Controller Design For Quad-rotor Unmanned Vehicle

The model of quad-rotor unmanned helicopter is under-actuated and of strong couplings,which has high demands for control system.To design a well-performance control system,one type of the helicopter is employed as research object.After proposing a dynamic model and analyzing its dynamical characteristics,three kinds of flight controllers are designed considering the external disturbance and structural uncertainties.Simulations demonstrate the efficiency of all proposed controllers.The flight theory of quad-rotor is presented at first.After detailed stress analysis,numerical model is established.To prepare for the following controller design,simulations are carried out,which reveal the coupling,controllability and other flight characteristics of the model.Consider designing a typical PID controller for the vehicle.On account of the fact that the model is under-actuated and of strong-coupling,nonlinear coefficients are omitted and the control system is separated into two parts: one for position demands tracking and another for controlling flight attitudes,which simplify the process of PID feedback laws design obviously.The efficiency and fragile robustness of the designed controllers is proved though simulation results.In order to enable quad-rotors with strong robustness against structural uncertainties,Lyapunov method is adopted to deduce robust control law.The control system is still consisted of double loop.Then large uncertainties are introduced into the model when simulation,the obtained results demonstrate that robust control system is able to track the position and attitude commands overcoming hybrid disturbances.However,the theory analysis show that the designed controller is of conservatism,and the loss of control performance occurs in the large attitude angle maneuver.For the control problem of quad-rotors with large attitude angle,NDI(Nonlinear Dynamic Inversion)method is accepted.The system of vehicles is divided into fast circuit,faster circuit and slow circuit based on Time-Scale Separation.The original NDI method is augmented to NMRDI(Nonlinear Model Reference Dynamic Inversion)method by bringing in a reference model to deal with structural uncertainties.What's more,a sliding mode observer is designed to observe the real-time disturbance during the flight and provide additional control quantities for the compensation of bad effects caused by the disturbance.Simulations show that sliding mode disturbance observer is able to observe the disturbance accurately and the nonlinear controller is of better performance compared with robust one under the condition of the same uncertainties,also there seems no loss of control efficiency occurring in large attitude angle flight,which implies the rationality of proposed NMRDI controller.