Flight Control Of Tilt Rotor UAV Basing On HOSVD During Transition Mode

A tilt rotor aircraft has the flight characteristics of both the fixed wing aircraft and the helicopter.It can not only fly a long distance at a high speed,but also has a low requirement of takeoff site.Moreover,it can switch flight mode among helicopter mode,fixed wing mode and transition mode to adapt to different fight condition.Because of these characteristics,people think highly of the application of tilt rotor aircraft in both military and civilian fields.It is a hot and difficult problem to design a controller for the transition mode of tilt rotor aircraft.Taking a small tilt rotor aircraft as a research object,the issues related to the transition mode are studied.The main work is as follows:Firstly,for abetter research about the transition mode of tilt rotor aircraft,we decouple the full scale model and obtain the longitudinal model of transition mode when the aircraft satisfies the requirement of non-sideslip flight.Secondly,aiming at the longitudinal model of transition mode,a series of LTI models which can reflect the behaviors of the original system within the entire flight envelope are obtained through Jacobian linearization.Then LPV systems are constructed respectively by direct fitting method and HOSVD method.By comparing the modeling process and the modeling accuracy of two different modeling methods,the conclusion that HOSVD(Higher Order Singular Value Decomposition)method has a less computational burden and a higher accuracy is drawn.Finally,through the SOS(sum of squares)technique and Lyapunov stability theory,a controller for the LPV approximate model of transition mode from fixed wing mode to helicopter mode is designed.The result of the simulation verifies the effectiveness of model transformation and controller design method.To enhance the robustness of the controller with respect to the uncertainties of the aerodynamic parameters and eliminate the influence of exogenous disturbance,a disturbance observer is designed.Moreover the simulation with sinusoidal perturbation and parameter perturbation verifies its effectiveness.