无人机集群近些年来在农业、工业和军事等多个领域取得长足发展，其编队安全问题日益受到重视，包括执行器故障和通信拓扑故障等在内的无人机集群装备故障引起广泛关注。为提高无人机集群编队安全性能，本论文以固定翼无人机集群为研究对象，将一般执行器故障与执行器失效导致的通信拓扑故障相结合，针对严重程度不同的执行器故障，从编队拓扑重构、执行器故障诊断与编队容错控制三个方面研究了解决方案。本文的主要研究内容包括：(1) 以固定翼无人机为对象完成了运动学建模、网络控制关系建模与执行器故障建模。在已有模型基础上，基于Matlab/Simulink 搭建了软件仿真平台，基于Pixhawk 飞控与Rflysim 开源系统搭建了硬件在环仿真平台。(2) 针对执行器失效引起的通信中断问题，分析了无人机集群网络对编队控制的影响，区分数据链网络与控制关系网络，分别基于无人机通信节点与连通边设计了编队拓扑重构算法和编队重构流程，完成了仿真与对比分析，对算法可行性进行了验证。(3) 针对无人机执行器故障问题，基于控制网络特性分析了无人机编队等效系统，设计了分布式执行器故障残差生成器，同时基于残差评价函数与诊断阈值，从编队整体角度实现执行器故障诊断，同时在软件仿真平台和硬件在环仿真平台上验证了算法的有效性。该故障诊断方法可通过设定连通关系使分布式执行器故障诊断算法应用于单无人机。(4) 针对不同严重程度的无人机集群执行器故障问题，基于一致性控制律和领从式无人机编队结构提出了在切换拓扑条件下的编队鲁棒容错控制方法，解决了编队重构拓扑切换问题与执行器故障问题。在给定故障场景和累积协同误差评价指标下，与现有编队控制律完成了对比仿真与分析，验证了算法的可行性。总结上述研究工作，提出了编队安全系统集成方案，为理论研究和工程应用提供充分可行的思路，有助于无人机集群系统装备的发展。
Unmanned aerial vehicle (UAV) swarm has made great progress in recent years inmany fields, including agriculture, industry and the military. The formation security problemof UAV swarm has been paid more and more attention in practice, including actuatorfaults and communication faults, etc. In order to improve the safety performance of UAVswarm formation, this paper takes fixed-wing UAV swarm as the research object, combineingthe general actuator fault with the communication topology fault caused by theactuator failure and aiming at the actuator faults with different severity, researches thesolution from three aspects: formation topology reconfiguration, actuator fault diagnosisand formation fault-tolerant control. The main research work of this paper includes:(1) The kinematic modeling, network control relationship modeling and actuatorfault modeling of fixed wing UAV were completed. Based on the existing model, a softwaresimulation platform is built based on Matlab/Simulink, and a hardware-in-the-loopsimulation platform was built based on Pixhawk flight controller and Rflysim open sourcesystems.(2) The influence of UAV swarm network on formation control is analyzed. Datalink network and control relationship network are distinguished, and formation topologyreconfiguration is considered under actuator failure. The formation topology reconfigurationalgorithm and the formation reconfiguration process is designed based on UAVcommunication node and connected edge. Simulation and comparative analysis is completed.(3) Aiming at the UAV actuator fault problem, the equivalent system of UAV formationis analyzed based on the characteristics of control network. The distributed actuatorfault residual generator is designed. At the same time, based on the residual evaluationfunction and diagnosis threshold, the actuator fault diagnosis is realized from the perspectiveof the whole formation, and the effectiveness of the algorithm is verified onboth software simulation platform and hardware-in-the-loop simulation platform. Thedistributed actuator fault diagnosis algorithm can be applied to single UAV by setting thecommunication relation.(4) Based on the fault of UAV swarm actuators with different severity, a formationrobust fault-tolerant control with switching topology is proposed to solve the topology switching problems of formation reconfiguration and actuator fault problems. Under thegiven fault scenario and the accumulative cooperative error evaluation index, the simulationand analysis are completed compared with the existing formation control law, wherethe feasibility of the algorithm is verified.Finally, the research work is summarized, and the scheme of formation safety systemintegration is proposed, which provides a feasible idea for theoretical research and engineeringapplication. It is helpful to the development of UAV swarm system equipment.
