A heterogeneous unmanned aerial vehicle multi-region coverage task allocation method

Abstract

This invention provides a method for allocating multi-region coverage tasks for heterogeneous unmanned aerial vehicles (UAVs), relating to the fields of combinatorial optimization and UAV scheduling. The method includes: building a parameter prediction model through multiple iterations based on simulated annealing; inputting task region information into the parameter prediction model to obtain the optimal parameters required for subsequent task packaging; running an improved density clustering algorithm under the optimal parameters to obtain packaged task clusters, each containing multiple task sets formed by partitioning all task regions; inputting the task clusters into an LKH solver to obtain cluster attribute information: scan area and path length; and establishing a linear programming solution framework based on UAV information and task cluster information to allocate task clusters. This invention addresses multi-region coverage task allocation for heterogeneous UAVs by packaging task clusters using intelligent clustering and optimizing task cluster allocation through linear programming, effectively reducing the completion time of multi-region coverage tasks and improving UAV task execution efficiency.

Description

Technical Field

[0001] This invention relates to the field of combinatorial optimization and UAV scheduling, and in particular to a method for multi-regional task allocation of heterogeneous UAVs. Background Technology

[0002] In recent years, unmanned aerial vehicles (UAVs) have demonstrated great potential due to their advantages such as low cost, high flexibility, and strong adaptability. However, limited by the range, payload, sensing, and computing capabilities of a single UAV, the use of multi-UAV swarms for mission execution has become a mainstream industry trend. While homogeneous UAV swarms, where all UAVs have identical performance and payload configurations, can improve mission execution capabilities through economies of scale, their single functional configuration is insufficient to adapt to diverse mission requirements and complex environmental constraints when facing complex multi-area coverage scenarios. In contrast, heterogeneous UAV swarms, which can be configured with UAVs of different flight performance, payload types, and operational capabilities, can significantly improve mission execution efficiency and system robustness, and have become a current research hotspot in UAV swarm technology.

[0003] Currently, there are two main technical approaches to the multi-region coverage task allocation problem of heterogeneous UAV swarms: one is mathematical programming methods, represented by mixed-integer linear programming and vehicle routing problem models. These methods can obtain the theoretically optimal solution by globally traversing the solution space, but their computational complexity increases exponentially with the number of task regions and the scale of the UAV swarm. They can only be adapted to small-scale task scenarios and cannot meet the engineering real-time solution requirements of large-scale multi-discrete region coverage tasks. The other is heuristic approximate solution methods, represented by genetic algorithms, ant colony algorithms, and particle swarm algorithms. These methods can reduce computational complexity through heuristic optimization strategies and obtain feasible solutions in a short time. However, their solution performance is heavily dependent on the manual tuning of algorithm hyperparameters. Improper parameter selection can easily lead to slow convergence, getting trapped in local optima, and unstable solution quality. Moreover, manual parameter tuning is highly dependent on the engineering experience of technical personnel, and its generalization and adaptability under different task scenarios is extremely poor. In addition, existing technical solutions still have many common technical defects: most solutions decouple task allocation from trajectory planning, completing task allocation first and then conducting single-aircraft trajectory planning, ignoring the direct impact of the spatial distribution of the task area on the flight path length and actual operation time, resulting in a significant deviation between the allocation result and the global optimal solution. Moreover, existing solutions are mostly designed for continuous coverage scenarios in a single large area, making it difficult to directly adapt to the full coverage operation requirements of multiple discrete areas. Existing density-based task clustering solutions require key parameters such as the core cutoff radius to be manually preset, which cannot adaptively optimize according to the spatial distribution characteristics of the task area, easily leading to problems such as uneven task load within the cluster and high spatial overlap between clusters, creating inherent defects in subsequent task allocation. At the same time, existing solutions are difficult to achieve precise matching between the performance differences of heterogeneous UAVs and the task load, which easily leads to the problem of unbalanced UAV load within the cluster, causing some UAVs to be idle and some UAVs to be overloaded, ultimately significantly lengthening the overall task completion time and failing to maximize the efficiency of heterogeneous UAV cluster operations. Summary of the Invention

[0004] The purpose of this invention is to provide a method for multi-region task allocation of heterogeneous UAVs. Through a three-stage step-by-step optimization architecture, it achieves adaptive optimization of clustering parameters, intelligent clustering and packaging of multi-task regions, and optimal task allocation of heterogeneous UAVs. This solves the technical problems in the prior art, such as load imbalance of heterogeneous UAVs, difficulty in manual parameter tuning, and difficulty in balancing solution efficiency and solution quality. It significantly improves the execution efficiency of multi-region coverage tasks of heterogeneous UAV clusters.

[0005] To achieve the above objectives, this invention proposes a multi-regional task allocation method for heterogeneous unmanned aerial vehicles (UAVs), comprising the following steps: Step S1, Parameter Optimization: Obtain task area information, input the task information into the parameter prediction model built based on simulated annealing algorithm, and obtain the optimal cutoff radius parameter required for the task packaging process through at least one iteration. Step S2, Task Packaging: Run the improved density clustering algorithm under the optimal cutoff radius parameter to cluster and package all task regions to obtain task clusters. Each task cluster contains at least one set of tasks formed after the division of all task regions. Input each task cluster into the LKH solver to calculate the attribute information of each task cluster, including the scan area and path length. Step S3, Task Allocation: Obtain the working capability information and task cluster information of heterogeneous UAVs, establish a linear programming solution framework, dynamically allocate task clusters with the optimization objective of minimizing the global maximum task completion time, and solve for the optimal allocation result of task clusters. The task cluster information includes the total scan area and total path length of all task regions in the task cluster, and the working capability information includes the flight speed and scanning speed of the UAV.

[0006] Preferably, step S1 specifically includes the following steps: Step S11: Obtain the coordinates of the task area and determine the cutoff radius of the parameter to be optimized for density clustering. Set the cutoff radius The search range is The initial solution is ; Step S12: Set the hyperparameters for the simulated annealing algorithm, including but not limited to the initial temperature. Cooling coefficient Termination temperature and maximum number of iterations ; Step S13: Set the current solution Substituting the values ​​into the density clustering algorithm, we obtain the clustering results and thus the number of drones. Calculate the average distance from all task regions within the cluster to the cluster center. , computation cluster with cluster Distance between centers The energy function is obtained. The calculation formula is: ; in, , Number the task clusters; Step S14: Set the initial solution Substituting the values ​​into the density clustering algorithm, we obtain the clustering results. Based on the formula for the energy function, we calculate the initial energy. ; Step S15: For the current solution By applying a random perturbation to the vicinity, a neighborhood solution is obtained. Based on the formula for the energy function, the neighborhood solution is calculated. energy ; Step S16: Determine and The size, if Accept neighborhood solutions and update , ;like Calculate the acceptance probability and with probability Accept neighborhood solutions, accept probability The calculation formula is: ; in, The current temperature; Step S17: Perform cooling operation ,in, For the first Temperature during round iteration Let the number of iterations be the cooling coefficient. Determine whether the iteration termination condition is met. The iteration termination condition is: the current temperature is satisfied. Current iteration round number If any one of the conditions is not met, return to step S14; if it is met, output the current solution. As the optimal cutoff radius .

[0007] Preferably, in step S11, the cutoff radius is set. The search range is The initial solution is Specifically, this involves calculating the distances between all task regions, sorting the distances in ascending order, and taking the distance ranking in the top 2% as the value. Take the distance value that ranks in the top 10%. Generate a random number between 0 and 1. , The calculation formula is: .

[0008] Preferably, in step S15, the current solution... By applying a random perturbation to the vicinity, a neighborhood solution is obtained. Specifically, this involves determining the perturbation step size. Generate a random number between 0 and 1. Through formula The neighborhood solution is calculated. .

[0009] Preferably, step S2 specifically includes the following steps: Step S21: Obtain the number of drones The basic operating parameters of heterogeneous UAVs, and based on the optimal cutoff radius. Calculate the local density of each task region and relative distance ; Step S22: Using the formula Calculate task area scores and categorize task areas according to scores. Sort in descending order by the number of drones. Select the corresponding top ranking The task region is used as the cluster center. ; Step S23: Assign the remaining non-central regions to the task clusters containing the task regions that are closest to the remaining non-central regions and have higher local density, and complete the clustering of all task regions to obtain all task clusters; Step S24: Solve using the LKH solver to obtain the optimal scan path length, sum the scan areas within each task cluster to obtain the total scan area of ​​each task cluster.

[0010] Preferably, step S21 specifically includes the following steps: Step S211: Calculate the task area The average scan time is calculated using the following formula: ; in, For the mission area Average scan time, For the mission area The scanning area, This represents the average scanning speed of all drones. Step S212: Calculate from the task region Transfer to The required average flight time is calculated using the following formula: ; in, To the mission area Transfer to Required average flight time, For the mission area With the mission area The distance between them This represents the average flight speed of all drones. Step S213: Calculate the Boolean value for all task regions using the following formula: ; in, For all task regions, Boolean value. To be the optimal cutoff radius, It is a Boolean function; Step S214: Specify the impact factor Calculate local density and relative distance The calculation formula is: ; ; in, Number the task area. The number of task areas. For the mission area The local density.

[0011] Preferably, step S3 specifically includes the following steps: Step S31: Obtain task cluster information and heterogeneous UAV operational capability information, specifically: Task cluster information includes: The sum of the scan areas of all task regions in the task cluster ; Optimal path length for all task regions in the scan cluster ; Information on the operational capabilities of heterogeneous drones includes: drones flight speed ; drones scanning speed ; Step S32: Specify the solver's constraints and the form of the output solution, specifically: Constraints: Each drone It can be assigned to one and only one task cluster. It satisfies the formula: ; in, As decision variables, Number the drone; Each task cluster It can only be assigned to one drone. It satisfies the formula: ; in, Number the task clusters; The output solution is in the form of a two-dimensional array. ,in, Indicates the task cluster Handed over to drones implement; Step S33: Define an objective function that minimizes the maximum task completion time among all drones. Specifically, calculate the drone's... Scanning task clusters The formula for calculating the time required for all regions is: ; in, For task clusters The sum of the lengths of all scan paths in the entire region. For task clusters The sum of all scanned areas in the entire region; Define the objective function as follows: ; Step S34: Use a mixed-integer linear programming solver to solve the problem and obtain the task cluster allocation results.

[0012] Preferably, mixed-integer linear programming solvers include, but are not limited to, the Python-PuLP open-source linear programming problem solver.

[0013] Therefore, this invention proposes a multi-regional task allocation method for heterogeneous unmanned aerial vehicles (UAVs), which has the following beneficial effects: (1) This invention constructs a parameter prediction model based on simulated annealing algorithm, realizes adaptive optimization of the core cutoff radius parameter of density clustering, and does not require manual parameter tuning. It solves the technical problems of traditional heuristic algorithm parameters relying on human experience and poor generalization ability in different task scenarios. At the same time, by designing a special energy function, it ensures that the optimized parameters can achieve the optimal clustering effect of "high intra-cluster density and large inter-cluster separation".

[0014] (2) This invention designs an improved density clustering algorithm, which calculates the local density by combining the scanning area of ​​the task region, the spatial distance, and the average operation capability of the UAV, and realizes the clustering and packaging of multiple task regions. At the same time, the LKH solver is introduced to pre-plan the optimal scanning path of each task cluster, and the task allocation and path planning are optimized in a coordinated manner, accurately quantifying the actual execution time of each task cluster, which greatly improves the accuracy of the subsequent allocation results. At the same time, by clustering and packaging, the large-scale multi-region allocation problem is transformed into a task cluster allocation problem that matches the number of UAVs, which greatly reduces the solution dimension of the problem, solves the pain point that the computational complexity of traditional mathematical programming methods increases exponentially with the task scale, and takes into account both the solution efficiency and the quality of the solution.

[0015] (3) In view of the performance differences of heterogeneous UAVs, the present invention constructs a mixed integer linear programming solution framework, with the optimization objective of minimizing the global maximum task completion time. By combining the flight speed and scanning speed of the UAV with the path length and scanning area of ​​the task cluster, the present invention achieves accurate matching between the task cluster and the heterogeneous UAVs, solves the problem of load imbalance of heterogeneous UAV clusters in traditional methods, effectively shortens the overall completion time of multi-area coverage tasks, and maximizes the operational efficiency of UAV clusters. Attached Figure Description

[0016] Figure 1 A flowchart of a multi-regional task allocation method for heterogeneous unmanned aerial vehicles; Figure 2 A flowchart of the parameter optimization step in a multi-regional task allocation method for heterogeneous unmanned aerial vehicles; Figure 3 A flowchart of the task packaging step in a multi-regional task allocation method for heterogeneous unmanned aerial vehicles; Figure 4 This is a flowchart of the task allocation step in a heterogeneous UAV multi-region task allocation method. Detailed Implementation

[0017] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0018] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0019] Example 1 For ease of description, this discussion takes the typical scenario of drone swarm task allocation, where multiple heterogeneous drones cover and scan multiple separate task areas within the task scenario. All environmental states within the task scenario are visible to each drone. Furthermore, in this embodiment, only the result of drone task allocation is considered, without considering how the drones scan after arriving at the task area.

[0020] like Figure 1 As shown, this invention proposes a multi-regional task allocation method for heterogeneous unmanned aerial vehicles (UAVs), specifically including the following steps: Step S1, Parameter Optimization: Obtain task region information, input the task information into the parameter prediction model built based on the simulated annealing algorithm, and obtain the optimal cutoff radius parameter required for the task packaging process through at least one iteration. Figure 2 As shown, the specific steps include: Step S11: Obtain task region information and set the required operating parameters for the simulated annealing algorithm, specifically: set... Given a set of task regions, obtain the coordinates of each task region. Its coordinates are ,in, For the mission area x-axis, For the mission area The ordinate is used to determine the cutoff radius of the density clustering parameter. Set the cutoff radius The search range is The initial solution is ,in: and This is achieved by calculating the distances between all task regions, sorting them in ascending order, and then taking the distance values ​​ranked at the 2% and 10% thresholds. (Initial solution) The calculation formula is as follows: ; in, A random number between 0 and 1; Step S12: Set the hyperparameters for the simulated annealing algorithm, including but not limited to the initial temperature. Cooling coefficient Termination temperature and maximum number of iterations ; Step S13: Set the current solution Substituting the values ​​into the density clustering algorithm, we obtain the clustering results and thus the number of drones. Calculate the average distance from all task regions within the cluster to the cluster center. , computation cluster with cluster Distance between centers The energy function is obtained. The calculation formula is: ; Step S14: Set the initial solution Substituting the values ​​into the density clustering algorithm, we obtain the clustering results. Based on the formula for the energy function, we calculate the initial energy. ; Step S15: For the current solution By applying a random perturbation to the vicinity, a neighborhood solution is obtained. Based on the formula for the energy function, the neighborhood solution is calculated. energy Specifically, this involves determining the perturbation step size. Generate a random number between 0 and 1. Through formula The neighborhood solution is calculated. ; Step S16: Determine and The size, if Then accept the neighborhood solution. and update , ,like Calculate the acceptance probability and with probability Accept neighborhood solutions, accept probability The calculation formula is: ; in, The current temperature; Step S17: Perform cooling operation ,in, For the first Temperature during round iteration Let the number of iterations be the cooling coefficient. Determine whether the iteration termination condition is met. The iteration termination condition is that the current temperature is satisfied. Current iteration round number If any one of the conditions is not met, return to step S14; if it is met, output the current solution. As the optimal cutoff radius .

[0021] Step S2, Task Packaging: Under the optimal cutoff radius parameter, run the improved density clustering algorithm to cluster and package all task regions, resulting in task clusters. Each task cluster contains at least one set of tasks formed after the division of all task regions. Input each task cluster into the LKH solver to calculate the attribute information of each task cluster, including the scan area and path length, such as... Figure 3 As shown, the specific steps include: Step S21: Obtain UAV parameters, let A collection of drones, For drones Information on work capabilities, including For the drone's flight speed, For the drone's scanning speed, and based on the optimal cutoff radius Calculate the local density of each task region and relative distance Specifically, it includes the following steps: Step S211: Calculate the task area The average scan time is calculated using the following formula: ; in, For the mission area Average scan time, For the mission area The scanning area, This represents the average scanning speed of all drones. Step S212: Calculate from the task region Transfer to The required average flight time is calculated using the following formula: ; in, To the mission area Transfer to Required average flight time, For the mission area With the mission area The distance between them This represents the average flight speed of all drones. Step S213: Calculate the Boolean value for all task regions using the following formula: ; in, For all task regions, Boolean value. To be the optimal cutoff radius, It is a Boolean function; Step S214: Specify the impact factor Calculate local density and relative distance The calculation formula is: ; ; in, Number the task area. The number of task areas. For the mission area The local density.

[0022] Step S22: Using the formula Calculate task area scores and categorize task areas according to scores. Sort in descending order by the number of drones. Select the corresponding top ranking The task area is used as the cluster center; Step S23: Assign other non-central regions to the same cluster as their nearest regions with higher local density to obtain all task clusters; Step S24: Pass the task cluster to the LKH solver, which automatically performs an adaptive neighborhood search with a penalty term to complete the task. Edge-swapping optimization is performed to obtain the optimal access sequence of the UAV to the task area. Based on this optimal access sequence, the optimal path length of all task areas in the scan cluster is calculated. Finally, the sum of the scan areas of all task areas in the task cluster is calculated to obtain all task information for the task cluster, completing the packaging of the task cluster. In this embodiment, the LKH solver version used is LKH 2.0.

[0023] Step S3, Task Allocation: Obtain heterogeneous UAV operational capability information and task cluster information, establish a linear programming solution framework, dynamically allocate task clusters with the optimization objective of minimizing the global maximum task completion time, and solve for the optimal allocation result of the task clusters. The task cluster information includes the total scan area and total path length of all task regions within the task cluster, and the operational capability information includes the UAV's flight speed and scanning speed, such as... Figure 4 As shown, it includes the following steps: Step S31: Obtain task cluster information and heterogeneous UAV operational capability information, and enable... A collection of task clusters, each task cluster Task volume information is Specifically: The task cluster information includes: The sum of the scan areas of all task regions in the task cluster ; Optimal path length for all task regions in the scan cluster ; Information on the operational capabilities of heterogeneous drones includes: drones flight speed ; drones scanning speed ; Step S32: Specify the solver's constraints and the form of the output solution, specifically: Constraints: Each drone can be assigned to one and only one task cluster, satisfying the formula: ; in, As decision variables, Number the task clusters. Number the drone; Each task cluster can only be assigned to one drone, satisfying the formula: ; The output solution is in the form of a two-dimensional array. ,in, Indicates the task cluster Handed over to drones implement; Step S33: Define an objective function that minimizes the maximum task completion time among all drones. Specifically, calculate the drone's... Scanning task clusters The formula for calculating the time required for all regions is: ; in, For task clusters The sum of the lengths of all scan paths in the entire region. For task clusters The sum of all scanned areas in the entire region; Define the objective function as follows: ; Step S34: Input the constraints, objective function, task cluster workload, and other information into the mixed-integer linear programming solver for solving. The solver obtains the task cluster allocation results. The task assignment is completed. In this embodiment, the Python-PuLP open-source solver is used for solving the problem. The solver versions used are: Python 3.10.15 and PuLP 2.9.0.

[0024] It is worth noting that all contents not described in detail in this invention are existing technologies and are well known to those skilled in the art.

[0025] This invention mitigates the load imbalance problem among UAVs by employing a mixed-integer linear programming method, combining task cluster attributes with UAV capability differences. Simultaneously, it improves the efficiency of task allocation by automatically predicting optimal parameters for improved density clustering through simulated annealing iteration. Furthermore, this invention collaboratively optimizes task allocation and path planning to minimize the maximum task completion time among all UAVs.

[0026] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for allocating multi-area coverage tasks for heterogeneous unmanned aerial vehicles (UAVs), characterized in that, Includes the following steps: Step S1, Parameter Optimization: Obtain task area information, input the task area information into the parameter prediction model built based on simulated annealing algorithm, and obtain the optimal cutoff radius parameter required for the task packaging process through at least one iteration. Step S2, Task Packaging: Run the improved density clustering algorithm under the optimal cutoff radius parameter to cluster and package all task regions to obtain task clusters. Each task cluster contains at least one set of tasks formed after the division of all task regions. Input each task cluster into the LKH solver to calculate the attribute information of each task cluster, including the scan area and path length. Step S3, Task Allocation: Obtain the working capability information and task cluster information of heterogeneous UAVs, establish a linear programming solution framework, dynamically allocate task clusters with the optimization objective of minimizing the global maximum task completion time, and solve for the optimal allocation result of task clusters. The task cluster information includes the total scan area and total path length of all task regions in the task cluster, and the working capability information includes the flight speed and scanning speed of the UAV.

2. The method for allocating multi-area coverage tasks for heterogeneous unmanned aerial vehicles according to claim 1, characterized in that: Step S1 specifically includes the following steps: Step S11: Obtain the coordinates of the task area and determine the cutoff radius of the parameter to be optimized for density clustering. Set the cutoff radius The search range is The initial solution is ; Step S12: Set the hyperparameters for the simulated annealing algorithm, including but not limited to the initial temperature. Cooling coefficient Termination temperature and maximum number of iterations ; Step S13: Set the current solution Substituting the values ​​into the density clustering algorithm, we obtain the clustering results and thus the number of drones. Calculate the average distance from all task regions within the cluster to the cluster center. , computation cluster with cluster Distance between centers The energy function is obtained. The calculation formula is: ； in, , Assign task cluster numbers; Step S14: Set the initial solution Substituting the values ​​into the density clustering algorithm, we obtain the clustering results. Based on the formula for the energy function, we calculate the initial energy. ; Step S15: For the current solution By applying a random perturbation to the vicinity, a neighborhood solution is obtained. Based on the formula for the energy function, the neighborhood solution is calculated. energy ; Step S16: Determine and The size, if Accept neighborhood solutions and update , ;like Calculate the acceptance probability and with probability Accept neighborhood solutions, accept probability The calculation formula is: ； in, The current temperature; Step S17: Perform cooling operation ,in, For the first Temperature during round iteration Let the number of iterations be the cooling coefficient. Determine whether the iteration termination condition is met. The iteration termination condition is: the current temperature is satisfied. Current iteration round number If any one of the conditions is not met, return to step S14; if it is met, output the current solution. As the optimal cutoff radius .

3. The method for allocating multi-area coverage tasks for heterogeneous unmanned aerial vehicles according to claim 2, characterized in that, In step S11, the cutoff radius is set. The search range is The initial solution is Specifically, this involves calculating the distances between all task regions, sorting the distances in ascending order, and taking the distance ranking in the top 2% as the value. Take the distance value that ranks in the top 10%. Generate a random number between 0 and 1. , The calculation formula is: .

4. The method for allocating multi-area coverage tasks for heterogeneous unmanned aerial vehicles according to claim 2, characterized in that: In step S15, the current solution is... By applying a random perturbation to the vicinity, a neighborhood solution is obtained. Specifically, this involves determining the perturbation step size. Generate a random number between 0 and 1. Through formula The neighborhood solution is calculated. .

5. The method for allocating multi-area coverage tasks for heterogeneous unmanned aerial vehicles according to claim 1, characterized in that: Step S2 specifically includes the following steps: Step S21: Obtain the number of drones The basic operating parameters of heterogeneous UAVs, and based on the optimal cutoff radius. Calculate the local density of each task region and relative distance ; Step S22: Using the formula Calculate task area scores and categorize task areas according to scores. Sort in descending order by the number of drones. Select the corresponding top ranking The task region is used as the cluster center. ; Step S23: Assign the remaining non-central regions to the task clusters containing the task regions that are closest to the remaining non-central regions and have higher local density, and complete the clustering of all task regions to obtain all task clusters; Step S24: Solve using the LKH solver to obtain the optimal scan path length, sum the scan areas within each task cluster to obtain the total scan area of ​​each task cluster.

6. The method for allocating multi-area coverage tasks for heterogeneous unmanned aerial vehicles according to claim 5, characterized in that: Step S21 specifically includes the following steps: Step S211: Calculate the task area The average scan time is calculated using the following formula: ； in, For the mission area Average scan time, For the mission area The scanning area, This represents the average scanning speed of all drones. Step S212: Calculate from the task region Transfer to The required average flight time is calculated using the following formula: ； in, To the mission area Transfer to Required average flight time, For the mission area With the mission area The distance between them This represents the average flight speed of all drones. Step S213: Calculate the Boolean value for all task regions using the following formula: ； in, For all task regions, Boolean value. To be the optimal cutoff radius, It is a Boolean function; Step S214: Specify the impact factor Calculate local density and relative distance The calculation formula is: ； ； in, Number the task area. The number of task areas. For the mission area The local density.

7. The method for allocating multi-area coverage tasks for heterogeneous unmanned aerial vehicles according to claim 1, characterized in that: Step S3 specifically includes the following steps: Step S31: Obtain task cluster information and heterogeneous UAV operational capability information, specifically: Task cluster information includes: The sum of the scan areas of all task regions in the task cluster ; Optimal path length for all task regions in the scan cluster ; Information on the operational capabilities of heterogeneous drones includes: drones flight speed ; drones scanning speed ; Step S32: Specify the solver's constraints and the form of the output solution, specifically: Constraints: Each drone It can be assigned to one and only one task cluster. It satisfies the formula: ； in, As decision variables, Number the drone; Each task cluster It can only be assigned to one drone. It satisfies the formula: ； in, Assign task cluster numbers; The output solution is in the form of a two-dimensional array. ,in, Indicates the task cluster Handed over to drones implement; Step S33: Define an objective function that minimizes the maximum task completion time among all drones. Specifically, calculate the drone's... Scanning task clusters The formula for calculating the time required for all regions is: ； in, For task clusters Optimal path length for all task regions For task clusters The total scanned area of ​​all task regions; Define the objective function as follows: ； Step S34: Use a mixed-integer linear programming solver to solve the problem and obtain the task cluster allocation results.

8. A method for allocating multi-area coverage tasks for heterogeneous unmanned aerial vehicles according to claim 7, characterized in that: Mixed-integer linear programming solvers include, but are not limited to, the Python-PuLP open-source linear programming problem solver.

Images