Modeling And Control System Design Of A Small Unmanned Helicopter

Unmanned helicopter can complete the tasks which manned aircraft and fixed-wing UAV could not finish because of its unique ability. It is becoming an important part of modern warfare. Some key technologies of modeling and flight control for unmanned helicopter were studied based on a RC helicopter.The helicopter is a Multiple-Input/Multiple-Output (MIMO) system with highly coupled, unsteady, and nonlinear characteristics. The dynamic model of the system was complex. The rigid equations was established in this paper by analyzing the relationship between the ground coordinate system and the body coordinate system, the main rotor flapping equations were established by analyzing the movement of main rotor, and the forces and moments balance equations were established by analyzing the forces of all parts of the body (main rotor, tail rotor, aerodynamic drag, fuselage, tail fin). Finally, the whole nonlinear model of the unmanned helicopter was established, and was linearized by using small perturbation theory and Taylor principle.Some parameters were unknown in the established model of unmanned helicopter, and need to be identified, so the parameter identification method based on flight data and the least square method was studied in this paper. Firstly, the attitude data collection system was designed, including the hardware design and software debugging of the digital signal processing chip, inertial measurement unit, wireless transmission, GPS, digital compass. And the input and output data of helicopter was obtained under manual fly. Then the identification method of linear model and nonlinear model based on least squares method was studied, and the identification results of two methods were compared.After system parameters were identified, the attitude control system for the hover state was studied in this paper. And the LQG controller and H∞controller based on Kalman filter were designed respectively, in which, LQG controller is the combine of LQR controller and Kalman filter, and can effectively deal with system dynamic problems and the gaussian noise in the data measurement. And H00 controller can reduce the impact of uncertain factor, such as system error and unmeasurable external disturbance on the control performance. Finally, the effectiveness of the above controller was shown through simulation.