Research On The Flight Control Method Of A Thrust-vectored Tailsitter UAV

Combining the features of fixed-wing aircraft and rotorcraft,tailsitter has the advantages of high cruise speed,long endurance and vertical takeoff and landing ability,which bring a wide application prospect.However,the strong coupling of dynamics and the singularity of Euler angles make the design of controller more difficult.In this paper,the thrust-vectored tailsitter is studied,and researches on system modeling,multi-sensor data fusion algorithm,attitude/altitude control algorithm and flight verification method of tailsitter are carried out.The main work is concluded as follows:Based on the dynamic characteristics of the thrust-vectored tailsitter,a six-degree-of-freedom nonlinear dynamic model containing feature of thrust vector is established;A simplified method of aerodynamic model is proposed,and then a calculation method of aerodynamic force coefficient is given;Modeling method of the thrust vector system is proposed,and the inverse solution equation of control input is obtained by quadratic fitting.In order to solve the problem of Euler angle singularity during hovering flight,a new vertical/horizontal Euler angles attitude solution method is proposed,which has the advantages of low computation cost and clear physical meaning;A data fusion attitude calibration algorithm is proposed,which combines the measurement data of various airborne sensors,and finally obtains the attitude of the tailsitter with good dynamic characteristics and high accuracy;A horizontal position estimation algorithm based on extended Kalman filter is proposed,as well as a data fusion altitude algorithm for indoor and outdoor flight environment,which can obtain more accurate horizontal position and altitude.According to the characteristics of thrust-vectored tailsitter flight states,the whole flight envelope state switching algorithm is designed,which can realize reliable separation from takeoff bracket,vertical hovering,state transition,horizontal flight and controllable forward landing;A new hybrid model reference adaptive attitude/altitude control algorithm is given,which can improve the dynamic response performance and robustness of the controllers by combining feedback controller and model reference adaptive controller.New platforms for attitude calibration,horizontal position estimation and altitude estimation data fusion test are designed to solve the problem of real value acquisition;Through the data fusion test,the effectiveness of multi-sensor data fusion attitude calibration algorithm,horizontal position estimation algorithm based on extended Kalman filter and multi-sensor data fusion altitude estimation algorithm is verified;A new aerodynamic configuration of tailsitter is proposed and the thrust-vectored tailsitter is developed;Through free-fly vertical takeoff and landing test and whole state flight test,the robustness of proposed state switching algorithm and attitude/altitude controllers is verified.