Unmanned ship cluster confrontation game method and system based on situation information

Abstract

The invention provides an unmanned ship cluster confrontation game method and system based on situation information. The survival ability and combat efficiency of an unmanned ship cluster in a complex confrontation environment are improved. The method comprises the following steps: firstly, collecting threat data of the enemy unmanned surface vehicle and quantitatively evaluating the threat degree, collecting performance data of the our unmanned surface vehicle and quantitatively evaluating the adaptability, and fusing the threat data and the performance data by using an analytic hierarchy process to obtain a comprehensive situation evaluation result; then, sub-task targets are determined according to the comprehensive situation assessment and the overall task target, and tasks are reasonably allocated to the unmanned surface vehicles by adopting an improved Hungary algorithm; and finally, performing local path planning on each unmanned ship by adopting an improved A * algorithm in combination with a local environment, and performing global path planning based on multi-agent collaborative consideration of cluster collaborative requirements and enemy threat distribution. According to the method, comprehensive situation assessment, reasonable task planning and scientific path planning are realized, and the unmanned ship cluster confrontation capability is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of unmanned ship cluster confrontation game, in particular to an unmanned ship cluster confrontation game method and system based on situation information. BACKGROUND

[0002] With the growing demand for ocean resource development, ocean rights maintenance and maritime security, unmanned ship technology has developed rapidly; when unmanned ship clusters perform tasks, they often face complex scenarios of confrontation with enemy unmanned ship clusters; in such a confrontation environment, how to effectively assess the situation and make reasonable task planning and path planning has become a key problem for the application of unmanned ship clusters.

[0003] Some traditional technologies focus on confrontation strategies based on simple rules, but such methods lack adaptability to dynamic battlefield situations and are difficult to deal with enemy tactics; some existing technologies attempt to use machine learning algorithms to make confrontation decisions for unmanned ship clusters; however, these methods usually require a large amount of labeled data for training, and it is difficult to obtain rich and accurately labeled training data in actual maritime confrontation scenarios; in addition, machine learning models may make mistakes when faced with complex situations that have not been trained.

[0004] At the same time, existing unmanned ship cluster confrontation technologies mostly only consider a single factor in situation assessment, such as only focusing on the position information or speed information of enemy unmanned ships, while ignoring important factors such as enemy weapon performance, differences in the performance of our unmanned ships, etc.

[0005] In summary, current unmanned ship cluster confrontation technologies have deficiencies in the comprehensiveness of situation assessment, the rationality of task planning, and the adaptability of path planning, and are difficult to meet the needs of complex and variable maritime confrontation environments. SUMMARY

[0006] To solve the technical problems mentioned in the background art, the present application proposes an unmanned ship cluster confrontation game method and system based on situation information, which can comprehensively consider the threat level of each individual in the enemy unmanned ship cluster and the performance of each individual in our unmanned ship cluster, achieve accurate situation assessment, make the best task planning, and give scientific and reasonable path planning results, thereby improving the response capability of unmanned ship clusters in confrontation scenarios.

[0007] To this end, the technical solution adopted by the present application is as follows:

[0008] 1. An unmanned ship cluster confrontation game system based on situation information, the system comprising:

[0009] M1, situation assessment module, construct threat assessment model and performance assessment model, the threat assessment model threat data of enemy unmanned ship cluster is threat degree assessment, the performance assessment model performance data of our unmanned ship cluster is performance evaluation; According to the result of threat degree assessment and performance evaluation, comprehensive situation assessment is carried out, and comprehensive situation assessment result is generated;

[0010] M2, unmanned ship module of task planning, according to the comprehensive situation assessment result, and set task target, generates sub task target, and based on performance evaluation result, the sub task target is distributed to each individual of our unmanned ship cluster by using intelligent task allocation algorithm;

[0011] M3, path planning module, including local path planning and global path planning, the local path planning adopts classical path planning algorithm, combines real-time local environment information, plans local safe path of individual unmanned ship of our party, and adjusts path parameters according to the performance of individual; The global path planning adopts global path planning algorithm based on multi-agent cooperation, and the local safe path is optimized and integrated.

[0012] Further, the threat data includes position threat, speed threat, weapon threat and communication threat,

[0013] According to the threat data, the threat assessment model is constructed, which is represented as:

[0014]

[0015] Among them, Indicates the threat degree score of enemy unmanned ship; Indicates position threat; Indicates speed threat; Indicates weapon threat; Indicates communication threat; 、 、 And Indicate the weight of position threat, speed threat, weapon threat and communication threat respectively.

[0016] Further, the performance data includes power performance, sensor performance, weapon carrying capacity and communication ability,

[0017] According to the performance data, the performance assessment model is constructed, which is represented as:

[0018]

[0019] Among them, Indicates the applicability score of our unmanned ship, and the applicability score is the quantitative embodiment of performance data; Indicates the power performance score; represents the sensor performance score; Indicates the weapon carrying capacity score; Indicates the communication ability score; 、 、 and They represent the weights of the power performance score, sensor performance score, weapon carrying capability score and communication capability score respectively.

[0020] Furthermore, the comprehensive situation assessment uses the analytic hierarchy process to construct a hierarchical structure model.

[0021] In the hierarchical model, the comprehensive situation assessment is the target layer, the threat assessment results and performance assessment results are the criterion layer, and the threat and performance assessment indicators are the solution layer.

[0022] Construct the judgment matrix of the criterion layer , and the judgment matrix of threat and performance evaluation indicators at the solution level and , calculate the judgment matrix 、 and The characteristic vector of , obtains the weight of threat degree and performance in comprehensive situation assessment and , and the weight of the threat assessment index and the weights of the performance evaluation indicators ,

[0023] According to the weights of the threat level and performance, as well as the weights of the evaluation indicators of the threat level and performance, a fusion calculation is performed to obtain a comprehensive situation assessment result, which is as follows:

[0024]

[0025] in, Indicates the comprehensive situation assessment results.

[0026] Furthermore, the intelligent task allocation algorithm adopts the improved Hungarian algorithm, and the specific steps are:

[0027] 1) Construct a bipartite graph , the subtask target set is , our unmanned boats are assembled as , as the two vertex sets of the bipartite graph, is an edge set, and the matching degree between the subtask goal and our unmanned boat is the edge weight.

[0028] 2) Set the initial match to an empty set solving the optimal matching of bipartite graph by Hungarian algorithm ,

[0029] generating a task allocation result according to the optimal matching.

[0030] Further, the local path planning is specifically,

[0031] 1) setting the current position of the unmanned surface vehicle as the starting point and the task target position as the ending point, constructing a cost function , which is expressed as:

[0032]

[0033] wherein, represents the actual cost from the starting point to point , represents the estimated cost from point to the ending point;

[0034] 2) taking the sea current influence and the obstacles on the path as real-time local environmental information; the sea current influence includes sea current velocity and direction , and setting the running direction of the unmanned surface vehicle at point as ,

[0035] when , the path cost in this direction is reduced, which is expressed as:

[0036]

[0037] when , the path cost is increased, which is expressed as:

[0038]

[0039] wherein, represents an adjustment coefficient;

[0040] the position of the encountered obstacle is taken as a new starting point, the task target position is kept unchanged, and the local safe path is re-planned;

[0041] 3) taking the power performance of the unmanned surface vehicle as the individual performance, setting the speed limit and acceleration limit , and in the process of planning the local safe path, when the speed and acceleration of the point on the path exceed or , the path planning direction is adjusted.

[0042] Further, in the global path planning, the multi-agent is the unmanned surface vehicle cluster, ​

[0043] Define the state variables of each individual unmanned surface vehicle in our unmanned surface vehicle swarm, denoted as:

[0044]

[0045] wherein, denotes the position of the unmanned surface vehicle ; denotes the velocity of the unmanned surface vehicle ; denotes the task type of the unmanned surface vehicle ;

[0046] Set the cooperative constraint conditions between individual unmanned surface vehicles, denoted as:

[0047]

[0048] wherein, denotes the position of the unmanned surface vehicle adjacent to the unmanned surface vehicle ; denotes the distance between the two; denotes the minimum safe distance between adjacent unmanned surface vehicles;

[0049] When the unmanned surface vehicle performs an attack task, and the unmanned surface vehicle performs a cover task, the difference between the distance of the two and the optimal cooperative distance between the attack and cover unmanned surface vehicles is minimized, denoted as The calculation formula is:

[0050]

[0051] wherein, denotes the optimal cooperative distance between the attack and cover unmanned surface vehicles;

[0052] Set the objective function , denoted as:

[0053]

[0054] wherein, denotes the number of unmanned surface vehicles of our side; denotes the threat degree of the enemy at the unmanned surface vehicle ; , and denote the weights;

[0055] Through an iterative optimization algorithm, the objective is to minimize the objective function, and finally obtain a globally optimal path planning result. According to the globally optimal path planning result, the local safe path is optimized and integrated.

[0056] A method for an unmanned boat swarm against a game system based on situation information, the method involving the following steps:

[0057] S1. Construct a threat assessment model and a performance assessment model, wherein the threat assessment model performs a threat assessment on the threat data of the enemy unmanned boat cluster, and the performance assessment model performs a performance assessment on the performance data of our unmanned boat cluster; perform a comprehensive situation assessment based on the results of the threat assessment and performance assessment to generate a comprehensive situation assessment result;

[0058] S2. Generate subtask objectives based on the comprehensive situation assessment results and the set mission objectives, and use an intelligent task allocation algorithm to assign the subtask objectives to each individual in our unmanned boat cluster based on the performance evaluation results;

[0059] S3. According to the sub-task target allocation, local path planning and global path planning are carried out. The local path planning adopts the classic path planning algorithm, combined with real-time local environmental information, to plan the local safe path of our individual unmanned boat, and adjust the path parameters according to the individual performance; the global path planning adopts the global path planning algorithm based on multi-agent collaboration to optimize and integrate the local safe path.

[0060] Compared with the prior art, the advantages of the present invention are:

[0061] 1. This invention integrates threat data such as the position, speed, weapon performance, and communication status of enemy UAVs with performance data such as the power, sensors, weapon loadout, and communication of our own UAVs. It uses the analytic hierarchy process and a multi-factor fusion algorithm to achieve real-time, comprehensive, and accurate situation assessment of complex battlefield environments, greatly improving the adaptability of UAV clusters to complex changes.

[0062] 2. This invention introduces an intelligent task allocation algorithm (such as the improved Hungarian algorithm) to efficiently match and allocate tasks to UAVs based on comprehensive situation assessment results and the individual performance of UAVs. This algorithm takes into account both task requirements and the synergistic relationship between UAVs, achieving optimal resource allocation and improving swarm combat effectiveness and task completion rate.

[0063] 3. This invention combines classic path planning methods such as the improved A* algorithm, taking into account local real-time environmental information (such as obstacles and currents) and the power performance of the unmanned boat to ensure navigation safety. At the same time, through multi-agent collaborative global path optimization, it rationally avoids high-threat areas of the enemy, maintains cluster coordinated operations, and effectively improves the survivability and execution efficiency of the unmanned boat cluster in confrontation scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0065] Figure 1 This is a flowchart of the unmanned boat swarm confrontation game of the present invention;

[0066] Figure 2 This is a flow chart of the situation assessment module of the present invention;

[0067] Figure 3 This is a flowchart of the task planning module of the present invention. DETAILED DESCRIPTION

[0068] To achieve the above objectives, the present invention is implemented through the following technical solutions, which include:

[0069] M1, situation assessment module, builds threat assessment model and performance assessment model. The threat assessment model evaluates the threat degree of the enemy's unmanned boat cluster threat data, and the performance assessment model evaluates the performance data of our unmanned boat cluster. Based on the results of threat degree assessment and performance assessment, a comprehensive situation assessment is performed to generate a comprehensive situation assessment result.

[0070] The threat data of the enemy UAV swarm includes position threat, speed threat, weapon threat, and communication threat. Based on these four threat data, a threat assessment model is constructed, which is expressed as:

[0071]

[0072] in, Indicates the threat score of the enemy unmanned boat; Indicates positional threat; Indicates a speed threat; Indicates a weapon threat; Indicates a communication threat; 、 、 and They represent the weights of position threat, speed threat, weapon threat and communication threat respectively, which are set according to the mission scenario, and their sum is equal to 1.

[0073] Position threat is calculated based on the distance between the enemy UAV and our key targets. The formula is:

[0074]

[0075] in, Indicates the position threat of the enemy unmanned ship; position coordinates of the enemy unmanned ship; position coordinates of the enemy unmanned ship; position coordinates of the enemy unmanned ship;

[0076] In the embodiment, the enemy unmanned ship is continuously monitored by using the radar, sonar, photoelectric sensor and other devices carried on the unmanned ship, and the position information of the enemy unmanned ship is obtained in real time. These sensors have high detection accuracy and can accurately capture the position coordinates of the enemy unmanned ship within a certain range.

[0077] The speed threat is calculated according to the relative relationship between the speed of the enemy unmanned ship and the average speed of the unmanned ship, and the formula is:

[0078]

[0079] wherein, speed threat of the enemy unmanned ship; speed of the enemy unmanned ship; speed of the enemy unmanned ship; average speed of the unmanned ship; In the embodiment, the speed of the enemy unmanned ship is measured by the sensor, which calculates the speed according to the position change of the enemy unmanned ship at different time points and the time interval.

[0080] The weapon threat is calculated according to the range and damage index of the weapon, and the comprehensive coefficient of the weapon threat is set as

[0081] The weapon threat calculation formula is:

[0082]

[0083] wherein,

[0084] weapon threat of the enemy unmanned ship; range of the weapon of the enemy unmanned ship; damage of the weapon of the enemy unmanned ship;

[0085] ​​​​​In the embodiment, the enemy weapon equipment type and performance are estimated through intelligence collection, historical data and analysis of enemy action mode; the intelligence collection can be through various channels such as satellite reconnaissance, information provided by the intelligence department, etc.; the historical data records the weapon equipment used by the enemy in the past; the analysis of the enemy action mode can start from the motion trajectory of the enemy unmanned ship, the attack mode, etc., to infer the possible weapon type, range, damage, etc. carried by the enemy unmanned ship.

[0086] The communication threat is calculated according to the communication signal strength and the communication frequency index, and the communication threat comprehensive coefficient is set as The calculation formula of the communication threat is:

[0087]

[0088] Among them, represents the communication threat of the i-th enemy unmanned ship; represents the communication signal strength of the i-th enemy unmanned ship; represents the communication frequency of the i-th enemy unmanned ship; if the enemy is in a high-intensity communication state, the threat degree evaluation value is increased by adjusting In the embodiment, electronic reconnaissance equipment is used to monitor the communication frequency band, signal strength and other communication state information of the enemy unmanned ship; the electronic reconnaissance equipment can scan the communication frequency band within a certain range, identify the communication signal of the enemy unmanned ship, and measure the signal strength, so as to judge the communication activity level of the enemy.

[0089] In the embodiment, the performance data of the unmanned ship cluster of our side includes power performance, sensor performance, weapon carrying capacity and communication ability, and according to the four performance data, a performance evaluation model is constructed, which is represented as:

[0090]

[0091]

[0092] Among them, represents the applicability score of the unmanned ship of our side, that is, the quantitative embodiment of the performance data; represents the power performance score; represents the sensor performance score; represents the weapon carrying capacity score; represents the communication ability score; , , and respectively represent the weight of the power performance score, the sensor performance score, the weapon carrying capacity score and the communication ability score, which are set according to the task scene, and and equal to 1. ​​​​

[0093] The power performance score calculation formula is:

[0094]

[0095] wherein, represents the maximum speed; represents the acceleration; and respectively represent the coefficients of the maximum speed and the acceleration; when performing a rapid reconnaissance or pursuit task, by adjusting the coefficients, the applicability of the unmanned ship with strong power performance is higher;

[0096] In the embodiment, the power performance parameters of the unmanned ship are obtained by sensors inside the unmanned ship; for example, the maximum speed is measured by a speed sensor, and the acceleration is measured by an acceleration sensor; these sensors collect the motion state data of the unmanned ship in real time, providing accurate information for subsequent performance evaluation.

[0097] The sensor performance score calculation formula is:

[0098]

[0099] wherein, represents the detection distance; represents the accuracy; and respectively represent the coefficients of the detection distance and the accuracy; when performing an intelligence collection task, by adjusting the coefficients, the applicability of the unmanned ship with good sensor performance is higher;

[0100] In the embodiment, the sensor performance parameters such as the detection distance and the accuracy are obtained through the system parameter configuration file; the system parameter configuration file records various performance indicators of the sensors of the unmanned ship, which can be directly read during performance evaluation.

[0101] The weapon carrying capacity score calculation formula is:

[0102]

[0103] wherein, represents the comprehensive coefficient of the weapon carrying capacity; represents the weight of the weapon type; represents the number of ammunition;

[0104] In the embodiment, the weapon management system records the weapon types and the number of ammunition carried by the unmanned ship, and these information can be obtained by interacting with the system.

[0105] The communication ability score calculation formula is:

[0106]

[0107] wherein, represents a communication distance; represents a bandwidth; and respectively represent a coefficient of a communication distance and a bandwidth.

[0108] The comprehensive situation assessment adopts an analytic hierarchy process to determine weights of the enemy threat degree and the performance of the own side in the comprehensive situation assessment; a hierarchical structure model is constructed, the comprehensive situation assessment is taken as a target layer, the threat degree assessment result and the performance assessment result are taken as a criterion layer, and the assessment indexes of the threat degree and the performance are taken as a scheme layer;

[0109] The criterion layer judgment matrix is set , and is represented as:

[0110]

[0111] wherein, represents an importance degree of the enemy threat degree relative to the performance of the own side;

[0112] In the embodiment, the judgment matrix is constructed by means of expert scoring or historical data statistics and the like aiming at the relative importance of the enemy threat degree and the performance of the own side in the criterion layer, and when the expert considers that the enemy threat degree is more important than the performance of the own side in the current task scene, the value of the enemy threat degree is taken as a number greater than 1.

[0113] According to the criterion layer judgment matrix setting mode, the judgment matrix of the assessment indexes of the threat degree and the performance in the scheme layer is constructed, wherein the threat degree assessment index judgment matrix is represented as:

[0114]

[0115] wherein, represents an importance degree of the position threat relative to the speed threat; represents an importance degree of the position threat relative to the weapon threat; represents an importance degree of the position threat relative to the communication threat; represents an importance degree of the speed threat relative to the weapon threat; represents an importance degree of the speed threat relative to the communication threat; represents an importance degree of the weapon threat relative to the communication threat; and the like, the judgment matrix of the performance assessment indexes in the scheme layer is constructed. .

[0116] After the judgment matrixes are constructed, the characteristic vectors of the judgment matrixes are calculated, and taking the matrix as an example, the solving equation is established and is represented as:

[0117]

[0118] in, represents the eigenvalue; represents the identity matrix; Represents the eigenvector; by solving the equation, the eigenvector corresponding to the maximum eigenvalue is obtained and normalized, which is the weight of threat degree and performance in the comprehensive situation assessment and Similarly, by solving the equation, we can get the weights of the threat and performance evaluation indicators in the solution layer. as well as ;

[0119] According to the weights of threat level, performance and threat level and performance evaluation indicators, the comprehensive situation assessment result is obtained by fusion calculation. The formula is:

[0120]

[0121] in, It represents the comprehensive situation assessment result; through fusion calculation, it integrates the enemy threat level and our performance assessment results to fully reflect the current battlefield situation.

[0122] M2, the mission planning module, generates sub-task objectives based on the comprehensive situation assessment results and the set mission objectives, and uses the intelligent task allocation algorithm to allocate the sub-task objectives to each individual in our unmanned boat cluster based on the performance evaluation results.

[0123] Analyze the comprehensive situation assessment results, combine with the pre-set overall mission goal, and determine the sub-task goals; in the embodiment, when the overall mission goal is to protect our offshore platform, and the comprehensive situation assessment results It shows that the enemy unmanned boat cluster is approaching our platform, and our defense in a certain area is relatively weak; set our offshore platform position to , the enemy UAV cluster position set is expressed as:

[0124]

[0125] Calculate the distance between the enemy unmanned boat and our platform using the following formula:

[0126]

[0127] in, Indicates the The distance between an enemy unmanned boat and our offshore platform; through calculation, determine the enemy unmanned boat that is closer and has a higher threat level;

[0128] At the same time, we analyze the defense situation of each area and divide the area of ​​our offshore platform into , the defense strength assessment value of each area is , obtain areas with relatively low defense strength; further determine the sub-task objectives as "dispatching some unmanned boats to intercept approaching enemy unmanned boats" and "strengthening the defense of a certain area of ​​our offshore platform."

[0129] The subtask target allocation adopts the improved Hungarian algorithm, and the specific steps are as follows:

[0130] 1) Construct a bipartite graph , the subtask target set is , our unmanned boats are assembled as , as the two vertex sets of the bipartite graph, is an edge set, and the matching degree between the mission goal and our unmanned boat is used as the edge weight;

[0131] 2) Set the initial match to an empty set ,Right now , prepare for the subsequent matching process, and then solve the optimal matching of the bipartite graph through the Hungarian algorithm;

[0132] 3) From the subtask goal set Starting from the unmatched vertices in the vertices, find augmenting paths through alternating paths (composed of alternating unmatched edges and matched edges) ,In the process of finding the augmenting path, a breadth-first search or depth-first search algorithm is used, starting from an unmatched subtask target vertex, to find an alternating path that can connect to another unmatched vertex of our unmanned boat;

[0133] 4) When an augmenting path is found , then The matched edges on the vertices become unmatched edges, and the unmatched edges become matched edges, resulting in a new matching Repeat the process of finding augmenting paths and updating matching until there is no augmenting path and the optimal matching is obtained Each update of the match will increase the number of matched sub-task targets and our unmanned boats. Generate task allocation results, and assign subtask goals to each individual in our unmanned boat cluster based on the task allocation results, ensuring that each task is assigned to the most suitable unmanned boat, and at the same time forming an effective collaborative relationship between the unmanned boats.

[0134] M3, path planning module, including local path planning and global path planning, the local path planning adopts a classical path planning algorithm, combining real-time local environment information, planning the local safe path of the individual unmanned surface vehicle, and adjusting the path parameters according to the performance of the individual; the global path planning adopts a global path planning algorithm based on multi-agent cooperation, optimizing and integrating the local safe path,

[0135] For each unmanned surface vehicle allocated to the task, taking the current position as the starting point and the task target position as the ending point, a classical path planning algorithm is adopted, combining real-time local environment information, planning the local safe path of the individual unmanned surface vehicle, and adjusting the path parameters according to the performance of the individual; in the embodiment, the classical path planning algorithm uses an improved A* algorithm, and the specific steps are,

[0136] 1) Set the cost function , which is expressed as:

[0137]

[0138] wherein, represents the actual cost from the starting point to point , and represents the estimated cost from point to the ending point; the actual cost can be calculated according to the distance, time and other factors of the unmanned surface vehicle, and the estimated cost is estimated using Manhattan distance or Euclidean distance;

[0139] 2) Add the sea current velocity and direction as dynamic weights to the search process of the local safe path, and the travel direction of the unmanned surface vehicle at point is ,

[0140] When , the path cost in this direction is reduced, which is expressed as:

[0141]

[0142] When , the path cost is increased, which is expressed as:

[0143]

[0144] wherein, represents an adjustment coefficient, which is adjusted according to the strength of the sea current and the influence degree on the unmanned surface vehicle;

[0145] 3) According to the power performance of the unmanned surface vehicle, set the speed limit and the acceleration limit In the search process of the local safety path, the speed and acceleration of each point on the path are calculated in real time to ensure that they do not exceed the limit; by calculating the distance and time interval between adjacent points on the path, the values of the speed and acceleration are obtained, and when the speed or acceleration exceeds the limit, the search direction is adjusted and other possible path points are selected to continue the search;

[0146] 4) In the search process, the obstacle information is detected in real time, and when a new obstacle is found, the path is planned again from the current position. In the embodiment, the unmanned ship can monitor the obstacles in the surrounding environment in real time through the sensors such as radar, sonar, etc. carried by itself. When a new obstacle is detected, the current position is taken as a new starting point, the task target position remains unchanged, and the A* algorithm is restarted for local safety path planning.

[0147] When the local safety path planning of the individual unmanned ship is completed, a global path planning algorithm based on multi-agent cooperation is used to optimize and integrate the local safety path, taking the unmanned ship cluster as the multi-agent,

[0148] The state variables of each individual unmanned ship in the unmanned ship cluster are defined and represented as:

[0149]

[0150] Among them, represents the position of the unmanned ship ; represents the speed of the unmanned ship ; represents the task type of the unmanned ship ;

[0151] The cooperative constraint conditions between individual unmanned ships are set and represented as:

[0152]

[0153] Among them, represents the position of the unmanned ship adjacent to the unmanned ship ; represents the minimum safety distance between adjacent unmanned ships; when the unmanned ship performs an attack task, the unmanned ship performs a cover task, the difference between the minimum distance and the optimal cooperative distance between the attack and cover unmanned ships is minimized, and the difference The calculation formula is:

[0154]

[0155] Among them, represents the optimal cooperative distance between the attacking and covering UUVs; all the cooperative constraints ensure the safe distance and cooperative effect between the UUVs of our side;

[0156] Setting the objective function , specifically, a cost function considering the cooperative constraints and the cost of avoiding the high-threat areas of the enemy, and adjusting and integrating the local safe paths through an iterative optimization algorithm, the objective function is expressed as:

[0157]

[0158] wherein, represents the number of UUVs of our side; represents the threat degree of the enemy at the position of the UUV; , , and represent the weights;

[0159] Through an iterative optimization algorithm (such as genetic algorithm, simulated annealing algorithm, etc.), the objective is to minimize the objective function, and finally obtain the globally optimal path planning result. According to the globally optimal path planning result, the local safe paths are optimized and integrated to ensure that the UUVs of our side can not only avoid obstacles and travel safely, but also form an effective cooperative combat path at the cluster level, thereby improving the survival ability and task execution success rate of the UUVs in the confrontation scenario,

[0160] In an embodiment, the genetic algorithm is used, a group of initial local safe paths are randomly generated as a population, each path represents an individual, then the fitness (i.e. the value of the objective function) of each individual is calculated, the individuals with higher fitness are selected for crossover and mutation operations to generate a new population, and the process is repeated until the termination condition (such as reaching the maximum number of iterations or the value of the objective function converging) is met, and finally the globally optimal path planning result is obtained.

[0161] A method of a UUV cluster confrontation game system based on situation information, the method involves the following operation steps:

[0162] S1, constructing a threat assessment model and a performance assessment model, the threat assessment model performs threat degree assessment on the threat data of the enemy UUV cluster, and the performance assessment model performs performance assessment on the performance data of the UUV cluster of our side; performing comprehensive situation assessment according to the results of the threat degree assessment and the performance assessment, and generating a comprehensive situation assessment result;

[0163] S2, generating sub-task targets according to the comprehensive situation assessment result and the set task target, and distributing the sub-task targets to each individual of the UUV cluster of our side based on the performance assessment result and using an intelligent task distribution algorithm;

[0164] S3. Based on the subtask target allocation, local path planning and global path planning are carried out. Local path planning uses the classic path planning algorithm, combined with real-time local environmental information, to plan the local safe path of our individual unmanned boats, and adjust the path parameters according to the individual performance; global path planning uses a global path planning algorithm based on multi-agent collaboration to optimize and integrate local safe paths.

[0165] The present invention proposes a situation information-based unmanned boat cluster confrontation game method and system, which conducts a comprehensive and accurate situation assessment by integrating multi-source threat information and the performance of the unmanned boats themselves, adopts an intelligent task allocation algorithm to achieve reasonable task planning, and combines a path planning algorithm with local and global collaborative optimization to improve the survivability and combat efficiency of the unmanned boat cluster in complex confrontation environments.

[0166] In summary, the present invention can significantly improve the situational awareness capability, rationality of task allocation and scientific path planning of unmanned boat clusters, enhance the overall combat effectiveness of the cluster, and adapt to the complex and changeable needs of maritime confrontation.

[0167] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A situation information-based unmanned boat swarm confrontation game system, characterized by: The system includes: M1, situation assessment module, constructs a threat assessment model and a performance assessment model. The threat assessment model performs a threat assessment on the threat data of the enemy unmanned boat cluster, and the performance assessment model performs a performance assessment on the performance data of our unmanned boat cluster; performs a comprehensive situation assessment based on the results of the threat assessment and performance assessment, and generates a comprehensive situation assessment result; M2, the mission planning module, generates subtask objectives based on the comprehensive situation assessment results and the set mission objectives, and uses an intelligent task allocation algorithm to allocate the subtask objectives to each individual in our unmanned boat cluster based on the performance evaluation results; M3, path planning module, including local path planning and global path planning. The local path planning adopts the classic path planning algorithm, combined with real-time local environmental information, to plan the local safe path of our individual unmanned boat, and adjust the path parameters according to the individual performance; the global path planning adopts the global path planning algorithm based on multi-agent collaboration to optimize and integrate the local safe path.

2. The unmanned boat swarm confrontation game system based on situation information according to claim 1 is characterized in that: The threat data includes location threat, speed threat, weapon threat and communication threat. A threat assessment model is constructed based on the threat data, which is expressed as: in, Indicates the threat score of the enemy unmanned boat; Indicates positional threat; Indicates a speed threat; Indicates a weapon threat; Indicates a communication threat; 、 、 and Represent the weights of position threat, speed threat, weapon threat and communication threat respectively.

3. The unmanned boat swarm confrontation game system based on situation information according to claim 1 is characterized in that: The performance data includes power performance, sensor performance, weapon carrying capacity and communication capability. A performance evaluation model is constructed based on the performance data, which is expressed as: in, Indicates the suitability score of our unmanned boat, which is a quantitative reflection of performance data; Indicates the power performance score; represents the sensor performance score; Indicates the weapon carrying capacity score; Indicates the communication ability score; 、 、 and They represent the weights of the power performance score, sensor performance score, weapon carrying capability score and communication capability score respectively.

4. The unmanned boat swarm confrontation game system based on situation information according to claim 2 or 3, characterized in that: The comprehensive situation assessment adopts the hierarchical analysis method to construct a hierarchical structure model. In the hierarchical model, the comprehensive situation assessment is the target layer, the threat assessment results and performance assessment results are the criterion layer, and the threat and performance assessment indicators are the solution layer. Construct the judgment matrix of the criterion layer , and the judgment matrix of threat and performance evaluation indicators in the solution layer and , calculate the judgment matrix 、 and The characteristic vector of , obtains the weight of threat degree and performance in comprehensive situation assessment and , and the weight of the threat assessment index and the weights of the performance evaluation indicators , According to the weights of the threat level and performance, as well as the weights of the evaluation indicators of the threat level and performance, a fusion calculation is performed to obtain a comprehensive situation assessment result, which is as follows: in, Indicates the comprehensive situation assessment results.

5. The unmanned boat swarm confrontation game system based on situation information according to claim 1 is characterized in that: The intelligent task allocation algorithm adopts the improved Hungarian algorithm, and the specific steps are: 1) Construct a bipartite graph , the subtask target set is , our unmanned boats are assembled as , as the two vertex sets of the bipartite graph, is an edge set, and the matching degree between the subtask goal and our unmanned boat is the edge weight. 2) Set the initial match to an empty set , solve the optimal matching of bipartite graph by Hungarian algorithm , A task allocation result is generated according to the optimal match.

6. The unmanned boat swarm confrontation game system based on situation information according to claim 1 is characterized in that: The local path planning is specifically as follows: 1) Set the current position of our unmanned boat as the starting point and the mission target position as the end point, and construct the cost function , expressed as: in, From the starting point to the point The actual cost, Indicates a point the estimated cost to reach the destination; 2) Use the influence of ocean currents and obstacles on the path as real-time local environmental information; ocean current influences include ocean current speed and direction , suppose our unmanned boat is at point The driving direction is , when , reducing the path cost in this direction, expressed as: when , improve the path cost, expressed as: in, represents the adjustment coefficient; The location of the encountered obstacle is used as the new starting point, the mission target location remains unchanged, and the local safe path is replanned; 3) Set a speed limit based on the power performance of our unmanned boat as an individual performance and acceleration limits , in the process of planning a local safe path, when the speed and acceleration of a point on the path exceeds or , adjust the path planning direction.

7. The unmanned boat swarm confrontation game system based on situation information according to claim 1 is characterized in that: In the global path planning, the multi-agent is our unmanned boat cluster, Define the state variables of each individual unmanned boat in our unmanned boat cluster, expressed as: in, Unmanned boat location; Unmanned boat speed; Unmanned boat The type of task; Set the coordination constraints between individual unmanned boats, which can be expressed as: in, Unmanned Boat Adjacent unmanned boat location; Indicates the distance between the two; Indicates the minimum safe distance between adjacent unmanned boats; When an unmanned boat Perform attack missions, unmanned boats To perform the cover mission, the difference between the distance between the two and the optimal coordination distance between the attack and cover unmanned boats is minimized. The calculation formula is: in, Indicates the optimal coordination distance between attacking and covering unmanned boats; Setting the objective function , expressed as: in, Indicates the number of our unmanned boats; Unmanned boat The enemy threat level at 、 and represents weight; Through the iterative optimization algorithm, the goal is to minimize the objective function and ultimately obtain a global optimal path planning result. Based on the global optimal path planning result, the local safe path is optimized and integrated.

8. A method for a swarm of unmanned boats to counter a game system based on situation information, the system disclosed in claims 1-7, the method involving the following steps: S1. Construct a threat assessment model and a performance assessment model. The threat assessment model performs threat assessment on the threat data of the enemy unmanned boat cluster, and the performance assessment model performs performance assessment on the performance data of our unmanned boat cluster. Performing a comprehensive situation assessment based on the results of the threat assessment and performance assessment to generate a comprehensive situation assessment result; S2. Generate subtask objectives based on the comprehensive situation assessment results and the set mission objectives, and use an intelligent task allocation algorithm to assign the subtask objectives to each individual in our unmanned boat cluster based on the performance evaluation results; S3. According to the sub-task target allocation, local path planning and global path planning are carried out. The local path planning adopts the classic path planning algorithm, combined with real-time local environmental information, to plan the local safe path of our individual unmanned boat, and adjust the path parameters according to the individual performance; the global path planning adopts the global path planning algorithm based on multi-agent collaboration to optimize and integrate the local safe path.

Images