Method for identifying collaborative pursuit action style of two-to-one pursuit game

Abstract

The invention discloses a method, a system and equipment for identifying a collaborative pursuit action pattern of a two-to-one pursuit game, and belongs to the technical field of spaceflight control. The method comprises the following steps: acquiring state information of a pursuit and an escaper according to a pursuit game track; calculating a relative position included angle and a relative speed included angle between the pursuit and the escaper according to the state information of the pursuit and the escaper; according to the relative position included angle and the relative speed included angle between the pursuit and the escaper, defining and classifying different pursuit action styles, and obtaining independent pursuit action styles of a one-to-one pursuit game; and performing combination and arrangement based on the independent pursuit action styles of the one-to-one pursuit game, and constructing a collaborative pursuit action style library of the two-to-one pursuit game. According to the method, one-to-one and two-to-one cooperative pursuit action pattern libraries are constructed, targeted action sets can be provided for different pursuit scenes, systematic understanding of the whole pursuit game process is achieved, and formulation and implementation of an efficient pursuit strategy are achieved.

Description

Technical Field

[0001] The present invention relates to the field of aerospace control technology, and particularly relates to a method, system, device, medium and program for identifying a cooperative pursuit action pattern in a two-on-one pursuit-evasion game. Background Art

[0002] In the field of pursuit-evasion games, which is full of strategies and challenges, its complexity and dynamics increase significantly with the increase in the number of participants. Especially in the scenario of a two-on-one pursuit-evasion game, the difficulty and depth of this strategic problem reach a new height. Each of the two pursuers not only needs to possess excellent pursuit-evasion skills and strategy-making abilities but also must be able to achieve efficient and close cooperation in real-time confrontation. This cooperation not only requires seamless communication and information sharing between the pursuers but also the ability to quickly and flexibly adjust their pursuit-evasion strategies and actions in the face of the constantly changing strategies and behaviors of the evader.

[0003] However, in the current research on two-on-one pursuit-evasion games of particles in a plane, there is a significant limitation: a complete and systematic action pattern library has not been established. As an important tool for understanding and optimizing the pursuit-evasion game process, the absence of the action pattern library makes it difficult to form a systematic understanding of the overall game behavior when analyzing the game results. Without such a library, it is impossible to comprehensively and deeply grasp various possible actions and strategy combinations in the pursuit-evasion game, nor can the evolution law of the game be accurately predicted and grasped.

[0004] The existence of this limitation not only hinders the understanding of the essence of the pursuit-evasion game but also severely restricts the formulation and implementation of efficient pursuit-evasion strategies. Summary of the Invention

[0005] Aiming at the problem in the prior art that a complete and systematic two-on-one pursuit-evasion game action pattern library has not been established, which restricts the formulation and implementation of efficient pursuit-evasion strategies. The present invention provides a method for identifying a cooperative pursuit action pattern in a two-on-one pursuit-evasion game, constructs an action pattern library for one-on-one and two-on-one cooperative pursuits, can provide a targeted action set for different pursuit-evasion scenarios, realizes a systematic understanding of the overall pursuit-evasion game process, and achieves the formulation and implementation of efficient pursuit-evasion strategies.

[0006] To achieve the above object, the present invention provides the following technical solutions.

[0007] In the first aspect, the present invention provides a method for identifying a cooperative pursuit action pattern in a two-on-one pursuit-evasion game, including: Obtaining the state information of the pursuer and the escapee according to the pursuit-evasion game trajectory; Calculating the relative position angle and relative velocity angle between the pursuer and the escapee according to the state information of the pursuer and the escapee; Define and classify different pursuit action patterns according to the relative position angle and relative velocity angle between the pursuer and the escapee, and obtain the individual pursuit action patterns of the one-on-one pursuit-evasion game; Based on the individual pursuit action patterns of the one-on-one pursuit-evasion game, perform combinatorial permutations to construct a collaborative pursuit action pattern library for the two-on-one pursuit-evasion game; Through the collaborative pursuit action pattern library of the two-on-one pursuit-evasion game, identify the behavioral characteristics of the pursuer and the escapee in the game, and obtain the behavioral patterns of the pursuer and the escapee in the game.

[0008] As a further improvement of the present invention, obtaining the state information of the pursuer and the escapee according to the pursuit-evasion game trajectory includes: Based on the existing pursuit-evasion game trajectory, obtain each point on the existing pursuit-evasion game trajectory; Based on each point on the existing pursuit-evasion game trajectory, obtain the state information of the pursuer and the escapee; The state information of the pursuer and the escapee includes: the pursuer 's position ; the pursuer 's speed ; the position of the escapee and the speed of the escapee .

[0009] As a further improvement of the present invention, calculating the relative position angle and relative velocity angle between the pursuer and the escapee according to the state information of the pursuer and the escapee includes: According to the state information of the pursuer and the escapee, calculate the position angle of the pursuer, the speed angle of the pursuer, and the speed angle of the escapee; According to the position angle of the pursuer, the speed angle of the pursuer, and the speed angle of the escapee, calculate the relative position angle and relative velocity angle between the pursuer and the escapee:

[0010] where is the relative position angle between the pursuer and the escapee; is the relative velocity angle between the pursuer and the escapee; is the speed angle of the escapee; is the speed angle of the pursuer; is the position angle of the pursuer.

[0011] As a further improvement of the present invention, different pursuit action styles are defined and classified according to the relative position angle and relative velocity angle between the pursuer and the escapee, and the individual pursuit action styles of the one-on-one pursuit-evasion game are obtained, including driving, chasing, intercepting, outflanking, meeting an attack, and ambushing;

[0012] In the formula, is the relative position angle between the pursuer and the escapee, is the relative velocity angle between the pursuer and the escapee.

[0013] As a further improvement of the present invention, the individual pursuit action styles based on the one-on-one pursuit-evasion game are combined and arranged to construct a cooperative pursuit action style library for the two-on-one pursuit-evasion game, including: Based on the individual pursuit action styles of the one-on-one pursuit-evasion game, the individual pursuit action styles are combined in pairs to construct a cooperative pursuit action style library for the two-on-one pursuit-evasion game; The cooperative pursuit action style library for the two-on-one pursuit-evasion game includes chasing, expelling, intercepting, sandwiching, surrounding, trapping, meeting an attack, blocking, pincer attack, ambushing, outflanking, outflanking, driving, blocking, surrounding and chasing, besieging, covering the rear, intercepting, cutting off, surrounding, and surrounding.

[0014] As a further improvement of the present invention, through the cooperative pursuit action style library for the two-on-one pursuit-evasion game, the behavioral characteristics of the pursuer and the escapee in the game are identified, and the behavioral patterns of the pursuer and the escapee in the game are obtained, including: Obtain the actions of the pursuer and the escapee during the game; Through the cooperative pursuit action style library for the two-on-one pursuit-evasion game, the actions of the pursuer and the escapee during the game are identified to form an action sequence during the game; Taking the action sequence during the game as a sample for clustering, the behavioral characteristics of the pursuer and the escapee in the game are analyzed.

[0015] In the second aspect, the present invention provides a recognition system for the cooperative pursuit action styles of the two-on-one pursuit-evasion game, including: A status information acquisition module: used to obtain the status information of the pursuer and the escapee according to the pursuit-evasion game trajectory; An angle calculation module: used to calculate the relative position angle and relative velocity angle between the pursuer and the escapee according to the status information of the pursuer and the escapee; An action style library acquisition module: used to define and classify different pursuit action styles according to the relative position angle and relative velocity angle between the pursuer and the escapee, and obtain the individual pursuit action styles of the one-on-one pursuit-evasion game; Build Style Library Module: It is used to combine and arrange individual pursuit action styles based on one-on-one pursuit-evasion games to build a cooperative pursuit action style library for two-on-one pursuit-evasion games; Behavior Pattern Recognition Module: It is used to identify the behavioral characteristics of pursuers and evaders in the game through the cooperative pursuit action style library of two-on-one pursuit-evasion games, and obtain the behavior patterns of pursuers and evaders in the game.

[0016] In a third aspect, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the method for identifying cooperative pursuit action styles of a two-on-one pursuit-evasion game are implemented.

[0017] In a fourth aspect, the present invention provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for identifying cooperative pursuit action styles of a two-on-one pursuit-evasion game are implemented.

[0018] In a fifth aspect, the present invention provides a computer program product, including computer instructions, and when the computer instructions are executed by a processor, the steps of the method for identifying cooperative pursuit action styles of a two-on-one pursuit-evasion game are implemented.

[0019] Compared with the prior art, the present invention has the following beneficial effects: By deeply analyzing the state information of pursuers and evaders in pursuit-evasion games, the relative position angle and relative speed angle between pursuers and evaders, the present invention accurately captures the real-time situation of both sides in dynamic games. On this basis, the present invention expands it from a simple one-on-one scenario to a complex cooperative situation of two-on-one. Through the refined sorting and classification of individual pursuit action styles in one-on-one pursuit-evasion games, and then combined with the flexible combination and arrangement of these basic action styles, a comprehensive and systematic cooperative pursuit action style library for two-on-one pursuit-evasion games is successfully constructed. The establishment of the cooperative pursuit action style library enriches the choice space of pursuit-evasion strategies, and at the same time makes the action selection for different pursuit-evasion scenarios more targeted and effective. More importantly, the action style library constructed by the present invention can also accurately identify and extract the behavioral characteristics of pursuers and evaders during the game process. This ability not only helps to deeply understand the internal mechanism of pursuit-evasion games, but also provides a scientific basis and strong support for formulating and implementing efficient pursuit-evasion strategies. Description of the Drawings

[0020] The drawings described herein are for illustrative purposes only and are not intended to limit the scope of the present invention disclosure in any way. In the drawings: Figure 1Schematic flow chart of the method for identifying the collaborative pursuit action patterns in a two-on-one pursuit-evasion game according to the present invention; Figure 2 Classification diagram of one-on-one pursuit patterns for the method for identifying the collaborative pursuit action patterns in a two-on-one pursuit-evasion game according to the present invention; Figure 3 Schematic diagram of one-on-one pursuit patterns for the method for identifying the collaborative pursuit action patterns in a two-on-one pursuit-evasion game according to the present invention; Figure 4 Schematic diagram of two-on-one collaborative pursuit patterns for the method for identifying the collaborative pursuit action patterns in a two-on-one pursuit-evasion game according to the present invention; Figure 5 Elbow method for the optimal K value in the embodiments of the present invention; Figure 6 Schematic diagram of behavior clustering and visualization in the embodiments of the present invention; Figure 7 Motion trajectories of the pursuer and the evader corresponding to each category in the embodiments of the present invention; Figure 8 Schematic structural diagram of the system for identifying the collaborative pursuit action patterns in a two-on-one pursuit-evasion game according to the present invention; Figure 9 Schematic diagram of the electronic device in the embodiments of the present invention. Detailed implementation manners

[0021] In order to enable those skilled in the art of the present technology to better understand the technical solutions in the present invention, the technical solutions in the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the present invention. The described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0023] Aiming at the problem in the prior art that a complete and systematic two-on-one pursuit-evasion game action pattern library has not been established, which restricts the formulation and implementation of efficient pursuit-evasion strategies. The present invention provides a method for identifying the collaborative pursuit action patterns in a two-on-one pursuit-evasion game, as Figure 1 shown, the method includes: S100: Obtain the state information of the pursuer and the evader according to the pursuit-evasion game trajectory; S200: Calculate the relative position angle and relative velocity angle between the pursuer and the escapee according to the status information of the pursuer and the escapee. S300: Define and classify different pursuit action patterns according to the relative position angle and relative velocity angle between the pursuer and the escapee, and obtain the individual pursuit action patterns of the one-on-one pursuit-evasion game. S400: Based on the combined arrangement of the individual pursuit action patterns of the one-on-one pursuit-evasion game, construct a cooperative pursuit action pattern library for the two-on-one pursuit-evasion game. S500: Identify the behavioral characteristics of the pursuer and the escapee in the game through the cooperative pursuit action pattern library of the two-on-one pursuit-evasion game, and obtain the behavioral patterns of the pursuer and the escapee in the game.

[0024] The present invention constructs a one-on-one and two-on-one cooperative pursuit action pattern library, which can provide a targeted action set for different pursuit-evasion scenarios, realize a systematic understanding of the overall pursuit-evasion game process, and achieve the formulation and implementation of an efficient pursuit-evasion strategy.

[0025] The following further explains the present invention with reference to specific drawings.

[0026] The present invention provides a method for identifying cooperative pursuit action patterns of a two-on-one pursuit-evasion game, including the following steps: S1: According to the existing pursuit-evasion game trajectory, obtain the status information of the pursuer and the escapee at each point on the trajectory, including the position and speed of the pursuer and the position and speed of the escapee, and then calculate the relative position angle and relative velocity angle between each pursuer and the escapee.

[0027] First, calculate the position angle , speed angle of the pursuer and the speed angle of the escapee as follows: (1) (2) (3) Then, calculate the relative position angle and relative velocity angle of the pursuer relative to the escapee from the following formula: (4) S2: Define the individual pursuit action patterns of the one-on-one pursuit-evasion game.

[0028] Define a single pursuit action pattern for one-on-one pursuit-evasion games according to the relative position angle and relative velocity angle between the pursuer and the evader, including six types: driving, chasing, intercepting, outflanking, meeting, and ambushing. As Figure 2 shown. The position of the evader is at the coordinate origin, and its velocity direction always points to the axis. The classification method of each pursuit action according to angle constraints is shown in formula (1): Among them, represents the angle between the position vector of the pursuer relative to the evader and the positive direction of the axis, represents the angle between the velocity direction vector of the pursuer and the positive direction of the axis. The specific schematic diagram of each pursuit action is as Figure 3 shown. The definition of each pursuit action is as follows: 1) Driving action: Force the target to retreat or be driven away from a certain area. The pursuer approaches the evader from the reverse half-plane of the evader's velocity direction, and the pursuer's velocity direction points to the rear of the evader.

[0029] 2) Chasing action: Quickly follow and approach the target. The pursuer approaches the evader from the reverse half-plane of the evader's velocity direction, and the pursuer's velocity direction always points to the evader within a certain error range.

[0030] 3) Intercepting action: Intercept or prevent the target from reaching a predetermined location on its movement path. The pursuer approaches the evader from the reverse half-plane of the evader's velocity direction, and the pursuer's velocity direction points to the front of the evader.

[0031] 4) Outflanking action: Move to the rear side of the target for attack. The pursuer approaches the evader from the half-plane where the evader's velocity direction is located, and the pursuer's velocity direction points to the rear of the evader.

[0032] 5) Meeting action: Actively approach the target and attack. The pursuer approaches the evader from the half-plane where the evader's velocity direction is located, and the pursuer's velocity direction always points to the evader within a certain error range.

[0033] 6) Ambushing action: Pre-ambush on the possible route of the target and suddenly launch an attack when the target arrives. The pursuer approaches the evader from the half-plane where the evader's velocity direction is located, and the pursuer's velocity direction points to the front of the evader.

[0034] S3: Construct a collaborative pursuit action pattern library for two-on-one pursuit-evasion games.

[0035] Based on individual pursuit action styles, pair them up to define two-on-one collaborative pursuit action styles, including chasing, expelling, intercepting, sandwiching, surrounding, trapping, meeting an attack, blocking, pincer attack, ambushing, outflanking, outmaneuvering, driving away, blocking, encircling in pursuit, besieging, covering the rear, intercepting, cutting off, surrounding, and besieging, a total of 21 kinds. As Figure 4 shown, there is no order between the two pursuers. When they swap roles, the resulting collaborative action remains unchanged.

[0036] Correspond the calculated pursuit actions of the two pursuers according to the collaborative pursuit style library, and the resulting action is the collaborative pursuit style in this state.

[0037] S4: Identify and analyze the behavioral characteristics emerging in the two-on-one pursuit-evasion game.

[0038] The trajectory evolution pattern of the pursuer and the evader throughout the game is called "behavior", that is, an ordered combination of multiple actions. By taking a series of complete game trajectories as statistical samples, statistical methods such as cluster analysis can be used to identify different behavioral patterns and collaborative strategies.

[0039] In each game scenario, the trajectories of the pursuer and the evader form a set of image data, and the image pixel points here can be represented by the position data of the three, so algorithms for image recognition and classification can be applied for cluster analysis. The K-means algorithm of unsupervised learning is used for clustering of pursuit behaviors. Its working principle is based on iterative optimization, dividing the data points into K clusters, so that each data point belongs to the nearest cluster, and the center (centroid) of each cluster is the average value of all data points belonging to that cluster. Finally, analyze the characteristics of each type of behavior after clustering.

[0040] The main steps of the Kmeans algorithm are as follows: (1) Determine the number of clusters: Before starting clustering, it is necessary to pre-specify the number K of clusters, and this value can be determined by the elbow method. Plot the curve of the sum of squared errors (SSE) within the clusters corresponding to different K values, and select the turning point similar to the "elbow" as the optimal K value. Suppose there are 𝑘 clusters, and the center of each cluster is , and the data points within the cluster are . Then the calculation method of SSE is as follows: (6) Among them, represents the th cluster, is the The center of each cluster. As the value of K increases, the SSE usually decreases because more clusters mean that the data points within each cluster are closer together. However, when the value of K increases beyond a certain point, the reduction in SSE becomes less significant, forming a turning point similar to an "elbow". This turning point is considered the optimal value of K.

[0041] (2) Initialize the center points: Randomly select K data points as the initial cluster centers to optimize the clustering results.

[0042] (3) Assign data points to the nearest cluster center: For each game trajectory data point, calculate its distance from each cluster center and assign the data point to the nearest cluster center to form K clusters.

[0043] (4) Update the cluster centers: For each cluster, calculate the mean of all data points within the cluster and then set this mean as the new cluster center.

[0044] (5) Repeat the assignment and update steps: Repeat steps (3) and (4) until a certain termination condition is met. Usually, the termination condition can be reaching a preset number of iterations, the change in cluster centers being less than a certain threshold, or the change in data points within the cluster no longer being significant.

[0045] After clustering is completed, perform result visualization using the t-SNE algorithm for visualizing the clustering results. The algorithm calculates the similarity between data points in a high-dimensional space and preserves this similarity in a low-dimensional space (usually two or three dimensions).

[0046] In summary, the present invention provides a method for identifying cooperative pursuit action patterns in a two-on-one pursuit-evasion game. By clearly defining pursuit actions, it can provide a clear theoretical basis for the result analysis of pursuit-evasion game problems. This framework helps researchers better understand and analyze the dynamic processes of pursuit and evasion, thereby designing more efficient pursuit strategies. The pursuit actions defined in the present invention contribute to the effective cooperation among pursuers. The present invention is no longer limited to a single pursuit-evasion scenario or a simplified model. In certain fields, such as bionics, aerospace, and military tactics, the definition of pursuit actions can promote the development of related technologies.

[0047] The following further explains and illustrates the present invention with specific embodiments.

[0048] Embodiment To demonstrate the use of the method for constructing a cooperative pursuit action pattern library based on the two-on-one pursuit-evasion game described in the present invention, take the scenario of a two-on-one pursuit-evasion game of mass points occurring in a grid model as an example to illustrate the use of the pattern library. Use an intelligent learning algorithm or an optimization algorithm to obtain 1000 sets of two-on-one pursuit trajectories as shown in Figure 5 the following.

[0049] The specific implementation steps of the present invention are given below: S1: According to the existing pursuit-evasion game trajectories, obtain the state information of the pursuer and the escapee at each point, including the position and speed of the pursuer and the position and speed of the escapee, and then calculate the relative position angle and relative speed angle between each pursuer and the escapee.

[0050] First, calculate the position angle , speed angle of the pursuer and the speed angle of the escapee as follows: (7) (8) (9) Then, calculate the relative position angle and relative speed angle between the pursuer and the escapee from the following formula: (10) Statistically analyze the position and speed information of the two pursuers and the escapee at each moment in the obtained 1000 groups of trajectories, and calculate the relative position angle and relative speed angle between the two pursuers and the escapee at each moment.

[0051] S2: Define the individual pursuit action styles for one-on-one pursuit-evasion games.

[0052] Define the individual pursuit action styles for one-on-one pursuit-evasion games according to the relative position angle and relative speed angle between the pursuer and the escapee, including six types: driving, chasing, intercepting, outflanking, meeting head-on, and ambushing. As Figure 2 shown. The position of the escapee is at the origin of the coordinate system, and its speed direction always points to the axis. The classification method of each pursuit action according to the angle constraint is shown in formula (1): (11) Among them, represents the angle between the position vector of the pursuer relative to the escapee and the positive direction of the axis, represents the angle between the speed direction vector of the pursuer and the positive direction of the axis. The specific schematic diagrams of each pursuit action are shown in Figure 3 . The definitions of each pursuit action are as follows: (1) Driving action: Forcibly drive the target away or retreat from a certain area. The pursuer approaches the escapee from the reverse half-plane of the direction of the escapee's velocity, and the direction of the pursuer's velocity points behind the escapee.

[0053] (2) Chasing action: Quickly follow and approach the target. The pursuer approaches the escapee from the reverse half-plane of the direction of the escapee's velocity, and the direction of the pursuer's velocity always points to the escapee within a certain error range.

[0054] (3) Interception action: Intercept or prevent the target from reaching a predetermined location on its movement path. The pursuer approaches the escapee from the reverse half-plane of the direction of the escapee's velocity, and the direction of the pursuer's velocity points in front of the escapee.

[0055] (4) Encirclement action: Move to the rear side of the target to attack. The pursuer approaches the escapee from the half-plane where the direction of the escapee's velocity is located, and the direction of the pursuer's velocity points behind the escapee.

[0056] (5) Confrontation action: Actively approach the target and attack. The pursuer approaches the escapee from the half-plane where the direction of the escapee's velocity is located, and the direction of the pursuer's velocity always points to the escapee within a certain error range.

[0057] (6) Ambush action: Lie in wait in advance on the route where the target may pass, and suddenly launch an attack when the target arrives. The pursuer approaches the escapee from the half-plane where the direction of the escapee's velocity is located, and the direction of the pursuer's velocity points in front of the escapee.

[0058] According to the calculated relative position angle between the two pursuers and the escapee and the relative velocity angle

[0059] which corresponds one-to-one with the one-on-one pursuit action patterns, determine the pursuit actions of the two pursuers.

[0060] S3: Construct a cooperative pursuit action pattern library for the two-on-one pursuit-evasion game. Figure 4 As shown, there is no order between the two pursuers. When the two swap roles, the resulting cooperative action remains unchanged.

[0061] Corresponding the calculated pursuit actions of the two pursuers according to the cooperative pursuit pattern library, the resulting action is the cooperative pursuit pattern in this state.

[0062] As shown in Table 1, the horizontal axis and the vertical axis are the individual pursuit actions of two pursuers respectively. The corresponding table is the two-on-one collaborative pursuit action. The numbers in the table are the occurrence times of this collaborative pursuit action. It can be seen that when using this algorithm for solution, the situation where one pursuer implements a driving action and the other pursuer implements an interception action, or both pursuers use driving or interception actions has a relatively high proportion. That is, the pursuit effect is better when the two pursuers use collaborative driving, collaborative interception or collaborative blocking actions. As shown in Table 1, it can be used more frequently during the game process.

[0063] Table 1 Action Statistics Table

[0064] S4: Classify behaviors and analyze behavioral characteristics.

[0065] Statistics of individual pursuit actions can show the situation of the game at a specific moment. For example, analyze the coordination of different collaborative pursuit actions and count the usage times of each action during the game process. However, analyzing only a single action often provides only a limited perspective and cannot deeply explore its impact on the evolution and result of the game. Because it ignores the evolution of strategies and the development of overall collaborative behaviors during the game process. Therefore, it is necessary to shift from the actions at a single moment to the behaviors throughout the game process. Here, the K-means algorithm of unsupervised learning is used to cluster pursuit behaviors. Finally, analyze the characteristics of each class of behaviors after clustering.

[0066] The main steps of the Kmeans algorithm are as follows: (1) Determine the number of clusters: Before starting clustering, it is necessary to pre-specify the number K of clusters, and this value can be determined by the elbow method. Draw a curve of the within-cluster error sum (Sum of Squared Errors, SSE) corresponding to different K values, and select the turning point similar to the "elbow" as the optimal K value. Suppose there are 𝑘 clusters, and the center of each cluster is , and the data points within the cluster are . Then the calculation method of SSE is as follows: (12) Among them, represents the th cluster, and is the center of the th cluster. Figure 5 is the change curve of SSE with the K value. When the K value increases to 4, the decrease of SSE becomes less significant, forming a turning point similar to the "elbow", and this turning point is the optimal K value.

[0067] (2) Initialize the center points: Randomly select K data points as the initial cluster centers to optimize the clustering results.

[0068] (3) Assign data points to the nearest cluster center: For each game trajectory data point, calculate its distance from each cluster center and assign the data point to the nearest cluster center to form K clusters.

[0069] (4) Update the cluster center: For each cluster, calculate the mean of all data points within the cluster and then set this mean as the new cluster center.

[0070] (5) Repeat the assignment and update steps: Repeat steps (3) and (4) until a certain termination condition is met. Usually, the termination condition can be reaching a preset number of iterations, the change in the cluster center being less than a certain threshold, or the change in the data points within the cluster no longer being significant.

[0071] After determining the number of clusters, clustering and visualization analysis of the pursuit behavior can be carried out. As Figure 6 shown, the data points of the same color in the figure represent an emerging type of behavior, and the separation of the positions where the data points are located represents the magnitude of the difference in behavior types. Data of the same color are assigned to nearby positions, and the boundary between different categories is relatively clear, indicating a good clustering effect. By plotting the game trajectories corresponding to the data points in each color, four scenarios with similar behaviors can be obtained, as Figure 7 shown. Summarize the behavioral characteristics in the four scenarios as follows: Phenomenon 1: The lazy behavior of the pursuers When lazy pursuers appear, they may keep moving forward in a certain direction or wander back and forth near a certain location, similar to random movement. Since the rewards of the two pursuers are shared in the set reward function, and the success of any pursuer in catching up represents the victory of the game, the emerging lazy behavior is reasonable. In teamwork, individuals may reduce their efforts because they feel that their contributions are not evaluated or rewarded separately, and this phenomenon is called the social loafing effect.

[0072] Phenomenon 2: The approaching and surrounding behavior at the boundary or corner The pursuers will drive the escapee to the boundary until they are captured. When the escapee is driven to the boundary, they will wander back and forth on the boundary but cannot escape from the boundary. If the escapee tries to escape towards the center, they will accelerate their capture by the pursuers. This phenomenon is similar to the way dolphins hunt in nature. Dolphins use sound waves and body movements to conduct team hunting. They often form a circle, drive the fish school towards the sea surface, and then catch them one by one. They also use the sediment or seaweed on the seabed as tools to trap the prey and increase the capture rate.

[0073] Phenomenon 3: The helplessness of the escapee When the escapee is forced to the boundary and faces a situation where there is no way to escape, they often fall into a state of helplessness. Near the boundary, the actions of the escapee become chaotic. They start to wander back and forth in a narrow space, trying to confuse the pursuers through irregular movements and delay the time of being captured. However, due to the limited space and the approaching of the pursuers, they have actually fallen into a disorderly and ineffective cycle. As time goes by, the pursuers gradually narrow the encirclement, and the range of choices for the escapee becomes more and more limited, and finally they are captured.

[0074] The second object of the present invention is to propose a recognition system for the collaborative pursuit action pattern of a two-on-one pursuit-evasion game, as Figure 8 shown, including: The status information acquisition module 100: used to acquire the status information of the pursuer and the escapee according to the pursuit-evasion game trajectory; The included angle calculation module 200: used to calculate the relative position included angle and the relative speed included angle between the pursuer and the escapee according to the status information of the pursuer and the escapee; The action pattern library acquisition module 300: used to define and classify different pursuit action patterns according to the relative position included angle and the relative speed included angle between the pursuer and the escapee, and acquire the individual pursuit action patterns of the one-on-one pursuit-evasion game; The pattern library construction module 400: used to perform combination permutations based on the individual pursuit action patterns of the one-on-one pursuit-evasion game to construct a collaborative pursuit action pattern library for the two-on-one pursuit-evasion game; The behavior pattern recognition module 500: used to identify the behavior characteristics of the pursuer and the escapee in the game through the collaborative pursuit action pattern library of the two-on-one pursuit-evasion game, and obtain the behavior patterns of the pursuer and the escapee in the game.

[0075] As Figure 9 shown, the third object of the present invention is to provide an electronic device, which includes: a processor 601, a memory 602, and a display screen 603. Among them, the memory 602 and the display screen 603 are both connected to the processor 601, such as through a bus 604. Optionally, the electronic device may further include a transceiver 605. It should be noted that in practical applications, the transceiver 605 is not limited to one, and the structure of the electronic device does not constitute a limitation to the embodiments of the present application.

[0076] The processor 601 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logical blocks, modules, and circuits described in connection with the disclosure of this application. The processor 601 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0077] The bus 604 may include a path for transmitting information between the above components. The bus 604 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus 604 may be divided into an address bus, a data bus, a control bus, etc.

[0078] The memory 602 may be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or it may also be an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory), or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), magnetic disk storage media, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

[0079] The memory 602 is used to store the application program code for implementing the solution of this application and is controlled by the processor 601 for execution. The processor 601 is used to execute the application program code stored in the memory 602 to implement the content shown in the foregoing method embodiments.

[0080] Figure 9 The illustrated electronic device is merely an example and should not impose any limitations on the functions and scope of use of the embodiments of the present application.

[0081] The fourth object of the present invention is to provide a computer-readable storage medium storing a computer program, on which a computer program is stored, and when the program is executed by a processor, it implements each process of the method embodiment as described above. Figure 1 For example, a memory including instructions, and the above instructions can be executed by the processor of the electronic device to complete the above method.

[0082] A computer-readable storage medium can be a tangible device that holds and stores instructions used by an instruction execution device. A computer-readable storage medium can be, but is not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination of the above. Specifically, a computer-readable storage medium can be a portable computer disk, a hard disk, a USB flash drive, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a podium random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, an optical disk, a magnetic disk, a mechanical encoding device, and any combination of the above.

[0083] The fifth object of the present invention is to provide a computer program product including computer instructions, and when the computer instructions are executed by a processor, they implement each process of the method embodiment as described above. Figure 1 and can achieve the same technical effects. To avoid repetition, it will not be elaborated here.

[0084] By reading the above description, many embodiments and many applications outside the provided examples will be obvious to those skilled in the art. Therefore, the scope of this teaching should not be determined with reference to the above description, but should be determined with reference to the full scope of the foregoing claims and the equivalents of these claims. For the sake of comprehensiveness, all articles and references, including patent applications and publications, are incorporated herein by reference. Omitting any aspect of the subject matter disclosed herein in the foregoing claims is not intended to abandon such subject matter, nor should it be considered that the applicant has not considered such subject matter as part of the disclosed inventive subject matter.

[0085] The above content is a further detailed description of the present invention. It cannot be determined that the specific embodiments of the present invention are limited to this. For those of ordinary skill in the technical field to which the present invention pertains, without departing from the concept of the present invention, several simple deductions or substitutions can still be made, and all should be regarded as belonging to the protection scope determined by the claims submitted by the present invention.

Claims

1. A method for identifying cooperative pursuit action patterns in a two-on-one pursuit-escape game, characterized in that: include: According to the pursuit-escape game trajectory, obtain the status information of the pursuer and the escaper; According to the status information of the pursuer and the escaper, the relative position angle and relative speed angle between the pursuer and the escaper are calculated; According to the relative position angle and relative speed angle between the pursuer and the escapee, different pursuit action styles are defined and classified to obtain a single pursuit action style for a one-on-one pursuit-escape game; Based on the combination and arrangement of the individual pursuit action styles in the one-to-one pursuit-escape game, a collaborative pursuit action style library for the two-to-one pursuit-escape game is constructed; Through the collaborative pursuit action style library of the two-on-one pursuit and escape game, the behavioral characteristics of the pursuer and the escaper in the game are identified, and the behavioral patterns of the pursuer and the escaper in the game are obtained.

2. The method for identifying the coordinated pursuit action pattern in a two-on-one pursuit-escape game according to claim 1, characterized in that: The state information of the pursuer and the escaper is obtained according to the pursuit-escape game trajectory, including: According to the existing pursuit-escape game trajectory, each point on the existing pursuit-escape game trajectory is obtained; Based on each point on the existing pursuit-escape game trajectory, obtain the status information of the pursuer and the escaper; The status information of the pursuer and the escaper includes: Location , Chaser Speed ; The location of the escapee and the speed of the escapee .

3. The method for identifying the coordinated pursuit action pattern in a two-on-one pursuit-escape game according to claim 1, characterized in that: The step of calculating the relative position angle and relative speed angle between the pursuer and the escapee according to the state information of the pursuer and the escapee comprises: According to the status information of the pursuer and the escaper, the position angle of the pursuer is calculated. , the speed angle of the pursuer The speed angle of the escaper ; According to the position angle of the pursuer , the speed angle of the pursuer The speed angle of the escaper , calculate the relative position angle between the pursuer and the escaper and the relative velocity angle : In the formula, is the relative position angle between the pursuer and the escaper; is the relative velocity angle between the pursuer and the escaper; is the velocity angle of the escaper; is the velocity angle of the pursuer; is the position angle of the pursuer.

4. The method for identifying the coordinated pursuit action pattern in a two-on-one pursuit-escape game according to claim 1, characterized in that: According to the relative position angle and relative speed angle between the pursuer and the escapee, different pursuit action styles are defined and classified to obtain a single pursuit action style for a one-on-one pursuit-escape game, including driving away, chasing, intercepting, detouring, confronting, and ambush; In the formula, is the relative position angle between the pursuer and the escaper, is the relative velocity angle between the pursuer and the escaper.

5. The method for identifying the coordinated pursuit action pattern in a two-on-one pursuit-escape game according to claim 1, characterized in that: The individual pursuit action styles based on the one-to-one pursuit-escape game are combined and arranged to construct a collaborative pursuit action style library for the two-to-one pursuit-escape game, including: Based on the individual pursuit action styles of the one-to-one pursuit-escape game, the individual pursuit action styles are combined in pairs to construct a collaborative pursuit action style library for the two-to-one pursuit-escape game; The collaborative pursuit action style library of the two-on-one pursuit and escape game includes chasing, expelling, intercepting, pinching, encircling, entrapment, counterattack, blocking, pincer attack, ambush, outflanking, detouring, driving out, blockade, encirclement, siege, cut off, intercept, blockade, encirclement, and encirclement.

6. The method for identifying the coordinated pursuit action pattern in a two-on-one pursuit-escape game according to claim 1, characterized in that: The cooperative pursuit action style library of the two-to-one pursuit-escape game is used to identify the behavioral characteristics of the pursuer and the escapee in the game, and obtain the behavioral patterns of the pursuer and the escapee in the game, including: Get the actions of the pursuer and the escaper during the game; Through the collaborative pursuit action style library of the two-on-one pursuit and escape game, the actions of the pursuer and the escaper in the game process are identified to form the action sequence in the game process; The action sequences in the game process are used as samples for clustering, and the behavioral characteristics of the pursuers and escapers in the game are analyzed.

7. A system for identifying cooperative pursuit action patterns in a two-on-one pursuit-escape game, characterized in that: include: Acquisition status information module: used to obtain the status information of the pursuer and the escaper according to the pursuit and escape game trajectory; Angle calculation module: used to calculate the relative position angle and relative speed angle between the pursuer and the escapee according to the status information of the pursuer and the escapee; Acquisition action style library module: used to define and classify different pursuit action styles according to the relative position angle and relative speed angle between the pursuer and the escapee, and obtain a single pursuit action style for a one-on-one pursuit-escape game; Building a style library module: used to combine and arrange the individual pursuit action styles of the one-to-one pursuit and escape game, and build a collaborative pursuit action style library for the two-to-one pursuit and escape game; Behavior pattern recognition module: It is used to identify the behavioral characteristics of the pursuer and the escapee in the two-on-one pursuit game through the collaborative pursuit action style library, and obtain the behavior patterns of the pursuer and the escapee in the game.

8. An electronic device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of a method for identifying a collaborative pursuit action pattern in a two-on-one pursuit-escape game as described in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for identifying the cooperative pursuit action pattern of a two-to-one pursuit and escape game described in any one of claims 1-6 are implemented.

10. A computer program product, characterized in that It includes computer instructions, which, when executed by a processor, implement the steps of a method for identifying a coordinated pursuit action pattern in a two-on-one pursuit and escape game as described in any one of claims 1-6.

Images