Case-based reasoning optimization method and system for unmanned cluster action strategy

Through the case reasoning optimization method, the similarity between the cases to be executed and historical cases in the unmanned cluster action strategy is calculated, and the key problem attributes and the second global similarity are selected, which improves the recommendation accuracy and effectiveness of the unmanned cluster action strategy. The problem of large errors in the prior art has been solved.

CN118917409APending Publication Date: 2024-11-08NAT UNIV OF DEFENSE TECH
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410827565.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-25
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

Due to the characteristics of autonomy, confrontation and uncertainty, the existing unmanned cluster action strategy recommendation method has large errors in the matching method, and the recommended action strategy is low in effectiveness.

Method used

Using case reasoning optimization method, by calculating the global similarity of multiple problem attributes between the case to be executed and the historical case, selecting the key problem attributes, and calculating the second global similarity, and finally selecting the history with the largest second global similarity Cases are used as recommended cases to guide the action strategies of cases to be implemented.

Benefits of technology

It improves the accuracy of recommended cases and the effectiveness of unmanned cluster action strategies, and enhances the accuracy of selecting similar historical cases.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118917409A_ABST
    Figure CN118917409A_ABST
Patent Text Reader

Abstract

The invention relates to a case-based reasoning optimization method and system for an unmanned cluster action strategy, and the method can quickly and reasonably find a similar historical case as an alternative recommendation case through the comparison of the first global similarity of a plurality of problem attributes with a historical case. Then case optimization is carried out on historical cases selected according to the first global similarities to improve the accuracy of recommended case selection, that is, when at least two first global similarities are in a preset interval, at least two historical cases can be used as alternative cases of to-be-executed cases, and the accuracy of recommendation case selection is improved. And selecting a key problem attribute from the plurality of problem attributes, calculating a second global similarity between the to-be-executed case and the plurality of historical cases on the key problem attribute, and then selecting the historical case with the maximum second global similarity as a recommended case of the to-be-executed case. According to the method, the accuracy of the searched recommended case can be improved, and then the effectiveness of an unmanned cluster action strategy is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of unmanned swarm action, and in particular, to a case-based reasoning optimization method and system for unmanned swarm action strategy. Background Art

[0002] An unmanned swarm refers to a cluster composed of a certain number of similar or heterogeneous unmanned systems / equipment, control systems and human-machine interfaces. An unmanned swarm achieves a leap in the overall capabilities of the swarm through the close collaboration of a single unmanned equipment (such as drones), and has outstanding advantages that are difficult to match with a single unmanned equipment.

[0003] After receiving a task, an unmanned swarm usually needs to determine its action strategy first and formulate an action plan based on the strategy. Currently, there is a method for recommending action strategies for unmanned swarms by matching them with historical action strategies. However, due to the autonomy, confrontation and uncertainty of the actions of unmanned swarms, this matching method has large errors, resulting in low effectiveness of the recommended action strategies. Summary of the invention

[0004] A first aspect of an embodiment of the present application proposes a case-based reasoning optimization method for an unmanned swarm action strategy, the case-based reasoning optimization method for an unmanned swarm action strategy comprising:

[0005] Obtain pending cases and multiple historical cases of the unmanned cluster;

[0006] Calculating a first global similarity between the case to be executed and the multiple historical cases on multiple problem attributes, wherein the problem attributes are used to characterize characteristic attributes of the problems of the cases;

[0007] If at least two of the first global similarities are within the interval between a preset first threshold and a second threshold, a key problem attribute is selected from the multiple problem attributes, and a second global similarity between the case to be executed and the multiple historical cases on the key problem attribute is calculated;

[0008] A historical case with the largest second global similarity is selected as a recommended case for the case to be executed, and an action strategy for the case to be executed is guided according to the action strategy corresponding to the recommended case.

[0009] In some implementations, before selecting the key problem attribute from the multiple problem attributes, the case-based reasoning optimization method for the unmanned swarm action strategy further includes:

[0010] Calculating normalized values ​​of the multiple historical cases on a first problem attribute, where the first problem attribute is any one of the multiple problem attributes;

[0011] Calculate the information entropy of the first question attribute according to the normalized value;

[0012] Calculating a weight value of the first question attribute according to the information entropy;

[0013] The step of selecting a key problem attribute from the multiple problem attributes includes:

[0014] Sorting the weight values ​​of the multiple problem attributes according to size to obtain a sorting result;

[0015] According to the ranking result, at least one of the question attributes with the largest weight value is selected as the key question attribute.

[0016] In some implementations, there are multiple key problem attributes, and calculating the second global similarity between the case to be executed and the multiple historical cases in the key problem attributes includes:

[0017] Calculating the similarity between the case to be executed and the multiple historical cases in each of the key problem attributes;

[0018] According to the similarity and weight values ​​corresponding to the multiple key problem attributes, a second global similarity between the to-be-executed case and the multiple historical cases in terms of the key problem attributes is calculated.

[0019] In some implementations, the second global similarity is calculated using the following formula:

[0020]

[0021] Among them, Sim(S i ,T j ) is the case to be executed T j With historical case S i The second global similarity between i,k ,T j,k ) is the case to be executed T j With historical case S i The similarity between them on the key problem attribute k, S i,k To characterize S i The key problem attribute k, T j,k Used to represent the case to be executed T j The key problem attribute k, ω k is the weight value of the key problem attribute, i≤I,i∈N + , I is the total number of the multiple historical cases, k≤K,k∈N + , K is the total number of the multiple key problem attributes, N + is the set of all positive integers.

[0022] In some implementations, the second threshold is less than the first threshold; after calculating the first global similarity between the case to be executed and the multiple historical cases on multiple problem attributes, the case reasoning optimization method for the unmanned cluster action strategy further includes:

[0023] If there is only one first global similarity greater than the first threshold, the historical case corresponding to the first global similarity greater than the first threshold is used as a recommended case.

[0024] In some implementations, the second threshold is less than the first threshold; after calculating the first global similarity between the case to be executed and the multiple historical cases on multiple problem attributes, the case reasoning optimization method for the unmanned cluster action strategy further includes:

[0025] If there is no case where the first global similarity is greater than the second threshold, no case is recommended to the case to be executed.

[0026] In some implementations, obtaining a plurality of historical cases includes:

[0027] extracting a plurality of historical cases from the historical case database;

[0028] After guiding the action strategy of the case to be executed according to the recommended case, the case reasoning optimization method of the unmanned cluster action strategy further includes:

[0029] The case to be executed is stored as a historical case in the historical case database.

[0030] A second aspect of an embodiment of the present application provides a case-based reasoning optimization device for an unmanned swarm action strategy, the case-based reasoning optimization device for an unmanned swarm action strategy comprising:

[0031] A data acquisition unit, used to acquire pending cases and multiple historical cases of the unmanned cluster;

[0032] A first global similarity calculation unit is used to calculate a first global similarity between the case to be executed and the multiple historical cases on multiple problem attributes, wherein the problem attributes are used to characterize characteristic attributes of the problems of the case;

[0033] A second global similarity calculation unit is configured to select a key problem attribute from the multiple problem attributes if at least two of the first global similarities are within the interval between the first threshold and the second threshold, and calculate a second global similarity between the case to be executed and the multiple historical cases on the key problem attribute;

[0034] The historical case recommendation unit is used to select a historical case with the largest second global similarity as a recommended case for the case to be executed, and guide the action strategy of the case to be executed according to the action strategy corresponding to the recommended case.

[0035] The third aspect of an embodiment of the present application proposes an electronic device, characterized in that it includes at least one controller and a memory for communicating with the controller; the memory stores instructions that can be executed by the at least one controller, and the instructions are executed by at least one of the controllers to enable at least one of the controllers to execute the case reasoning optimization method for the unmanned cluster action strategy as described above.

[0036] A fourth aspect of an embodiment of the present application proposes a computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, and when the computer program is executed, the case reasoning optimization method for the unmanned cluster action strategy as described above is implemented.

[0037] This application proposes a case reasoning optimization method for an unmanned cluster action strategy. This method can quickly and reasonably find similar historical cases as alternative recommended cases by comparing the first global similarity with historical cases on multiple problem attributes, and then optimize the historical cases selected based on the first global similarity to improve the accuracy of the recommended case selection, that is, when at least two first global similarities are within a preset interval, it means that there are at least two historical cases that can be used as alternative cases for the case to be executed, and the key problem attributes are selected from multiple problem attributes, and the second global similarity between the case to be executed and multiple historical cases on the key problem attributes is calculated, and then the historical case with the largest second global similarity is selected as the recommended case for the case to be executed. This method can improve the accuracy of the recommended cases found, and thus improve the effectiveness of the unmanned cluster action strategy.

[0038] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0040] Figure 1 It is a flowchart of case reasoning provided by an embodiment of the present application;

[0041] Figure 2 It is a flowchart of a case-based reasoning optimization method for an unmanned swarm action strategy provided by an embodiment of the present application;

[0042] Figure 3 It is a schematic diagram of a general framework of an unmanned swarm action strategy provided in an embodiment of the present application;

[0043] Figure 4 It is a flowchart of another case-based reasoning optimization method for unmanned swarm action strategy provided by an embodiment of the present application;

[0044] Figure 5 It is a schematic diagram of the recording result of the number of ground facilities destroyed during AT3 simulation provided in an embodiment of the present application;

[0045] Figure 6 It is an architecture diagram of a case-based reasoning optimization device for an unmanned swarm action strategy provided in an embodiment of the present application;

[0046] Figure 7 It is an architectural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0047] 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.

[0048] The following introduces the technical concepts used in this application:

[0049] (1) Case-Based Reasoning (CBR) technology is a knowledge reasoning method that solves problems by imitating human thinking and reasoning patterns. CBR mainly uses historical experience to solve similar current problems. It has unique advantages such as weak dependence on domain knowledge, intuitive case expression, efficient problem solving, and continuous knowledge accumulation.

[0050] Reference Figure 1 At present, the most commonly used case reasoning framework is the 4R framework, which consists of four parts: case retrieval (Retrieve), case reuse (Reuse), case revision (Revise) and case storage (Retain). When the framework is running, the problems to be solved in the real world are first represented as cases to be executed through cases. Then, case retrieval is used to match the historical case with the highest similarity to the current case from the historical cases. If the historical case can be directly used to solve the case to be executed, case reuse is adopted to complete the case reasoning process. Finally, according to the solution and effect of the case to be executed, the case to be executed can be stored as a new case in the historical case library for solving subsequent real-world problems.

[0051] From the perspective of knowledge reasoning, case representation is the knowledge modeling process of real-world problems and their solutions in the form of cases, which has a weak dependence on domain knowledge. Case retrieval, case reuse, and case correction can be understood as the knowledge application process of historical case data, which is both intuitive and efficient. Case preservation is the knowledge accumulation process of case data, thereby continuously enhancing case reasoning capabilities.

[0052] (2) Action strategy: The action strategy of unmanned swarm is an important prerequisite for the formulation of action plans. The action plans under the guidance of different action strategies vary greatly, and multiple different action plans can be formulated under the guidance of the same strategy.

[0053] This application adopts CBR technology to recommend unmanned cluster action strategies, and in the process of selecting recommended cases for cases to be executed, optimizes the historical cases selected based on the first global similarity, that is, when at least two first global similarities are in a preset interval, it means that there are at least two historical cases that can be used as alternative cases for the current case. In order to find recommended cases from the alternative cases, key problem attributes are selected from multiple problem attributes, and the second global similarities between the case to be executed and multiple historical cases on the key problem attributes are calculated, and then a historical case with the largest second global similarity is selected as the recommended case for the case to be executed.

[0054] Reference Figure 2 In one embodiment of the present application, a case-based reasoning optimization method for unmanned swarm action strategy is provided. The method includes steps S110 to S140:

[0055] Step S110: Obtain the to-be-executed cases and multiple historical cases of the unmanned cluster.

[0056] An unmanned swarm may be a swarm composed of drones, a swarm composed of unmanned devices such as robots and robot dogs, or a swarm composed of unmanned devices such as drones, robots and robot dogs. There is no specific limitation here.

[0057] The pending cases refer to the cases that need to be solved by the unmanned cluster. The historical cases are the cases that have been executed historically by the unmanned cluster. Each historical case has a corresponding action strategy. The purpose of this embodiment is to find the recommended cases from multiple historical cases, and then use their action strategies as guidance for the action strategies of the pending cases.

[0058] The action strategy of unmanned swarm is an important prerequisite for the formulation of action plans for unmanned swarm. The action plans under the guidance of different action strategies are quite different, and multiple different action plans can be formulated under the guidance of the same strategy. Generally, the action strategy of unmanned swarm can be designed based on the analysis of the purpose, method, and characteristics of unmanned swarm actions.

[0059] In some implementations, a general framework for unmanned swarm action strategies is designed from three dimensions: action effect, action time, and action cost. Figure 3 As shown. Among them, the action effect usually refers to the effect achieved by using unmanned swarms to perform tasks. It is the primary goal and main embodiment of the value of the use of unmanned swarms. It can be divided into reconnaissance effect, strike effect, and support effect according to the type of unmanned swarms. The action time can be expressed as the duration from the start to the end of the action task. The action time of the unmanned swarm is closely related to the task. Usually, the action time of the strike type is shorter, followed by the reconnaissance type, and the action time of support tasks such as communications is longer. The longer the action time, the higher the risk to the unmanned swarm itself. The action cost refers to the cost of resources consumed in the process of using unmanned swarms to carry out tasks, usually including the use cost and loss cost of unmanned swarms. The use cost includes the normal loss of unmanned swarm equipment and the consumption of ammunition, while the loss cost refers to the loss value of its own equipment due to the confrontational nature of the action. Generally, the low action cost of unmanned swarms is an important advantage over the use of traditional military forces.

[0060] In this embodiment, the general framework of the action strategy of the unmanned swarm designs 6 types of action strategies from three dimensions: action effect, action time, and action cost. Different strategies are suitable for different action types and different stages of the unmanned swarm. The specific corresponding relationships are shown in Table 1.

[0061]

[0062] Table 1

[0063] The effect-first strategy gives priority to the effectiveness of unmanned swarm operations and conducts close-in reconnaissance by increasing the swarm size or reconnaissance payload, thereby significantly improving the reconnaissance effect on key target objects.

[0064] The time priority strategy gives priority to shortening the action time and is mainly suitable for tracking and striking time-sensitive targets.

[0065] The cost-first strategy gives priority to reducing the cost of operations and is usually applicable to routine harassment of conventional targets during a stalemate to ensure the minimization of combat damage to one's own unmanned equipment.

[0066] The effect-averse strategy focuses on weighing the time and cost of the action. It is essentially a combination of the time-priority strategy and the cost-priority strategy, and seeks to achieve limited action effects at the lowest cost and in a very short time. For example, in the early stages of an action, a high-altitude reconnaissance drone can be used to conduct a low-threat survey of the target area.

[0067] The time-aversion strategy focuses on weighing the effect and cost of the action. It is also a combination of the effect-first strategy and the cost-first strategy, seeking to achieve the greatest effect at the lowest cost. For example, ground drones can be used to assist in carrying out cost-effective strike operations against key targets.

[0068] The cost-averse strategy focuses on weighing the effect and time of the action, which can be understood as the joint use of the effect-first strategy and the time-first strategy, in order to achieve the maximum effect of the action in the shortest time without considering the cost consumption. For example, a large-scale reconnaissance and strike drone is used to carry out a saturation strike on time-sensitive targets to ensure the complete destruction of the target.

[0069] It should be noted that the six strategies described above are only one of the implementation methods. In practical applications, they are not limited to the above strategies.

[0070] In some implementations, the case usually includes three parts: problem description, solution, and solution effect. In the unmanned swarm action strategy recommendation problem, the problem description is a standardized representation of the mission, force, scenario, and other information of the unmanned swarm action, the solution is mainly the aforementioned six types of action strategies, and the solution effect includes the goal achievement and the combat damage of the party.

[0071] To facilitate the explanation of case representation, let there be I historical cases in total, and let S = {S i |i≤I,i∈N +}, indicating that there are I historical cases in S, where S i is the i-th historical case. The set of cases to be executed is T = {T j |i≤J,j∈N +}, then there are J problems to be solved, among which T j is the jth case to be executed.

[0072] The problem description part can be expressed in a standardized way by extracting the characteristic attribute set of the problem. Let the problem attribute set be A = {a k |k≤K,k∈N +}, indicating that the number of problem attributes is K, where a k In terms of the feature attribute value, record the historical case S i The problem attribute value is in is the kth problem attribute value of the i-th historical case, and the corresponding problem attribute is recorded as S i,k The pending case T j The problem attribute value is in is the kth problem attribute value of the jth case to be executed, and the corresponding problem attribute is recorded as T j,k .

[0073] The solution part is represented by action strategy. Let the solution attribute set be R = {r l |l≤L,l∈N +}, where r l is the lth attribute of the solution, and the total number of attributes is L. As for the action strategy of the unmanned swarm, the solution attribute is only one action strategy, and the attribute value is the corresponding action strategy. i The scheme attribute value is expressed as That is, one of the six action strategies, including the effect priority strategy.

[0074] The solution effect part represents the actual operation effect of the historical case. The effect attribute set is E = {e u |u≤U,u∈N +}, where e u is the u-th attribute of the scheme effect, and the total number of attributes is U. In the problem of unmanned swarm action, the scheme effect attributes include goal achievement and cost-effectiveness, so the historical case S i The effect attribute value is expressed as In summary, the action cases of unmanned swarm are shown in Table 2.

[0075]

[0076] Table 2

[0077] Step S120: Calculate the first global similarity between the case to be executed and multiple historical cases in multiple problem attributes.

[0078] In the step, the first global similarity is the similarity considering multiple problem attributes. Since the components of the case include: problem description, solution and solution effect. Problem attributes are characteristic attributes of the problem used to characterize the case, the problem description part can be standardized by extracting the characteristic attributes of the problem.

[0079] In some implementations, the similarity between the case to be executed and multiple historical cases in a problem attribute may be calculated first, then a corresponding weight may be set for each problem attribute, and finally a first global similarity may be calculated based on the weight combined with the similarity.

[0080] For example, here we divide the attributes of the unmanned swarm action case into four categories, namely deterministic (such as the number of attacking drones), fuzzy (such as the strength of target facilities), interval (such as target reaction radius) and sub-type (such as battlefield physical weather), and design corresponding attribute similarity calculation models for each type. The following is a detailed introduction:

[0081] Step S1210: Calculate the similarity between the case to be executed and multiple historical cases in a problem attribute.

[0082] For different problem attributes, there are different similarity calculation methods between the case to be executed and multiple historical cases, such as bulldozer distance, Jaccard coefficient, etc., which are not specifically limited here.

[0083] Step S1220: Calculate question attribute weights using the entropy weight method.

[0084] Step S1221, calculating the normalized values ​​of multiple historical cases on the problem attributes, including:

[0085]

[0086] in, For historical case S i The normalized value on problem attribute k, For historical case S i The value of the problem attribute k, min i,k is the minimum value of multiple historical cases on the problem attribute k, max i,k is the maximum value of multiple historical cases on the problem attribute k, i≤I,i∈N + , I is the total number of historical cases, k≤K,k∈N + , K is the total number of problem attributes, N + is the set of all positive integers.

[0087] Step S1222, calculating the information entropy of the problem attribute according to the normalized value, including:

[0088]

[0089] Among them, E k is the information entropy of problem attribute k.

[0090] Step S1223, calculating the weight value of the question attribute according to the information entropy, including:

[0091]

[0092] Among them, ω k is the weight value of the problem attribute k.

[0093] Step S1230: Calculate the first global similarity:

[0094] Here, after obtaining the similarity between the problem attributes of the case to be executed and multiple historical cases, as well as the weight values ​​corresponding to different problem attributes, the first global similarity between the case to be executed and multiple historical cases can be calculated by weighted summation. For example, the first global similarity is calculated by the following formula:

[0095]

[0096] Among them, Sim(S i ,T j ) is the case to be executed T j With historical case S i The first global similarity between i,k ,T j,k ) is the case to be executed T j and historical case S i The similarity between the problem attributes k, S i,k For characterizing historical cases S i The target problem attributes k, T j,k Used to represent the case to be executed T j The problem attribute k.

[0097] Step S130: If at least two first global similarities are within the interval between the preset first threshold and the second threshold, a key problem attribute is selected from the multiple problem attributes, and a second global similarity between the case to be executed and the multiple historical cases on the key problem attribute is calculated.

[0098] In some implementations, after the first global similarity is calculated, a historical case with the highest first global similarity may be found as a recommended case.

[0099] In some embodiments, an interval can be set by two thresholds. When the first global similarity is in this interval, the historical cases corresponding to the first global similarity can be used as candidates for recommended cases. The first threshold and the second threshold can be set according to actual conditions and are not limited here. Any candidate historical case can be selected as a recommended case because its corresponding first global similarity is in a preset interval and meets the requirements. However, considering the strong confrontation and high risk of unmanned swarm actions, a judgment condition is added to the recommended case to select a more reasonable historical case from the candidate historical cases as a recommended case.

[0100] If at least two first global similarities are within the interval between the preset first threshold and the second threshold, the local case optimization operation can be used to select the key problem attributes for secondary retrieval of alternative historical cases to determine relatively better recommended cases. In this embodiment, the key problem attributes are selected from multiple problem attributes, and the second global similarities between the case to be executed and the multiple historical cases on the key problem attributes are calculated, and then the corresponding recommended cases are found according to the second global similarities.

[0101] In some implementations of the present application, the key question attributes are selected in the following manner, including step S210 and step S220:

[0102] Step S210: sort the weight values ​​of multiple question attributes according to their size to obtain a sorting result.

[0103] Step S220: According to the sorting result, at least one question attribute with the largest weight value is selected as the key question attribute.

[0104] This embodiment selects key problem attributes based on the size of the weight value. In the process of obtaining the weight value using formulas (1) to (3), the information entropy of different problem attributes is fully considered, and the weight ratio of different problem attributes is determined by the information entropy. When the first global similarity of multiple historical cases is within the set interval, by narrowing the range of the problem attribute, that is, selecting a certain key problem attribute, the gap between multiple alternative cases can be better judged, and it is convenient to find the most similar historical case from the alternative cases.

[0105] In some implementations of the present application, there are multiple key problem attributes. Step S130 calculates the second global similarity between the case to be executed and multiple historical cases in terms of the key problem attributes, including the following S1310-S1320:

[0106] Step S1310: Calculate the similarity between the case to be executed and multiple historical cases in each key problem attribute.

[0107] Calculate the similarity between the case to be executed and multiple historical cases on each key issue attribute.

[0108] Step S1320: Calculate a second global similarity between the case to be executed and the multiple historical cases in terms of the key problem attributes according to the similarity and the weight values ​​corresponding to the multiple key problem attributes.

[0109] The corresponding weight values ​​can be calculated using the above formulas (1) to (3), and then the second global similarity can be calculated using formula (48), which will not be described in detail here.

[0110] Step S140: Select a historical case with the largest second global similarity as a recommended case for the case to be executed, and guide the action strategy of the case to be executed according to the action strategy corresponding to the recommended case.

[0111] In this embodiment, a historical case with the largest second global similarity is selected as a recommended case for the case to be executed, and the action strategy corresponding to the recommended case can be reused as the action strategy for the case to be executed.

[0112] In this embodiment, case reasoning technology is applied to the recommendation process of unmanned cluster action strategy, and a new action strategy recommendation method is provided. By comparing the first global similarity with historical cases on multiple problem attributes, similar historical cases can be quickly and reasonably found as alternative recommended cases, and then the historical cases selected based on the first global similarity are optimized to improve the accuracy of recommended case selection, that is, when at least two first global similarities are in a preset interval, it means that there are at least two historical cases that can be used as alternative cases for the case to be executed, and the key problem attributes are selected from multiple problem attributes, and the second global similarity between the case to be executed and multiple historical cases on the key problem attributes is calculated, and then the historical case with the largest second global similarity is selected as the recommended case for the case to be executed. This method can improve the accuracy of the recommended cases found, and thus improve the effectiveness of the unmanned cluster action strategy.

[0113] In some embodiments of the present application, the second threshold is less than the first threshold; after calculating the first global similarity between the case to be executed and the multiple historical cases on multiple problem attributes, the case reasoning optimization method of the unmanned cluster action strategy further includes S150:

[0114] Step S150: If there is only one first global similarity greater than a first threshold, the historical case corresponding to the first global similarity greater than the first threshold is used as a recommended case.

[0115] Different from the above embodiment, here a correction condition is added in the case correction process, that is, if only one first global similarity is greater than the first threshold, the historical case corresponding to the first global similarity greater than the first threshold is used as the recommended case. This embodiment does not need to calculate the second global similarity, which reduces the calculation complexity.

[0116] In some embodiments of the present application, the second threshold is less than the first threshold; after calculating the first global similarity between the case to be executed and the multiple historical cases on multiple problem attributes, the case reasoning optimization method of the unmanned cluster action strategy further includes step S160:

[0117] Step S160: If there is no first global similarity greater than the second threshold, the case is not recommended to the case to be executed.

[0118] After the first global similarity is calculated as above, a historical case with the highest similarity can generally be selected as a recommended case. However, if the highest similarity still cannot meet the preset conditions, that is, there is no situation where the first global similarity is greater than the second threshold, then it is considered that the historical case with the highest first global similarity still does not meet the recommendation requirements. In this case, the case is not recommended to the case to be executed, but can be recommended through other means.

[0119] In this embodiment, when the first global similarity is not greater than the second threshold, no recommendation is made to avoid executing an action strategy that deviates greatly from the ideal action strategy.

[0120] Reference Figure 4 In some implementations, three conditions for case modification are provided:

[0121] Condition 1: If the historical case S i Cases to be executed j The first global similarity exceeds the first threshold Right now Then select historical case S i Recommended case.

[0122] Condition two, if It shows that there are at least two historical cases (S i and S i′ ) and the first global similarity of the case to be executed is greater than the second threshold At this time, local case optimization operations can be used to select key problem attributes for secondary retrieval of candidate cases, that is, to calculate the second global similarity (see the introduction of the above embodiment for details) and determine relatively better recommended cases. If necessary, domain experts can adjust and optimize the problem attributes and solution attributes of the recommended cases.

[0123] Condition three, if If the difference between historical cases and cases to be executed is too large, other methods should be adopted to generate unmanned swarm action strategies.

[0124] Usually,

[0125] Summary: The first global similarity with the case to be executed is less than and greater than All historical cases are a set of candidate cases If |BS| = 1, the case is determined to be the final recommended case. Otherwise, the h key problem attributes with the largest weight among all problem attributes are calculated, h <J,h∈N +Then, only h key problem attributes are used to recalculate the second global similarity between the case to be executed and all the candidate cases. Finally, the candidate case with the largest second global similarity is selected as the recommended case. Usually, the local optimization performance of a case is closely related to the number of key problem attributes. Too few key problem attributes cannot reflect the core characteristics of the problem, while too many key problem attributes may lead to unclear local optimization effects. Therefore, h∈[3,5] is recommended.

[0126] A set of experimental examples is provided below:

[0127] (1) Experimental background;

[0128] The scenario is that a Red Army unmanned swarm attacks a Blue Army ground facility. The Red Army's unmanned equipment includes several attack drones and reconnaissance drones, which can form groups to carry out swarm operations according to mission requirements. The Blue Army has deployed several anti-drone equipment for ground facilities. Now it is required to recommend the most appropriate Red Army unmanned swarm action strategy for several typical Blue Army ground facilities and guide the formulation of the Red Army unmanned swarm action plan.

[0129] (2) Experimental analysis;

[0130] Based on the experimental assumptions, we first sorted out the problem attributes of the unmanned swarm action case, which mainly include the red side equipment information, the blue side target information and the action environment information. Specifically, the red side equipment information represents the main force of our side to carry out combat operations, and the selected attributes mainly include the number of offensive drones (a1), the number of reconnaissance drones (a2) and the action radius (a3). The blue side target information reflects the basic situation of the action target object, and the selected attributes mainly include the facility solidity (a4), the number of anti-drone equipment (a5), the number of targets (a6) and the reaction distance (a7). The action environment information reflects the environmental factors that have a greater impact on the unmanned swarm action, and the selected attributes mainly include physical weather (a8) and electromagnetic environment (a9). Among them, a1, a2, a5, and a6 are deterministic problem attributes, and their values ​​are exact values; a3 and a7 are essentially interval distances from 0 to the maximum radius, and should be regarded as interval problem attributes, and their values ​​are interval values; a4 and a9 are usually described qualitatively and their values ​​have a certain continuity, and should be regarded as fuzzy problem attributes, and their values ​​are triangular fuzzy numbers; a8 has classification characteristics, such as sunny (5), cloudy (4), rainy (3), snowy (2), and foggy (1), and should be processed as a categorical problem attribute, and its value should be one of the types. For ease of processing, a4 and a9 need to be graded first. The values ​​of a4 can be fragile (1), poor (2), good (3), and strong (4), and the values ​​of a9 can be poor (1), poor (2), average (3), good (4), and good (5).

[0131] Secondly, the solution attribute of the historical case is the action strategy taken (r1), which is one of multiple action strategies. Here, it is assumed that there are 6 action strategies (effect priority strategy AT1, time priority strategy AT2, cost priority strategy AT3, effect aversion strategy AT4, time aversion strategy AT5, cost aversion strategy AT6), that is, r1∈{AT1,AT2,AT3,AT4,AT5,AT6}. Finally, the operation effect of the solution includes goal achievement (e1) and cost-effectiveness ratio (e2).

[0132] Among them, the target achievement attribute value may be success or failure, but all cases in this experiment are taken as success, that is, e1 = Yes; the cost-effectiveness attribute value is represented by the number of red party drone losses divided by the number of blue party anti-drone equipment and ground facilities, that is, e2∈[0,1].

[0133] To verify the performance of the method, we searched for public drone action cases on the Internet, sorted out the problem and solution attribute values ​​and execution effects, and manually supplemented some of the unpublished data to construct a dataset with 10 historical cases (S1-S 10 ) historical case library. In terms of cases to be executed, four typical blue side ground facility scenarios (T1-T4) are designed, as shown in Table 3.

[0134]

[0135]

[0136] Table 3

[0137] In order to calculate the first global similarity between the four cases to be executed and all historical cases, it is necessary to explain the attribute similarity calculation process. First, a1, a2, a3, a5, a6 and a7 are deterministic and interval problem attributes, which can be calculated according to the above formula. Secondly, a8 is a classification problem attribute. According to the impact of weather on drone operations, the subspace is divided into B1={5,4}, B2={3,2}, B3={1}. Finally, a4 and a9 belong to fuzzy problem attributes, and the calculation process is relatively complicated. Here, a4 is taken as an example, and its value space is a4∈{1,2,3,4}, then t=3. The value of the a4 attribute is represented by a triangular fuzzy number, and the historical case S is given. i Cases to be executed j The similarity calculation results for all possible values ​​of the a4 attribute are shown in Table 4.

[0138]

[0139] Table 4

[0140] Based on this, the first global similarity calculation results between all pending cases and historical cases are given as shown in Table 5. Taking T1 as an example, the similarity results between pending cases and historical cases are given as shown in Table 6. Among them, the problem attribute weight calculated based on the entropy weight method is ω k =(0.1741,0.015,0.174,0.094,0.102,0.164,0.201,0.035,0.04).

[0141]

[0142] Table 5

[0143]

[0144]

[0145] Table 6

[0146] Considering the high risk of unmanned swarm operations, a relatively high threshold is given. Analysis of the first global similarity results shows that the execution cases T1, T2, T3 and T4 have the highest similarity with S9, S7, S2 and S8 respectively, and all of them are greater than But only the first global similarity between T3 and S2 is greater than Therefore, case-by-case optimization of T1, T2, and T4 is required.

[0147] First, construct the candidate case sets BS1 = {S9}, BS2 = {S1, S7}, and BS4 = {S8} respectively. Since |BS1| = |BS4| = 1, it is only necessary to continue to optimize T2 locally. Secondly, given h = 3, select the three key problem attributes with the largest weights from all problem attributes, namely a1, a3, and a7. Then, based on the key problem attributes, recalculate the attributes and the second global similarity of the alternative cases in T2 and BS2. The results are shown in Table 7.

[0148]

[0149] Table 7

[0150] Finally, according to the second global similarity result after local optimization of the case, the best recommended case of T2 is determined to be S1, and the corresponding recommendation strategy is also adjusted from the time-averse strategy to the cost-first strategy. The reason is that T2 and S7 have too large differences in physical weather attributes, and unmanned swarm actions are easily affected by adverse weather such as fog, which may restrict the effectiveness of the action. Therefore, it is more appropriate to adopt a cost-first strategy.

[0151] (3) Experimental conclusions;

[0152] Finally, the recommendation results of the four cases to be executed are summarized in Table 8.

[0153] Cases to be executed Recommended Cases Recommended strategies Policy Type Name <![CDATA[T1]]> <![CDATA[S9]]> AT3 Cost-first strategy <![CDATA[T2]]> <![CDATA[S1]]> AT3 Cost-first strategy <![CDATA[T3]]> <![CDATA[S2]]> AT1 Effect-first strategy <![CDATA[T4]]> <![CDATA[S8]]> AT2 Time priority strategy

[0154] Table 8

[0155] Reference Figure 5 , a set of simulation embodiments are provided below;

[0156] To further verify the practicability of the proposed method, based on AnyLogic simulation platform (8.7.6 University Edition), the Agent modeling method was adopted to establish a simulation model of attacking drones and missiles with the case to be executed T1 as the simulation scenario. Figure 5 shown.

[0157] The main simulation parameters are set as follows:

[0158] (1) Red team simulation parameters: The base deployment position is (0, 0, 0). There are 8 attack drones, the combat radius is 300 km, the flight speed is 60 km / h, each drone can carry up to 4 bombs, and the bomb flight speed is 50 km / h. There are 4 reconnaissance drones, the flight altitude is 8 km, and the flight speed is 60 km / h.

[0159] (2) Blue team simulation parameters: The number of ground facilities is 4, and the deployment positions are (580, 200, 0), (680, 200, 0), (580, 340, 0) and (680, 340, 0). The number of anti-UAV equipment is 2 sets, the reaction distance is 4000m, the deployment positions are (460, 200, 0) and (460, 340, 0), the number of concurrent guided missiles is 2, and the missile flight speed is 80km / h.

[0160] In order to facilitate the simulation of unmanned swarm action, the unmanned swarm action is decomposed according to the typical modes and characteristics of unmanned swarm action and combined with the simulation parameter settings. Then, it is combined in a targeted manner according to the six types of action strategies given in Table 1 to form an action plan under the guidance of the corresponding strategy, as shown in Table 9.

[0161] Strategy Number Action combination Strategy Number Action combination AT1 ①③④⑤⑦⑧⑦ AT4 ④⑥②③⑦ AT2 ④⑤②③⑦ AT5 ①③④⑥⑦⑧⑦ AT3 ①③④⑥⑦ AT6 ④⑤②③⑦⑧⑦

[0162] Table 9

[0163] Among them, the unmanned swarm action includes: ① 4 reconnaissance drones go to the target area for high-altitude reconnaissance. ② Each reconnaissance drone guides a group of attack drones. ③ The reconnaissance drones return to the base. ④ 8 attack drones form groups of 2. ⑤ The attack drone carries 4 bombs. ⑥ The attack drone carries 2 bombs. ⑦ The group returns to the base. ⑧ The group conducts additional strikes on the remaining ground facilities.

[0164] On this basis, 500 Monte Carlo simulations were run and the experimental results are shown in Table 10. The cost-effectiveness ratio is the ratio of the average number of attacking drones lost to the average number of ground facilities destroyed.

[0165]

[0166] Table 10

[0167] In the simulation results, the cost priority strategy (AT3) recommended for T1 performed best in terms of performance indicators such as the average number of lost attacking drones, cost-effectiveness, and average number of bombs consumed, outperforming other action strategies. This result is also significantly better than the average result of all action strategies (representing the possible results of randomly selected strategies), verifying the actual performance of the unmanned swarm action strategy recommendation method proposed in this paper. The record results of the 400th-500th simulation of AT3 on the number of ground facilities destroyed are given as follows Figure 7 As shown, the number of ground facilities destroyed in a single attack was at least 2 and at most 4 (all destroyed), with an average of 3.496.

[0168] In addition, by comparing and analyzing the simulation results in Table 9, it can be found that different action strategies have obvious advantages in focusing on performance indicators. For example, the effect priority strategy (AT1) has the largest number of ground facilities destroyed on average, close to all destroyed (3.986 / 4), and has the best action effect. The effect aversion strategy (AT4) has the shortest average task completion time and the lowest bomb consumption in terms of action cost. The average number of attack drones lost is only 4.082 (ranked third). This result shows that adopting different action strategies has a great impact on the results of unmanned swarm action, and it is of great practical value to recommend unmanned swarm action strategies.

[0169] Reference Figure 5 In one embodiment of the present application, a case-based reasoning optimization device for an unmanned swarm action strategy is provided. The case-based reasoning optimization device for an unmanned swarm action strategy includes:

[0170] A data acquisition unit, used to acquire pending cases and multiple historical cases of the unmanned cluster;

[0171] A first global similarity calculation unit is used to calculate a first global similarity between the case to be executed and the multiple historical cases on multiple problem attributes, wherein the problem attributes are used to characterize characteristic attributes of the problems of the case;

[0172] A second global similarity calculation unit is used to select a key problem attribute from multiple problem attributes if at least two first global similarities are within the interval between the first threshold and the second threshold, and calculate a second global similarity between the case to be executed and the multiple historical cases on the key problem attribute;

[0173] The historical case recommendation unit is used to select a historical case with the largest second global similarity as a recommended case for the case to be executed, and guide the action strategy of the case to be executed according to the action strategy corresponding to the recommended case.

[0174] It should be noted that this embodiment and the above method embodiment are based on the same inventive concept, so the relevant contents of the above method embodiment are also applicable to this embodiment and will not be described in detail here.

[0175] Reference Figure 7 The embodiment of the present application further provides an electronic device, the electronic device comprising:

[0176] at least one memory;

[0177] at least one processor;

[0178] at least one program;

[0179] The program is stored in the memory, and the processor executes at least one program to implement the case-based reasoning optimization method for implementing the above-mentioned unmanned swarm action strategy disclosed in the present invention.

[0180] The electronic device may be any intelligent terminal including a mobile phone, a tablet computer, a personal digital assistant (PDA), a vehicle-mounted computer, etc.

[0181] The electronic device according to the embodiment of the present application is described in detail below.

[0182] The processor 1600 may be implemented by a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application;

[0183] The memory 1700 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1700 can store an operating system and other applications. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the program code is stored in the memory 1700, and the processor 1600 calls and executes the case-based reasoning optimization method for the unmanned cluster action strategy of the embodiment of this application.

[0184] Input / output interface 1800, used to implement information input and output;

[0185] The communication interface 1900 is used to realize the communication interaction between the device and other devices. The communication can be realized through a wired manner (such as USB, network cable, etc.) or a wireless manner (such as mobile network, WIFI, Bluetooth, etc.);

[0186] Bus 2000 , which transmits information between various components of the device (e.g., processor 1600 , memory 1700 , input / output interface 1800 , and communication interface 1900 );

[0187] The processor 1600 , the memory 1700 , the input / output interface 1800 , and the communication interface 1900 are connected to each other in communication within the device via the bus 2000 .

[0188] An embodiment of the present application also provides a storage medium, which is a computer-readable storage medium. The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the case-based reasoning optimization method for the unmanned cluster action strategy described above.

[0189] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0190] The embodiments described in this application are intended to more clearly illustrate the technical solutions of the embodiments of this application and do not constitute a limitation on the technical solutions provided by the embodiments of this application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0191] Those skilled in the art will appreciate that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0192] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0193] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0194] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0195] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0196] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0197] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0198] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0199] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions to enable an electronic device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of each embodiment method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store programs.

[0200] The above is a specific description of the preferred implementation of the embodiments of the present application, but the embodiments of the present application are not limited to the above-mentioned implementation methods. Technical personnel familiar with the field can also make various equivalent modifications or substitutions without violating the spirit of the embodiments of the present application. These equivalent modifications or substitutions are all included in the scope defined by the claims of the embodiments of the present application.

Claims

1. A case-based reasoning optimization method for unmanned swarm action strategy, It is characterized in that The case-based reasoning optimization method for the unmanned swarm action strategy includes: Obtain pending cases and multiple historical cases of the unmanned cluster; Calculating a first global similarity between the case to be executed and the multiple historical cases on multiple problem attributes, wherein the problem attributes are used to characterize characteristic attributes of the problems of the cases; If at least two of the first global similarities are within the interval between a preset first threshold and a second threshold, a key problem attribute is selected from the multiple problem attributes, and a second global similarity between the case to be executed and the multiple historical cases on the key problem attribute is calculated; A historical case with the largest second global similarity is selected as a recommended case for the case to be executed, and an action strategy for the case to be executed is guided according to the action strategy corresponding to the recommended case.

2. According to claim 1, the case-based reasoning optimization method for unmanned swarm action strategy, It is characterized in that Before selecting the key problem attribute from the multiple problem attributes, the case-based reasoning optimization method for the unmanned swarm action strategy further includes: Calculating normalized values ​​of the multiple historical cases on a first problem attribute, where the first problem attribute is any one of the multiple problem attributes; Calculate the information entropy of the first question attribute according to the normalized value; Calculating a weight value of the first question attribute according to the information entropy; The step of selecting a key problem attribute from the multiple problem attributes includes: Sorting the weight values ​​of the multiple problem attributes according to size to obtain a sorting result; According to the ranking result, at least one of the question attributes with the largest weight value is selected as the key question attribute.

3. The case-based reasoning optimization method for unmanned swarm action strategy according to claim 2, It is characterized in that There are multiple key problem attributes, and calculating the second global similarity between the case to be executed and the multiple historical cases on the key problem attributes includes: Calculating the similarity between the case to be executed and the multiple historical cases in each of the key problem attributes; According to the similarity and weight values ​​corresponding to the multiple key problem attributes, a second global similarity between the to-be-executed case and the multiple historical cases in terms of the key problem attributes is calculated.

4. The case-based reasoning optimization method for unmanned swarm action strategy according to claim 3, It is characterized in that The second global similarity is calculated by the following formula: Among them, Sim(S i ,T j ) is the case to be executed T j With historical case S i The second global similarity between i,k ,T j,k ) is the case to be executed T j With historical case S i The similarity between them on the key question attribute k, S i,k To characterize S i The key problem attribute k, T j,k Used to represent the case to be executed T j The key problem attribute k, ω k is the weight value of the key problem attribute, i≤I,i∈N + , I is the total number of the multiple historical cases, k≤K,k∈N + , K is the total number of the multiple key problem attributes, N + is the set of all positive integers.

5. The case-based reasoning optimization method for unmanned swarm action strategy according to claim 1, It is characterized in that The second threshold is less than the first threshold; after calculating the first global similarity between the case to be executed and the multiple historical cases on multiple problem attributes, the case reasoning optimization method of the unmanned cluster action strategy also includes: If there is only one first global similarity greater than the first threshold, the historical case corresponding to the first global similarity greater than the first threshold is used as a recommended case.

6. The case-based reasoning optimization method for unmanned swarm action strategy according to claim 1, It is characterized in that The second threshold is less than the first threshold; after calculating the first global similarity between the case to be executed and the multiple historical cases on multiple problem attributes, the case reasoning optimization method of the unmanned cluster action strategy also includes: If there is no case where the first global similarity is greater than the second threshold, no case is recommended to the case to be executed.

7. The case-based reasoning optimization method for unmanned swarm action strategy according to claim 1, It is characterized in that Get multiple historical cases including: extracting a plurality of historical cases from the historical case database; After guiding the action strategy of the case to be executed according to the recommended case, the case reasoning optimization method of the unmanned cluster action strategy further includes: The case to be executed is stored as a historical case in the historical case database.

8. A case-based reasoning optimization device for unmanned swarm action strategies. It is characterized in that The case-based reasoning optimization device for the unmanned swarm action strategy includes: A data acquisition unit, used to acquire pending cases and multiple historical cases of the unmanned cluster; A first global similarity calculation unit is used to calculate a first global similarity between the case to be executed and the multiple historical cases on multiple problem attributes, wherein the problem attributes are used to characterize characteristic attributes of the problems of the case; A second global similarity calculation unit is configured to select a key problem attribute from the multiple problem attributes if at least two of the first global similarities are within the interval between the first threshold and the second threshold, and calculate a second global similarity between the case to be executed and the multiple historical cases on the key problem attribute; The historical case recommendation unit is used to select a historical case with the largest second global similarity as a recommended case for the case to be executed, and guide the action strategy of the case to be executed according to the action strategy corresponding to the recommended case.

9. An electronic device, It is characterized in that It includes at least one controller and a memory for communicating with the controller; the memory stores instructions that can be executed by the at least one controller, and the instructions are executed by at least one of the controllers to enable at least one of the controllers to execute the case reasoning optimization method for the unmanned swarm action strategy as described in any one of claims 1 to 7.

10. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, which, when executed, implements the case-based reasoning optimization method for the unmanned swarm action strategy according to any one of claims 1 to 7.

Citation Information

Cited By

  • Action scheme recommendation method and system based on action dependence partial order

    CN119939162A

  • Machine learning-based case reasoning method and related equipment

    CN120743966A