Design And Implement Of Fault Tolerant Control System Based On Quadrotor

Flying Control System (FCS) is the key system which determines the safety ofUnmanned Aerial Vehicles (UAV). The system which measures attitude and rate ofattitude is the key element that determines the safty and reliability of FCS. It is the focusof research to design fault-tolerant fly control system to solve sensor faults, and it’sdifficult for researchers and designers. Many researches are on fault-tolerant theory andapproaches, there is lack of excellent classifying plant.Quadrotor is a low-cost and portable aerial-vehicle which can vertical take off andland (VTOL) and move in6DOF. After all, quadrotor can play an excellent role toclassify these algorithms on fault-tolerant control.In order to design a fault-tolerant FCS and classify it on quadrotor, this thesiscompleted follow works:1. Completed the control law designed under normal situation with the kinematicand dynamic model. On the basis of the analysis of the quadrotor flightprinciple, established the kinematic and dynamic models; and then designedProportional-Integral-Derivative (PID) contoller, Mode-Reference-Adaptive–Contol (MRAC) controller, Backsteping contoller to Quadrotor’s attitude. Atthe end, analysed the robustness of the three controllers.2. Analyze threatness from the common faults such as sensor bias, loss ofaccuracy of sensor, sensor drift, frozen of sensor of the aerial vehicle whichwas controlled by PID controller and MRAC controller.3. Design the fault tolerant flying control system including a superfluous sensorssystem and fault tolerant controllers using fault observers based on PIDcontroller and MRAC controller.4. Implement the fault tolerant FCS and experiment. The implement included theattitude and heading system, the flying control computer. The experimentincluded classifications of superfluous sensors system, fault tolerant controllaw.This dessertation gets the fellow lines upon the work: 1. Using a proportional component and a differentiation component can stabilizethe attitude system of Quadrotor, and a damper feedbacking with rate of angleincreased the stability of the Quadrotor’s attitude.2. MRAC controller is more robust than PID controller with disturbance torque.3. Backstepping controller is designed based on nonlinear system. But the crosscoupling effect sensor noise is awfully negative for the flying control system.4. A superfluous system can effectively reduce the negative effect from the sensorfaults include sensor bias, sensor accuracy change, sensor drift, frozen sensor.5. The fault tolerant control law can effectively decrease the negtive effect fromthe sensor faults (especially sensor bias).