Research On Game Decision-making And Cooperative Communication Methods For Multi-UAV

With the advantages of low cost,high flexibility,convenient deployment and no casualties,unmanned aerial vehicle(UAV)has been received extensive attention in military and civil applications.As multi-UAV coordination could make up for the limitation of single-UAV,it plays an important role in complex mission scenarios.However,the multi-UAV coordination can't do without the support of decision-making and communication technologies.Among them,optimal task decision-making is the core to achieve cooperative control.Adaptive swarm clustering is the basis of improving network scalability.Efficient routing selection is the key to realize long-distance communication.And reasonable resource and trajectory optimization is the premises to guarantee the communication performance.Therefore,this thesis studies the task decision-making,swarm clustering,routing selection,resource allocation and trajectory optimization problems for the purpose of realizing efficient missions of multi-UAV.The main research contents are as follows:(1)The multi-UAV task decision-making problem under complex scenarios is investigated.Firstly,large-scale UAV task decision-making for non-adversarial scenarios is studied.Considering the importance of tasks,the number of UAVs and fuel consumption,a non-adversarial task decision-making method based on coalition game is proposed.Further considering the dynamic and uncertain nature of the UAVs and the tasks,as well as the influence of the limited communication range of the UAVs,a task reallocation method is proposed to realize the dynamic adjustment of the decision result.This is used to solve the problem of slow convergence and poor adaptability of traditional task decision-making methods.Secondly,multi-UAV task decision-making for adversarial scenarios is studied.Considering the complex game and dynamic stochastic characteristics of the adversarial process,an adversarial task decision-making method based on stochastic game is proposed to maximize the benefits of own side.This is used to provide the optimial strategy under the adversarial environment and solve multi-stage game decision problems.(2)The UAV swarm clustering problem is investigated with a cooperative relationship.Considering that the rapid movement of UAVs will cause frequent updates of communication network topology and unstable communication links.And UAVs performing the same task have highly similar movement patterns and frequent information exchange requirements.A UAV clustering method based on coalition game is proposed.In this method,a coalition game model of swarm clustering is established.After that,a three-stage solution algorithm including UAV switching set generation-switch operation establishment-optimal switch operation selection is designed to divide the UAV network into multiple connected areas.This is used to solve the difficult problem of large-scale network management and improve the network stability.(3)To meet the long-distance information transmission requirement,the multi-UAV routing selection problem is investigated.Condering that some UAVs are far away from the ground station,which leads to the limitation of the communication range and rate of the direct transmission link.A multi-UAV routing selection method based on network formation game is proposed.In this method,a network formation game model of routing selection is established.Then,a solution algorithm to find the pure-strategy and mixed-strategy for UAVs is designed to plan a the information transmission path to the ground station for each UAV.This is used to solve the problem of remote information transmission between the UAVs and the ground station.(4)The resource allocation and trajectory optimization problem for the UAV communication is investigated.Considering that there is no route from UAV to ground station,relay UAV is deployed to assist communication.Under the communication requirements and energy constraints of mission UAVs as well as the energy and mobility constraints of relay UAV,in order to maximize the network throughput,an iterative method based on convex optimization is designed to reasonably optimize not only the time and power of the mission UAVs,but also the power and flying trajectory of the relay UAV.This is used to solve the remote communication problem between UAVs and ground station without routing.