UAV aggregation and mission planning method based on adaptive clustering contract network

Abstract

The present application relates to a method for drone aggregation and task planning based on an adaptive clustering contract network. The method comprises: clustering a drone cluster according to the location information of drone units in a drone cluster to obtain multiple drone subclusters; constructing a cluster task planning contract network between drone subclusters and a unit task planning contract network of drone units within a drone subcluster according to the task attributes and unit attributes of drone units in the drone cluster; allocating tasks to be executed to each drone subcluster according to the cluster task planning contract, and within each drone subcluster, allocating tasks of the drone subcluster to each drone unit according to the unit task planning contract network. The use of this method can improve the efficiency of multi-task allocation.

Description

Technical Field

[0001] The present application relates to the technical field of unmanned aerial vehicle mission planning, and in particular to a method for unmanned aerial vehicle aggregation and mission planning based on an adaptive clustering contract network. Background Art

[0002] UAV mission planning has developed from simple single-task allocation to multi-task allocation, involving a variety of complex task models, including network flow model, traveling salesman model, collaborative multi-task allocation model, etc., and there is also a certain research foundation on the methods of solving these models, such as evolutionary algorithms and market-based methods.

[0003] In terms of contract network protocols, the classic approach is for drone units to bid directly to obtain tasks. In terms of the deployment mode of task allocation, it is divided into centralized and distributed. Centralized task allocation relies on the control center, which makes overall task planning based on drone units and task information. Distributed task allocation relies on each drone unit to make autonomous decisions based on the task environment and its own status. Although centralized task allocation can achieve better results by using complete information when processing tasks with larger granularity, it has poor adaptability to complex and changing environments and relies on the processing power of the control center and the timeliness of information transmission. Although distributed task allocation gives drone units autonomous decision-making power and is suitable for small-granularity and non-communication environments, it may not achieve the optimal task allocation effect when processing large-granularity tasks due to the lack of global information. Since most drone units bid directly in the classic contract network protocol, the advantages of the cluster are not fully utilized, which may lead to unreasonable and inefficient task allocation and the inability to fully tap the potential of drone cluster collaboration. Summary of the invention

[0004] Based on this, it is necessary to provide a drone aggregation and task planning method based on adaptive clustering contract network to address the above technical problems.

[0005] A method for unmanned aerial vehicle aggregation and task planning based on an adaptive clustering contract network, the method comprising:

[0006] Clustering the drone cluster according to the location information of the drone units in the drone cluster to obtain a plurality of drone subclusters;

[0007] According to the task attributes and unit attributes of the drone units in the drone cluster, a cluster task planning contract network between drone sub-clusters and a unit task planning contract network of drone units within a drone sub-cluster are constructed;

[0008] The tasks to be executed are allocated to each drone sub-cluster according to the cluster task planning contract. Within each drone sub-cluster, the tasks of the drone sub-cluster are allocated to each drone unit according to the unit task planning contract network.

[0009] In one embodiment, the method further includes: performing position data analysis on the drone units in the drone cluster. standardization;

[0010] From the set number of clusters Take the values ​​in turn and calculate the mean square error of the K-means model , calculate the first-order relative rate of change and the second-order relative rate of change of the mean square error;

[0011] Select an optimal K value to maximize the second-order relative rate of change, and use the optimal K value as the number of drone subclusters;

[0012] Assign the drone units in the drone cluster to each drone sub-cluster.

[0013] In one embodiment, the method further includes: calculating the first-order relative rate of change of the mean square error as follows:

[0014] ;

[0015] in, is the mean square error The first-order relative rate of change of

[0016] The second-order relative rate of change of the mean square error is calculated as:

[0017] ;

[0018] in, is the mean square error The second-order relative rate of change of .

[0019] In one of the embodiments, the method further includes: calculating the net efficiency value of the drone cluster in executing the task to be planned according to the task attributes and unit attributes of the drone units in the drone cluster;

[0020] The net efficiency value is used as the bidding value to construct a cluster task planning contract network between drone sub-clusters.

[0021] In one embodiment, the unit attributes include: mission execution capability, stability, and value;

[0022] The task attributes include: task execution difficulty value, task consumption and task reward value; the calculation steps of the net efficiency value are:

[0023] Obtain the average mission execution capability, average stability, and average value of drone units in a drone cluster;

[0024] Calculate the task reward value of the drone cluster according to the average task execution capability, the task execution difficulty value and the task reward value;

[0025] Calculate the task cost value of the drone cluster according to the task consumption, the average value, the average stability, the distribution information of the drone units in the drone cluster, and the number of existing tasks in the drone cluster;

[0026] According to the task reward value, the task cost value and the pre-acquired task priority, a net efficiency value of the drone cluster executing the task to be planned is obtained.

[0027] In one of the embodiments, the method further includes: calculating the unit net efficiency value of the drone cluster executing the task to be planned according to the task attribute and the unit attribute of the drone unit;

[0028] Calculate the load factor of the drone unit based on the workload of the drone unit;

[0029] The unit net efficiency value and the load factor are used to calculate the bidding value, and a unit task planning contract network of the drone units in the drone sub-cluster is constructed.

[0030] In one embodiment, the unit attributes include: mission execution capability, stability, and value;

[0031] The task attributes include: task execution difficulty value, task consumption and task reward value;

[0032] The calculation steps of the unit net efficiency value are:

[0033] The benefit value of the drone unit is calculated based on the drone unit's mission execution capability, mission execution difficulty value, and mission reward value;

[0034] Calculate the cost value according to the task consumption, the value, the stability, and the distance between the UAV unit and the task execution point;

[0035] Calculating a unit net efficiency value according to the benefit value, the cost value, and the pre-acquired task priority;

[0036] The steps to calculate the load factor are:

[0037] The workload of the drone unit is calculated based on the time consumption of all tasks in the drone unit's contract list;

[0038] The load factor is calculated based on the workload of the drone unit and the average workload of the drone units in the drone cluster.

[0039] In one of the embodiments, it also includes: when there are unassigned tasks, the unassigned tasks are reallocated to other drone units whose task loads are not full according to the unit task planning contract network.

[0040] In one of the embodiments, it further includes: when all drone units have a task execution contract with a negative net effectiveness value for the task to be executed, exchanging the task execution contract with the task execution contracts of other drones;

[0041] If the sum of the net efficiency values ​​of the two drone units after the exchange is greater than the sum of the net efficiency values ​​of the two drone units before the exchange, the task exchange is performed, otherwise the task exchange is not performed.

[0042] A drone aggregation and mission planning device based on an adaptive clustering contract network, the device comprising:

[0043] A clustering module, used for clustering the drone cluster according to the location information of the drone units in the drone cluster to obtain a plurality of drone sub-clusters;

[0044] A contract network construction module is used to construct a cluster task planning contract network between drone sub-clusters and a unit task planning contract network of drone units within a drone sub-cluster according to the task attributes and unit attributes of drone units in the drone cluster;

[0045] The planning module is used to allocate the tasks to be executed to each drone sub-cluster according to the cluster task planning contract, and within each drone sub-cluster, allocate the tasks of the drone sub-cluster to each drone unit according to the unit task planning contract network.

[0046] The above-mentioned UAV aggregation and task planning method based on adaptive clustering contract network can form multiple UAV sub-clusters by clustering according to the location information of UAV units, which can not only improve the management efficiency of large-scale UAV clusters and make regional task execution easier to coordinate and command, but also optimize resource allocation and enhance resource sharing effects such as local communication. The cluster and unit task planning contract network constructed based on task and unit attributes not only comprehensively considers multiple factors to achieve reasonable task allocation, gives full play to the advantages of each UAV, but also improves the flexibility and adaptability of task allocation. Whether it is a single or mixed task scenario, efficient allocation can be achieved by adjusting the contract network parameter rules, providing strong support for the application of UAV task planning in multiple fields and enhancing the quality and efficiency of overall task execution. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1It is a flowchart of a method for UAV aggregation and task planning based on an adaptive clustering contract network in one embodiment;

[0048] Figure 2 is a schematic diagram of a process of an adaptive clustering algorithm in one embodiment;

[0049] Figure 3 A schematic diagram of a process of planning a cluster task contract network and a unit task contract network in one embodiment;

[0050] Figure 4 A schematic diagram of a flow chart of task reallocation and task exchange steps in one embodiment;

[0051] Figure 5 It is a flowchart of a method for UAV aggregation and task planning based on an adaptive clustering contract network in one embodiment;

[0052] Figure 6 This is a diagram of the distribution position information of each drone unit at the initial moment in an embodiment;

[0053] Figure 7 A clustering effect index diagram for different cluster numbers in one embodiment;

[0054] Figure 8 Schematic diagram of adaptive grouping of drone units in one embodiment;

[0055] Fig. 9 A schematic diagram of task allocation in one embodiment;

[0056] Fig.10 It is a structural block diagram of a drone aggregation and mission planning device based on an adaptive clustering contract network in one embodiment. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0058] In one embodiment, Figure 1 As shown, a method for UAV aggregation and task planning based on an adaptive clustering contract network is provided, comprising the following steps:

[0059] Step 102: cluster the drone cluster according to the location information of the drone units in the drone cluster to obtain a plurality of drone sub-clusters.

[0060] Step 104, based on the task attributes and unit attributes of the drone units in the drone cluster, construct a cluster task planning contract network between drone sub-clusters and a unit task planning contract network of drone units within the drone sub-clusters.

[0061] Step 106, assign the tasks to be executed to each drone sub-cluster according to the cluster task planning contract, and within each drone sub-cluster, assign the tasks of the drone sub-cluster to each drone unit according to the unit task planning contract network.

[0062] In the above-mentioned UAV aggregation and task planning method based on adaptive clustering contract network, multiple UAV sub-clusters are formed by clustering according to the location information of UAV units, which can not only improve the management efficiency of large-scale UAV clusters and make regional task execution easier to coordinate and command, but also optimize resource allocation and enhance resource sharing effects such as local communication. The cluster and unit task planning contract network constructed based on task and unit attributes not only comprehensively considers multiple factors to achieve reasonable task allocation, gives full play to the advantages of each UAV, but also improves the flexibility and adaptability of task allocation. Whether it is a single or mixed task scenario, efficient allocation can be achieved by adjusting the contract network parameter rules, providing strong support for the application of UAV task planning in multiple fields and enhancing the quality and efficiency of overall task execution.

[0063] In one embodiment, the adaptive clustering algorithm flow chart is as follows Figure 2 As shown, the position data of the drone units in the drone cluster are analyzed. Normalization; from the set number of clusters Take the values ​​in turn and calculate the mean square error of the K-means model , calculate the first-order relative change rate and the second-order relative change rate of the mean square error; select the optimal K value to maximize the second-order relative change rate, and use the optimal K value as the number of drone sub-clusters; allocate the drone units in the drone cluster to each drone sub-cluster.

[0064] Specifically, the K-Means clustering algorithm is a commonly used unsupervised learning algorithm used to divide sample data into non-overlapping clusters. In the classic K-Means clustering algorithm, Usually, the decision is made based on the background knowledge of the problem and human experience, which is contrary to the original intention of liberating manpower and realizing process automation. Value, use Value and Determine the best relationship between Value. Based on The elbow shape of the value graph corresponds to The value is considered to be the optimal number of clusters, that is, the number of drone cluster divisions. The calculation formula is:

[0065] ;

[0066] in, Indicates The location of the drone unit, express The center point of the drone cluster to which it belongs. In order to calculate the corresponding value.

[0067] In another embodiment, the first-order relative rate of change of the mean square error is calculated as:

[0068] ;

[0069] in, is the mean square error The first-order relative rate of change of the mean square error is calculated as follows:

[0070] ;

[0071] in, is the mean square error The second-order relative rate of change of .

[0072] In one of the embodiments, the net performance value of the drone cluster executing the task to be planned is calculated based on the task attributes and unit attributes of the drone units in the drone cluster; the net performance value is used as the bid value to construct a cluster task planning contract network between the drone sub-clusters.

[0073] Specifically, in the adaptive allocation of tasks by drone units, a drone cluster can be regarded as a collection of adjacent drone units. After the adaptive drone cluster division of drone units, the task bidding of the drone cluster should be carried out by comprehensively considering the attributes, status and task attributes of all drone units in the drone cluster. The attributes of all drone units in the drone cluster, such as value, task execution capability and stability average, are used as the bidding attributes of the drone cluster. That is, the average task execution capability of all drone units , average stability , average value , the indicator calculation formula can be modified according to the actual task, as well as the weight adjustment of the indicator value;

[0074] ;

[0075] ;

[0076] ;

[0077] in, , , Indicates the mission execution capability, stability, and value of a single drone unit. is the number of drone units in the drone cluster. Consider the task reward value of the task:

[0078] ;

[0079] in, It is an adjustable parameter. is the task reward value, represents the probability of successfully executing the task. Next, consider the total cost of the drone cluster to perform the task :

[0080] ;

[0081] The above formula comprehensively considers the task consumption, the average value, the average stability, the distribution information of the drone units in the drone cluster, and the number of existing tasks in the drone cluster to bid for the task. , , is an adjustable parameter, is the difficulty value of task execution, When regarded as , Indicates the distance between the center of gravity of all drone units in the drone cluster and the mission area. is the number of existing tasks in the drone cluster;

[0082] ;

[0083] In the above formula The priority of the task reflects the importance of the task. The net efficiency of the drone cluster in performing this task is used as the bidding value by the drone cluster. All drone clusters calculate the net efficiency of the task as the bidding value according to the same process and bid for the task.

[0084] In one of the embodiments, the unit net efficiency value of the drone cluster executing the task to be planned is calculated based on the task attributes and unit attributes of the drone unit; the load factor of the drone unit is calculated based on the workload of the drone unit; the unit net efficiency value and the load factor are calculated as bid values ​​to construct a unit task planning contract network for the drone units in the drone sub-cluster.

[0085] Specifically, after the drone cluster bids, there are usually multiple tasks in the drone cluster. At this time, the drone units in the drone cluster need to bid against each other. The bidding of a single drone unit needs to consider the benefit value of executing the task. The calculation method is similar to the calculation of the benefit value of the drone cluster. The benefit value of the drone unit for:

[0086] ;

[0087] in, Indicates the mission execution capability of the drone unit, Indicates the difficulty value of task execution. is an adjustable parameter. At the same time, the overhead cost of the drone unit to perform this task also needs to be considered to measure the cost value of the drone unit to perform this task:

[0088] ;

[0089] in, , It is an adjustable parameter. Represents the distance between the drone unit and the task point position of the bidding task. If the bidding task is not the first task of the drone unit at this time, then calculate the distance between this task and the last task in the drone unit's task list at this time. represents the probability of failure of the drone unit memory task. Similarly, when When ;

[0090] ;

[0091] In the above formula, is the priority of the task. According to the above formula, the net efficiency of the drone unit in performing the task can be obtained. Considering the workload of the drone unit, assuming that the drone unit has n missions, the payload of a single drone unit The time consumption of all tasks in its contract list Sum:

[0092] ;

[0093] The drone unit needs to bid with other drone units in the same drone cluster. According to the load of the drone unit and the load of other drone units in the drone cluster, the workload coefficient of the drone unit can be approximately calculated. for:

[0094] ;

[0095] in, is the average time consumed by all drone units in the drone cluster to perform their tasks. If there is no task in the drone cluster, .if , indicating that the execution time of the task in the drone unit task list is long, and it is not suitable for accepting new tasks, and the bid value will become negative. The final bid value of the drone unit for the task in its drone cluster for:

[0096] ;

[0097] The drone units in the drone cluster calculate the bid value for the tasks in the drone cluster according to the above calculation process. The drone units with the highest bids obtain the corresponding tasks and join their contract network. After all drone units in the drone cluster bid for all tasks in the drone cluster, the process ends.

[0098] In summary, if Figure 3 As shown in the figure, a schematic diagram of the UAV cluster task negotiation and UAV unit group negotiation process is given. The second-order relative change rate of the indicator is used to calculate the number of categories that the drone units should be divided into, determine the drone clusters to which multiple drone units belong, and form multiple drone clusters; in the contract network negotiation between drone clusters, all drone clusters perform performance calculations on all tasks and make pre-bids based on this, and determine the tasks of each drone cluster based on the bidding results of the drone cluster; secondly, bidding is conducted at the drone unit level within the drone cluster, and the bidding object is the tasks within the drone cluster, and the tasks of the drone units within the cluster are determined based on the bidding results of the drone units within the drone cluster.

[0099] In one embodiment, when there are unassigned tasks, the unassigned tasks are reallocated to other drone units whose task loads are not full according to the unit task planning contract network. When all drone units have task execution contracts with negative net performance values ​​for the tasks to be executed, the task execution contracts are exchanged with the task execution contracts of other drones; if the sum of the net performance values ​​of the two drone units after the exchange is greater than the sum of the net performance values ​​of the two drone units before the exchange, the task exchange is performed, otherwise the task exchange is not performed.

[0100] Specific as Figure 4As shown, each drone unit performs task redistribution, finds unassigned tasks, and assigns tasks to drone units whose task capacity is not full. In order to ensure that the overall efficiency of the task allocation scheme is better, the drone units with negative task bid values ​​are adjusted, and the change in the sum of the effectiveness of the two drone units before and after the exchange of tasks is calculated after pre-exchanging tasks with any other drone units. If the sum of the effectiveness of the two becomes larger after the exchange, the task is exchanged.

[0101] In summary, the present invention constructs the cluster task planning contract network between drone sub-clusters and the unit task planning contract network of drone units within the drone sub-cluster as a first-layer negotiation of the contract network, and task exchange and task reallocation are the second-layer negotiation of the contract network, such as Figure 5 As shown, the task planning is finally completed.

[0102] The present invention compares the proposed multi-layer contract network algorithm for adaptive UAV cluster division with the random assignment method and the classic contract network algorithm to perform planning effect statistics on task allocation cases in the same scenario. The experimental scenario is set as follows: the UAV cluster has 12 UAV units and 20 tasks to be executed. The attributes of the UAV units are shown in Table 1, and the attributes of the tasks are shown in Table 2. The algorithm in this study adaptively divides the 12 UAV units of the UAV cluster into multiple UAV clusters as shown in Table 1.

[0103] Table 1 UAV unit attributes

[0104]

[0105] Table 2 Task attributes

[0106]

[0107] In order to better represent the initial position of the drone unit, the initial position of the drone unit is displayed on a map. The longitude and latitude in this map can also be replaced by the two-dimensional coordinates of the ground plane in the local area mission. Figure 6 shown.

[0108] The task is divided through the multi-layer contract network task allocation solution algorithm of adaptive drone cluster division. The effect of the first process of the algorithm drone unit adaptive drone cluster division algorithm is demonstrated. First, calculate the number of categories within a certain range. Value Figure 7 As shown, the optimal number of drone cluster divisions for each drone unit in the drone unit distribution scenario is calculated. The drone units are divided into drone clusters according to the location information of each drone unit.

[0109] Figure 8The squares of different colors represent the location distribution of drone units in different drone clusters, and the red triangles represent the cluster centers of each drone cluster. The adaptive drone cluster division function can realize the reasonable division of drone units in drone clusters without the need to input the number of pre-divided categories, which is convenient for the reasonable management of drone units when there are a large number of drone units and their distribution is not concentrated. The 12 drone units in this scene belong to 3 different subclusters under the same drone cluster. According to Figure 8 As shown, the drone cluster is divided into three groups according to regions, and the three drone sub-clusters have 3, 4, and 5 drone units respectively.

[0110] The task allocation of drone units is as follows: Fig. 9 As shown. The red circle represents the mission point you need to go to, and the blue triangle represents the drone unit. Lines of the same color represent the missions of the same drone unit. The connecting arrows reflect the mission sequence information: starting from the drone unit, pointing to the missions in the drone unit contract list in turn. If there are multiple missions in the drone unit contract list, they are connected in order of the contracts. Fig. 9 The mission scenario shown is not based on the principle of proximity, as the UAV unit has to consider many factors besides the mission location.

[0111] Table 3 shows the matching of each drone unit and the task in the three drone subsets, as well as the benefit value of each drone unit performing the task. Since the overall benefit needs to be maximized, just like Tian Ji's horse racing, the benefit value of some drone units can be negative. In the experiment, the maximum task load of a single drone unit is set to 2. It is worth noting that drone unit 12 is not assigned any task. By looking up Table 1, it can be seen that the execution capability of this drone unit is 18 and the stability is 7. Both are relatively small, so no tasks are assigned in the adaptive task allocation process.

[0112] Table 3 Statistical results after 6 experiments using two methods

[0113]

[0114] In order to verify the robustness of the model, the experimental scenario was changed. The attributes of the drone unit and the task attributes were changed, and six comparative experiments were conducted under random task allocation, single-layer contract network, and the contract network of this study.

[0115] Table 4 Statistical results after 6 experiments using two methods

[0116]

[0117] As shown in the statistical results of the experiments in the six scenarios in Table 4, the multi-layer contract network task allocation algorithm based on adaptive UAV cluster division has better task allocation effect than random task allocation and traditional contract network algorithm. The UAV unit group has gained greater benefits at the group level, improved the utilization value of UAV unit resources, and achieved better results on the mission objectives.

[0118] It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0119] In one embodiment, Fig.10 As shown, a UAV aggregation and task planning device based on an adaptive clustering contract network is provided, comprising: a clustering module 1002, a contract network construction module 1004 and a planning module 1006, wherein:

[0120] A clustering module 1002 is used to cluster the drone cluster according to the location information of the drone units in the drone cluster to obtain multiple drone sub-clusters;

[0121] The contract network construction module 1004 is used to construct a cluster task planning contract network between drone sub-clusters and a unit task planning contract network of drone units within a drone sub-cluster according to the task attributes and unit attributes of the drone units in the drone cluster;

[0122] The planning module 1006 is used to allocate the tasks to be executed to each drone sub-cluster according to the cluster task planning contract, and within each drone sub-cluster, allocate the tasks of the drone sub-cluster to each drone unit according to the unit task planning contract network.

[0123] In one embodiment, the clustering module 1002 is also used to cluster the location data of the drone units in the drone cluster. Normalization; from the set number of clusters Take the values ​​in turn and calculate the mean square error of the K-means model , calculate the first-order relative change rate and the second-order relative change rate of the mean square error; select the optimal K value to maximize the second-order relative change rate, and use the optimal K value as the number of drone sub-clusters; allocate the drone units in the drone cluster to each drone sub-cluster.

[0124] In one embodiment, the clustering module 1002 is further configured to calculate the first-order relative rate of change of the mean square error as:

[0125] ;

[0126] in, is the mean square error The first-order relative rate of change of the mean square error is calculated as follows:

[0127] ;

[0128] in, is the mean square error The second-order relative rate of change of .

[0129] In one of the embodiments, the contract network construction module 1004 is also used to calculate the net performance value of the drone cluster executing the task to be planned based on the task attributes and unit attributes of the drone units in the drone cluster; and use the net performance value as the bid value to construct a cluster task planning contract network between the drone sub-clusters.

[0130] In one embodiment, the unit attributes include: task execution capability, stability and value; the task attributes include: task execution difficulty value, task consumption and task reward value; the contract network construction module 1004 is also used to obtain the average task execution capability, average stability and average value of the drone units in the drone cluster; according to the average task execution capability, the task execution difficulty value and the task reward value, the task reward value of the drone cluster is calculated; according to the task consumption, the average value, the average stability, the distribution information of the drone units in the drone cluster and the number of existing tasks in the drone cluster, the task overhead value of the drone cluster is calculated; according to the task reward value, the task overhead value and the pre-acquired task priority, the net efficiency value of the drone cluster in executing the task to be planned is obtained.

[0131] In one of the embodiments, the contract network construction module 1004 is also used to calculate the unit net efficiency value of the drone cluster to perform the planned task based on the task attributes and unit attributes of the drone unit; calculate the load factor of the drone unit based on the workload of the drone unit; calculate the bid value of the unit net efficiency value and the load factor, and construct the unit task planning contract network of the drone units in the drone sub-cluster.

[0132] In one embodiment, the unit attributes include: task execution capability, stability and value; the task attributes include: task execution difficulty value, task consumption and task reward value; the contract network construction module 1004 is also used to calculate the benefit value of the drone unit based on the task execution capability, task execution difficulty value and task reward value of the drone unit; calculate the cost value based on the task consumption, the value, the stability and the distance between the drone unit and the task execution point; calculate the unit net efficiency value based on the benefit value, the cost value and the pre-acquired task priority; calculate the workload of the drone unit based on the time consumption of all tasks in the contract list of the drone unit; calculate the load factor based on the workload of the drone unit and the average workload of the drone units in the drone cluster.

[0133] In one of the embodiments, a task allocation and exchange module is also included, which is used to reallocate the unallocated tasks to other drone units with insufficient task loads according to the unit task planning contract network when there are unallocated tasks.

[0134] In one embodiment, the task allocation and exchange module is used to exchange the task execution contract with the task execution contracts of other drones when all drone units have a task execution contract with a negative net performance value for the task to be executed; if the sum of the net performance values ​​of the two drone units after the exchange is greater than the sum of the net performance values ​​of the two drone units before the exchange, the task exchange is performed, otherwise the task exchange is not performed.

[0135] For the specific definition of the drone aggregation and task planning device based on the adaptive clustering contract network, please refer to the definition of the drone aggregation and task planning method based on the adaptive clustering contract network mentioned above, which will not be repeated here. Each module in the above-mentioned drone aggregation and task planning device based on the adaptive clustering contract network can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0136] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0137] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for UAV aggregation and mission planning based on adaptive clustering contract network, characterized in that: The method comprises: Clustering the drone cluster according to the location information of the drone units in the drone cluster to obtain a plurality of drone subclusters; According to the task attributes and unit attributes of the drone units in the drone cluster, a cluster task planning contract network between drone sub-clusters and a unit task planning contract network of drone units within a drone sub-cluster are constructed; Allocate the tasks to be performed to each drone sub-cluster according to the cluster task planning contract, and within each drone sub-cluster, allocate the tasks of the drone sub-cluster to each drone unit according to the unit task planning contract network; According to the mission attributes and unit attributes of various drone units in the drone cluster, the drone cluster is clustered to obtain multiple drone subclusters, including: The position data of the drone units in the drone cluster are analyzed standardization; From the set number of clusters Take the values ​​in turn and calculate the mean square error of the K-means model , calculate the first-order relative rate of change and the second-order relative rate of change of the mean square error; Calculating the first-order relative rate of change and the second-order relative rate of change of the mean square error includes: The first-order relative rate of change of the mean square error is calculated as: ； in, is the mean square error The first-order relative rate of change of The second-order relative rate of change of the mean square error is calculated as: ； in, is the mean square error The second-order relative rate of change of Select an optimal K value to maximize the second-order relative rate of change, and use the optimal K value as the number of drone subclusters; Assign the drone units in the drone cluster to each drone sub-cluster.

2. The method according to claim 1, characterized in that According to the mission attributes and unit attributes of the drone units in the drone cluster, a cluster mission planning contract network between drone sub-clusters is constructed, including: According to the task attributes and unit attributes of the drone units in the drone cluster, the net efficiency value of the drone cluster in executing the task to be planned is calculated; The net efficiency value is used as the bidding value to construct a cluster task planning contract network between drone sub-clusters.

3. The method according to claim 2, characterized in that The unit attributes include: task execution capability, stability, and value; The task attributes include: task execution difficulty value, task consumption and task reward value; The calculation steps of the net efficiency value are: Obtain the average mission execution capability, average stability, and average value of drone units in a drone cluster; Calculate the task reward value of the drone cluster according to the average task execution capability, the task execution difficulty value and the task reward value; Calculate the task cost value of the drone cluster according to the task consumption, the average value, the average stability, the distribution information of the drone units in the drone cluster, and the number of existing tasks in the drone cluster; According to the task reward value, the task cost value and the pre-acquired task priority, a net efficiency value of the drone cluster executing the task to be planned is obtained.

4. The method according to claim 1, characterized in that: According to the mission attributes and unit attributes of the drone unit, a unit mission planning contract network of the drone units in the drone sub-cluster is constructed, including: According to the task attributes and unit attributes of the drone unit, the unit net efficiency value of the drone cluster executing the task to be planned is calculated; Calculate the load factor of the drone unit based on the workload of the drone unit; The unit net efficiency value and the load factor are used to calculate the bidding value, and a unit task planning contract network of the drone units in the drone sub-cluster is constructed.

5. The method according to claim 4, characterized in that The unit attributes include: task execution capability, stability, and value; The task attributes include: task execution difficulty value, task consumption and task reward value; The calculation steps of the unit net efficiency value are: The benefit value of the drone unit is calculated based on the drone unit's mission execution capability, mission execution difficulty value, and mission reward value; Calculate the cost value according to the task consumption, the value, the stability, and the distance between the UAV unit and the task execution point; Calculating a unit net efficiency value according to the benefit value, the cost value, and the pre-acquired task priority; The steps to calculate the load factor are: The workload of the drone unit is calculated based on the time consumption of all tasks in the drone unit's contract list; The load factor is calculated based on the workload of the drone unit and the average workload of the drone units in the drone cluster.

6. The method according to any one of claims 3 to 5, characterized in that: The method further comprises: When there are unassigned tasks, the unassigned tasks are reallocated to other drone units whose task loads are not full according to the unit task planning contract network.

7. The method according to claim 6, characterized in that The method further comprises: When all drone units have a mission execution contract with a negative net effectiveness value for the mission to be performed, the mission execution contract is exchanged with the mission execution contracts of other drones; If the sum of the net efficiency values ​​of the two drone units after the exchange is greater than the sum of the net efficiency values ​​of the two drone units before the exchange, the task exchange is performed, otherwise the task exchange is not performed.

Images