Research On Fault Diagnosis And Fault Tolerant Control Of Quadrotor UAV

As researchers continue to conduct in-depth research,the application technology level of the four-rotor UAV in control performance and fault tolerance has been improved,the functions are more diversified,and the handling is getting better and better.In this context,the Quadrotor UAV is more widely used in military,agricultural and film and television shooting fields.Since the quadrotor UAV is an under-driven,strongly coupled,non-linear system,it puts forward higher requirements on the stability of the control system.Moreover,due to the complex structure of the four-rotor UAV and the light weight of the material,the failure caused by the excessive burden of parts has also attracted the attention of many scholars.The actuator of the Quadrotor UAV has a particularly heavy workload and is more prone to failure due to its high intensity of use.If effective fault-tolerant control cannot be performed after the actuator fails,it will result in incomplete flight missions,damage to the airframe and even serious consequences for loss of life and property.Therefore,in this paper,a faulttolerant controller is designed for the problem of partial failure of the actuator,so as to make the Quadrotor UAV maintain normal flight or performance loss within an acceptable range after the failure of the actuator.Based on the above problems,the following research was conducted:Firstly,the flight principle of the Quadrotor UAV and the operating mechanism of the actuator under each flight action are introduced.The frame coordinate system and ground coordinate system of the Quadrotor UAV and the conversion mode between the two coordinate systems are established.According to the dynamic principle of the Quadrotor UAV,the nonlinear models of angular motion and linear motion are established.Finally,the mathematical model of the actuator failure of the Quadrotor UAV is presented.Second,the controller parameters of the quadcopter UAV are optimized to improve the control performance.Aiming at the failure problem of the partial failure of the actuator of the four-rotor UAV,a fault diagnosis mechanism based on deep belief network is designed to classify different fault states of the actuator.Then,a faulttolerant controller based on fault compensation strategy is designed,which is to design corresponding fault-tolerant control laws for different fault situations,and multiple control laws are combined into a fault-tolerant controller.Through the fault information obtained by the fault diagnosis subsystem,and by calling the corresponding fault-tolerant control law,the partial failure fault of the actuator is processed.Finally,Based on the deep belief network fault diagnosis,the failure of diagnosis algorithm is optimized.In order to further improve the accuracy and practicality of fault diagnosis,considering the situation of single actuator failure or double actuator simultaneous failure,a support vector machine fault diagnosis system based on genetic algorithm optimization is designed.On the basis of designing a fault-tolerant controller using a fault compensation mechanism,a fault-tolerant controller that combines fault compensation and sliding mode control is designed to ensure the fault-tolerant capability after a fault and improve the robustness of the system.