ZAM Quadrotor VTOL UAV: Prototype Development and Control Translator Dynamic Modelling

Motivated by the need to control a quadrotor tail-sitter unmanned aerial vehicle (QTUAV), the thesis presents the development of a method to convert user inputs and estimated measurements of QTUAV states into control set-points. Pinder suggests control of the QTUAV using an optimal controller to achieve zero angular momentum (ZAM) and desired set-points [1]. The QTUAV dynamic model is dependent on previous states and has several phases of flight, which range from hover to forward flight. An Euler dynamic model relating desired set-points, system states, propeller velocities and angular momentum is developed in the work. A pilot familiar with control of a Cessna-172 airplane is to command the craft. Cessna-172 airplane user inputs are transformed into desired set-points using a dynamic model of a Cessna-172 airplane. The user is provided a single intuitive concept of the system and is able to provide commands to the quadrotor without knowledge of the quadrotor dynamics. However command limitations exist due to differences between the user’s concept of the flight envelope and the actual flight envelope. Alternative dynamic models that represent the user-system concept can be chosen to allow control translation and accommodate flight envelope differences. Components for a prototype quadrotor helicopter were selected, and a quadrotor helicopter prototype constructed. Joystick inputs were converted into desired control states, transformed into a pulse-period-modulated (PPM) signal and output to a remote controller through a stereo sound port, using MATLAB. This basic control translator was used to control a tethered quadrotor however the inertia of the quadrotor was not sufficiently high to allow adequate control. Thrust testing is completed to determine variation in thrust coefficients and equipment capabilities. The propeller assembly was attached to a base which was placed upon a digital balance to measure rotational speed and thrust at various controller settings. A coefficient of thrust of 2.94 x 10-6 N/rpm2 was experimentally determined for the APC-E 7x5 propeller. The selected electronic speed controller (ESC) did not provide speed control independent of power supply, showing that controller setting control is inadequate to control the motor rotational speeds . Measurements of either propeller rotational speeds or thrusts are required in the control system to provide feedback of control set-points. This work has accomplished further development of a method of control for a QTUAV. A dynamic model was developed which provides the relationships between current and future system states. It reviews current developments in unmanned aerial vehicles and presents several limitations to this method of control. Future development of the dynamic model is required to enable complete attitude representation using the unit quaternion. Additionally, thrust and drag parameters must be determined for the wing. Furthermore, the dynamic models are to be modified to attain optimising functions. Implementation of the dynamic model will require the development of an observer to provide information about the craft states