A farm heterogeneous unmanned aerial vehicle cooperative planning method and system based on double-layer auction

Abstract

The application provides a farm heterogeneous unmanned aerial vehicle cooperative planning method based on a double-layer auction, which comprises the following steps: 1. Based on an improved k-means clustering algorithm, heterogeneous unmanned aerial vehicles are divided into different types; 2. Based on the improved k-means clustering algorithm, adaptive task clusters are generated for different types of heterogeneous unmanned aerial vehicles, and the task cluster size is dynamically adjusted to match the types of heterogeneous unmanned aerial vehicles; 3. Through a two-stage mechanism of same-type bidding and cross-type compensation bidding, tasks are allocated to each unmanned aerial vehicle, wherein the same-type bidding only allows the heterogeneous unmanned aerial vehicles matching the task cluster type to bid, and the cross-type compensation bidding allows other types of heterogeneous unmanned aerial vehicles to participate in the bidding of the remaining tasks; 4. Based on a simulated annealing algorithm, the paths of the unmanned aerial vehicles in the allocation of tasks are optimized, and a globally optimal path is generated. The application effectively improves the utilization rate of unmanned aerial vehicle resources and the task completion rate, and provides a more intelligent and efficient solution for intelligent agricultural data monitoring.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of wireless communication, and particularly relates to a farm heterogeneous unmanned aerial vehicle cooperative planning method and system based on a double-layer auction. BACKGROUND

[0002] With the accelerated development of agricultural modernization, large-scale intelligent agricultural scenes are increasingly popular. Such intelligent farms not only have a wide area, but also have fine internal functional area division. Different regions plant different types of crops, and the planting area has significant differences. The soil characteristics are also different due to geographical location, historical cultivation and other factors. For example, the soil in some areas is acidic, which is suitable for the growth of economic crops such as blueberries; the soil in some areas is fertile, which is more suitable for the planting of food crops.

[0003] In this complex and variable agricultural environment, data monitoring is a key link to achieve precision agriculture. Unmanned aerial vehicles, with their efficiency, flexibility and multi-sensor fusion capabilities, have become an important tool for agricultural data monitoring. Compared with traditional manual monitoring methods, unmanned aerial vehicles can efficiently cover the entire farm in a short time, quickly complete preliminary data collection, and significantly reduce time costs. At the same time, unmanned aerial vehicles can flexibly adjust flight paths and monitoring strategies according to topographic features and crop distribution, achieving accurate data collection. In addition, unmanned aerial vehicles equipped with high-resolution cameras, multispectral sensors and other advanced equipment can obtain multi-dimensional data such as crop growth conditions, soil fertility levels, and pest occurrence, providing accurate and forward-looking decision support for agricultural production.

[0004] However, when assigning tasks to heterogeneous unmanned aerial vehicles, traditional auction algorithms often have problems such as uneven resource utilization, insufficient task allocation, and some tasks left over, making it difficult to meet the needs of intelligent agriculture for efficient, comprehensive and secure data monitoring. SUMMARY

[0005] The application provides a farm heterogeneous unmanned aerial vehicle cooperative planning method and system based on a double-layer auction. First, an improved k-means clustering algorithm is used to divide the heterogeneous unmanned aerial vehicles into three types, and task clusters are generated for different types. Then, a double-layer auction mechanism is introduced for task allocation to ensure that tasks are reasonably matched with unmanned aerial vehicles of different types. Finally, a simulated annealing algorithm is used to optimize the flight path, realizing local obstacle avoidance and global path cooperation, thereby improving the resource utilization rate, task completion rate and monitoring safety of unmanned aerial vehicles.

[0006] To solve the above technical problems, the application provides the following technical solutions:

[0007] A farm heterogeneous unmanned aerial vehicle cooperative planning method based on a double-layer auction, comprising the following steps:

[0008] Step 1. Based on the improved k-means clustering algorithm, heterogeneous UAVs are divided into different types;

[0009] Step 2 Based on the improved k-means clustering algorithm, adaptive task clusters are generated for different types of heterogeneous UAVs, and the task cluster size is dynamically adjusted to match the type of heterogeneous UAVs;

[0010] Step 3. Assign tasks to each heterogeneous UAV through a two-stage mechanism of same-type bidding and cross-type compensation bidding, where same-type bidding only allows heterogeneous UAVs matching the task cluster type to bid, and cross-type compensation bidding allows other types of heterogeneous UAVs to participate in the bidding for the remaining tasks;

[0011] Step 4. Based on the simulated annealing algorithm, the UAV path in the task allocation of each heterogeneous UAV is optimized to generate the globally optimal path.

[0012] Further, the type of heterogeneous UAV in step 1 includes: high-range-low-capacity type, balanced type, and low-range-high-capacity type.

[0013] Further, step 1 includes:

[0014] Based on the normalized range and capacity feature vector, the initial center of different types of heterogeneous UAVs is selected;

[0015] Calculate the range-capacity weighted distance of each heterogeneous UAV to the initial center of different types, and assign each heterogeneous UAV to the center cluster with the smallest range-capacity weighted distance;

[0016] After each iteration, check the center cluster attributes and force to meet the type constraints by removing or adjusting members until the clustering converges.

[0017] Further, the calculation formula of the range-capacity weighted distance is:

[0018]

[0019] wherein, represents the range-capacity weighted distance of UAV individual i to the initial center of the kth type of heterogeneous UAV, represents the UAV individual, represents the three types of heterogeneous UAVs, represents the range weight of the kth type of heterogeneous UAV, represents the capacity weight of the kth type of heterogeneous UAV, and represents the range mean and capacity mean of the kth type of heterogeneous UAV, and represents the normalized range and normalized capacity, is the range of UAV individual i, data capacity of the individual i drone; maximum range of all heterogeneous drones, maximum data capacity of all heterogeneous drones.

[0020] Further, the step 2 comprises:

[0021] According to the heterogeneous drone type, set the weighted features of the task cluster, and direct selection of the task initial center as the task cluster center;

[0022] Calculate the distance-data volume weighted distance of the task point to the task cluster center, and assign each task point to the task cluster with the minimum distance-data volume weighted distance;

[0023] Check the task cluster attributes after each iteration, and force to meet the type constraints by removing or absorbing the task points until the clustering converges.

[0024] Further, the calculation formula of the distance-data volume weighted distance is:

[0025]

[0026] wherein, Dij represents the distance-data volume weighted distance of the task point j to the task cluster center of the i-th type, Dij represents the distance-data volume weighted distance of the task point j to the task cluster center of the i-th type, is the distance feature weight, is the data volume feature weight, is the original distance of the task point j to the drone base, is the data volume size of the task point j, represents the original distance mean of the task cluster center of the i-th type, represents the original distance mean of the task cluster center of the i-th type, represents the data volume mean of the task cluster center of the i-th type. represents the data volume mean of the task cluster center of the i-th type.

[0027] Further, the step 3 comprises:

[0028] In the same type bidding stage, calculate the bid price of the heterogeneous drone matching the task cluster type, and select the bidder with the lowest bid price among the bid prices of the heterogeneous drone matching the task cluster type as the winner;

[0029] In the cross-type compensation auction stage, the remaining tasks are bid by other types of heterogeneous drones, the bid price of other types of heterogeneous drones is calculated, and the bidder with the lowest bid price among the bid prices of other types of heterogeneous drones is selected as the winner.

[0030] Further, the calculation formula of the bid price of the heterogeneous drone matching the task cluster type is:

[0031]

[0032] wherein, denotes the bidding price of the heterogeneous UAV matching the task cluster type, is the cost of the heterogeneous UAV matching the task cluster type to perform the task , is a random perturbation term, and denote the cost weight of the flight distance and the capacity, respectively, is a random perturbation threshold value;

[0033] The bidding price calculation formula of the other type of heterogeneous UAV is:

[0034]

[0035] wherein, denotes the bidding price of the other type of heterogeneous UAV, is the cost of the other type of heterogeneous UAV to perform the task ,

[0036] The constraint condition is: the maximum flight distance constraint ; the maximum capacity constraint .

[0037] Further, the step 4 comprises:

[0038] The initial path is a closed route, and the candidate path is generated by inserting, deleting or moving the detour point;

[0039] The energy of the generated candidate new path is calculated based on the annealing energy function , which includes a length term, an obstacle avoidance penalty term and a flight distance constraint penalty term;

[0040] The path is updated according to the energy difference and the probability acceptance criterion until the temperature decreases to a threshold value or the maximum number of iterations is reached.

[0041] In another aspect, the present application provides a farm heterogeneous UAV cooperative planning system based on double-layer auction, comprising:

[0042] A UAV pre-classification module is used to divide the heterogeneous UAVs into different types based on an improved k-means clustering algorithm;

[0043] An adaptive task cluster generation module is used to generate adaptive task clusters for different types of heterogeneous UAVs based on an improved k-means clustering algorithm, and dynamically adjust the task cluster size to match the heterogeneous UAV types;

[0044] a double-layer auction task allocation module for allocating tasks for each unmanned aerial vehicle through a two-stage mechanism of same-type bidding and cross-type compensation bidding, wherein the same-type bidding only allows heterogeneous unmanned aerial vehicles matching the task cluster type to bid, and the cross-type compensation bidding allows other types of heterogeneous unmanned aerial vehicles to participate in bidding for remaining tasks;

[0045] a simulated annealing path optimization module for optimizing the path of each unmanned aerial vehicle in the allocated task based on a simulated annealing algorithm to generate a globally optimal path.

[0046] Compared with the prior art, the present application has the following beneficial effects:

[0047] 1. Compared with the traditional k-means clustering algorithm which randomly selects an initial center and has no constraint processing, the present application directionally selects a type initial center while using a hard constraint to forcibly adjust after each iteration, and accurately represents the core features of different types.

[0048] 2. Compared with the traditional k-means clustering which cannot distinguish the importance of features when using Euclidean distance, the present application defines different weights for each type of unmanned aerial vehicle and task cluster, calculates a weighted Euclidean distance, amplifies the influence of type core features, and guides the clustering result to accelerate towards the optimal solution.

[0049] 3. Compared with the traditional auction algorithm which is prone to local optimum and causes problems of waste of part of unmanned aerial vehicle resources and omission of tasks, the present application uses a clustering algorithm and a deep cooperation mechanism of double-layer auction, forcibly allocates tasks to suitable type unmanned aerial vehicles, and allows other types of unmanned aerial vehicles to take over unallocated tasks when meeting the capability constraint.

[0050] 4. The simulated annealing algorithm is used to plan the path, and the path length index, obstacle avoidance penalty and voyage constraint penalty are introduced to optimize the path length while balancing the obstacle avoidance demand and global path consistency. BRIEF DESCRIPTION OF DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the present application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0052] Figure 1 The flowchart of the embodiment of the present application;

[0053] Figure 2 The unmanned aerial vehicle pre-classification process schematic diagram in the embodiment of the present application;

[0054] Figure 3A flowchart for generating an adaptation task cluster in an embodiment of the present application is shown.

[0055] Figure 4 A flowchart for a double-layer auction mechanism in an embodiment of the present application is shown.

[0056] Figure 5 A flowchart for simulated annealing optimization path in an embodiment of the present application is shown.

[0057] Figure 6 A schematic diagram of the system architecture in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0058] To make the objectives, technical solutions, and advantages of the present application clearer, the technical solutions in the present application will be described below in conjunction with the accompanying drawings in the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0059] Embodiment 1

[0060] As shown in the figure, the present embodiment provides a heterogeneous farm unmanned aerial vehicle cooperative planning method based on a double-layer auction, assuming that heterogeneous unmanned aerial vehicles are dispatched to different areas in a farm to complete data monitoring tasks, the tasks are divided into three stages, first, the heterogeneous unmanned aerial vehicles need to be pre-classified, then each unmanned aerial vehicle needs to be assigned a task, and finally the path of each unmanned aerial vehicle needs to be planned. In this process, the unmanned aerial vehicle needs to optimize the task allocation mechanism to improve resource utilization and task completion rate. Figure 1 Specifically, the following steps are included:

[0061] Step 1. Based on the improved k-means clustering algorithm, the heterogeneous unmanned aerial vehicles are divided into different types.

[0062] Step 1. Based on the improved k-means clustering algorithm, the heterogeneous unmanned aerial vehicles are divided into different types.

[0063] In the present embodiment, the heterogeneous unmanned aerial vehicles are divided into Type1 (high range-low capacity), Type2 (balanced), and Type3 (low range-high capacity). This process mainly uses the improved k-means clustering algorithm for pre-classification. The specific implementation is as follows:

[0064] First, determine the task point set , where each task point attribute is represented as , including the coordinates of the task point and the data volume of the task point .

[0065] The unmanned aerial vehicle set ​, the initial coordinates of the UAV , the maximum flight range of the UAV , and the maximum data capacity of the UAV . Define the type feature vector , which represents the normalized range and normalized capacity, ensuring that the range and capacity are comparable in numerical range and avoiding the dominance of a certain dimension in the clustering result. Next, an improved k-means clustering algorithm is used to divide the heterogeneous UAVs into types, which is implemented as follows:

[0066] (1) First, select the initial center for the three types of UAVs:

[0067] where is the normalized range mean, is the normalized capacity mean. The above formula indicates that for Type1 ( ), select the UAV with the maximum (range-to-capacity ratio) and the minimum as the initial center ; for Type2 ( ), select the UAV with the closest range and capacity to the mean as the initial center ; for Type3 ( ), select the UAV with the maximum and the minimum as the initial center ;

[0068] (2) Calculate the range-capacity weighted distance of each UAV to different type centers:

[0069]

[0070] In the above formula: denotes the range weight and capacity weight of the three types of UAVs, respectively, and denote the range and capacity mean of the kth type of UAV.

[0071] (3) According to the range-capacity weighted distance, assign each UAV to the cluster with the minimum weighted distance:

[0072] , denotes the type label of the UAV assigned to the cluster with the minimum distance, and the mean of the UAV features in the cluster is dynamically updated as the new center:

[0073] where denotes the kth type of UAV Classes of drones.

[0074] (4) Next, ensure that the characteristics of each class meet the predefined type, requiring checking drone type constraints: (in the following formula , , , Threshold value

[0075]

[0076] If the average capacity of Type 1 exceeds the threshold, move the drone with the largest capacity in the cluster to Type 3; if the average range of Type 3 exceeds the threshold, move the drone with the longest range in the cluster to Type 1; if Type 2 is unbalanced, remove the drone that deviates the most from the average and reassign it to another cluster. After determining convergence or reaching the maximum number of iterations, terminate the iteration, i.e., complete the pre-classification of drones.

[0077] As shown in Figure 2 , the specific steps are as follows:

[0078] Step 101: Start the process, input heterogeneous drone data .

[0079] Step 102: Calculate the feature vector of the heterogeneous drone.

[0080] Step 103: Select the initial center of three types of drones.

[0081] Step 104: Calculate the dynamic weighted distance of each drone to the initial center.

[0082] Step 105: Assign the drone to the type with the closest dynamic weighted distance.

[0083] Step 106: Update the cluster center.

[0084] Step 107: Check the type constraints of the drones in this class; if the type constraints are met, go to step 109.

[0085] Step 108: Correct the members of this class of drones.

[0086] Step 109: Determine whether to converge or reach the maximum number of iterations.

[0087] Step 110: Output the classification results of the heterogeneous drones.

[0088] Step 111: End the process.

[0089] Adaptive task cluster generation: Generate three adaptive task clusters for the three types of drones that have been classified. This process mainly uses the improved k-means clustering algorithm to generate the three task clusters.

[0090] Step 2 generates adaptive task clusters for different types of heterogeneous UAVs based on the improved k-means clustering algorithm, and dynamically adjusts the task cluster size to match the heterogeneous UAV types;

[0091] After pre-classifying the drones into three types, it is necessary to generate an adaptive task cluster for each type of drone. First, for each task point , according to its coordinates and the initial coordinates of the drone , calculate the distance between the mission point and the UAV base , and according to the type of drone Calculate weighted features: ,in is the distance feature weight, is the feature weight of data volume. Next, we use the improved k-means clustering algorithm to generate three adaptive task clusters. The algorithm is implemented as follows:

[0092] (1) First, select the initial mission center for different types of UAVs :

[0093]

[0094] For Type 1 UAVs, the features of long-distance mission points are directly amplified by high distance weights to ensure that the first mission center represents long-distance characteristics; for Type 2 UAVs, the mission point with the highest comprehensive score of distance and data volume is selected; for Type 3 UAVs, the features of high-data volume mission points are directly amplified by high data volume weights to ensure that the first mission center represents high-data volume characteristics.

[0095] (2) Calculation task points To cluster center The weighted original distance - data volume distance, Respectively represent the mean of the k-th task cluster center in the original distance and data volume dimensions:

[0096]

[0097] (3) According to the original distance minus the weighted distance of the data volume, each task point is assigned to the cluster with the smallest weighted distance:

[0098] , Indicates that the task Assign the type label to the task cluster with the smallest distance, and dynamically update the center of each cluster to the feature mean of the task points in the cluster: wherein represents the type task cluster.

[0099] (4) After each iteration, the task cluster cluster properties need to be checked to meet the UAV type definition, where the type constraint is: , , , is a threshold value

[0100]

[0101] The expression low-k task cluster is used for (1) Task cluster, if the distance is not up to standard, remove the nearest 10% of task points in the cluster, if the data volume is over-standard, remove the 10% of task points with the largest data volume; for (2) The task cluster needs to remove the task point that deviates the most from the balance; for (3) If the data volume of the task cluster does not meet the standard, absorb the high-data-volume task points from other clusters. After judging the convergence of clustering or reaching the maximum number of iterations, terminate the iteration, that is, complete the generation of task clusters adapted to different types of UAVs.

[0102] Further introduce step 2, as shown in Figure 3 , the specific steps are:

[0103] Step 201: Start the process, input task point data .

[0104] Step 202: Calculate the weighted features of the task point data according to the three types of UAVs .

[0105] Step 203: According to the corresponding type of UAV feature, select three task cluster initial centers.

[0106] Step 204: Calculate the weighted distance of each task point to the task cluster center.

[0107] Step 205: Assign the task point to the task cluster center with the smallest weighted distance.

[0108] Step 206: Update the task cluster center.

[0109] Step 207: Check the type constraint of different task clusters. If it meets the type constraint, go to step 209, otherwise execute the next step.

[0110] Step 208: Correct the cluster members.

[0111] Step 209: Determine whether to converge or reach the maximum number of iterations.

[0112] Step 210: End the process.

[0113] Step 3. Assign tasks to each UAV through a two-stage mechanism of same-type auction and cross-type compensation auction, where the same-type auction only allows heterogeneous UAVs matching the task cluster type to bid, and the cross-type compensation auction allows other types of heterogeneous UAVs to participate in the bidding of the remaining tasks;

[0114] The double-layer auction algorithm used is as follows:

[0115] A. Same-type auction

[0116] This stage only allows UAVs matching the task cluster type to participate in bidding, balancing data volume and voyage. Each UAV performs the task at a cost calculated by the following formula:

[0117]

[0118] and represent the cost weights of voyage and capacity, satisfying ; represents the maximum voyage of UAV , represents the maximum capacity of UAV ; the bid price submitted by the UAV is the cost plus a random disturbance term : . The constraint conditions are: maximum voyage constraint ; maximum capacity constraint .

[0119] For each task , the UAV with the lowest bid price is selected as the winner: , represents the winner of task , and the winner pays its bid price .

[0120] B. Cross-type compensation auction

[0121] This layer of auction aims to handle tasks that are not assigned by same-type UAVs, allowing other types of UAVs to participate in bidding to improve task completion rate. First, for the remaining tasks , other types of UAVs ( ) submit bid prices: The constraint conditions are: maximum voyage constraint ; maximum capacity constraint​ The drone with the lowest bid price is selected as the winner .

[0122] Further introduce step 3, as shown in Figure 4 , the specific steps are:

[0123] Step 301: Start the process.

[0124] Step 302: Submit bids for drones of the same type.

[0125] Step 303: Determine whether the range and capacity constraints are met. If not, return to step 302; if yes, proceed to the next step.

[0126] Step 304: Select the lowest bidder as the winner.

[0127] Step 305: Determine whether there are remaining drones of this type. If yes, return to step 302; if no, proceed to the next step.

[0128] Step 306: Determine whether all tasks have been assigned. If yes, go to step 311.

[0129] Step 307: Bid for drones of other types.

[0130] Step 308: Determine whether the range and capacity constraints are met.

[0131] Step 309: Select the lowest bidder as the winner.

[0132] Step 310: Determine whether there are remaining drones.

[0133] Step 311: If there are no remaining drones, the process ends.

[0134] Simulated annealing path planning and obstacle avoidance: After each drone is assigned a task, it needs to plan a path and avoid obstacles. This process mainly needs to optimize the global path length and coordinate local obstacle avoidance.

[0135] Step 4. Optimize the path of each drone assigned to the task based on the simulated annealing algorithm to generate the globally optimal path.

[0136] After assigning tasks to each drone, use the simulated annealing algorithm to optimize the path and achieve obstacle avoidance. The simulated annealing algorithm used is as follows:

[0137] (1) The initial path is a closed route , where B represents the coordinates of the drone base, represents the coordinates of the task points that the drone needs to execute.

[0138] (2) In the iterative optimization process, the candidate path is generated by the domain operation of inserting / deleting / moving the detour point, and the annealing energy function is used to calculate With , the annealing energy of the generated candidate path, the annealing energy of the initial path, and the difference is The annealing energy function includes the path length term, the obstacle avoidance penalty term and the voyage constraint penalty term:

[0139]

[0140] In the above formula: denotes the Euclidean distance of the path segment , is the obstacle avoidance penalty weight, which forces the path to avoid obstacles, denotes each edge of the obstacle polygon, is the voyage constraint penalty term weight denotes the obstacle indicator function, which calculates the intersection of the path segment and each edge of the obstacle polygon. If there is an intersection, it is marked as a conflict between the path and the obstacle, and the indicator function value is 1, otherwise it is 0. denotes the maximum voyage of the UAV.

[0141] Update the path according to the following rules, and then cool down .

[0142] (3) When the temperature or reaches the maximum number of iterations, output the optimal path as is the UAV base, the detour point inserted in the optimal path, is the task point.

[0143] Further introduce step 4, as shown in Figure 5 , the specific steps are:

[0144] Step 401: Start the flow.

[0145] Step 402: Plan the initial path as a straight line path.

[0146] Step 403: Set the annealing initial temperature.

[0147] Step 404: Apply the domain operation to generate a new path.

[0148] Step 405: Calculate the energy difference between the new path and the original path.

[0149] Step 406: Determine whether the energy difference is less than 0; if the energy difference is less than 0, go to step 409.

[0150] Step 407: If the energy difference is greater than 0, accept the new path with a probability. If the new path is accepted, go to step 409.

[0151] Step 408: If the new path is not accepted, keep the original path, and go to step 410.

[0152] Step 409: Accept the new path.

[0153] Step 410: Update the current best path.

[0154] Step 411: Cool down.

[0155] Step 412: Determine whether the temperature meets the termination condition (reduces to the lowest temperature or reaches the maximum number of iterations). If not, return to step 404.

[0156] Step 413: Output the optimal path.

[0157] Step 414: End the process.

[0158] Then output the final allocation and path.

[0159] Embodiment 2

[0160] As Figure 6 described, the embodiment provides a heterogeneous unmanned aerial vehicle cooperative planning system based on a double-layer auction farm, which comprises:

[0161] An unmanned aerial vehicle pre-classification module for dividing heterogeneous unmanned aerial vehicles into different types based on an improved k-means clustering algorithm;

[0162] An adaptive task cluster generation module for generating adaptive task clusters for different types of heterogeneous unmanned aerial vehicles based on the improved k-means clustering algorithm, and dynamically adjusting the size of the task cluster to match the type of the heterogeneous unmanned aerial vehicle;

[0163] A double-layer auction task allocation module for allocating tasks to each unmanned aerial vehicle through a two-stage mechanism of same-type bidding and cross-type compensation bidding, wherein the same-type bidding only allows heterogeneous unmanned aerial vehicles matching the type of the task cluster to bid, and the cross-type compensation bidding allows other types of heterogeneous unmanned aerial vehicles to participate in the bidding for the remaining tasks;

[0164] A simulated annealing path optimization module for optimizing the path of each unmanned aerial vehicle in the allocated task based on the simulated annealing algorithm, and generating a globally optimal path.

[0165] It should be understood that parts not elaborated in the specification are all prior art.

[0166] It should be understood that the above description is merely a detailed example of the preferred embodiment and is not to be taken in a limiting sense. There can be many variations to the embodiments described herein without departing from the spirit of the application. The scope of the application should be determined by a fair reading of the appended claims, along with the full text of the specification.

Claims

1. A collaborative planning method for heterogeneous drones in farms based on double-layer auction, characterized by: The following steps are involved: Step 1. Based on the improved k-means clustering algorithm, the heterogeneous UAVs are divided into different types. The initial centers of different types of heterogeneous UAVs are selected according to the normalized range and capacity feature vectors. Calculate the range-capacity weighted distance of each heterogeneous UAV to different types of initial centers, and assign each heterogeneous UAV to the center cluster with the smallest range-capacity weighted distance; After each iteration, the central cluster attributes are checked and the drone type constraints are enforced by removing or adjusting members until the clustering converges; Step 2: Based on the improved k-means clustering algorithm, adaptive task clusters are generated for different types of heterogeneous UAVs, and the task cluster size is dynamically adjusted to match the heterogeneous UAV types. The weighted features of the task cluster are set according to the heterogeneous UAV types, and the initial center of the task is selected as the task cluster center. Calculate the distance between the task point and the task cluster center - the data volume weighted distance, and assign each task point to the task cluster with the smallest distance - data volume weighted distance; After each iteration, the task cluster attributes are checked and the task cluster type constraints are forced to be satisfied by removing or absorbing task points until the clustering converges; Step 3. Assign tasks to each heterogeneous UAV through a two-stage mechanism of same-type bidding and cross-type compensatory auction. The same-type auction only allows heterogeneous UAVs matching the task cluster type to bid, while the cross-type compensatory auction allows heterogeneous UAVs of other types to bid for the remaining tasks. This includes: In the same-type bidding stage, the bid prices of heterogeneous drones matching the task cluster type are calculated, and the bidder with the lowest bid price among the heterogeneous drones matching the task cluster type is selected as the winner; In the cross-type compensation auction stage, the remaining tasks are bid by other types of heterogeneous drones. The bid prices of other types of heterogeneous drones are calculated, and the bidder with the lowest bid price among the bid prices of other types of heterogeneous drones is selected as the winner. Step 4. Optimize the UAV paths in each heterogeneous UAV assignment task based on the simulated annealing algorithm to generate the global optimal path.

2. The method for collaborative planning of heterogeneous drones in farms based on double-layer auction according to claim 1 is characterized in that: The heterogeneous UAV types in step 1 include: high-range-low-capacity type, balanced type, and low-range-high-capacity type.

3. The method for collaborative planning of heterogeneous drones in farms based on double-layer auction according to claim 1 is characterized in that: The calculation formula of the range-capacity weighted distance is: in, represents the range-capacity weighted distance of UAV individual i to the initial center of the kth heterogeneous UAV, Represents individual drones, Represents three types of heterogeneous drones, represents the range weight of the kth heterogeneous UAV, represents the capacity weight of the kth type of heterogeneous UAVs, and represents the mean range and capacity of the kth type of heterogeneous UAVs, and represents the normalized range and normalized capacity, is the range of drone individual i, is the data capacity of drone individual i; is the maximum range among all heterogeneous UAVs, It is the maximum data capacity among all heterogeneous drones.

4. The method for collaborative planning of heterogeneous drones in farms based on double-layer auction according to claim 1 is characterized in that: The calculation formula of the distance-data volume weighted distance is: in, Indicates that task point j is The distance between the center of the task cluster and the weighted distance of the data volume, is the distance feature weight, is the data feature weight, is the original distance between mission point j and the UAV base, is the data size of task point j, Indicates the The original mean distance between the centers of the task clusters, Indicates the The mean amount of data at the center of the task cluster.

5. The method for collaborative planning of heterogeneous drones in farms based on double-layer auction according to claim 1 is characterized in that: The calculation formula for the bidding price of heterogeneous UAVs matching mission cluster types is: in, represents the bidding price of heterogeneous UAVs matching the task cluster type, Heterogeneous UAVs matching the mission cluster type Execute the task the cost, is a random disturbance term, and denote the cost weights of range and capacity respectively, is the random disturbance threshold; The formula for calculating the bidding price of other types of heterogeneous drones is: in, represents the bidding price of other types of heterogeneous drones, For other types of heterogeneous drones Execute the task the cost, The constraints are: Maximum range constraint ; Maximum capacity constraint .

6. The method for collaborative planning of heterogeneous drones in farms based on double-layer auction according to claim 2 is characterized in that: The step 4 comprises: The initial path is a closed route, and candidate paths are generated by inserting, deleting or moving detour points; The energy of candidate new paths generated based on annealing energy function , the annealing energy function includes a length term, an obstacle avoidance penalty term, and a range constraint penalty term; The path is updated according to the energy difference and probability acceptance criteria until the temperature drops to a threshold or the maximum number of iterations is reached.

7. A farm heterogeneous drone collaborative planning system based on double-layer auction, characterized by: include: UAV pre-classification module, which is used to classify heterogeneous UAVs into different types based on the improved k-means clustering algorithm; Adaptive task cluster generation module, which is used to generate adaptive task clusters for different types of heterogeneous UAVs based on the improved k-means clustering algorithm, and dynamically adjust the task cluster size to match the heterogeneous UAV types; A two-tiered auction task allocation module, which is used to assign tasks to each drone through a two-stage mechanism of same-type bidding and cross-type compensatory auction. The same-type auction only allows heterogeneous drones matching the task cluster type to bid, while the cross-type compensatory auction allows other heterogeneous drones to participate in the bidding for the remaining tasks. The simulated annealing path optimization module is used to optimize the UAV paths in each UAV assignment task based on the simulated annealing algorithm to generate the global optimal path; The farm heterogeneous drone collaborative planning system based on double-layer auction is used to execute the steps in the farm heterogeneous drone collaborative planning method based on double-layer auction as described in any one of claims 1-6.

Images