Research On UAV Swarm Confrontation Based On Complex Network And Colonel Blotto Game

All military powers are developing UAV swarm combat technology.UAV swarm combat will become an important combat pattern in the future.Distributed control has become the mainstream method for UAV swarms due to its advantages of robustness.However,due to the highly dynamic situation and communication interference under confrontation conditions,the previous distributed confrontation tactical design method based on God’s perspective cannot adapt to today’s swarm confrontation.The swarm confrontation game method is a feasible way to solve the current technical bottleneck of swarm confrontation.The first is to build a swarm confrontation model to establish a full understanding of the elements of swarm confrontation.The second is to analyze the vulnerability of swarms.The third is to coordinate the internal control of the swarm to give full play to the advantages of swarm consensus consistency in distributed control.The fourth is to choose the optimal strategy for swarm confrontation,and use advantages to attack the enemy’s weaknesses.This paper takes the UAV swarm combat as the military background,takes the swarm confrontation strategy game as the main line,guides the design of the UAV swarm confrontation strategy through the analysis of the swarm vulnerability,and improves the response speed within the swarm through the rapid consensus of the swarm consensus.Provide key technical support for future UAV swarm operations.The main work and innovations are as follows:1.Based on complex network and game theory,a swarm confrontation game model is constructed.In the UAV swarm confrontation,under the condition of limited communication,the situation of both sides changes highly dynamically,and the dynamic change of the swarm is described and analyzed by using the powerful complex network and the convenient measurement tools.Using the idea of game theory,a model of swarm confrontation game is constructed.2.Based on the capability generation mechanism,the swarm vulnerability is analyzed.From the perspective of the mechanism of swarm capability generation,the complexity of the swarm is discussed,and the vulnerability of the swarm is analyzed from three aspects: the swarm command structure,the swarm control method and the invulnerability of the topology structure.In terms of the complexity of the swarm system,the analysis and verification of the number of individuals,the complexity of individuals,the form of individual interaction,the heterogeneity of individuals,and the interaction between humans and the swarm have an impact on the complexity of the swarm.The group size effect,structure effect and vulnerability summarized are applied to the strategy design of the confrontation,and it is concluded that when our forces are small,we will ”divide and attack” the opponent,and when our forces are sufficient,choose to encircle and destroy them.3.Based on the characteristics of the small world network,a swarm consensus algorithm is designed.In the swarm confrontation scenario,when the swarm communication capability is limited and the topology structure is highly dynamic,the consensus consistency of the swarm is difficult to achieve or the convergence is slow.Focusing on accelerating the convergence of the swarm consensus,relying on the Vicsek basic model,based on biological swarm intelligence and small World network characteristics,swarm fast consensus algorithm is designed.Based on the method of graph theory,it is analyzed that the factor affecting the convergence speed of swarm consensus is the algebraic connectivity of the graph.Aiming at the real needs of distributed control within the swarm,a new method of building a distributed small-world network based on edge connection is proposed,which can adapt to the dynamic communication topology.Inspired by the swarm intelligence of birds,a method is proposed to interact with only six neighbors,which greatly reduces the communication cost.In this paper,five swarm consensus algorithms,including the basic comparison algorithm,are proposed,and the efficiency and economy of the algorithm are verified by experiments.The fifth swarm consensus algorithm based on swarm intelligence and small-world network characteristics can improve the overall response speed of the swarm,more than double the convergence speed,and reduce the communication cost by more than one-fifth,enabling the UAV swarm to adapt to the rhythm faster swarm strategies.4.The effectiveness of the designed tactical strategy is verified based on the XTdrone swarm tactical simulation platform.First,based on the group size effect and structure effect summarized in the swarm vulnerability analysis,a tactical strategy of ”divide and strike” is designed.Secondly,the research on swarm consensus consistency realizes the rapidity of swarm decision-making under the condition of swarm distributed control,which is conducive to swarm execution of time-sensitive strategies,so the tactical strategy of blitz is designed.The above two tactical strategies are applied to the swarm tactical simulation,forming an overwhelming advantage over the opponent,and quantitatively analyzes the correlation between the win rate and Lanchester’s law.5.Based on the Colonel Blotto game model,a group confrontation method for UAV swarms is proposed.Due to the huge number of UAVs in the swarm confrontation,it is impossible to configure the UAV confrontation strategy separately.It is proposed to divide the UAV swarms into groups,so that the two opposing swarms can fight on different battlefields.A model based on the Colonel Blotto game is constructed to improve the efficiency of swarm confrontation.At the same time,space constraints of battlefield are introduced,that is,a fixed-size battlefield space can only meet the confrontation requirements of a limited number of UAVs.The concept of boundary contact rate is proposed.Based on the nonlinear optimization method,combined with Lanchester’s law and double oracle algorithm,a strategy game equilibrium solution method for UAV swarm confrontation is designed.The optimal strategy of boundary contact rate and group confrontation is revealed.Through the experimental analysis,the law of optimal configuration of the grouping strategy in the swarm confrontation is obtained.All these provide important technical support for future UAV swarm confrontation.