UAV Swarm Formation And Obstacle Avoidance Strategy Based On Time-varying Network Topologies

Unmanned aerial vehicles(UAVs)have the ability to perform tasks that cause danger or difficulty to humans,so they are becoming more and more popular in practice.Currently,UAVs have entered a period of rapid development,where UAV swarms are seen as the future development trend.The UAV swarm composed of multiple UAVs is a very flexible multi-UAV system.It has strong collaboration capabilities and the ability to perform tasks autonomously,which can meet the needs of increasingly complex tasks in future.So it has received much attention.The traditional control methods for multi-UAV systems are usually based on pre-designed instructions,in which each UAV is fully controlled,and the UAV cannot move autonomously,facing the challenge of autonomous decision-making and processing based on real-time environment and operating conditions.For example,when the UAV swarm encounters unknown obstacles,it is difficult for the UAVs to avoid the obstacles in time,causing the UAVs to crash or fail the mission.This thesis aims to propose an efficient UAV swarm formation and obstacle avoidance strategy.Under this strategy,the UAV swarm has the advantages of both orderliness and autonomy,so that the navigation of the UAV swarm can adapt to the changing and complex environment.Then,the flight efficiency and survival rate of UAVs cna be guaranteed.The contributions in this thesis are summarised as follows:1.The UAV swarm formation controller is designed.Based on the movement characteristics of the UAV,a second-order UAV system model is established.In view of the poor network communication environment and system delay,a formation controller based on time-varying network topology and a formation controller based on time-varying delay are designed,respectively.We give a rigorous analysis of system stability,and obtained the appropriate feedback control matrix and the sufficient conditions for system convergence.The correctness of the algorithm is verified through simulation experiments,and the speed of system convergence is clearly illustrated.2.The UAV swarm obstacle avoidance controller is proposed.By studying the biological intelligence swarm algorithm that simulates the behavior of animal swarms in nature,the UAV swarm avoidance controller is designed,which can simulate the behavior of animal swarms to avoid predators,so that the UAVs in the UAV swarm can also autonomously avoid dynamic obstacles while avoiding internal conflicts.Simulation experiments demonstrates that the obstacle avoidance controller responses to dynamic obstacles sensitively,and the system is stable during the process of avoiding obstacles.3.A switching scheme between two kinds of controllers is designed,according to the characteristics of the distributed system.When any UAV detects an obstacle,it will switch to the obstacle avoidance controller,without considering the status of the surrounding UAVs.An appropriate obstacle detection distance and the internal safe distance are set according to the size of UAVs,so that the system can respond in time.When the UAV does not detect an obstacle,it will switch to the formation controller for efficient flight.This switching scheme ensures that the UAV system is still stable during the switching process.