| The tilt-rotor dual Unmanned Aerial Vehicle(UAV)flight has the characteristics of short response time and can fly flexibly,it completes the power vectoring function of the UAV through tilt-rotor servo rotation.And the small tilt-rotor dual UAV can navigate in narrow passages and has greater advantages in surveying and delivering light objects.In terms of UAV control,the tilt-rotor dual UAV is an underdriven system with four inputs and six outputs,and the UAV is often subject to forces from turbulent wind fields and internally coupled moments during actual flight.These disturbances seriously increase the difficulty of stable flight of the UAV,and the self-anti-disturbance controller observes the system disturbances and compensates for the disturbances in the model,so as to achieve accurate control of the UAV.This thesis takes the developed tilting twin-rotor UAV as the object,and carries out the study of the control system under the rotor mode of the tilting twin-rotor UAV as well as the experiments of vertical take-off and landing(VTOL)and low-speed navigation,which are divided into four parts.In the first part,the force analysis of the simple tilting twin-rotor UAV was carried out,including rotor force and moment and wind field drag.A mathematical model of the UAV under turbulent wind field was established.Parameters such as windward area and mass under the coordinate plane of the UAV were obtained by instrumental measurements;The second part analyzes the characteristics of the UAV,designs the active disturbance rejection controller(ADRC).In this thesis,the second-order expansive state observer is improved by analyzing the ADRC,and proves the completion of the proof of convergence,and increases the observation speed of the expansive state observer at the same time under the premise of convergence.To address the problem of the wide range of parameters of the ADRC,a particle swarm algorithm(PSO)is designed to perform optimization in the case of difficulties in tuning the parameters using empirical methods,which solves the difficulties in manually tuning the parameters to obtain more accurate disturbance estimates,while the disturbance estimates are passed to the nonlinear error feedback for real-time compensation and output.In the concluding part of the designed controller,the controllers of the attitude inner and outer loop positions are designed;In the third part,starting from turbulent wind field simulation,MATLAB is used to validate the UAV control and verify the effectiveness of the ADRC against wind,and set up comparative control system simulations to test the disturbance observation capability and control accuracy of the improved and optimized ADRC before and after.The results show that the used ADRC can follow the fast observation speed and accurate control output,and the improved controller has better interference observation effect.The fourth part is the hardware selection of the UAV,the completion of the design of the tilting twin-rotor UAV and the completion of the outdoor flight experiment verification.In the actual flight process,a ADRC design with attitude inner loop and position outer loop is used.The outer loop monitors the position information in real time by sensors,and the attitude controls the attitude output by position decoupling.Flight experiments and simulations show that the improved ADRC can quickly track disturbances and achieve stable and accurate target functions,while the PSO solves the problem of complex empirical tuning parameters,thus designing an implementation of simple and accurate tracking UAV control. |