【作者】 刘欢；

【导师】 宋永端； 代东林；

【作者基本信息】 重庆大学 ， 工程（控制工程）（专业学位）， 2022， 硕士

【摘要】 四旋翼无人机系统具有控制较灵活、结构比较简单、成本低等特点,并且能够垂直起降和自主悬停等,在各个领域都具有至关重要的作用。但无人机的工作环境比如电力巡检等,都存在着气流、阻力等对飞行器的外界干扰因素。此外,四旋翼无人机系统是比较复杂且耦合性强的欠驱动系统,执行器若发生故障,那么无人机将大大降低飞行质量从而无法完成预定任务。因此,为减少由外界干扰、执行器故障等带来的安全隐患,对无人机飞行过程进行容错控制具有重要的现实意义。另外,由于无人机自身的物理特性,若因超负荷作业导致驱动饱和从而有可能发生安全事故,因此其位置和姿态的控制输入输出也需要约束在一定的范围内,以避免安全隐患的发生。因此,基于以上难点,本文主要针对四旋翼无人机执行器故障和驱动饱和的问题,综合考虑系统模型的不确定性、外界环境扰动、输入受限等影响因素,对四旋翼无人机系统进行建模,并利用鲁棒性或RBF神经网络开发一种新的PID自适应容错控制算法,主要工作包括:(1)针对四旋翼无人机的系统模型进行数学建模,将经典的无人机本体模型扩展成综合性的系统模型。综合考虑无人机受各种环境扰动等问题,同时将四旋翼无人机的驱动故障以及执行器输入饱和的问题考虑其中,建立了考虑各种不确定性因素以及驱动故障和输入受限的四旋翼无人机欧拉模型。(2)针对上述考虑了各种因素的无人机欧拉系统模型,研究了基于经典PID算法的系统控制方案,分析了在考虑执行器故障和驱动饱和的系统情况下,出现外界扰动时,控制器的调节方式和容错控制能力,以及系统的跟踪效果。(3)针对四旋翼无人机的欧拉系统,开发了一种基于执行器故障和驱动饱和的新的自适应PID容错控制解决方案,该算法考虑系统输出受限的情况下使用鲁棒性研究了两种控制算法,使其同时具有结构简单的PID形式,且方案中的PID时变增益由分析算法自动更新,无需手动调节三个PID增益参数,研究的控制方案不仅对外部干扰具有鲁棒性,对未知虚拟参数具有自适应能力,而且对不可预知的驱动故障具有容错能力。(4)通过理论推导、仿真分析充分证明了本文中提出的改进式PID的控制方案的有效性。该方案能够保证整个无人机闭环系统内的信号都有界,能够使状态结果输出有界,并且可以进一步通过实验仿真验证所设计的控制方案的可行性。并最终通过三维仿真进行飞行测试验证算法的有效性。更多还原

【Abstract】 The quadrotor UAV system has the characteristics of simple structure,flexible control,low cost,and can take off and land vertically and autonomously hover,which plays a vital role in various industrial and social fields.However,the working environment of the UAV,such as power inspection,environmental monitoring,etc.,has external interference factors such as airflow and resistance to the aircraft.In addition,the quadrotor UAV is a complex under-actuated system.If it fails,the UAV will greatly reduce the flight quality and cannot complete the intended mission.Therefore,in order to reduce the potential safety hazards caused by external interference and actuator failure,it is of great practical significance to carry out fault-tolerant control of the UAV flight process.In addition,considering the safety and the physical characteristics of the UAV itself,its location and the control input and output of attitude also need to be limited within a certain constraint range.Therefore,based on the above difficulties,this paper mainly focuses on the problem of quadrotor UAV drive failure and actuator input saturation,and comprehensively considers the uncertainty of the system model,external environmental disturbances,input limitations and other influencing factors.The dynamic modeling of the system is carried out,and a new PID adaptive fault-tolerant control algorithm is developed by using the robustness or RBF neural network.The main work includes:(1)Carry out mathematical modeling for the system model of the quad-rotor UAV,expand the classic UAV ontology model into a comprehensive dynamic model,analyze the force of the quad-rotor UAV structure,and comprehensively consider the UAV’s stress various environmental disturbances and other issues,while considering the drive failure of the quadrotor UAV and the problem of actuator input saturation.A dynamic model of the quadrotor UAV considering various uncertain factors,drive failures and input constraints is established.(2)Aiming at the above-mentioned Euler system model of the UAV considering various factors,the system control scheme based on the classical PID algorithm is studied.And the control scheme of the control system is analyzed considering the drive fault and actuator input saturation,and the external environmental disturbance occurs.The adjustment method and fault-tolerant control ability of the controller,as well as the tracking effect of the system.(3)For the Euler system of the quadrotor UAV,a new adaptive PID control solution based on drive failure and actuator input saturation was developed,and two control algorithms were studied using robustness to make it structured at the same time.Simple PID structure,and the PID time-varying gain in the scheme is automatically updated by the analysis algorithm,and there is no need to manually adjust the three PID gain parameters.The researched control scheme is not only robust to external disturbances,but also adaptive to unknown virtual parameters.It is also fault tolerant against unpredictable drive failures.(4)The effectiveness of the improved PID control scheme proposed in this paper is fully proved through theoretical derivation and simulation analysis.The control scheme can ensure that all signals in the closed-loop system are bounded,and the output of state results can be bounded,and the feasibility of the designed control scheme can be further verified by experimental simulation.And finally through the physical flight test to verify the effectiveness of the algorithm.更多还原