Design and control of a quad-rotor UAV

Abstract

In order to develop a Quad-Rotor UAV, a successive task from conceptual design to control design was performed. First, the dynamic model for control system design was derived by Euler-Lagrange and Newton’s $2^{nd}$ law. In order to satisfy the system requirements, the conceptual design of the Quad-Rotor UAV has been performed. The best issue for designing a MAV is the design of propulsion group. As any set of propulsion group, that is, battery, propeller and motor, there is a big difference of endurance, so we use an optimizer which searches the best combination. According to the result of conceptual design, electronic system design is performed. In order to test control algorithm and hardware system, the 3DoF flying mill was made with two microcontrollers. In system control, we use conventional adaptive sliding mode control that estimates the norm bound of lumped uncertainty. However, if we use the original approach, adaptation parameters can be drifted in real implementation, because restriction to a sliding surface can not always be achieved. In order to prevent this phenomenon, we apply modified sliding surface technique to the conventional adaptive sliding mode control. In addition, integral term is added to the sliding surface for preventing initial chattering and high gain. As the update law of the integral term, classical control method such as PD control can be used for equivalent control law of sliding mode control. Experiments using the PD control on the 3DoF flying mill, and numeral simulations using the proposed approach for real flight are performed.