Unmanned aerial vehicle group multi-area coverage path planning method and equipment

Abstract

The invention provides an unmanned aerial vehicle group multi-area coverage path planning method and device, and relates to the technical field of unmanned aerial vehicle coverage path planning, and the method comprises the steps: S1, building a problem model based on an unmanned aerial vehicle group multi-area coverage task, and the problem model comprises an environment model, a multi-target cost function and a constraint condition; and S2, performing full-coverage path planning by taking the multi-target cost function of the optimization problem model as a target, and calculating to obtain an optimal path of the unmanned aerial vehicle group. According to the method, the problem model of the multi-area coverage task of the unmanned aerial vehicle group is constructed based on the environment model, the multi-target cost function and the constraint condition, the priorities of the number of unmanned aerial vehicles, the total path distance and the cooperation efficiency of the unmanned aerial vehicles and the cluster cooperation constraint in the task process are considered, and the three targets are optimized at the same time when the problem model is optimized. And finally, a modeling effect meeting task requirements is formed.

Description

Technical Field

[0001] The present invention relates to the technical field of coverage path planning of unmanned aerial vehicles (UAVs), and in particular to a method and device for planning multi-area coverage paths of a swarm of unmanned aerial vehicles (UAVs). Background Art

[0002] With the rapid development of drone technology, it has been widely used in many fields and has also brought new opportunities for tasks such as monitoring and assessment in disaster emergency management. For example, disasters such as earthquakes, floods or fires can cause damage, and rapid and accurate assessments are needed to guide emergency response and recovery efforts. Drones equipped with onboard sensors have the ability to quickly cover large areas, access otherwise difficult-to-reach locations, and capture high-resolution images. They are a safe and reliable technology for disaster emergency management and obtaining information about disaster areas, which is of great significance to subsequent decision-making processes and rescue operations.

[0003] Coverage path planning is an essential step in planning drone visits and covering all target areas. Drones have significant advantages in regional full coverage tasks, and most current research focuses on the problem of drone coverage path planning for a single area. However, in actual scenarios, the consistency of coverage tasks often needs to be considered, that is, to perform multi-area coverage tasks. Compared with a single drone, a drone swarm has obvious advantages in covering a larger range and more areas. At the same time, as the range increases and the number of areas increases, the path planning problem becomes more complicated. The collaborative work of drone swarms also brings about complex regional allocation and path optimization problems. In our actual scenarios, unreasonable planning may result in waste of drone resources, repeated visits or non-visits to certain areas, etc.

[0004] The coverage problem of drone swarms in large-scale multi-target regional scenarios is an integration of the MTSP problem and the coverage path planning (CPP) problem. The MTSP problem is an NP-Hard problem, while the integration problem is more challenging. Although the path planning methods of single drones and swarm drones have achieved certain research results and have shown potential in practical applications, when facing multi-region collaborative path planning for large-scale post-disaster assessments, there are still challenges such as multi-region allocation, efficient collaboration of multiple drones, and comprehensive optimization of paths within and outside the region that need to be solved urgently, and the collaboration efficiency and coverage efficiency of multi-drone multi-region tasks need to be further improved. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide a method and device for multi-area coverage path planning of a drone swarm, which is used to solve the problem of inaccurate path planning when existing drone swarms perform multi-area tasks.

[0006] The present invention provides a method for planning a multi-area coverage path of a drone swarm, comprising:

[0007] S1: Construct a problem model based on the multi-area coverage task of drone swarms. The problem model includes: environmental model, multi-objective cost function and constraints;

[0008] S2: Perform full coverage path planning based on the multi-objective cost function of the optimization problem model and calculate the optimal path for the drone swarm.

[0009] Preferably, step S1 specifically comprises:

[0010] S11: Based on the multi-area coverage mission of the drone swarm, multiple drones are used to scan multiple mission areas and build an environmental model;

[0011] S12: construct a multi-objective cost function based on the number of drones, total path distance and drone cooperation efficiency, and set constraints for the multi-objective cost function;

[0012] S13: Using environmental model, multi-objective cost function and constraints as the problem model of multi-area coverage task of UAV swarm.

[0013] Preferred:

[0014] The expression of the multi-objective cost function F is:

[0015] F=min{aF1+bF2+cF3}

[0016] In the formula, a, b and c are weight coefficients, F1, F2 and F3 are the number of drones, total path distance and drone cooperation efficiency respectively. The formula is as follows:

[0017] F1=m

[0018] In the formula, m is the number of all drones used when performing the mission;

[0019]

[0020] Where D i is the path distance of the i-th UAV, and the formula is as follows:

[0021]

[0022] Where j and k are regions, j,k∈{0,1,…,n}, region 0 is the drone base; x i,j,k is a decision variable. If the i-th drone moves from area j to k, then x i,j,k =1, otherwise x i,j,k =0; H irepresents the mobile height assigned to the i-th UAV; α is the contribution coefficient, which adjusts the contribution of the mobile height to the total path distance; h represents the mission height of all UAVs; C j,k represents the path distance of the drone from area j to area k at the moving height; r k represents the path distance of the UAV at the mission altitude within region k;

[0023]

[0024] In the formula, is the mean path distance of the i-th UAV.

[0025] Preferably, step S2 specifically comprises:

[0026] S21: Set the parameters of the particle population, including: the number of particles, the maximum number of iterations Q, the inertia weight coefficient and the weight coefficient, construct the particle position set and the particle velocity set through the index number of the drone base and each area, initialize the particle population, the particle position set and the particle velocity set, and set q = 0;

[0027] S22: let q=q+1;

[0028] S23: Extract the u-th particle from the particle population and update the particle velocity V of the u-th particle u ;

[0029] S24: Obtain the unvisited region set of the u-th particle. If the unvisited region set is not an empty set, proceed to step S25; otherwise, proceed to step S26;

[0030] S25: Update the particle position X of the uth particle u , the particle position X is swept by the sweep function u Translated into full coverage waypoints for each drone;

[0031] S26: Update the fitness, local optimal value and global optimal value of all particles with the goal of optimizing the multi-objective cost function;

[0032] S27: If q=Q, the full coverage path point of each drone corresponding to the global optimal particle at this time is used as the optimal path of the drone group, otherwise return to step S22.

[0033] Preferred:

[0034] The expression for the particle position is:

[0035]

[0036] Among them, the particle position X u A particle with n+1 dimensions The access group element set, the particle of the dth dimension The access group elements obtained by the dth region and the previous access region a and the next access region b respectively<a,d> and<d,b> The access group elements are composed of two groups of area index numbers.

[0037] Preferred:

[0038] The expression for particle velocity is:

[0039]

[0040] Among them, V u represents the particle velocity with n+1 dimensions; p(a,d)∈[0,1] is a probability function, which represents the probability of each pair of access group elements being selected during the particle velocity update process.

[0041] A storage medium, wherein the storage medium stores instructions and data for implementing the multi-area coverage path planning method of a drone swarm.

[0042] A multi-area coverage path planning device for a drone swarm includes: a processor and a storage medium; the processor loads and executes instructions and data in the storage medium to implement the multi-area coverage path planning method for a drone swarm.

[0043] The present invention has the following beneficial effects:

[0044] Based on the environmental model, multi-objective cost function and constraint conditions, a problem model of the multi-area coverage task of a drone swarm was constructed. The priorities of the three objectives of the number of drones, total path distance and drone cooperation efficiency during the task and the constraints of the swarm cooperation were taken into account. When optimizing the problem model, the three objectives were optimized at the same time, and finally a modeling effect that met the task requirements was formed.

[0045] Full coverage path planning is performed with the multi-objective cost function of the optimization problem model as the goal, so as to improve the performance of the regional allocation selection mechanism and the comprehensive optimization of paths inside and outside the region, improve the coverage efficiency and collaboration efficiency, which is conducive to the actual production application of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is a flow chart of a method according to an embodiment of the present invention;

[0047] Figure 2 A schematic diagram to record the camera's footprint size;

[0048] Figure 3 The experimental results of full coverage path planning in a scenario with a range of 500m×500m, a target area of ​​50, and a maximum path distance constraint of 3 kilometers for the UAV;

[0049] Figure 4 The following is the experimental result of multi-area full coverage path planning of three algorithms when the number of target areas is 50;

[0050] Figure 5 The figure is a comparison of the balance rates of different algorithms in full coverage path planning;

[0051] Figure 6 This is a structural diagram of the device according to an embodiment of the present invention;

[0052] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0053] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0054] Reference Figure 1 The present invention provides a method for planning a multi-area coverage path of a drone swarm, comprising:

[0055] S1: Construct a problem model based on the multi-area coverage task of drone swarms. The problem model includes: environmental model, multi-objective cost function and constraints;

[0056] Furthermore, step S1 is specifically as follows:

[0057] S11: Based on the multi-area coverage mission of the drone swarm, multiple drones are used to scan multiple mission areas and build an environmental model;

[0058] Specifically, a group of drones equipped with downward-facing cameras are used to cover multiple scattered convex polygonal areas. All drones take off and land from the same base. There are no restricted areas in the target area that are prohibited from access. Only the flight and scanning performance of the drones are focused on, and other functions and properties are ignored. The footprint size (coverage grid) of the onboard camera is determined based on the drone height and camera characteristics. The specific grid coverage area is determined by the grid length and width, such as Figure 2 shown.

[0059] S12: construct a multi-objective cost function based on the number of drones, total path distance and drone cooperation efficiency, and set constraints for the multi-objective cost function;

[0060] Furthermore, the expression of the multi-objective cost function F is:

[0061] F=min{aF1+bF2+cF3}

[0062] In the formula, a, b and c are weight coefficients, F1, F2 and F3 are the number of drones, total path distance and drone cooperation efficiency respectively. The formula is as follows:

[0063] F1=m

[0064] In the formula, m is the number of all drones used when performing the mission;

[0065]

[0066] Where D i is the path distance of the i-th UAV, and the formula is as follows:

[0067]

[0068] Where j and k are regions, j,k∈{0,1,…,n}, region 0 is the drone base; x i,j,k is a decision variable. If the i-th drone moves from area j to k, then x i,j,k =1, otherwise x i,j,k =0; H i represents the mobile height assigned to the i-th UAV; α is the contribution coefficient, which adjusts the contribution of the mobile height to the total path distance; h represents the mission height of all UAVs; C j,k represents the path distance of the drone from area j to area k at the moving height; r k represents the path distance of the UAV at the mission altitude within region k;

[0069] Specifically, the speed of the drone when climbing and descending is slower than the horizontal movement speed, and the coefficient α is used to adjust the contribution of the moving height to the total path distance;

[0070]

[0071] In the formula, is the mean path distance of the i-th UAV.

[0072] Specifically, F3 is the path distance variance value of the drone group, and the variance is used to represent the collaboration efficiency (collaboration synchronization rate). When a multi-drone system coordinates operations, the smaller the distance variance of the drones, the smaller the path distance difference between the drones, indicating a higher collaboration efficiency.

[0073] Furthermore, based on the research problem, the task-related constraints of multiple UAVs covering multiple areas are obtained: such as all UAVs departing from and returning to the base, restrictions on non-overlapping area visits, and energy consumption constraints when UAVs perform tasks.

[0074] S13: Using environmental model, multi-objective cost function and constraints as the problem model of multi-area coverage task of UAV swarm.

[0075] S2: Perform full coverage path planning based on the multi-objective cost function of the optimization problem model and calculate the optimal path for the drone swarm.

[0076] Furthermore, step S2 is specifically as follows:

[0077] S21: Set the parameters of the particle population, including: the number of particles, the maximum number of iterations Q, the inertia weight coefficient and the weight coefficient, construct the particle position set and the particle velocity set through the index number of the drone base and each area, initialize the particle population, the particle position set and the particle velocity set, and set q = 0;

[0078] Specifically, information of multiple target areas to be covered and the footprint size of the onboard camera are imported, the vertices of the areas to be covered are determined, the path lengths of the areas to be covered along different sweeping directions, and the access point set corresponding to the lengths are determined;

[0079] Configure the parameters of the particle population based on the access point set;

[0080] The access relationship between multiple drones and multiple regions is represented by a special encoding method based on access point sets. The particle position can access all target areas, and the speed is the possibility of the current access area.

[0081] Furthermore, the particle positions that can access all areas are initialized and constructed while satisfying the energy consumption limit of the drone;

[0082] The expression for the particle position is:

[0083]

[0084] Among them, the particle position X u A particle with n+1 dimensions The access group element set, the particle of the dth dimension The access group elements obtained by the dth region and the previous access region a and the next access region b respectively<a,d> and<d,b> The access group elements are composed of two groups of area index numbers.

[0085] Furthermore, the initialization speed uses a random method to set the speed of each dimension of each particle;

[0086] The expression for particle velocity is:

[0087]

[0088] Among them, V u represents the particle velocity with n+1 dimensions; p(a,d)∈[0,1] is a probability function, which represents the probability of each pair of access group elements being selected during the particle velocity update process.

[0089] S22: let q=q+1;

[0090] S23: Extract the u-th particle from the particle population and update the particle velocity V of the u-th particle u ;

[0091] Specifically, the particle velocity update adopts the comprehensive learning particle swarm optimization algorithm, which increases the population diversity of the group by learning different dimensions of other particles, thereby preventing premature convergence and falling into a local solution, as shown below:

[0092]

[0093] in They are the position and velocity of the particle, w is the inertia weight coefficient, c is the acceleration coefficient, and rand d is a d-dimensional random number uniformly distributed in the range [0,1], Represented as the dth dimension of the current best kth particle, index d Returns the index value of the particle with the best performance in the current d-th dimension, where index d The decision depends on the probability p s , if rand d Greater than the selection probability p s , the particle's corresponding dimension will be learned from its own pBest; otherwise, it will compare and judge the pBest of other particles to learn.

[0094] Speed-related operations are performed by updating probabilities. The particle speed update operation rules based on the set are as follows:

[0095]

[0096] Among them, the speed update is divided into two parts. The first part is the operation of converting the speed set into a possibility set, and the second part is the operation of converting the position clarity set into a possibility set. By adding the two possibility sets, the maximum probability of the current d-th dimension access group is obtained, and the speed possibility set used for subsequent position updates is obtained.

[0097] S24: Obtain the unvisited region set of the u-th particle. If the unvisited region set is not an empty set, proceed to step S25; otherwise, proceed to step S26;

[0098] S25: Update the particle position X of the uth particle u , the particle position X is swept by the sweep function u Translated into full coverage waypoints for each drone;

[0099] Specifically, the particle position update formula is as follows:

[0100]

[0101] Creating a new generation of particle positions is done by updating each individual dimension, selecting from three clear sets according to the priority of the set, in the following order: and init_X d When none of the selectable access groups meets the constraints, it returns to the base to send additional drones and create a new access path for them.

[0102] The first step is to get the speed set, as follows:

[0103]

[0104] in By using The probability of accessing the group element in the possible set is used to determine the element. When the probability of accessing the group element <a, d> in the possible set is greater than rand, its possible element is saved in the clear set. Then the position is clear set All access group elements in are contained in init_X d And comes from the previous generation of particles; finally, the universal clear set init_X d , init_X d Contains all possible access group elements.

[0105] S26: Update the fitness, local optimal value and global optimal value of all particles with the goal of optimizing the multi-objective cost function;

[0106] Specifically, the number of drones, path length, and collaboration efficiency are comprehensively considered, and the linear weighted summation method is used to set priorities for the three objectives and sum them up to calculate the fitness of the current individual;

[0107] S27: If q=Q, the full coverage path point of each drone corresponding to the global optimal particle at this time is used as the optimal path of the drone group, otherwise return to step S22.

[0108] Comparison of simulation experiment results:

[0109] The first set of experiments is to set up multiple groups of scenarios with different ranges, use different algorithms to dispatch drones to perform multi-area full coverage path planning, and compare and evaluate the performance of the drone swarm multi-area coverage path planning method (SPSO) of the present invention according to the evaluation indicators; then, in the second set of experiments, the path planning is performed for the algorithms using the greedy strategy and the algorithms not using the greedy strategy in the same scenario to evaluate the feasibility and effectiveness of the greedy strategy. The simulation results are shown in Figure 2. Figure 3 , 4 and 5 and shown in Tables 1, 2 and 3.

[0110] The present invention conducts two sets of simulation experiments to illustrate the feasibility, effectiveness and superiority of the proposed SPSO algorithm. The first set of experiments is to set up multiple sets of scenarios with different ranges, and use different algorithms to dispatch drones to perform multi-area full coverage path planning; then, in the second set of experiments, the feasibility of the greedy strategy is compared and evaluated for the same scenario using the greedy strategy and the non-adopted algorithm.

[0111] The experimental scenes are 500m×500m and 1km×1km. A certain number of target areas are randomly deployed in different scenes. These target areas are convex polygons with random shapes and sizes and are spatially dispersed. The base coordinates of the drone swarm are (0,0), the total number of target areas n is set between 10 and 200, the sweep range of the drone equipped with a camera is l×w=1.5m×3m, the mission height h of the drone swarm is 8m, the moving height H1 is 10m, α is 5, and the maximum path distance of the drone is s. At the same time, the result evaluation indicators are divided into three: 1) Total number of drones: m; 2) Total path distance: 3) Balance rate of drone swarm: Higher values ​​of the balance ratio indicate higher rates of collaborative synchronization.

[0112] In the first set of simulation experiments, the SPSO algorithm was experimentally compared with the APPA algorithm and the Fast NN algorithm to compare the multi-objective optimization path planning of minimizing the total number of drones, minimizing the total path distance of the drone swarm, and maximizing the balance rate of the drone swarm under multiple identical scenarios.

[0113] First, consider a scenario with a range of 500m×500m, a target area of ​​50, and a maximum path distance of 3km. The full coverage path planning of the drone swarm is as follows: Figure 3 As shown, Figure 3 (a) is the full coverage path planned by SPSO, Figure 3 (b) is the full coverage path planned by APPA, Figure 3 (c) is the full coverage path planned by Fast NN. The yellow square at coordinate (0,0) is the UAV base.

[0114] At the same time, we compared the planning results of the three algorithms based on the evaluation indicators proposed above, including the number of dispatched drones, the total path distance to complete the task, and the balance rate of the drone swarm. The results are shown in Table 1.

[0115] Table 1. Regional full coverage path planning data of three algorithms

[0116]

[0117] Then the scenario is 1km×1km, the path distance of the drone is limited to 10 kilometers, and the number of target areas is increased from 15, 20, 25, 30, 35, 40, 45 to 50. Based on this, simulation experiments were carried out on the three algorithms for different numbers of target areas. Figure 4 The experimental results of multi-area full coverage path planning of three algorithms are shown when the number of target areas is 50. Figure 4 (a) is the full coverage path planned by SPSO, Figure 4 (b) is the full coverage path planned by APPA, Figure 4 (c) Full coverage path planned by Fast NN.

[0118] At the same time, the corresponding data of the experimental results of the current scenario are shown in Table 2.

[0119] Table 2. Regional full coverage path planning data of three algorithms

[0120]

[0121] In the second set of simulation experiments, in order to verify the effectiveness of the greedy strategy introduced in the SPSO algorithm proposed in this paper, we still consider conducting multi-area path planning experiments for drone swarms in multiple identical scenarios and compare them based on the proposed evaluation indicators.

[0122] This group of experiments considers a scenario with a range of 1.5km×1.5km, the number of random target areas increases from 10 to 50, and the maximum path distance constraint of the drone is set to 15 kilometers. The results are shown in Table 3 and Figure 5 ,The experimental simulation results are the same in the number of drones dispatched by the drone swarm. ,Here we compare the total path distance and balance rate of the drone swarm.

[0123] Table 3. Regional coverage path planning data of SPSO and PSO

[0124]

[0125] See also Figure 6 , Figure 6 It is a schematic diagram of the working of the hardware device of an embodiment of the present invention, and the hardware device specifically includes: a multi-area coverage path planning device 401 for a drone swarm, a processor 402 and a storage medium 403.

[0126] A drone swarm multi-area coverage path planning device 401: The drone swarm multi-area coverage path planning device 401 implements the drone swarm multi-area coverage path planning method.

[0127] Processor 402: The processor 402 loads and executes the instructions and data in the storage medium 403 to implement the multi-area coverage path planning method of the drone swarm.

[0128] Storage medium 403: The storage medium 403 stores instructions and data; the storage medium 403 is used to implement the multi-area coverage path planning method of the drone swarm.

[0129] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0130] The serial numbers of the embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments. In a unit claim that lists several means, several of these means may be embodied by the same hardware item. The use of the words first, second, and third, etc. does not indicate any order and these words may be interpreted as identifiers.

[0131] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for planning a multi-area coverage path for a drone swarm, characterized in that: include: S1: Construct a problem model based on the multi-area coverage task of drone swarms. The problem model includes: environmental model, multi-objective cost function and constraints; S2: Perform full coverage path planning based on the multi-objective cost function of the optimization problem model and calculate the optimal path for the drone swarm.

2. The method for planning a multi-area coverage path for a drone swarm according to claim 1, characterized in that: Step S1 is specifically as follows: S11: Based on the multi-area coverage mission of the drone swarm, multiple drones are used to scan multiple mission areas and build an environmental model; S12: construct a multi-objective cost function based on the number of drones, total path distance and drone cooperation efficiency, and set constraints for the multi-objective cost function; S13: Using environmental model, multi-objective cost function and constraints as the problem model of multi-area coverage task of UAV swarm.

3. The method for planning a multi-area coverage path for a drone swarm according to claim 2, characterized in that: The expression of the multi-objective cost function F is: F=min{aF1+bF2+cF3} In the formula, a, b and c are weight coefficients, F1, F2 and F3 are the number of drones, total path distance and drone cooperation efficiency respectively. The formula is as follows: F1=m In the formula, m is the number of all drones used when performing the mission; Where D i is the path distance of the i-th UAV, and the formula is as follows: Where j and k are regions, j,k∈{0,1,…,n}, region 0 is the drone base; x i,j,k is a decision variable. If the i-th drone moves from area j to k, then x i,j,k =1, otherwise x i,j,k =0; H i represents the mobile height assigned to the i-th UAV; α is the contribution coefficient, which adjusts the contribution of the mobile height to the total path distance; h represents the mission height of all UAVs; C j,k represents the path distance of the drone from area j to area k at the moving height; r k represents the path distance of the UAV at the mission altitude within region k; In the formula, is the mean path distance of the i-th UAV.

4. The method for planning a multi-area coverage path for a drone swarm according to claim 1, characterized in that: Step S2 is specifically as follows: S21: Set the parameters of the particle population, including: the number of particles, the maximum number of iterations Q, the inertia weight coefficient and the weight coefficient, construct the particle position set and the particle velocity set through the index number of the drone base and each area, initialize the particle population, the particle position set and the particle velocity set, and set q = 0; S22: let q=q+1; S23: Extract the u-th particle from the particle population and update the particle velocity V of the u-th particle u ; S24: Obtain the unvisited region set of the u-th particle. If the unvisited region set is not an empty set, proceed to step S25; otherwise, proceed to step S26; S25: Update the particle position X of the uth particle u , the particle position X is swept by the sweep function u Translated into full coverage waypoints for each drone; S26: Update the fitness, local optimal value and global optimal value of all particles with the goal of optimizing the multi-objective cost function; S27: If q=Q, the full coverage path point of each drone corresponding to the global optimal particle at this time is used as the optimal path of the drone group, otherwise return to step S22.

5. The method for planning a multi-area coverage path for a drone swarm according to claim 4, characterized in that: The expression for the particle position is: Among them, the particle position X u A particle with n+1 dimensions The access group element set, the particle of the dth dimension The access group elements obtained by the dth region and the previous access region a and the next access region b respectively<a,d> and<d,b> The access group elements are composed of two groups of area index numbers.

6. The method for planning a multi-area coverage path for a drone swarm according to claim 4, characterized in that: The expression for particle velocity is: Among them, V u represents the particle velocity with n+1 dimensions; p(a,d)∈[0,1] is a probability function, which represents the probability of each pair of access group elements being selected during the particle velocity update process.

7. A storage medium, characterized in that: The storage medium stores instructions and data for implementing the multi-area coverage path planning method for a drone swarm as described in any one of claims 1 to 6.

8. A multi-area coverage path planning device for a drone swarm, characterized by: include: Processor and storage medium; the processor loads and executes instructions and data in the storage medium to implement the multi-area coverage path planning method for a drone swarm as described in any one of claims 1 to 6.

Images