SUBJECT: M.S. Thesis Presentation
BY: Brady Anderson
TIME: Tuesday, November 14, 2017, 11:00 a.m.
PLACE: MARC Building, 101
TITLE: Adaptive Control for Modular Vertical Lift Air Vehicles
COMMITTEE: Dr. Jonathan Rogers, Chair (ME)
Dr. Aldo Ferri (ME)
Dr. Jun Ueda (ME)


Modular vertical lift air vehicles are comprised of individual vehicles, such as quadrotors, connected to a payload for the purposes of cooperative aerial manipulation and transportation. This method of payload manipulation has unique advantages in terms of scalability, flexibility, and robustness to vehicle failure due to redundancy. However, there are numerous difficulties with designing and controlling such systems ranging from mechanical complexity of attachment devices to control allocation and robustness to uncertainty.

Various aspects are explored in regards to modular air vehicle design and control. A simulation model is first described which supports high-fidelity simulation of vehicle-payload attachment, payload transportation, and vehicle-payload detachment. A control algorithm is designed to enhance cooperative payload manipulation. Simulation studies using the new control method examine flight performance and energy consumption in example scenarios where modular vertical lift vehicles cooperatively transport various payloads, including bluff bodies and airfoil shapes. Further studies analyze the effectiveness of an extended Kalman filter in providing key values to the model inversion based controller.

The mechanical design of a modular vertical lift vehicle is provided, including experimentally-derived estimates of thrust and energy consumption. Overall, results show that payload manipulation and transportation via modular vertical lift vehicles is a feasible idea for certain classes of payloads.