Multi-UAV Flight Control Based On Real-time Capacity Of Communication Channel

With the rapid development of multi-UAV technology in recent years,people gradually realize that communication is the critical factor restricting the effectiveness of multiUAV system.Under a distributed control framework,each UAV needs to interact with its neighbor UAV in real-time to achieve the goal.However,interferences between UAVs frequently occur due to the limited communication bandwidth,which leads to transmission delays and communication outages.As a result,the control performance is reduced dramatically.Since the control and communications are highly coupled,a cross-field perspective is necessary to handle this issue.More specifically,analyzing the fundamental relationship between multi-UAV control and communications performances plays a key role.In this thesis,we design the communication topology and the formation flight controller based on the physical layer communication model of multi-UAV systems to ensure effective communications and control.The main work includes:Considering the constraint of real-time channel capacity,we establish a communication model of multi-UAV,which is a basis for designing the communication topology.Based on the real-time communication model,the extreme communication region of the multi-UAV system is established,and the properties of the communication topology determined by the number and position of UAVS in the formation pattern are analyzed.According to the consensus protocol,we study three distance-based sufficient conditions for existing a spanning tree in the communication topology.A discrete-time controller of UAV formation flight based on the second-order integrator model and the unicycle model are designed under the constraint of realtime channel capacity.By utilizing the derived properties of the extreme communication region,we propose a method for developing robust and scalable communication topologies.Then a consensus-based discrete-time controller of multi-UAV formation flight was provided.Furthermore,we design a communication topology such that the multi-UAV formation flight problem is converted into decoupled trajectory tracking problems of a single UAV.To solve these problems,we give an accurate discretization method of the unicycle model and propose a globally asymptotically tracking controller based on optimizing a designed Lyapunov function.The effectiveness of both these controllers are proved rigorously.We design and establish a digital twin experimental system for multiple UAVs and verify the algorithms.The system includes three parts: the simulation subsystem,the physical subsystem,and the ground station subsystem.Note that the simulation subsystem is based on the XTDrone platform,and the physical subsystem is formed by a group of quadrotors and the Vicon motion capture system.As these three subsystems exchange information in a real-time manner,the real and virtual UAVs are strongly connected.Based on the experimental system,we corroborate the derived properties of the extreme communication region and the effectiveness of discrete-time controller of multiUAV formation flight.