A method for task allocation of unmanned swarms based on expected probability of success

Abstract

The application provides a method for task allocation of unmanned swarm based on success expectation probability, which first decomposes the task of unmanned swarm into a plurality of serial formation tasks, each of which is completed by a plurality of unmanned platforms in parallel cooperation; then screens the unmanned platform resources according to the success expectation probability threshold constraint of the formation task, so as to reduce the search space of task resource allocation; finally, a heuristic exploration strategy is adopted to efficiently explore and optimize the task resource allocation scheme, so as to quickly form a suboptimal scheme meeting the success expectation probability threshold. The method does not take the optimal task allocation as the target but is driven by the success expectation probability of the task, can support real-time allocation of the task of unmanned swarm, prevent over-allocation of resources of a single formation task, effectively improve the success rate and efficiency of task execution and reduce the risk and cost.

Description

Technical Field

[0001] This invention relates to the field of unmanned swarm task allocation technology, and in particular to an unmanned swarm task allocation method based on the expected probability of success. Background Technology

[0002] Drone swarms have become an inevitable trend in the future development of drones, and task allocation within drone swarms is a key technology, requiring the completion of tasks such as intelligence gathering, surveillance, reconnaissance, and multi-target attacks in the shortest possible time. Traditional task allocation methods mainly fall into three categories: mathematical programming methods, task allocation methods based on contract networks, and heuristic intelligent optimization algorithms.

[0003] Mathematical programming is a deterministic algorithm dedicated to finding the optimal solution to the task allocation problem (see ALIDAEE B, WANG H, LANDRAM F. A note on integer programming formulations of thereal-time optimal scheduling and flight path selection of UAVs [J]. IEEE Trans on Control Systems Technology, 2009, 17(4): 839-843). However, with the increase of task complexity, mathematical programming is time-consuming and cannot find the optimal solution in a finite time. Contract-based methods use a bidding process between buyers and sellers to compete for the right to execute tasks (see BOGDANOWICZ Z R. A new efficient algorithm for optimal assignment of smart weapons to targets [J]. Computers & Mathematics with Applications, 2009, 58(10): 1965-1969). However, due to the need for multiple rounds of negotiation between the task execution entities, complex task allocation suffers from problems such as large communication volume, long time consumption, and uneven allocation. Heuristic intelligent optimization algorithms, such as genetic algorithms and ant colony algorithms, seek the global optimal solution through a large number of random search attempts (see ROBERGE V, TARBOUCHI M, LABONTE G. Comparison of parallel genetic algorithm and particle swarm optimization for realtime uav path planning [J]. IEEE Trans on Industrial Informatics, 2013, 9(1): 132-141), which have the problems of low search efficiency and long time consumption. Summary of the Invention

[0004] Objective of the Invention: The technical problem this invention aims to solve is to address the shortcomings of existing technologies by providing a method for unmanned swarm task allocation based on the expected success probability. This method uses the expected success probability as a constraint to filter unmanned platform resources, effectively improving the timeliness and rationality of unmanned resource allocation and better adapting to the collaborative needs of unmanned operations in complex and highly confrontational battlefield environments. Specifically, this invention includes the following:

[0005] The unmanned swarm task T is configured to be decomposed into serial executions. The formation tasks (including detection, tracking, jamming, and attack) are respectively denoted as follows: , … , The total number of unmanned platforms is N, and the mission of each formation is... No more than The unmanned platforms work in parallel and collaboratively to complete the task. , Each unmanned platform can participate in a maximum of one formation mission. A swarm mission allocation scheme T needs to be generated to ensure that each formation mission... The expected probability of success is no less than a pre-set threshold. Therefore, the expected probability of success for the unmanned swarm mission T is no less than Then perform the following steps:

[0006] Step 1: Construct the capability representation matrix of the unmanned platform;

[0007] Step 2: Analyze the feasibility of the formation mission;

[0008] Step 3: Calculate the minimum number of platforms required for the formation task;

[0009] Step 4: Initialize the formation task allocation matrix;

[0010] Step 5: Set the current formation task number =1;

[0011] Step 6: Assess the current formation mission;

[0012] Step 7: Filter out resources for the current formation mission;

[0013] Step 8: Filter the current formation mission resources;

[0014] Step 9: Update the current formation mission number. , If satisfied Return to step 6; otherwise, continue to step 10.

[0015] Step 10: Count the number of values ​​taken by the unmanned platform: for each formation mission The number of unmanned platforms with values ​​of 0, 1, and null in the formation task allocation matrix is ​​counted and denoted as follows: , and ;

[0016] Step 11: Allocate resources for the formation task;

[0017] Step 12: Optimize the allocation of resources for formation tasks;

[0018] Step 13: Output the resource allocation plan for each formation task: If the resource allocation plan for each formation task is found... The expected probability of task success is no less than the threshold. Optimization scheme Output ;

[0019] Step 14, end the process.

[0020] Step 1 includes: targeting N unmanned platforms and Formation task construction Dimensional unmanned platform capability representation matrix ,matrix The Line 1 Column units represent unmanned platforms Complete formation mission success rate , , .

[0021] Step 2 includes: tasks for each formation. N unmanned platforms will complete their formation missions according to their respective tasks. The success probabilities are sorted from high to low, i.e., according to the unmanned platform capability representation matrix. Sort the cell values ​​from largest to smallest, denoted as , Indicates completion of formation mission The unmanned platform ranked Nth in capability determines the formation task according to formula (1). Is it possible to satisfy the threshold? If the conditions are not met, then the formation task is judged. The expected probability threshold for success is not met. If not, skip to step 14; otherwise, continue to step 3.

[0022] (1).

[0023] Step 3 includes: tasks for each formation. Calculate the threshold according to formula (2) Minimum number of unmanned platforms required :

[0024] (2).

[0025] Step 4 includes: targeting An unmanned platform and Formation task construction Dimensional formation task assignment matrix ,matrix The Line 1 Column cell Characterization Formation Mission Should it be assigned to an unmanned platform? , A value of zero indicates a formation task. Not allocated to unmanned platforms , A value of 1 indicates a formation task. Assigned to unmanned platforms ,matrix All cell values ​​are initialized to null.

[0026] Step 6 includes: for the current formation task , Each unmanned platform first allocates tasks according to the formation matrix. The values ​​are sorted, and the first part of the unmanned platforms are assigned to the matrix. The value in the matrix is ​​either 1 or null; the latter part of the unmanned platform is assigned a matrix. If the value is 0, then the unmanned platform completes the current formation task. The success probabilities are ranked from highest to lowest to obtain the updated formation task. The unmanned platforms are sorted and denoted as follows: , Indicates completion of the current formation mission. The Nth unmanned platform in the success probability ranking is updated. The number of unmanned platforms that satisfy the constraint that the task allocation matrix can take the value of 1 or null, i.e. Determine the formation task according to formula (3) Is it possible to meet the threshold? If it is impossible to satisfy the condition, then the formation task is judged. Threshold not met If not, skip to step 14; otherwise, continue to step 7.

[0027] (3).

[0028] Step 7 includes: for the current formation mission According to the threshold Constraints were used to eliminate all unmanned platforms with an excessively low probability of success in formation missions. The unmanned platform that will soon satisfy formula (4) In the corresponding unit of the formation task allocation matrix Set the value to zero:

[0029] (4).

[0030] Step 8 includes: based on the current formation mission. Sorting of unmanned platforms, setting filter variables Set the current unmanned platform sorting number to 1. The value is 1, indicating the current unmanned platform. , This indicates the current formation mission. The sorting number is The unmanned platform shall perform the following steps:

[0031] Step 8-1, Assess the unmanned platform: If the current unmanned platform If no assignment is made, then the team will be assigned a task. The success rate cannot meet the threshold That is, it satisfies formula (5):

[0032] (5)

[0033] If the filter variable is set to 1, then the formation task is determined. Unmanned platforms must be allocated Unmanned platforms In the corresponding unit of the formation task allocation matrix Set the value to 1, and set the unmanned platform Corresponding to other formation missions Formation task allocation matrix unit The value is set to 0. Update the unmanned platform sorting number Update the current unmanned platform If the current unmanned platform If no assignment is made, then the team will be assigned a task. The completion probability still meets the threshold. If formula (5) is not satisfied, then the formation task... Unmanned platforms may not be allocated. Update filter variables =0;

[0034] Step 8-2, Determine the iteration condition: If the selected variables... The value is 1, and the current unmanned platform's sorting number. Not greater than ,Right now If the condition is met, the unmanned platform assessment is iterated and the process returns to step 8-1; otherwise, the formation task resource screening ends and the process continues to step 9.

[0035] Step 11 includes: based on the filtered and selected formation task allocation matrix, calculating each formation task according to formula (6). Assigned to unmanned platforms Allocation probability :

[0036] (6)

[0037] in, Indicates unmanned platform In formation mission The probability of each unmanned platform is calculated using formula (6) based on the sorting number in the formula. Assignments are made, and the number of heuristic exploration schemes is set to [number]. Then explore solutions for each allocation. , The corresponding formation task allocation matrix Initialize to , Tasks will be assigned Corresponding formation task allocation matrix unit The value is set to 1 for other grouping tasks. The corresponding value is set to 0. Until the mission of each formation The number of unmanned platforms allocated has reached its minimum number. ,in .

[0038] Step 12 includes: assigning a formation task matrix to each heuristic exploration scheme. Make optimizations and adjustments. If formation mission The expected probability of success has been met. Then the task will not be adjusted. Unmanned platforms; if formation mission The expected probability threshold for success is not met. And the number of unmanned platforms allocated to it shall not exceed Further optimization and adjustments are made: While ensuring that each unmanned platform can participate in a maximum of one formation mission, one approach is to add a new unmanned platform. Secondly, the tasks assigned to the formation will be... An unmanned platform With assigned to another formation mission unmanned platform Exchange, missions of each formation In the current plan Income As shown in formula (7),

[0039] (7)

[0040] According to the profit function shown in formula (8), the adjusted profit increase is... maximum:

[0041] (8)

[0042] against A heuristic exploration scheme, explored sequentially, satisfies a success rate of no less than [percentage missing] for all formation tasks. The optimization solution is to stop searching once a solution is found.

[0043] Beneficial effects: Compared with traditional path planning methods, this invention has the following advantages: This invention does not aim at the globally optimal task allocation but is driven by the expected probability of task success, realizing efficient suboptimal allocation of unmanned swarm tasks. It can effectively improve the real-time performance of task allocation and effectively prevent the over-allocation of task resources in a single formation, while significantly improving the success rate and efficiency of task execution and reducing the cost of task execution. Attached Figure Description

[0044] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention in the above and / or other aspects will become clearer.

[0045] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0046] This invention provides a method for task allocation in unmanned bee colonies based on expected success probability, including the following:

[0047] The unmanned swarm mission T is configured to be decomposed into serially executed tasks. ( The formation missions include detection, tracking, jamming, and attack, respectively denoted as... , … The total number of unmanned platforms is N, and the mission of each formation is... ( ) by no more than ( ( ) A swarm of unmanned platforms works in parallel and collaboratively, with each platform participating in at most one formation task. A task allocation scheme for the unmanned swarm task T needs to be generated to ensure that each formation task... The expected probability of success is no less than a pre-set threshold. ( Therefore, the expected probability of success for the unmanned swarm mission T is no less than [a certain value]. .

[0048] Step 1: Construction of the unmanned platform capability representation matrix: For N unmanned platforms and Formation task construction Dimensional unmanned platform capability representation matrix ,matrix The ( ) line number Column units represent unmanned platforms Complete formation mission ( The probability of success of ) is denoted as , ;

[0049] Step 2, Feasibility Analysis of Formation Missions: For each formation mission... Each unmanned platform will complete its formation mission according to its assigned task. The success probabilities are sorted from high to low, i.e., according to the unmanned platform capability representation matrix. Sort the cell values ​​from largest to smallest, denoted as Determine the formation task according to formula (1) Is it possible to satisfy the threshold? If the conditions are not met, then the formation task is judged. The expected probability threshold for success is not met. If yes, proceed to step 14; otherwise, continue to step 3.

[0050] (1)

[0051] Step 3: Calculation of the minimum number of platforms for each formation task: For each formation task Calculate the threshold according to formula (2) The minimum number of unmanned platforms required :

[0052] (2)

[0053] Step 4, Initialize the formation task allocation matrix: For An unmanned platform and Formation task construction Dimensional formation task assignment matrix ,matrix The ( ) line number Column cell Characterization Formation Mission Should it be assigned to an unmanned platform? A value of zero indicates a formation task. Not allocated to unmanned platforms A value of 1 indicates a formation task. Assigned to unmanned platforms ,matrix All cell values ​​are initialized to null.

[0054] Step 5: Set the current formation task number: Set the current formation task number. It is 1, that is ;

[0055] Step 6, Current Formation Task Evaluation: For the current formation task... Each unmanned platform first follows the task allocation matrix. Sort the values, with the first part being 1 or null and the second part being 0, and then complete the current formation task according to these values. The success probabilities are ranked from highest to lowest to obtain the updated formation task. The unmanned platforms are sorted and denoted as follows: ,renew The number of unmanned platforms that satisfy the constraint that the task allocation matrix can take the value of 1 or null, i.e. Determine the formation task according to formula (3) Is it possible to meet the threshold? If it is impossible to satisfy the condition, then the formation task is judged. Threshold not met If yes, skip to step 14; otherwise, continue to step 7.

[0056] (3)

[0057] Step 7: Filter resources for the current formation task: For the current formation task... According to the threshold Constraints were used to eliminate all unmanned platforms with an excessively low probability of success in formation missions. The unmanned platform that will soon satisfy formula (4) The corresponding cell in the formation task allocation matrix is ​​set to zero. ,

[0058] (4)

[0059] Step 8: Filter resources for the current formation mission: Based on the current formation mission The unmanned platforms are sorted, with the filter variable set to 1, i.e. Set the current unmanned platform sorting number to 1, that is... Current unmanned platforms The formation mission resource selection includes two sub-steps: unmanned platform assessment and iterative condition judgment.

[0060] Step 8-1 Unmanned Platform Assessment: If the current unmanned platform If no assignment is made, then the team will be assigned a task. The success rate cannot meet the threshold That is, it satisfies formula (5).

[0061] (5)

[0062] If the filter variable is set to 1, then the formation task is determined. Unmanned platforms must be allocated Unmanned platforms In the formation task allocation matrix, the corresponding cell is set to 1, that is... and will unmanned platforms Corresponding to other formation missions ( The value of the formation task allocation matrix element is set to 0, that is... Update the unmanned platform sorting number Update the current unmanned platform If the current unmanned platform If no assignment is made, then the team will be assigned a task. The completion probability still meets the threshold. If formula (5) is not satisfied, then the formation task... Unmanned platforms may not be allocated. Update the filter variable to 0, that is ;

[0063] Step 8-2 Iteration condition judgment: If the filter variable takes the value 1, that is... And the current unmanned platform sorting number Not greater than ,Right now If the unmanned platform assessment is successful, return to step 8-1; otherwise, end the formation task resource screening and continue to step 9.

[0064] Step 9: Update the current formation mission number: Update the current formation mission number. , If satisfied Return to step 6; otherwise, continue to step 10.

[0065] Step 10: Statistics on the number of values ​​collected by the unmanned platform: For each formation mission The number of unmanned platforms with values ​​of 0, 1, and null in the formation task allocation matrix is ​​counted and denoted as follows: , and ;

[0066] Step 11, Heuristic exploration of formation task allocation: Based on the formation task allocation matrix after filtering and selection, calculate the formation tasks according to formula (6). Assigned to unmanned platforms Allocation probability :

[0067] (6)

[0068] in, Indicates unmanned platform In formation mission The probability of each unmanned platform is calculated using formula (6) based on the sorting number in the formula. Assignments are made, assuming the number of heuristic exploration schemes is... Then explore solutions for each allocation. , The corresponding formation task allocation matrix Initialize to , Based on this, tasks will be assigned. The corresponding formation task allocation matrix cell is set to 1. Other formation missions ( The corresponding value is set to 0. Until the mission of each formation The number of unmanned platforms allocated has reached its minimum number. ;

[0069] Step 12, Optimize Formation Task Resource Allocation: Optimize the formation task allocation matrix for each heuristic exploration scheme. ( ) will be optimized and adjusted, if the formation task The expected probability of success has been met. Then the unmanned platform for this mission will not be adjusted; if it is a formation mission The expected probability threshold for success is not met. And the number of unmanned platforms allocated to it shall not exceed Further optimization and adjustments are made: While ensuring that each unmanned platform can participate in a maximum of one formation mission, one approach is to add a new unmanned platform. Secondly, the tasks assigned to the formation will be... An unmanned platform With assigned to another formation mission unmanned platform Exchange, missions of each formation In the current plan The returns are shown in formula (7).

[0070] (7)

[0071] According to the profit function shown in formula (8), the adjusted profit increase is maximized:

[0072] (8)

[0073] against A heuristic exploration scheme, explored sequentially, satisfies a success rate of no less than [percentage missing] for all formation tasks. The optimization solution is to stop searching once found;

[0074] Step 13, Output the resource allocation plan for the formation task: If each formation is found The expected probability of task success is no less than Optimization scheme Output ;

[0075] Step 14, process ends.

[0076] The "Unmanned Swarm Task Allocation Method Based on Success Expectation Probability" proposed in this invention uses the success expectation probability of swarm task collaboration as a constraint to screen subsequent unmanned platforms for unmanned swarm tasks, and then adopts a heuristic exploration strategy to efficiently search for swarm task allocation schemes that meet the success expectation probability threshold. Compared with traditional path planning methods, the "Unmanned Swarm Task Allocation Method Based on Success Expectation Probability" of this invention has the following advantages: (1) It does not aim at optimal task allocation but is driven by the success expectation probability of tasks, which can support the real-time allocation of unmanned swarm tasks; (2) It can prevent the over-allocation of resources for individual swarm tasks, effectively improve the success rate and efficiency of task execution, and reduce risks and costs.

[0077] Example 1: Capability characterization of unmanned platforms.

[0078] The unmanned swarm mission T is configured to be decomposed into serially executed tasks. The formation missions include coordinated detection, coordinated jamming, and coordinated attack, respectively denoted as... , and If the total number of unmanned platforms is N=6, then the unmanned platform capability representation matrix is... The structure is as follows:

[0079] Step 1, Construction of the unmanned platform capability representation matrix: This involves constructing a capability representation matrix for six unmanned platforms and... Formation task construction Dimensional unmanned platform capability representation matrix ,matrix The ( ) line number Column units represent unmanned platforms Complete formation mission The probability of success is denoted as . , As shown in Table 1 (Unmanned Platform Capability Characterization Matrix):

[0080] Table 1

[0081] Unmanned Platform 1 0.8 0.2 0.2 Unmanned Platform 2 0.3 0.3 0.9 Unmanned Platform 3 0.9 0.9 0.9 Unmanned Platform 4 0.3 0.9 0.9 Unmanned Platform 5 0.9 0.1 0.2 Unmanned Platform 6 0.3 0.1 0.8

[0082] Example 2: Task allocation and threshold for unmanned bee swarms Maximum number of platforms =1.

[0083] Based on the implementation of Case 1, assume that each formation has a mission ( ) by no more than ( ( ) A swarm of unmanned platforms works in parallel and collaboratively, with each platform participating in at most one formation task. A task allocation scheme for the unmanned swarm task T needs to be generated to ensure that each formation task... The expected probability of success is no less than a pre-set threshold. =0.98 ( Therefore, the expected probability of success for the unmanned swarm mission T is no less than [a certain value]. .

[0084] Step 1, Construction of the unmanned platform capability representation matrix: Same as in Implementation Case 1;

[0085] Step 2, Feasibility analysis of formation tasks: Taking formation tasks as an example. Taking collaborative detection as an example, each unmanned platform will first complete its formation mission according to its assigned task. The success probabilities are sorted from highest to lowest, then further sorted by serial number from smallest to largest, and denoted as... According to formula (1), Determine the formation task Under constraints The threshold is not met. Skip to step 14 to end the process.

[0086] Example 3: Task allocation and threshold for unmanned bee swarms Maximum number of platforms =2.

[0087] Based on the implementation of Case 1, assume that each formation has a mission ( ) by no more than ( ( ) A swarm of unmanned platforms works in parallel and collaboratively, with each platform participating in at most one formation task. A task allocation scheme for the unmanned swarm task T needs to be generated to ensure that each formation task... The expected probability of success is no less than a pre-set threshold. =0.98 ( Therefore, the expected probability of success for the unmanned swarm mission T is no less than [a certain value]. .

[0088] Step 1, Construction of the unmanned platform capability representation matrix: Same as in Implementation Case 1;

[0089] Step 2, Feasibility analysis of formation tasks: Taking formation tasks as an example. Taking collaborative detection as an example, each unmanned platform is assigned a task based on its formation mission. The success probabilities are sorted from high to low, i.e., according to the unmanned platform capability representation matrix. Sort the cell values ​​from largest to smallest, denoted as According to formula (1), Determine the formation task It is possible to meet the threshold. Similarly, determining the formation task and All of them may meet the threshold, so continue to step 3.

[0090] Step 3, Calculation of the minimum number of platforms for each formation task: For each formation task Calculate the threshold according to formula (2) The minimum number of unmanned platforms required ,get ;

[0091] Step 4, Initialize the formation task allocation matrix: For An unmanned platform and Formation task construction Dimensional formation task assignment matrix ,matrix All unit values ​​are initialized to null, as shown in Table 2 (initialized formation task assignment matrix X):

[0092] Table 2

[0093] Unmanned Platform 1 null null null Unmanned Platform 2 null null null Unmanned Platform 3 null null null Unmanned Platform 4 null null null Unmanned Platform 5 null null null Unmanned Platform 6 null null null

[0094] Step 5, Setting the Current Formation Task Number: Set the current formation task number. It is 1, that is ;

[0095] Step 6, Current formation task evaluation: Determine the formation task according to formula (3). It is possible to meet the threshold. Continue to step 7;

[0096] Step 7, Current Formation Task Resource Filtering: For the current formation task According to the threshold Constraints are applied to eliminate unmanned platforms 2, 4, and 6 with excessively low success rates in formation missions. Specifically, the values ​​of unmanned platforms 2, 4, and 6 that satisfy formula (4) are set to zero in the corresponding cells of the formation mission allocation matrix, as shown in Table 3 (Formation Missions). The formation task allocation matrix X after resource filtering is shown below:

[0097] Table 3

[0098] Unmanned Platform 1 null null null Unmanned Platform 2 0 null null Unmanned Platform 3 null null null Unmanned Platform 4 0 null null Unmanned Platform 5 null null null Unmanned Platform 6 0 null null

[0099] Step 8, Current Squadron Task Resource Filtering: Based on the current squadron task The unmanned platforms are sorted, with the filter variable set to 1, i.e. Set the current unmanned platform sorting number to 1, that is... Current unmanned platforms ,

[0100] Step 8-1, Unmanned Platform Assessment: If the current unmanned platform If no assignment is made, then the team will be assigned a task. The success rate is still sufficient to meet the threshold. If formula (5) is not satisfied, the filter variable is updated to 0. ;

[0101] Step 8-2, Iteration condition judgment: If the value of the filter variable is 0, the formation task resource filtering ends and proceeds to step 9;

[0102] Step 9, Update the current formation mission number: Update the current formation mission number. , ,satisfy Continue with step 6 below;

[0103] Step 6, Current formation task evaluation: Determine the formation task according to formula (3). It is possible to meet the threshold. Continue to step 7;

[0104] Step 7, Current Formation Task Resource Filtering: For the current formation task According to the threshold Constraints are used to eliminate all unmanned platforms with a low success rate in formation tasks. Specifically, the values ​​of unmanned platforms 1, 2, 5, and 6 that satisfy formula (4) are set to zero in the corresponding cells of the formation task allocation matrix, as shown in Table 4 (formation task). The formation task allocation matrix X after resource filtering is shown below:

[0105] Table 4

[0106] Unmanned Platform 1 null 0 null Unmanned Platform 2 0 0 null Unmanned Platform 3 null null null Unmanned Platform 4 0 null null Unmanned Platform 5 null 0 null Unmanned Platform 6 0 0 null

[0107] Step 8, Current Squadron Task Resource Filtering: Based on the current squadron task The unmanned platforms are sorted, with the filter variable set to 1, i.e. Set the current unmanned platform sorting number to 1, that is... Current unmanned platforms ,

[0108] Step 8-1, Unmanned Platform Assessment: Determine the current status of the unmanned platform. If no assignment is made, then the team will be assigned a task. The success rate cannot meet the threshold If the conditions are met (5) and the filter variable is 1, then the formation task is determined. Unmanned platforms must be allocated Unmanned platforms In the formation task allocation matrix, the corresponding cell is set to 1, that is... and will unmanned platforms Corresponding to other formation missions The values ​​of the task allocation matrix units for formation 3 are set to 0, and the sorting numbers of the unmanned platforms are updated. Update the current unmanned platform ;

[0109] Step 8-2, Iteration condition judgment: The filter variable is 1 and the current unmanned platform sorting number. Not greater than If so, iteratively assess the unmanned platform and continue with step 8-1 below;

[0110] Step 8-1 Unmanned Platform Assessment: Determine the current unmanned platform If no assignment is made, then the team will be assigned a task. The success rate cannot meet the threshold If the conditions are met (5) and the filter variable is 1, then the formation task is determined. Unmanned platforms must be allocated Unmanned platforms In the formation task allocation matrix, the corresponding cell is set to 1, that is... and will unmanned platforms Corresponding to other formation missions The values ​​of the task allocation matrix units for formation 3 are set to 0, and the sorting numbers of the unmanned platforms are updated. Update the current unmanned platform The formation task allocation matrix is ​​shown in Table 5 (formation tasks). The formation task allocation matrix X after resource filtering is shown below:

[0111] Table 5

[0112] Unmanned Platform 1 null 0 null Unmanned Platform 2 0 0 null Unmanned Platform 3 0 1 0 Unmanned Platform 4 0 1 0 Unmanned Platform 5 null 0 null Unmanned Platform 6 0 0 null

[0113] Step 8-2, Iteration condition judgment: Current unmanned platform sorting number Greater than End the formation task resource screening and continue to step 9;

[0114] Step 9, Update the current formation mission number: Update the current formation mission number. , ,satisfy Continue with step 6 below;

[0115] Step 6, Current formation task evaluation: Determine the formation task according to formula (3). It is possible to meet the threshold. Continue to step 7;

[0116] Step 7, Current Formation Task Resource Filtering: For the current formation task According to the threshold Constraints were used to eliminate all unmanned platforms with an excessively low probability of success in formation missions. And 5, that is, the values ​​of unmanned platforms 1 and 5 in the corresponding cells of the formation task allocation matrix are set to zero;

[0117] Step 8, Current Formation Task Resource Filtering: Filter out unmanned platforms 2 and 6, and update the formation task allocation matrix unit, as shown in Table 6 (Formation Tasks). The formation task allocation matrix X after resource filtering is shown below:

[0118] Table 6

[0119] Unmanned Platform 1 null 0 0 Unmanned Platform 2 0 0 1 Unmanned Platform 3 0 1 0 Unmanned Platform 4 0 1 0 Unmanned Platform 5 null 0 0 Unmanned Platform 6 0 0 1

[0120] Step 9, Update the current formation mission number: Update the current formation mission number. , =4, not satisfied. Continue with step 10 below;

[0121] Step 10, Statistics on the Number of Values ​​Taken by Unmanned Platforms: Based on Formation Tasks For example, the number of unmanned platforms with values ​​of 0, 1, and null in the formation task allocation matrix is ​​counted, and denoted as follows: , and Similarly, , , , ;

[0122] Step 11, heuristic exploration of formation task resource allocation: Based on the formation task allocation matrix after filtering and selection, calculate each formation task according to formula (6). Assigned to unmanned platforms The allocation probabilities are shown in Table 7 (Assignment probabilities of unmanned platforms for each formation mission). As shown in the image:

[0123] Table 7

[0124] Unmanned Platform 1 100% 0 0 Unmanned Platform 2 0 0 100% Unmanned Platform 3 0 100% 0 Unmanned Platform 4 0 100% 0 Unmanned Platform 5 100% 0 0 Unmanned Platform 6 0 0 100%

[0125] Set the number of heuristic exploration schemes to be The heuristic exploration scheme is output, as shown in Table 8 (Heuristic Exploration Scheme Y):

[0126] Table 8

[0127] Unmanned Platform 1 1 0 0 Unmanned Platform 2 0 0 1 Unmanned Platform 3 0 1 0 Unmanned Platform 4 0 1 0 Unmanned Platform 5 1 0 0 Unmanned Platform 6 0 0 1

[0128] Step 12, Optimize resource allocation for formation tasks: Formation tasks The expected probability of success has been met. Then the unmanned platform for this mission will not be adjusted. ;

[0129] Step 13, Output the formation task resource allocation plan: Output ;

[0130] Step 14, Process End: The process ends.

[0131] In complex unmanned swarm missions, corresponding expected success probability thresholds can be set for different formation tasks based on expert experience. Unmanned platform resources can then be screened based on these thresholds, and allocation schemes can be heuristically explored and optimized in conjunction with specific task requirements. This invention does not aim for optimal task allocation but is driven by the expected success probability of the task. It can support real-time allocation of unmanned swarm tasks and prevent over-allocation of resources to individual formation tasks, effectively improving task execution success rate and efficiency while reducing risks and costs.

[0132] The research work of this invention was supported by the Collaborative Innovation Center for New Software Technologies and Industrialization.

[0133] This invention provides a method for task allocation in unmanned bee colonies based on the expected probability of success. Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment of the invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. A method for task allocation in unmanned bee colonies based on expected success probability, characterized in that, include: The unmanned swarm task T is set to be decomposed into serial executions. The formation tasks are respectively denoted as , … , The total number of unmanned platforms is N, and the mission of each formation is... No more than The unmanned platforms work in parallel and collaboratively to complete the task. , Each unmanned platform can participate in a maximum of one formation mission. A swarm mission allocation scheme T needs to be generated to ensure that each formation mission... The expected probability of success is no less than a pre-set threshold. Therefore, the expected probability of success for the unmanned swarm mission T is no less than Then perform the following steps: Step 1: Construct the capability representation matrix of the unmanned platform; Step 2: Analyze the feasibility of the formation mission; Step 3: Calculate the minimum number of platforms required for the formation task; Step 4: Initialize the formation task allocation matrix; Step 5: Set the current formation task number =1; Step 6: Assess the current formation mission; Step 7: Filter out resources for the current formation mission; Step 8: Filter the current formation mission resources; Step 9: Update the current formation mission number. , If satisfied Return to step 6; otherwise, continue to step 10. Step 10: Count the number of values ​​taken by the unmanned platform: for each formation mission The number of unmanned platforms with values ​​of 0, 1, and null in the formation task allocation matrix is ​​counted and denoted as follows: , and ; Step 11: Allocate resources for the formation task; Step 12: Optimize the allocation of resources for formation tasks; Step 13: Output the resource allocation plan for each formation task: If the resource allocation plan for each formation task is found... The expected probability of task success is no less than the threshold. Optimization scheme Output ; Step 14, end the process; Step 1 includes: targeting N unmanned platforms and Formation task construction Dimensional unmanned platform capability representation matrix ,matrix The Line 1 Column units represent unmanned platforms Complete formation mission success rate , , ; Step 2 includes: tasks for each formation. N unmanned platforms will complete their formation missions according to their respective tasks. The success probabilities are sorted from high to low, i.e., according to the unmanned platform capability representation matrix. Sort the cell values ​​from largest to smallest, denoted as , Indicates completion of formation mission The unmanned platform ranked Nth in capability determines the formation task according to formula (1). Is it possible to satisfy the threshold? If the conditions are not met, then the formation task is judged. The expected probability threshold for success is not met. If not, skip to step 14; otherwise, continue to step 3. (1)。 2. The method according to claim 1, characterized in that, Step 3 includes: tasks for each formation. Calculate the threshold according to formula (2) Minimum number of unmanned platforms required : (2)。 3. The method according to claim 2, characterized in that, Step 4 includes: targeting An unmanned platform and Formation task construction Dimensional formation task assignment matrix ,matrix The Line 1 Column cell Characterization Formation Mission Should it be assigned to an unmanned platform? , A value of zero indicates a formation task. Not allocated to unmanned platforms , A value of 1 indicates a formation task. Assigned to unmanned platforms ,matrix All cell values ​​are initialized to null.

4. The method according to claim 3, characterized in that, Step 6 includes: for the current formation task , Each unmanned platform first allocates tasks according to the formation matrix. The values ​​are sorted, and the first part of the unmanned platforms are assigned to the matrix. The value in the matrix is ​​either 1 or null; the latter part of the unmanned platform is assigned a matrix. If the value is 0, then the unmanned platform completes the current formation task. The success probabilities are ranked from highest to lowest to obtain the updated formation task. The unmanned platforms are sorted and denoted as follows: , Indicates completion of the current formation mission. The Nth unmanned platform in the success probability ranking is updated. The number of unmanned platforms that satisfy the constraint that the task allocation matrix can take the value of 1 or null, i.e. Determine the formation task according to formula (3) Is it possible to meet the threshold? If it is impossible to satisfy the condition, then the formation task is judged. Threshold not met If not, skip to step 14; otherwise, continue to step 7. (3)。 5. The method according to claim 4, characterized in that, Step 7 includes: for the current formation mission According to the threshold Constraints were used to eliminate all unmanned platforms with an excessively low probability of success in formation missions. The unmanned platform that will soon satisfy formula (4) In the corresponding unit of the formation task allocation matrix Set the value to zero: (4)。 6. The method according to claim 5, characterized in that, Step 8 includes: based on the current formation mission. Sorting of unmanned platforms, setting filter variables Set the current unmanned platform sorting number to 1. The value is 1, indicating the current unmanned platform. , This indicates the current formation mission. The sorting number is The unmanned platform shall perform the following steps: Step 8-1, Assess the unmanned platform: If the current unmanned platform If no assignment is made, then the team will be assigned a task. The success rate cannot meet the threshold That is, it satisfies formula (5): (5) If the filter variable is set to 1, then the formation task is determined. Unmanned platforms must be allocated Unmanned platforms In the corresponding unit of the formation task allocation matrix Set the value to 1, and set the unmanned platform Corresponding to other formation missions Formation task allocation matrix unit The value is set to 0. Update the unmanned platform sorting number Update the current unmanned platform If the current unmanned platform If no assignment is made, then the team will be assigned a task. The completion probability still meets the threshold. If formula (5) is not satisfied, then the formation task... Unmanned platforms may not be allocated. Update filter variables =0; Step 8-2, Determine the iteration condition: If the selected variables... The value is 1, and the current unmanned platform's sorting number. Not greater than ,Right now If the condition is met, the unmanned platform assessment is iterated and the process returns to step 8-1; otherwise, the formation task resource screening ends and the process continues to step 9.

7. The method according to claim 6, characterized in that, Step 11 includes: based on the filtered and selected formation task allocation matrix, calculating each formation task according to formula (6). Assigned to unmanned platforms Allocation probability : (6) in, Indicates unmanned platform In formation mission The probability of each unmanned platform is calculated using formula (6) based on the sorting number in the formula. Assignments are made, and the number of heuristic exploration schemes is set to [number]. Then explore solutions for each allocation. , The corresponding formation task allocation matrix Initialize to , Tasks will be assigned Corresponding formation task allocation matrix unit The value is set to 1 for other grouping tasks. The corresponding value is set to 0. Until the mission of each formation The number of unmanned platforms allocated has reached its minimum number. ,in .

8. The method according to claim 7, characterized in that, Step 12 includes: assigning a formation task matrix to each heuristic exploration scheme. Make optimizations and adjustments. If formation mission The expected probability of success has been met. Then the task will not be adjusted. Unmanned platforms; if formation mission The expected probability threshold for success is not met. And the number of unmanned platforms allocated to it shall not exceed Further optimization and adjustments are made: While ensuring that each unmanned platform can participate in a maximum of one formation mission, one approach is to add a new unmanned platform. Secondly, the tasks assigned to the formation will be... An unmanned platform With assigned to another formation mission unmanned platform Exchange, missions of each formation In the current plan Income As shown in formula (7), (7) According to the profit function shown in formula (8), the adjusted profit increase is... maximum: (8) against A heuristic exploration scheme, explored sequentially, satisfies a success rate of no less than [percentage missing] for all formation tasks. The optimization solution is to stop searching once a solution is found.

Images