV-subtask pairs to generate a complete adaptation value matrix, which is the traversal adaptation value. The rows of the matrix represent idle UAVs, the columns represent subtask monomers, and each element is the adaptation value of the corresponding UAV to perform the corresponding subtask. The higher the adaptation value, the more the performance of the UAV matches the task requirements and the stronger its ability to complete the task.

[0052] At the same time, the power matching unit extracts task parameters related to the endurance requirement from the attribute set of N subtask monomers, such as task execution time, flight distance, load requirement, etc. The power matching unit estimates the power requirement of each subtask for the drone based on the task parameters and the drone energy consumption model, and determines the task energy consumption required to complete each task. Then, the power matching unit obtains the power status information of each idle drone from the feature set, including parameters such as battery capacity, current power, and discharge curve, and compares the determined task energy consumption with the available power of the drone to determine whether the drone meets the endurance required for task execution. For drones with insufficient power to support task execution, they are marked as unexecutable tasks, thereby generating power elimination constraints. The power elimination constraint is a Boolean matrix. If the element in the matrix is ​​1, it means that the corresponding drone is eliminated due to insufficient power; if it is 0, it means that the corresponding drone meets the endurance requirements and can be used as an alternative drone.

[0053] At the same time, the environmental impact unit extracts environmental information related to each subtask from the basic environment library, such as terrain, meteorology, electromagnetic and other environmental parameters in the task execution area, reflecting the environmental challenges and constraints that may be encountered when executing the task. Then, the environmental impact unit obtains the environmental adaptation characteristics of each candidate UAV from the feature set, such as airworthy wind speed, dust and water resistance level, anti-interference ability, etc., to characterize the ability of the UAV to cope with different environmental conditions, such as stability and reliability in complex environments such as strong winds, dust, and electromagnetic interference. Next, in the design of the environmental adaptability evaluation function of the environmental impact unit, the mission environment information is matched and calculated with the environmental adaptation characteristics of the UAV. Among them, the environmental adaptability evaluation function can be constructed based on different methods such as fuzzy logic reasoning and machine learning to measure the degree to which the environmental adaptability of the UAV meets the mission environment conditions. Through the environmental adaptability evaluation function, the environmental adaptability of each UAV for each subtask is quantified into a numerical value, which is the environmental adaptation value. The environmental adaptation value is also represented by a matrix, where the rows of the matrix represent idle UAVs, the columns represent subtask monomers, and each element is the environmental adaptation value of the corresponding UAV in the corresponding mission environment. The higher the adaptation value, the more the UAV's environmental adaptability matches the mission environment conditions, and the higher the reliability of completing tasks in complex environments.

[0054] After obtaining the traversal adaptation value, power elimination constraint and environmental adaptation value, the obtained traversal adaptation value, power elimination constraint and environmental adaptation value are normalized and converted into a unified quantitative index, so that the evaluation results of different dimensions can be comprehensively compared to obtain the normalized processing results. Subsequently, based on the normalized processing results, weighted summation or other combination strategies are used to obtain a quantitative description of the matching quality of each UAV with each subtask, thereby achieving feature matching.

[0055] By building an intelligent matching model, the adaptability of drones and subtasks is evaluated from three perspectives: mission, energy, and environment. This overcomes the limitations of single feature matching, quantifies mission requirements, power constraints, and environmental impacts, and comprehensively and objectively characterizes the drone's ability to perform tasks, thereby improving the accuracy and efficiency of feature matching.

[0056] Furthermore, the embodiments of the present application include:

[0057] Screening matching solutions using the power elimination constraint, and establishing a basic matching set based on the screening results;

[0058] Setting weighted coefficients of the ergodic adaptation value and the environmental adaptation value, and performing weighted calculation of the ergodic adaptation value and the environmental adaptation value of each solution in the basic matching set by using the weighted coefficients to generate a weighted calculation result;

[0059] The task levels of the N subtask entities are set, the weighted calculation results are compensated with the task levels, and the normalization process is completed based on the compensation results.

[0060] In a preferred embodiment, a method for normalizing ergodic adaptation values, power elimination constraints, and environmental adaptation values ​​is provided.

[0061] First, an initial set containing all possible drone-task matching schemes is established. The set is a two-dimensional table, where each row corresponds to an idle drone and each column corresponds to a subtask monomer. Each element in the table represents a scheme in which the corresponding drone is assigned to perform the corresponding subtask. The initial set covers all possible task assignment combinations. Then, the power elimination constraint generated by the power matching unit is extracted. The power elimination constraint is in the form of a matrix. The dimension of the matrix is ​​consistent with that of the initial set, and the elements in the matrix are 0 or 1. If an element is 1, it means that the corresponding drone-task matching scheme is eliminated due to insufficient power; if it is 0, it means that the scheme passes the power constraint test. Subsequently, each scheme in the initial set is traversed to check the value of its corresponding position in the power elimination constraint. If the value is 1, the scheme is removed from the initial set; if the value is 0, the scheme is retained. Through this traversal and screening process, the schemes that do not meet the power requirements are eliminated one by one, leaving only the schemes whose power meets the task requirements. The results of traversal and screening form a basic matching set, which is a subset of the initial set and contains all drone-task matching solutions that pass the power constraint test.

[0062] Next, two weighting coefficients are introduced, corresponding to the traversal adaptation value and the environmental adaptation value respectively. Among them, the value range of the weighting coefficient is a real number between 0 and 1, reflecting the relative importance of the two indicators in the comprehensive evaluation. When the task feature matching is more important, the traversal adaptation value is given a larger weight; when the environmental adaptability is more critical, the environmental adaptation value is given a larger weight. The sum of the two weighting coefficients is 1 to ensure the numerical stability of the weighted calculation results. Then, for each drone-task matching scheme in the basic matching set, its adaptation value and environmental adaptation value in the traversal adaptation value are extracted. Subsequently, the traversal adaptation value and the environmental adaptation value are linearly combined using the weighting coefficient to calculate the comprehensive matching degree of each scheme and obtain the weighted calculation result.

[0063] Subsequently, based on the weighted calculation results, the task level factor is introduced to give differentiated resource inclination to subtasks of different importance. First, according to the attributes such as task objectives, time requirements, and priorities, the importance of N subtask monomers is divided, and each subtask is assigned a task level. Among them, the task level is represented by a number, and the larger the value, the higher the importance of the task. For example, the task level is divided into 1 to 5 levels, with level 1 representing the lowest priority and level 5 representing the highest priority. The division of task levels comprehensively considers factors such as the urgency of the task and the impact on the overall goal, and is automatically generated by task planners or intelligent decision-making algorithms. Then, the weighted calculation result matrix is ​​compensated and adjusted based on the task level. Specifically, for each drone-task matching scheme, its comprehensive matching degree is multiplied by the task level coefficient of the corresponding subtask. Among them, the task level coefficient is an increasing function that is positively correlated with the task level, such as a linear function, an exponential function, etc. By multiplying the task level coefficient, the comprehensive matching metric of high-level tasks is amplified, and the comprehensive matching metric of low-level tasks is reduced, so that the task allocation strategy is inclined to important tasks, and more high-quality resources are allocated to them first, so as to compensate the weighted calculation results and obtain compensation results. After that, the compensated weighted calculation results are normalized, and the comprehensive matching degrees of different numerical ranges are mapped to a unified interval to facilitate comparison and sorting between schemes. The result of the normalization process is a matrix with the same dimension as the basic matching set. The elements in the matrix represent the final evaluation value of each drone-task matching scheme. Based on this matrix, all drone-task matching schemes can be sorted to complete the task pairing of idle drones with N subtask monomers.

[0064] Through steps such as power elimination constraint screening, weighted calculation fusion, task level compensation, and normalization, quantitative evaluation and optimization of task allocation are achieved under the premise of multiple constraints and multiple attributes. Factors such as energy limitations, task-resource matching, and task importance are fully considered to complete the task pairing of idle drones with N sub-task units, so as to give full play to the intelligent collaborative advantages of heterogeneous drone clusters and significantly improve the task planning and execution efficiency in complex environments.

[0065] Furthermore, the embodiments of the present application include:

[0066] Parse the subtasks corresponding to the search cluster and obtain the target range area through the parsing results;

[0067] Performing trajectory optimization of area search on the target range area, establishing a first trajectory through the position of the departure starting point and the target range area, and adjusting the trajectory optimization result with the first trajectory to generate a flight trajectory fused with the first trajectory.

[0068] In a feasible implementation, after completing the task allocation of the civil heterogeneous unmanned swarm, the flight trajectories of the search swarm and the following swarm are further planned and optimized. The trajectory planning of the search swarm is to explore and cover the target area, while the trajectory planning of the following swarm needs to be coordinated with the real-time perception results of the search swarm.

[0069] First, the subtasks assigned to the search cluster are parsed, the target area information contained in the task is extracted, and the target range area is determined. The target range area is defined by a series of geographic coordinate points or regional boundary lines, indicating the spatial range that the search task needs to explore. By extracting the target range area, spatial constraints are provided for subsequent trajectory planning. The flight trajectory of the search cluster needs to cover the area as completely and efficiently as possible to ensure the thoroughness of the search and detection.

[0070] Then, with the target range area as a constraint, the trajectory optimization algorithm is used to optimize the flight trajectory of the regional search. The trajectory optimization problem is a multi-objective optimization problem, which comprehensively considers multiple performance indicators such as regional coverage, flight distance, repeated detection rate, and information gain. Among them, the regional coverage rate measures the spatial coverage integrity of the search trajectory to the target area. The higher the coverage rate, the lower the probability of missing the target; the flight distance measures the path length of the search trajectory. The shorter the distance, the smaller the flight time and energy consumption; the repeated detection rate measures the frequency of repeated searches of the same area. Moderate repetition helps to improve the search reliability, but too much repetition reduces the search efficiency; the information gain measures the richness of the information content of the area passed by the search path. The area with high information gain should be detected first to obtain more target information. The evaluation weights of various performance indicators are set according to the UAV collaborative flight mission to form an objective function of comprehensive performance. The trajectory optimization algorithm uses this objective function as the optimization criterion and gradually optimizes the various performance indicators of the search trajectory through intelligent search and iterative update. Among them, the trajectory optimization algorithm includes heuristic search algorithms such as genetic algorithm, particle swarm optimization algorithm, and simulated annealing algorithm. By repeatedly iterating, comparing, and screening candidate trajectories, the global optimal solution is gradually approached, and finally a regional search trajectory with high comprehensive performance is obtained as the trajectory optimization result. However, the starting position of the regional search trajectory obtained by trajectory optimization is random, and the drone needs to fly from its current position to the starting point, which will cause additional flight overhead. In order to further optimize the overall flight path, the construction mechanism of the first trajectory is introduced, that is, a straight line trajectory is planned as the first trajectory through the geometric center of the current position of the drone and the target range area, which represents the shortest path for the drone to fly directly from the current position to the target range area.

[0071] Next, based on the first trajectory, the regional search trajectory obtained by trajectory optimization is adjusted and fused, and the starting point of the regional search trajectory is connected with the end point of the first trajectory to form a complete flight trajectory from the current position to the search area and conduct regional search. Through this trajectory fusion mechanism, the flight path of the drone from the current position to the target range area is optimized into a straight line, shortening the flight distance and time; at the same time, the search trajectory forms a seamless connection with the first trajectory while maintaining the original coverage efficiency and detection performance, ensuring the continuity and smoothness of the entire flight trajectory.

[0072] In practical applications, the trajectory optimization algorithm and the planning method of the first trajectory are flexibly adjusted according to the mission requirements and environmental characteristics. For example, under complex terrain conditions, the first trajectory considers introducing constraints such as threat avoidance and terrain following to improve flight safety; in dynamic target search tasks, the trajectory optimization algorithm can add a target motion prediction model to optimize the search trajectory in real time; in multi-UAV collaborative search tasks, constraints such as formation maintenance and communication connection are added to the trajectory optimization to achieve synchronous search and information sharing of the cluster, thereby further improving the environmental adaptability, mission adaptability and cluster collaboration of the search trajectory.

[0073] Furthermore, the embodiments of the present application include:

[0074] Establish serialized tasks for searching clusters and following clusters, and set the interval between serial tasks;

[0075] A pre-flight trajectory of the following cluster is set with the target range area, the departure starting point and the task interval, wherein the pre-flight trajectory includes a take-off time node.

[0076] In a feasible implementation, a serialized task planning method for a search cluster and a follow cluster is proposed. By setting task intervals and take-off time nodes, flight coordination of the two clusters in time and space is achieved, thereby improving the overall task execution efficiency.

[0077] First, the task execution process of the search cluster and the following cluster is described in a serialized manner, and the tasks of the two clusters are divided into a series of ordered subtask units. Each subtask unit includes a complete process of takeoff, cruising, target search, and return. By establishing serialized tasks, the order and time arrangement of the search cluster and the following cluster in the task execution process are clearly defined. Then, based on the serialized tasks, the serial task interval is further set to represent the time interval between the takeoff times of two adjacent serial task units, reflecting the takeoff time difference between the search cluster and the following cluster, ensuring that the following cluster starts the flight mission with an appropriate time delay after the search cluster completes the real-time environmental data collection and information feedback, avoiding flight conflicts between the two clusters, and ensuring that the following cluster can obtain the information provided by the search cluster in a timely manner and adjust the pre-flight trajectory within the time interval.

[0078] Then, the flight trajectory of the search cluster is used as the pre-flight trajectory of the following cluster, and the routes of the two clusters within the target range are kept consistent, so as to maximize the flight synchronization and coordination efficiency between the clusters and reduce the computational overhead caused by the separate trajectory planning of the following cluster. However, since the starting points of the following cluster and the search cluster are different, adaptive adjustments need to be made in the take-off phase of the pre-flight trajectory. Similar to the search cluster, a straight take-off route is planned as the first trajectory of the following cluster based on the starting point of the following cluster and the geometric center of the target area, and the straight trajectory is spliced ​​with the route within the target range to form a complete pre-flight trajectory of the following cluster from take-off to the search area. After determining the spatial direction of the pre-flight trajectory, the take-off time node of the following cluster is set according to the task interval, indicating the moment when the following cluster starts to execute the serialized task unit, and its setting is coordinated with the task execution progress of the search cluster. Specifically, taking the execution time of each serialized task unit of the search cluster as a reference and adding the task interval, the take-off time node of the corresponding serialized task unit of the following cluster can be determined, thereby effectively controlling the take-off timing of the following cluster so that it can intervene in subsequent tasks with an appropriate time delay after the search cluster completes the necessary real-time environmental data collection.

[0079] The flight synchronization within the target range is achieved by spatially reusing the pre-flight trajectory of the following cluster with the flight trajectory of the searching cluster. Adaptive adjustment of the departure starting point is achieved through a differentiated first trajectory during the take-off phase. Task coordination with the searching cluster is achieved through the setting of the take-off time node based on the task interval. This simplifies the coordination between clusters while taking into account the take-off constraints and task requirements of the following cluster, thus achieving efficient cluster collaborative path planning.

[0080] It should be noted that the pre-flight trajectory is only a planning scheme for the cluster at the beginning of the mission. In subsequent flights, the pre-flight trajectory can be adaptively adjusted and optimized according to the real-time environmental data collection results of the search cluster, deviations can be corrected in time, flight control parameters can be updated, and decisions can be dynamically optimized to improve the mission execution efficiency and environmental adaptability of the entire cluster. At the same time, the setting of the take-off time node is not static, but a dynamically updated time reference, which is flexibly adjusted according to the actual mission execution progress of the search cluster.

[0081] Furthermore, the embodiments of the present application include:

[0082] Performing location mapping based on the real-time environment data collection results to establish real-time location environment features;

[0083] Extracting target feature data from the real-time environmental data collection result, and when the target feature data meets a preset trigger threshold, obtaining the real-time device position of the drone in the associated tracking cluster;

[0084] The path is replanned according to the real-time device position and the position of the target feature data, the pre-flight trajectory adjustment is completed, and the flight control parameters are updated according to the association between the replanned path and the real-time position environmental characteristics.

[0085] In a preferred embodiment, after obtaining the real-time environmental data collection results of the search cluster, the pre-flight trajectory of the following cluster is dynamically optimized and adjusted, and the flight control parameters are updated to adapt to the real-time changes in environmental characteristics and improve the cluster's mission execution accuracy and flight safety.

[0086] First, the real-time environmental data collected by the search cluster is processed by location mapping. The search cluster continuously collects multi-source heterogeneous environmental data during flight, such as images, videos, radar echoes, etc. These data carry rich location information and reflect the environmental characteristics of different regions. By labeling and georeferencing these data, they are accurately mapped to the real geographic coordinate system to generate a real-time updated location environment feature map, dynamically characterizing the environmental attributes of different locations in the mission area, such as terrain undulations, distribution of objects, meteorological conditions, etc., providing a location perception basis for target extraction and trajectory optimization.

[0087] Then, the target feature is extracted from the real-time environment data collection results to obtain the target feature data. Among them, the target feature refers to the significant representation of the target object or event that the search cluster is concerned about in the real-time environment data collection results, such as the image appearance characteristics of the target, the radar scattering characteristics of the target, the movement mode of the target, etc. After the target feature data is extracted, it is judged whether the threshold for triggering the following cluster task is reached. Among them, the trigger threshold is a series of judgment conditions set according to task requirements and experience, such as the confidence of the target feature, the target threat level, the target continuous tracking time, etc. When the extracted target feature data meets the trigger threshold, it can be considered that the search cluster has found the target. At this time, according to the task planning and cluster formation, the drone in the following cluster responsible for the current target is determined, and the real-time device position is obtained through the drone's own navigation and positioning system. Subsequently, the real-time device position of the drone in the following cluster is used as the optimization starting point, and the spatial position of the target feature data is used as the optimization end point. The flight trajectory of the drone in the following cluster is replanned to generate the drone flight path from the optimization starting point to the optimization end point. Then, the newly generated UAV flight path will replace the subsequent route in the original pre-flight trajectory from the current moment, forming a seamless flight path from the current position to the target position, so that the UAVs in the following cluster can adjust the flight plan in real time according to the latest target information, respond to changes in the target position and threat environment in a timely manner, and improve the flexibility and adaptability of task execution. While adjusting the trajectory, based on the re-planned flight trajectory and real-time location environment characteristics, the control variables such as the speed, altitude, and heading of the UAV are optimized to keep them consistent with the route status, and the flight control parameters such as the curvature of the route, the climb angle, and the heading change rate are updated, thereby completing the path coordination of the UAVs in the following cluster.

[0088] Further, such as Figure 2 As shown, the embodiment of the present application includes:

[0089] Record flight abnormalities and mark them as abnormalities;

[0090] Establish the risk factor of the regional location based on the abnormal identification results;

[0091] Planning constraints for trajectory planning are performed according to the risk factor to complete real-time update of the trajectory of the following cluster.

[0092] In a feasible implementation, during the UAV swarm mission, due to the complex and changeable environment and numerous threat factors, the UAVs may encounter various flight anomalies, such as equipment failure, sudden weather changes, enemy interference, etc. In order to enhance the robustness and survivability of the swarm in dealing with anomalies and improve the mission success rate, a trajectory dynamic adjustment method based on abnormal event analysis is proposed.

[0093] First, the flight telemetry data, equipment working status, task execution progress and other information are obtained in real time through airborne sensors and cluster communication networks, and various abnormal events that occur during the flight of the UAV are recorded. Among them, abnormal events refer to unexpected situations that deviate from the normal flight status and mission plan, which may come from internal system failures or external environmental interference. After obtaining the abnormal event data, it is classified and identified, and the type, cause, and degree of danger of the abnormal event are determined. The abnormal pattern and attribution are identified, and the corresponding danger level label is given according to the severity and impact range of the abnormality as the abnormal identification result. Subsequently, the spatial location of the abnormal event is extracted and mapped to the target range area to obtain the abnormal distribution map. On this basis, the danger level label in the abnormal identification result is bound to the abnormal position corresponding to the abnormal distribution map, and the danger levels of all abnormal events at each abnormal position are superimposed, and then the danger level superposition results of all positions are mapped to the [0, 1] interval to obtain the danger coefficient of the regional position. Subsequently, the danger factor of the established regional location is introduced into the UAV trajectory planning. By introducing risk avoidance items in the planning objectives, the generated flight trajectory can automatically stay away from high-risk areas, complete the real-time update of the trajectory of the following cluster, avoid potential threats, and ensure the flight safety and mission continuity of the UAV.

[0094] Through trajectory planning constraints based on the risk factor, a risk-aware trajectory optimization mechanism is constructed. Through the integration of spatial risk distribution and trajectory search algorithms, active isolation of flight paths and dangerous areas is achieved, fundamentally improving the survivability and environmental adaptability of drones. Compared with traditional behavioral avoidance methods, based on regional risk assessment, the safety of flight paths is proactively predicted and optimized from a global perspective, avoiding the traps and risk accumulation problems that may be caused by local obstacle avoidance, and improving the safety level of drone operations.

[0095] In summary, the civilian heterogeneous unmanned cluster path coordination method based on unmanned collaboration provided by the embodiment of the present application has the following technical effects:

[0096] Acquire the cooperative flight mission of UAVs, analyze the cooperative flight mission of UAVs, establish N subtask monomers, divide the complex cooperative flight mission, form multiple subtasks that can be executed by a single UAV, and prepare for the subsequent task allocation. Carry out initial environment collection according to the cooperative flight mission of UAVs, establish a basic environment library, perform self-check of idle UAVs, establish a feature set of idle UAVs, and the feature set includes performance characteristics and power characteristics. Perform feature matching with the basic environment library, N subtask monomers and feature sets, complete the task pairing of idle UAVs and N subtask monomers, establish N UAV-subtask mappings, fully consider the characteristics and task requirements of heterogeneous UAVs, realize the reasonable division of labor within the cluster, and improve the efficiency of task execution. Classify N UAVs according to the N UAV-subtask mappings, establish search clusters and follow-up clusters, group UAVs according to task types, and build clusters with different functions to lay the foundation for cooperative execution of tasks. Plan the flight trajectory of the search cluster and the pre-flight trajectory of the follow-up cluster according to the corresponding subtasks, and use the search cluster to perform real-time environmental data collection, use the search cluster to detect the environment, obtain dynamic environmental information, and provide data support for the path optimization of the follow-up cluster. The pre-flight trajectory is adjusted based on the real-time environmental data collection results and the sub-task units of the following cluster, and the flight control parameters are updated to complete the path coordination of the UAV collaborative flight mission. The pre-planned path is dynamically optimized based on real-time environmental feedback, which improves the ability of the following cluster to cope with complex environments and ensures the smooth completion of the mission, thereby improving the environmental adaptability and collaborative efficiency of heterogeneous UAV clusters.

[0097] Embodiment 2

[0098] Based on the same inventive concept as the civilian heterogeneous unmanned cluster path coordination method based on unmanned collaboration in the aforementioned embodiment, Figure 3 As shown, the embodiment of the present application provides a civilian heterogeneous unmanned cluster path coordination system based on human-machine collaboration, the system comprising:

[0099] The flight mission acquisition module 11 is used to acquire the UAV collaborative flight mission, analyze the UAV collaborative flight mission, and establish N subtask monomers;

[0100] The task mapping establishment module 12 is used to perform initial environment collection according to the UAV collaborative flight mission, establish a basic environment library, perform idle UAV self-check, establish a feature set of idle UAVs, and the feature set includes performance characteristics and power characteristics. The basic environment library, N subtask monomers and feature sets are used to perform feature matching, complete the task pairing of idle UAVs and N subtask monomers, and establish N UAV-subtask mapping;

[0101] The drone classification module 13 is used to classify N drones according to the N drone-subtask mapping and establish a search cluster and a follow cluster;

[0102] An environmental data collection module 14 is used to plan the flight trajectory of the search cluster and the pre-flight trajectory of the following cluster according to the corresponding subtasks, and perform real-time environmental data collection with the search cluster;

[0103] The trajectory adjustment module 15 is used to perform pre-flight trajectory adjustment based on the real-time environmental data collection results and the sub-task monomers of the tracking cluster, and update the flight control parameters to complete the path coordination of the UAV cooperative flight mission.

[0104] Furthermore, the task mapping establishment module 12 also includes the following execution steps:

[0105] The basic environment library, N subtask monomers and feature sets are input into the intelligent matching model to perform matching. The steps are as follows:

[0106] The task feature matching unit performs performance adaptation matching on N subtask monomers and feature sets to obtain traversal adaptation values;

[0107] The power matching unit performs power adaptation matching of N subtask monomers and feature sets to generate power elimination constraints;

[0108] Performing feature set adaptation evaluation under the basic environment library through the environment impact unit to generate an environment adaptation value, wherein the task feature matching unit, the power matching unit and the environment impact unit are all processing units of the intelligent matching model;

[0109] The traversal adaptation value, power elimination constraint and environmental adaptation value are normalized, and matching is completed based on the processing results.

[0110] Furthermore, the task mapping establishment module 12 also includes the following execution steps:

[0111] Screening matching solutions using the power elimination constraint, and establishing a basic matching set based on the screening results;

[0112] Setting weighted coefficients of the ergodic adaptation value and the environmental adaptation value, and performing weighted calculation of the ergodic adaptation value and the environmental adaptation value of each solution in the basic matching set by using the weighted coefficients to generate a weighted calculation result;

[0113] The task levels of the N subtask entities are set, the weighted calculation results are compensated with the task levels, and the normalization process is completed based on the compensation results.

[0114] Furthermore, the environmental data collection module 14 includes the following execution steps:

[0115] Parse the subtasks corresponding to the search cluster and obtain the target range area through the parsing results;

[0116] Performing trajectory optimization of area search on the target range area, establishing a first trajectory through the position of the departure starting point and the target range area, and adjusting the trajectory optimization result with the first trajectory to generate a flight trajectory fused with the first trajectory.

[0117] Furthermore, the environmental data collection module 14 further includes the following execution steps:

[0118] Establish serialized tasks for searching clusters and following clusters, and set the interval between serial tasks;

[0119] A pre-flight trajectory of the following cluster is set with the target range area, the departure starting point and the task interval, wherein the pre-flight trajectory includes a take-off time node.

[0120] Furthermore, the trajectory adjustment module 15 includes the following execution steps:

[0121] Performing location mapping based on the real-time environment data collection results to establish real-time location environment features;

[0122] Extracting target feature data from the real-time environmental data collection result, and when the target feature data meets a preset trigger threshold, obtaining the real-time device position of the drone in the associated tracking cluster;

[0123] The path is replanned according to the real-time device position and the position of the target feature data, the pre-flight trajectory adjustment is completed, and the flight control parameters are updated according to the association between the replanned path and the real-time position environmental characteristics.

[0124] Furthermore, the embodiment of the present application also includes a trajectory real-time update module, which includes the following execution steps:

[0125] Record flight abnormalities and mark them as abnormalities;

[0126] Establish the risk factor of the regional location based on the abnormal identification results;

[0127] Planning constraints for trajectory planning are performed according to the risk factor to complete real-time update of the trajectory of the following cluster.

[0128] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.

[0129] Furthermore, the first or second mentioned above may not only represent an order relationship, but may also represent a specific concept, and / or refer to multiple elements that can be selected individually or in full. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A civil heterogeneous unmanned cluster path coordination method based on unmanned collaboration, characterized in that: The method comprises: Obtaining a UAV collaborative flight mission, parsing the UAV collaborative flight mission, and establishing N subtask monomers; Perform initial environment collection according to the UAV collaborative flight mission, establish a basic environment library, perform idle UAV self-check, establish a feature set of idle UAVs, the feature set includes performance features and power features, perform feature matching with the basic environment library, N subtask monomers and feature set, complete task pairing of idle UAVs and N subtask monomers, and establish N UAV-subtask mapping; Classify N drones according to the N drone-subtask mapping, and establish search clusters and follow clusters; Plan the flight trajectory of the search cluster and the pre-flight trajectory of the following cluster according to the corresponding subtasks, and perform real-time environmental data collection with the search cluster; The pre-flight trajectory is adjusted based on the real-time environmental data collection results and the sub-task units of the following cluster, and the flight control parameters are updated to complete the path coordination of the UAV collaborative flight mission.

2. The method according to claim 1, characterized in that The feature matching based on the basic environment library, N subtask monomers and feature sets includes: The basic environment library, N subtask monomers and feature sets are input into the intelligent matching model to perform matching. The steps are as follows: The task feature matching unit performs performance adaptation matching on N subtask monomers and feature sets to obtain traversal adaptation values; The power matching unit performs power adaptation matching of N subtask monomers and feature sets to generate power elimination constraints; Performing feature set adaptation evaluation under the basic environment library through the environment impact unit to generate an environment adaptation value, wherein the task feature matching unit, the power matching unit and the environment impact unit are all processing units of the intelligent matching model; The traversal adaptation value, power elimination constraint and environmental adaptation value are normalized, and matching is completed based on the processing results.

3. The method according to claim 2, characterized in that The traversal adaptation value, the power elimination constraint and the environment adaptation value are normalized, and matching is completed according to the processing result, including: Screening matching solutions using the power elimination constraint, and establishing a basic matching set based on the screening results; Setting weighted coefficients of the ergodic adaptation value and the environmental adaptation value, and performing weighted calculation of the ergodic adaptation value and the environmental adaptation value of each solution in the basic matching set by using the weighted coefficients to generate a weighted calculation result; The task levels of the N subtask entities are set, the weighted calculation results are compensated with the task levels, and the normalization process is completed based on the compensation results.

4. The method according to claim 1, characterized in that The method of planning the flight trajectory of the search cluster and the pre-flight trajectory of the following cluster according to the corresponding subtasks, and performing real-time environmental data collection with the search cluster, includes: Parse the subtasks corresponding to the search cluster and obtain the target range area through the parsing results; Performing trajectory optimization of area search on the target range area, establishing a first trajectory through the position of the departure starting point and the target range area, and adjusting the trajectory optimization result with the first trajectory to generate a flight trajectory fused with the first trajectory.

5. The method according to claim 4, characterized in that The method further comprises: Establish serialized tasks for searching clusters and following clusters, and set the interval between serial tasks; A pre-flight trajectory of the following cluster is set with the target range area, the departure starting point and the task interval, wherein the pre-flight trajectory includes a take-off time node.

6. The method according to claim 5, characterized in that The pre-flight trajectory adjustment is performed based on the real-time environmental data collection results and the sub-task monomers of the following cluster, and the flight control parameters are updated, including: Performing location mapping based on the real-time environment data collection results to establish real-time location environment features; Extracting target feature data from the real-time environmental data collection result, and when the target feature data meets a preset trigger threshold, obtaining the real-time device position of the drone in the associated tracking cluster; The path is replanned according to the real-time device position and the position of the target feature data, the pre-flight trajectory adjustment is completed, and the flight control parameters are updated according to the association between the replanned path and the real-time position environmental characteristics.

7. The method according to claim 1, characterized in that The method comprises: Record flight abnormalities and mark them as abnormalities; Establish the risk factor of the regional location based on the abnormal identification results; Planning constraints for trajectory planning are performed according to the risk factor to complete real-time update of the trajectory of the following cluster.

8. A civilian heterogeneous unmanned cluster path coordination system based on human-machine collaboration, used to implement the civilian heterogeneous unmanned cluster path coordination method based on human-machine collaboration as described in any one of claims 1 to 7, characterized in that: The method comprises: A flight mission acquisition module, which is used to acquire the UAV collaborative flight mission, analyze the UAV collaborative flight mission, and establish N subtask monomers; A task mapping establishment module, the task mapping establishment module is used to perform initial environment collection according to the UAV collaborative flight mission, establish a basic environment library, perform idle UAV self-check, establish a feature set of idle UAVs, the feature set includes performance features and power features, perform feature matching with the basic environment library, N subtask monomers and feature set, complete task pairing of idle UAVs with N subtask monomers, and establish N UAV-subtask mapping; A drone classification module, wherein the drone classification module is used to classify N drones according to the N drone-subtask mapping, and establish a search cluster and a follow cluster; An environmental data collection module, the environmental data collection module is used to plan the flight trajectory of the search cluster and the pre-flight trajectory of the follow-up cluster according to the corresponding subtask, and perform real-time environmental data collection with the search cluster; The trajectory adjustment module is used to perform pre-flight trajectory adjustment based on the real-time environmental data collection results and the sub-task monomers of the follow-up cluster, and update the flight control parameters to complete the path coordination of the UAV collaborative flight mission.

Images