CN113283124A - Multi-agent-based autonomous USoS participation model construction method and system - Google Patents
Multi-agent-based autonomous USoS participation model construction method and system Download PDFInfo
- Publication number
- CN113283124A CN113283124A CN202110822426.5A CN202110822426A CN113283124A CN 113283124 A CN113283124 A CN 113283124A CN 202110822426 A CN202110822426 A CN 202110822426A CN 113283124 A CN113283124 A CN 113283124A
- Authority
- CN
- China
- Prior art keywords
- model
- investigation
- communication
- rule
- rules
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000010276 construction Methods 0.000 title claims abstract description 12
- 238000000034 method Methods 0.000 claims abstract description 36
- 230000003993 interaction Effects 0.000 claims abstract description 4
- 238000004891 communication Methods 0.000 claims description 53
- 238000011835 investigation Methods 0.000 claims description 42
- 239000003795 chemical substances by application Substances 0.000 claims description 40
- 230000033001 locomotion Effects 0.000 claims description 26
- 230000006399 behavior Effects 0.000 claims description 13
- 238000001514 detection method Methods 0.000 claims description 13
- 239000013598 vector Substances 0.000 claims description 10
- 230000015572 biosynthetic process Effects 0.000 claims description 7
- 230000005484 gravity Effects 0.000 claims description 7
- 238000009527 percussion Methods 0.000 claims description 7
- 239000000126 substance Substances 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 4
- 238000004422 calculation algorithm Methods 0.000 claims description 2
- 230000007123 defense Effects 0.000 claims description 2
- 241000764238 Isis Species 0.000 claims 1
- 238000010009 beating Methods 0.000 claims 1
- 230000002123 temporal effect Effects 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 18
- 230000007246 mechanism Effects 0.000 abstract description 11
- 238000012550 audit Methods 0.000 abstract 2
- 238000004088 simulation Methods 0.000 description 16
- 238000010586 diagram Methods 0.000 description 13
- 238000013461 design Methods 0.000 description 9
- 230000000875 corresponding effect Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 230000000007 visual effect Effects 0.000 description 5
- 230000002776 aggregation Effects 0.000 description 4
- 238000004220 aggregation Methods 0.000 description 4
- 238000012800 visualization Methods 0.000 description 4
- 241001465754 Metazoa Species 0.000 description 3
- 230000003542 behavioural effect Effects 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 230000035772 mutation Effects 0.000 description 3
- 238000012163 sequencing technique Methods 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 244000144992 flock Species 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 241000251468 Actinopterygii Species 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003054 catalyst Substances 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000007717 exclusion Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000002431 foraging effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000005215 recombination Methods 0.000 description 1
- 230000006798 recombination Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000010206 sensitivity analysis Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- HOWHQWFXSLOJEF-MGZLOUMQSA-N systemin Chemical compound NCCCC[C@H](N)C(=O)N[C@@H](CCSC)C(=O)N[C@@H](CCC(N)=O)C(=O)N[C@@H]([C@@H](C)O)C(=O)N[C@@H](CC(O)=O)C(=O)OC(=O)[C@@H]1CCCN1C(=O)[C@H]1N(C(=O)[C@H](CC(O)=O)NC(=O)[C@H](CCCN=C(N)N)NC(=O)[C@H](CCCCN)NC(=O)[C@H](CO)NC(=O)[C@H]2N(CCC2)C(=O)[C@H]2N(CCC2)C(=O)[C@H](CCCCN)NC(=O)[C@H](CO)NC(=O)[C@H](CCC(N)=O)NC(=O)[C@@H](NC(=O)[C@H](C)N)C(C)C)CCC1 HOWHQWFXSLOJEF-MGZLOUMQSA-N 0.000 description 1
- 108010050014 systemin Proteins 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/11—Complex mathematical operations for solving equations, e.g. nonlinear equations, general mathematical optimization problems
- G06F17/13—Differential equations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/448—Execution paradigms, e.g. implementations of programming paradigms
- G06F9/4482—Procedural
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Business, Economics & Management (AREA)
- Software Systems (AREA)
- Computational Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Tourism & Hospitality (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Marketing (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- Educational Administration (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Operations Research (AREA)
- Algebra (AREA)
- Primary Health Care (AREA)
- Databases & Information Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an autonomous USoS participation model construction method based on multiple agents, which comprises the following steps: constructing a main body model; establishing an exchange relationship between a main body model of the system and the system; and introducing behavior rules to constrain the interaction relation to form an autonomous USoS participation model based on multiple agents. The participation model can autonomously control each independent system in the model, and the validity of the proposed audit participation model is checked by displaying the dynamic process of audit participation. The invention adopts a modeling method based on logic drive to express the judgment and inference rules of system evolution, constructs a cooperative mechanism of reconnaissance-communication-attack rules, focuses more on the behavior modeling of the system than the prior art, and leads the unmanned system to better accord with the actual combat requirements.
Description
Technical Field
The invention relates to the field of autonomous control, in particular to the crossing field of multi-agent cooperative control, system confrontation and combat simulation, and particularly relates to a multi-agent-based autonomous USoS participation model construction method and a multi-agent-based autonomous USoS participation model construction system.
Background
The OODA (Objective observed decision Attack) battle loop is a closed loop formed by reconnaissance equipment, decision equipment, striking equipment and other objects of an enemy and is used for completing a specific battle task. The process is not completed once, but the sharing and updating of information and the continuous execution of equipment actions are completed in a continuous loop until the goal of fighting (generally referred to as the goal of elimination) is reached. It essentially conforms to the basic behavioral specification of an agent in a complex world, namely the cyclic reciprocating process of environment-perception-decision-action-environment.
Most of the agents in the traditional USoS are homogeneous, i.e., similar in function and behavior. However, in actual combat, different types of agents have different capabilities, and the coordination mechanism among the agents is more complicated. The invention constructs a cooperative mechanism of reconnaissance-communication-attack rules based on an OODA (object oriented data acquisition) operational cycle theory, so that an unmanned system can better meet actual operational requirements.
Disclosure of Invention
The invention aims to provide a multi-agent-based autonomous USoS participation model construction method and a multi-agent-based autonomous USoS participation model construction system, so as to solve the technical defects in the prior art.
By using the theory of operational rings, three classes are constructed for the operational systemModels, respectively investigationAnd make a decisionAnd striking. In addition, the other system is ownedWill be all by my partyAre regarded as targets and therefore, the targets are not designed separately. All kinds ofThe models have different capability attributes and thus can perform different functions. For different purposesThe model is constructed by two methods: for each type ofSpecially designing a model; designing a general model, and filtering condition identification by setting attribute valuesType (if it has investigation capability, it can be identified as investigation). According to practical experience, it is recommended to use the second method, on the one hand avoidingDifferent types ofThe repeated setting of the common attribute, on the other hand, the identification can be realized only by simple judgment conditionsAnd thus implements the functionality of the first method. Thus, inIn solid modeling, the invention firstly designs the generalModels, then identified by design screening conditionsType (b).
The design key of (1) is the setting of attributes, on one hand, the requirements of a scene to be simulated can be met, and on the other hand, the design is refined enough to avoid redundant and redundant design. Designing the general purpose of equipment nodes according to the characteristics of a combat systemThe model is shown in table 1:
TABLE 1
Lines 14-21 of Table 1 areIs the most important attribute in the subsequent modeling process. Further, the aboveRather than being completely independent between attributes, the following lists the dependencies that exist between attributes:
And the number of the command channels is less than or equal to the number of the communication channels.
In satisfying the aboveOn the basis of attribute constraint, the identification can be realized by setting rulesType (c) of the cell. Suppose an equipmentOnly corresponding to one equipment type, one equipment type is not considered at allThere are cases where there are multiple equipment types. According to equipmentThe attribute judgment type identification rule is as follows:
The combat system accomplishes the strike to the target through the nimble cooperation between the equipment node, according to the combat ring theory, has 4 kinds of different even limit types in the combat system, and the difference is: the three types of nodes form the whole fighting capacity through four connected edges. Class 4 with edgesThe model is as follows:
(1) investigation edge
The functions are as follows: the detection edge can store the detected target information into a target information list, and the two are usedFor example, if present, fromToThe detection edge of (1) is 。
(2) Communication edge
The functions are as follows: the communication edge can share the target information to otherTwo of the aboveFor example, if present, fromToOn the communication side of。
(3) Finger control edge
(since the command issuing relies on communication, the distance between the parties is less thanCommunication range of); 、。
the functions are as follows: the control edge can process the target information and then distribute different target information to different attack nodes according to a certain rule, namely。
(4) Striking edge
the functions are as follows: the strike edge can strike the target according to the target information distributed by the command control node, whether the strike is successful or not is processed according to a certain probability rule, and if the strike is successful, the strike edge is deletedAnd all the connecting edges connected with the connecting edge.
Equipment behavior rulesModel primary description equipment nodeHow to take corresponding action in the face of environmental changes. Multiple corresponding in the battle systemIn the model, equipmentThe behavior of (1) is mainly reflected in two aspects, one isAnother is a plurality ofChange of network structure (new node introduction, edge reconnection).
The main characteristic of the battle system is that the equipment can be dynamically recombined to realize the overall optimal battle effect. One important embodiment of dynamic recombination is to reconnect the edges of the battle network according to the dynamic environment change, so that more high-quality battle rings are formed, and the whole battle network keeps the optimal performance. Ideally, each equipment could theoretically form a particular type of link with other equipment, but since the number of links forming any type is not infinite in practice, a decision needs to be made as to which equipment to close the loop when supply and demand is not met (i.e., the limit on the number of links is not sufficient to meet all the link requirements). The following closed-loop rule design is respectively carried out for the detection edge, the communication edge, the decision edge and the attack edge:
threat-first reconnaissance rule of continuous border:
and sequencing the threat degrees of the targets which can be detected by the detection equipment, and sequentially detecting the targets according to the sequencing. The threat can be defined as the distance of the target from the reconnaissance equipment, the distance is just inversely related to the reconnaissance side quality, the closer the distance, the higher the threat degree, and the higher the reconnaissance side quality at the moment. Of course, other calculation methods may be used for the threat level. Designing the edge connecting rule of the reconnaissance edge according to the principle of threat priority, as follows:
for investigationAccording to its scope of investigationThe inner target threat levels rank the targets. If the number of objects in the investigation range is less thanDetecting all the targets, namely forming detection edges with all the targets; if the number of targets in the investigation range is larger thanThen get the front of the rank of the target threatIndividual target and investigationConstituting a scout edge.
Rule of communication side connection nearby:
the quality of the communication edge is in negative correlation with the distance between the communication equipment, so that the communication can be sequentially carried out with other equipment according to the sequencing of the distance, and the quality of the communication edge can be ensured to be larger. Accordingly, the rules for near communication connection are designed as follows:
all ofAll have certain communication capabilityTo aim atTo its communication rangeInter-my equipment node distanceThe distances of (a) are sorted. If the number of the equipment nodes of my party in the communication range is less thanForming a communication edge with all the equipment nodes of our party; if the number of the equipment nodes of my party in the communication range is larger thanThen take the distanceMinimum frontIndividual my equipment nodes form a communication edge.
Weakest preferred strike alignment rule:
as long as the target is within the striking range, the probability of striking an edge is not related to the distance, but to the average radius of the target. In this context, it is considered thatThe attribute includes information related to a life value, and in order to destroy the target of the other party as soon as possible, the target with the lowest life value should be selected for the prior attack. Therefore, the weakest preferred strike alignment rule is designed as follows:
the rule bagThe continuous edge rule design of the directing edge and the striking edge is included because the directing node potentially tells the striking node how the striking edge should be formed when the directing node commands the striking node to strike a certain target. To is directed atThe information it grasps is as follows: 1) strike from my partyAll attribute information (mainly depending on the position, attack power, defense power, cost of one attack, etc.); 2) enemies in a target listLocation, offensive power, defensive power, and value information. Based on the information, hitting our party by using a stable matching algorithmAnd enemies in the target listOne-to-one or many-to-one matching (multiple hits by my party can be made)Attack an enemy target).
Equipment nodeIs the basis for the dynamic countermeasure implemented. In the battle system countermeasure, the equipment set is a system, and can also be regarded as a cluster, and the moving direction and speed of the equipment at each step in the system countermeasure can be determined by taking reference to a cluster motion model.
In nature, individual animals reach a certain number of groups, can be well self-organized to form forward, turning, avoiding … … and other matrix types, and can be seamlessly switched among different matrix types, as shown in fig. 1, a cluster motion model is used for simulating various impressive animal cluster motions in nature, wherein fig. 1 (a) is a schematic diagram of bird cluster motion, and fig. 1 (b) is a schematic diagram of fish cluster motion.
The bird swarm model is one of the cluster motions of a complex system theory study. By observation, the behavior mechanism of a bird group mainly comprises three types, as shown in fig. 2: FIG. 2 (a) shows the aggregation propensity: individuals tend to move closer to surrounding individuals to avoid isolation; FIG. 2 (b) shows velocity alignment, with the velocities of the individual and surrounding neighbors remaining synchronized; fig. 2 (c) shows density repulsion, and when the distance between individuals is too close, the individuals move toward each other to avoid collision.
Above-mentioned action mechanism can guarantee to avoid the collision between the individuality in the bird crowd on the one hand, and on the other hand can guarantee certain aggregative nature between the bird crowd to can reach the sharing of information between the individuality under the prerequisite of avoiding the collision (like the bird crowd is looking for the in-process of food, and an individual seeks food, can attract other individuality to look for food in the future).
Equipment for fighting against combat systems by using moving rules of animal clustersFor reference purposes, e.g. for scoutingThe principle of finding the reconnaissance target is similar to that of foraging of the bird flock, and on one hand, the target of the other party is found more quickly by the bird flock, and on the other hand, individual reconnaissance is avoidedToo much aggregation. By using bird group behavior mechanism for reference and combining the characteristics of fighting system, 3 equipment types are constructedThe movement rules of (2) are as follows:
the detection Agent's "spring type" movement rule:
for equipment systems, the aim of reconnaissance is to find enemy targets as quickly, as early and as complete as possible, which requires that sufficient areas be reconnaissance. Thus, for investigationIn other words, the moving principle is as follows: 1) need to be dispersed to reduce overlap of investigation ranges, and 2) need to be aggregated to some extent to avoid missing investigation regions. Thus, when investigatingFrom other investigationWhen the distance is too small, the two mutually repel and attract each other, the process is similar to the force in different directions and different magnitudes generated by different stretching of the spring, and accordingly, the spring type investigation is definedAnd (4) moving rules. First, defining a reconnaissance based on bird group clustering and collision avoidance rejection mechanismThe elastic model ofFangzhi investigationThe elasticity calculation is illustrated for the example shown in fig. 3.
Wherein the content of the first and second substances,is thatAll others in the scope of the scoutSquare scoutThe center of gravity of the vehicle,is thatThe position of (a).
Firstly, according to fig. 3, the magnitude of the elastic force is calculated in three steps:
Wherein the content of the first and second substances,is composed ofIs determined by the coordinate of (a) in the space,is composed ofToThe vector of (a) is determined,as vectorsThe 2-norm of (a), the length,is composed ofToThe vector of (a) is normalized,define the length and forward and reverse directions of the spring: greater than 0 means thatAndis less thanIn the investigation range ofIs a repulsive force, orientedPoint of directionAnd is andthe larger the value, the larger the repulsive force; less than 0 means thatAndis greater thanThe range of investigation of (a) is,represents an attractive force, the direction isPoint of directionAnd is andthe larger the value, the larger the attraction force.
Secondly, based on the magnitude of the elastic force, it can be calculatedThe direction and distance of movement of. Order toIs composed ofIs obtained by obtainingThe relationship between them is as follows:
thus, there are approximated:andi.e. byThe position at the next time is the vector sum of the current position and the current velocity, and the velocity at the next time is the vector sum of the current velocity and the current elastic force.
for the battle system, in order to guarantee investigationAnd control by fingerThe communication between the two devices needs to make each command control as possibleWith a certain number of investigationsRemain within communication range. The communication range is equivalent to the visual range of the birds in the bird swarm model, and for birds outside the communication rangeAnd information such as the position and the speed cannot be acquired. In addition, for multiple commandersCommunication redundancy may be caused if the aggregation level is too high, i.e. multiple fingersAnd the same investigationCommunication is performed, thereby causing a waste of communication resources. Therefore, by using the aggregation tendency and density exclusion mechanism in the bird group behavior model for reference, the finger control is constructedThe close communication uniform movement rule of (2), as shown in fig. 4.
Wherein the study object in the figure is,Is thatAll others within communication rangeSquare scoutThe center of gravity of the vehicle,is thatAll other commanders within communication rangeThe center of gravity of the vehicle,is thatThe position of (a).
Respectively calculateAttractive force ofAnd repulsive forceFrom which the resultant of the two forces is calculated as. Then, can be based onThe current position, velocity and resultant calculate the velocity and position at its next time.
for the fighting system fighting task, the order is requiredAccording to investigationDerived information, and commandAnd the issued command moves to the target and executes the striking task. At the same time, it is necessary to ensure that the signals do not exceed and control the control as much as possibleThe communication distance therebetween. Thus, strikingThe movement rules of (1) need to be both target-oriented movement and finger-oriented controlMove as shown in fig. 5. Wherein the study object in the figure is,Is the position of the object or objects,is all covered in the communication rangeDirection A ofThe center of gravity of the vehicle,is thatThe position of (a).
Respectively calculateRespectively of two attractive forcesForce of mixingFrom this, the resultant of the two forces is calculated as:then can be based onThe current position, speed and resultant force are calculated to calculate the speed and position at the next moment,is the location of the target.
The combat system is a typical open system, and both sides can add new equipment nodes to the system according to actual conditions so as to achieve the purpose of winning. If the introduction of a new equipment node is not arranged, the whole combat system is closed, and finally an equilibrium state is reached. Suppose there are two partiesAndthrough the mutual influence between the two parties, the final equilibrium state has three types: only remain in the systemLeft in the systemA small amount of the catalyst remains in the systemAndand the two do not interfere with each other, as shown in fig. 6.
However, in actual combat, node introduction is not unlimited due to resource limitations (otherwise, one party introduces an unlimited number of own nodes instantaneously), and therefore, a certain node introduction mechanism needs to be set.
Firstly, a base is setAnd equipmentIn contrast to this, the present invention is,having an individual set of attributes, respectivelyRespectively representing the existing resource amount, the unit time resource recovery amount and adding a reconnaissanceResource consumption, adding a commandResource consumption and a new hitThe resource consumption of. Then, 3 node introduction mechanisms were designed as follows.
Introducing a new equipment node rule in a saturation mode:
according to the new increaseAnd make a decisionStrikingOnce the current amount of resources reaches the order of (2) generating the correspondingThe amount of resources required, i.e. to generate one suchThe specific rule is as follows:
introducing a new equipment node rule as required:
in-system investigationAnd make a decisionAnd strikingWhen the existing quantity does not meet the demand quantity, the corresponding quantity is generatedThe specific rule is as follows:
regular introduction of new equipment node rules
Introducing 3 types of nodes at intervals of fixed time, or sequentially introducing investigation at intervals of fixed time according to sequenceAnd make a decisionAnd strikingIn this loop, in the following manner of introducing 3 kinds of nodes simultaneously, the rule for introducing new equipment nodes at regular time is designed as follows:
based on the method, the invention also provides a multi-agent-based autonomous USoS participation model construction system, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor realizes the steps of any method when executing the computer program.
The invention has the following beneficial effects:
1. the multi-Agent-based autonomous USoS participation model construction method provided by the invention does not require a modeler to establish a detailed mathematical model for a collaborative process, and can describe interaction behaviors such as cooperation, countermeasure and the like among the agents only by paying attention to behavior modeling of a system.
2. According to the multi-Agent-based autonomous USoS participation model construction method, the constructed entity Agent model and the behavior rules can support and construct a dynamic simulation model of an equipment system.
3. The method for constructing the autonomous USoS participation model based on the multi-Agent realizes the USOS engagement model through the multi-Agent programmable modeling simulation environment NetLoco. The dynamic course of the engagement can be visually demonstrated and the validity of the proposed model can be verified.
The present invention will be described in further detail below with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the invention and, together with the description, serve to explain the invention and not to limit the invention. In the drawings:
FIG. 1 is a schematic diagram of a cluster motion phenomenon provided by the present invention;
FIG. 2 is a schematic diagram of a bird group behavior mechanism provided by the present invention;
FIG. 3 is a diagram of Agent movement rules provided by the present invention;
FIG. 4 is a schematic diagram of a movement rule of an Agent according to the present invention;
FIG. 5 is a schematic diagram of the movement rules of the striking Agent according to the present invention;
FIG. 6 is a schematic diagram of 3 equilibrium states of the closed combat system provided by the present invention;
FIG. 7 is a framework diagram of a combat system collaborative countermeasure simulation system according to a preferred embodiment of the present invention;
FIG. 8 is a visualization interface of a combat system collaborative confrontation simulation model according to a preferred embodiment of the present invention;
FIG. 9 is a schematic diagram of the countermeasure process of the unmanned combat system according to the different steps provided in the preferred embodiment of the present invention;
FIG. 10 is a schematic diagram of the 1-time combat fight process of the unmanned combat system according to the preferred embodiment of the present invention;
FIG. 11 is a schematic diagram of the 10 combat systems of the preferred embodiment of the present invention;
FIG. 12 is a schematic diagram comparing Agent capability provided by the preferred embodiment of the present invention;
FIG. 13 is a flow chart of a multi-agent based autonomous USoS participation model construction method provided by the present invention.
Detailed Description
The embodiments of the invention will be described in detail below with reference to the drawings, but the invention can be implemented in many different ways as defined and covered by the claims.
The embodiment exemplifies the equipment node and the equipment edge connecting entity constructed according to the researchModels and associated behavioral rules, relying onThe simulation platform constructsThe battle system of (2) confronts the simulation model, and the design idea and the whole framework of the model are shown in FIG. 7.
The example of the present embodiment can be fully reproduced, and the operation method is as follows: mounting ofTools (open source software); downloads the program and executes the code. Because the simulation has a certain randomness, each run does not necessarily produce the same dynamic process and result.
The framework shown in FIG. 7 has 4 layers from the lowest logical layer to the highest visual layer, wherein the rule layer and the data layer are usedThe language is realized by code programming, and the visual layer isThe logic layer is not editable and invisible. According to the framework, corresponding equipment is constructed for each typeModels, behavioral rules models, and visualization models.
The embodiment example mainly aims at research on autonomous confrontation of unmanned equipment system, and equipment is arranged in the process of autonomous confrontationAnd forming the battle ring connecting edges between the equipment according to the constructed behavior rule model so as to form the battle ring. And (4) taking the introduction mechanism of the new equipment node into consideration, carrying out sensitivity analysis of different capabilities on the influence of the countermeasure result.
Firstly, according to the constructed simulation model, comparing the process and result data of the cooperative countermeasure of the combat system under different conditions.
In the visualization module, the visualization module is used,the color of (c) indicates its formation (the red party is my party). The pentagons of the lower left corner and the upper right corner are bases,“”、“"and""represents the types of percussion, command and reconnaissance bases, tank, plane and pentagon respectively representing percussion, reconnaissance and commandType, red, brown, blue and black bordered represent the scout, communication, command and strike types, respectively.
Suppose that the unmanned combat system is divided into two parts of red and blue, each of which has 3 basesRespectively, is a detection baseFinger control baseAnd striking baseCan consume certain resources to generate corresponding types. Each base stationCertain variables need to be set as shown in table 2. Generated by basesInherit the attributes of the base, but generatedThe mobility attribute needs to be changed from 0 to 1. The new node generation rule of the present embodiment introduces a new device node rule with reference to a saturation formula.
TABLE 2 base information
First, a visual interface of the combat system versus the simulation model is designed, as shown in fig. 8.
FIG. 8 (a) is a visual interface for visually displaying the dynamic changes of the tactical system; fig. 8 (b) and 8 (c) are partial data records of the history and current state of the operation of the combat system, respectively.
The end condition of the confrontation simulation model of the unmanned combat system is set as: when the combat system has only the equipment of the same battleOr when the simulation step reaches the maximum step of the constraint, the simulation is stopped, and the maximum step is set to 2000 in this embodiment.
Of both partiesThe initial capabilities were set according to table 2. The fighting process of the unmanned combat system is shown in the figure. The whole countermeasure process comprises 427 steps, and 1 time is taken for each 40 steps of the countermeasure process from step 0 to step 160. Due to space limitation, the final sub-graph gives the final state, and only the blue equipment is arranged in the battle systemAnd the result shows that the blue matrix wins.
The unmanned system combat process embodies cooperation among the same pieces of equipment for formation and confrontation among different pieces of equipment for formation. The cooperation is embodied at the communication side and the instruction control side, the communication side transmits the detected target information among the equipments in the same formation, and the instruction control side sends the task to the striking side. Antibodies are now reconnaissance (acquiring target information) and hit (hitting target). The corresponding challenge results of the above challenge process are shown in fig. 10.
FIG. 10 (a) records the current remaining equipment of both parties at each stepQuantity, (b) equipment recording loss of both red and blue until current stepThe number of the cells. The results show that the opponent's battle wins, consistent with the confrontation process of fig. 9. Wherein, FIG. 9 (a) -FIG. 9 (f) are step 0 and step 0, respectively40. The confrontation results in steps 80, 120, 160 and 427 are shown schematically. Since my party is investigatingThe detection range of the detector is only 10km, and the other party detectsThe reconnaissance range of 100km results in that the number of equipment of our party is always at a disadvantage and the equipment of our partyIs always on the rising trend.
In addition, by collecting data of 10 times of simulation, equipment was plottedThe trace heatmap of (a), is shown in fig. 11.
EquipmentA trace of color is left when passing a location, and the colors can be superimposed. Thus, the moreThe heavier the color is at the position passed. Of both partiesMeeting in the middle of the battlefield is also the most aggressive area of confrontation.
Through many experiments, under different capability settings, three equilibrium states are found in the embodiment. The first is a state where only the red system agent is alive. The second is a state where only the blue system agent is alive. Third, both red and blue system proxies exist, with no interaction between them. The first two equilibrium states are understandable and occur when all system agents of a single campaign are destroyed. For the third equalization, it may result from two cases: (1) the base of a camp is totally destroyed, resulting in no new system agents being added to the battlefield. (2) The number of the rest special workers and bases is less than three, and an effective hunting chain of observer- > judge- > executor cannot be formed.
Next, this example designs 6 capability schemes, as shown in table 3. Two-by-two of the 6 capacity schemes were challenged 20 times, and in order to eliminate random errors that may occur due to different camps, the challenge between any two schemes would be "red-to-blue" 10 times and "blue-to-red" 10 times, respectively.
TABLE 3 Capacity schemes
Make the red and blue left after the countermeasure simulation ends equipNumber difference betweenIs an index of performance efficiency, whereinThe winning of the direction of the red is shown,the indication of the win-win of the blue,indicating a flat hand. The statistical results are shown in table 4.
TABLE 4 statistics of results of systematic challenge under different capacity scenarios
Then, the average of the system challenge results under different capacity schemes was calculated as shown in table 5.
Then, 6 capacity scenarios were evaluated overall. Taking case of scenario 1, its evaluation score is the average of the results of the scenario compared to the other scenarios, i.e., [12.4, 0.4, 18.1, 18.1, 9.4, 8.4, -1.9, 16, 16.3, 9.9], thus the scenario 1 score is 5.355. From this, the total evaluation values of the 6 capability schemes are calculated as: 5.355, 0.815, 5, -4.965, -6.4 and 0.195. Thus, scheme 1> scheme 3> scheme 2> scheme 6> scheme 4> scheme 5.
When the fighting capabilities of the two parties in the fighting system are not very different, the result can be changed by only slightly improving some capabilities of the parties. Therefore, the critical point of the acquisition capability is important. If one breaks this critical point, the countermeasures can be reversed. This critical point is commonly referred to as a discontinuity point, according to the system theory, and is used to describe the discontinuity and abruptness of the change. Next, the scouting will be done in redFor example, mutation points under the two-party current capability scheme shown in Table 6 were analyzed.
TABLE 6
In table 6, the different capability indexes between the two capability schemes are mutually good and bad, and are relatively balanced, and the bold shows that the capability of the array where the column is located is higher than that of the other arrayThe capacity of the tank, the result of the fight depending on the amount of the fight remaining when the fight reaches the end conditionThe number of the cells. Make our party reconnaissanceHas a scouting ability of,. To "remainLoss of quantity "andquantity in responseStatistical indicators of changes. Each confrontation simulation willThe value of (c) is increased by 1 and each simulation is repeated 10 times to eliminate the random effect, and the statistical result is shown in fig. 12.
The dark-colored curve and the dark-colored region of the graph (a) in FIG. 12 represent respectively the remaining equipment in our party's camp after the completion of the confrontationThe mean and 95% confidence intervals of the quantities, respectively, the light-colored lines and areas represent the remaining equipment of the basket formationMean value of the amount and 95% confidence interval. As can be inferred from FIG. 12 (a), when my party reconnaissanceAbility to reconnaissanceLess than 50, the equilibrium state of the confrontation is I, whenAbove 70, the equilibrium state of the confrontation is my winning. Therefore, the temperature of the molten metal is controlled,has a mutation point of [50,70 ]]Within the interval.
For more detailed analysis, interval [50,70 ] is used]The area of (c) is enlarged as shown in (b). (b) The figure is a close-up of the cross-over area of figure (a), showing only the remaining equipmentMean curve of quantities. As can be seen from the graph (b), the depth curve has two intersections, and the 2 nd intersection is regarded as an abnormal value. From the 1 st cross pointIn the interval [63,64 ]]In between, a median value of 63.5 in the interval can be assigned as a mutation point.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (8)
1. The multi-agent-based autonomous USoS participation model construction method is characterized by comprising the following steps:
constructing a master model that is genericModels, including investigationAnd make a decisionAnd striking(ii) a The method for constructing the main body model comprises the following steps:
identifying a subject model type by setting an attribute value filter condition, the attribute value including a reconnaissance rangeScout channelCommunication rangeCommunication channelCapability of controlling by fingerAnd a finger control channelRange of percussionStriking channelMaximum moving speedMobility of the mobile terminalResidual oil amountAvailability, availability;
if and only if、、The time main body model belongs to the strike class; establishing an exchange relationship between the main body model of the system and the system;
and introducing behavior rules to constrain the interaction relation to form an autonomous USoS participation model based on multiple agents.
2. The multi-agent based autonomous USoS participation model building method according to claim 1, wherein the following constraints are satisfied between the subject models:
3. The multi-agent based autonomous USoS participation model building method as claimed in claim 1, wherein the exchange relationship between the subject model and the system comprises 4 types of connection edge: investigation, communication, command and strike sides:
wherein the content of the first and second substances,is a type of the subject model, and is,is used for the formation of the main model,is the distance between the nodes and is,in order to determine the scope of investigation of the node,in order for a node to execute a state,is the communication range of the node or nodes,is the hit range of the node.
4. The multi-agent based autonomous USoS participation model building method of claim 1, wherein the behavior rules comprise battle closed-loop rules, movement rules and body model introduction rules.
5. The multi-agent based autonomous USoS participation model building method of claim 4, wherein the closed-loop rules comprise a threat-first reconnaissance edge rule, a near communication edge rule, and a weakest-first strike edge rule;
investigation edge connecting gaugeThen: for investigation classAccording to its scope of investigationThe threat degree of the inner target orders the targets, if the number of the targets in the detection range is less thanDetecting all the targets, namely forming detection edges with all the targets; if the number of targets in the investigation range is larger thanThen get the front of the rank of the target threatIndividual target and investigationForm a scout edge, whereinThe number of target nodes can be simultaneously detected for the detection channels of the nodes;
rule of communication side connection nearby: all ofAll have certain communication capacity aiming atTo its communication rangeInner distance of my square body modelIf the number of the principal models of our party in the communication range is less thanThen, forming a communication edge with all the principal models of our party; if the number of the principal models of our party in the communication range is larger thanThen take the distanceMinimum frontThe main body models of the individual parties form a communication edge;
weakest preferred strike alignment rule: class of current decisionCommand strikesAttacking a certain target and simultaneously attacking classes based on the own partyAll attribute information in the object listPosition, attack power, defense power and value information, and striking the class of our party by using a stable matching algorithmAnd enemies in the target listPerforming one-to-one or many-to-one matching to make the striking classA striking edge is formed between the two.
6. The method as claimed in claim 4, wherein the movement rules include "spring-loaded" movement rules, finger controlClose communication uniform movement rules, attacks and the likeA task point movement-oriented rule;
the "spring type" movement rule: defining investigation classesThe elastic force model of (2): c is toAll other scouts within the scout rangeThe center of gravity of the vehicle,is thatThe position of (a); findingAll others within the scope of investigation ofIs recorded as a set(ii) a Computing collectionsAll ofCenter of gravity of,
Wherein the content of the first and second substances,is composed ofIs determined by the coordinate of (a) in the space,is composed ofThe vector of the C is obtained by the method,as vectorsThe 2-norm of (a), the length,is composed ofThe vector to C is normalized,define the length and forward and reverse directions of the spring: greater than 0 means thatAnd the distance between C is less thanIn the investigation range ofBelonging to repulsive force, the direction is directed from CAnd is andthe larger the value, the larger the repulsive force; less than 0 means thatAnd the distance between C is larger thanThe range of investigation of (a) is,represents an attractive force, the direction isIs directed to C, anThe larger the value, the larger the attraction force;
secondly, based on the magnitude of the elastic forceThe direction and distance of movement of; order toIs composed ofIs obtained by obtainingThe relationship between them is as follows:
wherein the content of the first and second substances,is a unit mass, and therefore, is approximated by:andi.e. byThe position at the next moment is the vector sum of the current position and the current speed, and the speed at the next moment is the vector sum of the current speed and the current elasticity;
finger controlAnd (3) closing communication uniform movement rules: respectively calculateAttractive force ofAnd repulsive forceFrom which the resultant of the two forces is calculated as(ii) a Then, according toCalculating the speed and the position of the current position, the speed and the resultant force at the next moment;
percussion typeTask point-oriented movement rules: respectively calculateRespectively of two attractive forcesForce of mixingFrom this, the resultant of the two forces is calculated as:then according toThe current position, velocity and resultant force calculate the velocity and position at the next time,is the location of the target.
7. The multi-agent based autonomous USoS participation model building method of claim 4, wherein the subject model introduction rules comprise a saturation type new equipment node introduction rule, a new equipment node introduction rule according to requirements, and a new equipment node introduction rule at regular time;
introducing a new main body model rule by a saturation formula: according to newly added investigation classesDecision making classAnd strikesOnce the current amount of resources reaches the order of (2) generating the correspondingThe amount of resources required for the system,namely to generate one of the;
Introducing new subject model rules as required: class of investigation in the systemDecision making classAnd strikesWhen the existing quantity does not meet the demand quantity, the corresponding quantity is generated;
8. A multi-agent based autonomous USoS participation model building system comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the computer program implements the steps of the method of any of the preceding claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110822426.5A CN113283124B (en) | 2021-07-21 | 2021-07-21 | Multi-agent-based autonomous USoS participation model construction method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110822426.5A CN113283124B (en) | 2021-07-21 | 2021-07-21 | Multi-agent-based autonomous USoS participation model construction method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113283124A true CN113283124A (en) | 2021-08-20 |
CN113283124B CN113283124B (en) | 2021-10-15 |
Family
ID=77286824
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110822426.5A Active CN113283124B (en) | 2021-07-21 | 2021-07-21 | Multi-agent-based autonomous USoS participation model construction method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113283124B (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105893656A (en) * | 2016-03-24 | 2016-08-24 | 中国人民解放军装甲兵工程学院 | Platform-level Agent interactive simulation-oriented army unit combat test method |
CN109636699A (en) * | 2018-11-06 | 2019-04-16 | 中国电子科技集团公司第五十二研究所 | A kind of unsupervised intellectualized battle deduction system based on deeply study |
CN112347636A (en) * | 2020-11-04 | 2021-02-09 | 北京电子工程总体研究所 | Equipment guarantee simulation modeling method based on Multi-Agent technology |
CN112364500A (en) * | 2020-11-09 | 2021-02-12 | 中国科学院自动化研究所 | Multi-concurrency real-time countermeasure system oriented to reinforcement learning training and evaluation |
CN112749496A (en) * | 2020-11-25 | 2021-05-04 | 中国人民解放军国防科技大学 | Equipment system combat effectiveness evaluation method and system based on time sequence combat ring |
-
2021
- 2021-07-21 CN CN202110822426.5A patent/CN113283124B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105893656A (en) * | 2016-03-24 | 2016-08-24 | 中国人民解放军装甲兵工程学院 | Platform-level Agent interactive simulation-oriented army unit combat test method |
CN109636699A (en) * | 2018-11-06 | 2019-04-16 | 中国电子科技集团公司第五十二研究所 | A kind of unsupervised intellectualized battle deduction system based on deeply study |
CN112347636A (en) * | 2020-11-04 | 2021-02-09 | 北京电子工程总体研究所 | Equipment guarantee simulation modeling method based on Multi-Agent technology |
CN112364500A (en) * | 2020-11-09 | 2021-02-12 | 中国科学院自动化研究所 | Multi-concurrency real-time countermeasure system oriented to reinforcement learning training and evaluation |
CN112749496A (en) * | 2020-11-25 | 2021-05-04 | 中国人民解放军国防科技大学 | Equipment system combat effectiveness evaluation method and system based on time sequence combat ring |
Non-Patent Citations (6)
Title |
---|
AYA FUSANO等: "《Study of multi-agent based combat simulation for grouped OODA Loop》", 《SICE ANNUAL CONFERENCE 2011》 * |
BINGFENG GE等: "《Surrogate Model-Based Rapid Optimization for the Deployment of Combat System of Systems》", 《2020 IEEE INTERNATIONAL SYSTEMS CONFERENCE (SYSCON)》 * |
XIAOYU SONG等: "《A Cooperative Aerial Interception Model based on Multi-Agent System for UAVs》", 《2021 IEEE 5TH ADVANCED INFORMATION TECHNOLOGY, ELECTRONIC AND AUTOMATION CONTROL CONFERENCE (IAEAC)》 * |
叶国青等: "《基于使命和方法框架的武器装备体系结构建模》", 《火力与指挥控制》 * |
王晓君等: "《基于Multi-Agent的通信战场自主攻击》", 《太赫兹科学与电子信息学报》 * |
陈嘉勋等: "《基于Agent的建模与仿真技术在军事系统中的应用综述》", 《飞航导弹》 * |
Also Published As
Publication number | Publication date |
---|---|
CN113283124B (en) | 2021-10-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Ontanón et al. | A survey of real-time strategy game AI research and competition in StarCraft | |
Robertson et al. | A review of real-time strategy game AI | |
CN112269396A (en) | Unmanned aerial vehicle cluster cooperative confrontation control method for eagle pigeon-imitated intelligent game | |
Armstrong | General purpose intelligence: arguing the orthogonality thesis | |
Curran et al. | Using PCA to efficiently represent state spaces | |
CN112600795A (en) | Method and system for collapsing combat network under incomplete information | |
CN114326822B (en) | Unmanned aerial vehicle cluster information sharing method based on evolutionary game | |
Kamrani et al. | Learning objective agent behavior using a data-driven modeling approach | |
Tang et al. | Networking the boids is more robust against adversarial learning | |
Hoang et al. | Near-optimal adversarial policy switching for decentralized asynchronous multi-agent systems | |
Ma et al. | A survey of research on the distributed cooperation method of the uav swarm based on swarm intelligence | |
Yang et al. | WISDOM-II: A network centric model for warfare | |
CN113283124B (en) | Multi-agent-based autonomous USoS participation model construction method and system | |
Koopmanschap et al. | Tailoring a cognitive model for situation awareness using machine learning | |
de Campos Jr et al. | Evaluation of asynchronous multi‐swarm particle optimization on several topologies | |
Wu et al. | A game-based approach for designing a collaborative evolution mechanism for unmanned swarms on community networks | |
Rogers et al. | A micromanagement task allocation system for real-time strategy games | |
Sher | Dxnn platform: the shedding of biological inefficiencies | |
Albert et al. | Competitive evolution of tactical multiswarm dynamics | |
Yang | A networked multi-agent combat model: Emergence explained | |
Jiang et al. | Anti-drone policy learning based on self-attention multi-agent deterministic policy gradient | |
Wang et al. | A graph based simulation of reorganization in multi-agent systems | |
Chia et al. | Evolving swarm of UAVs | |
Yu et al. | Simulation of collective pursuit‐evasion behavior with runtime situational awareness | |
Frantz | Swarm intelligence for autonomous UAV control |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |