Abstract: 
The growing domain of aerial robotics necessitates advancements in the control strategies and robustness of over-actuated and fully-actuated aerial vehicles. This thesis proposal makes contributions to this endeavor by providing in-depth analysis and methodologies concerning these vehicles, control allocation strategies during actuator failures, high-fidelity simulations, and a unified control framework. Our completed work has primarily revolved around the design and control of two categories of aerial vehicles: tilted hexarotors and tiltrotor Vertical Takeoff and Landing (VTOL) vehicles. This thesis proposal presents a fault-tolerant design for a tiltrotor VTOL, demonstrating an efficient, dynamic control allocation that can adapt to actuator failures. This work also presents a comprehensive simulation of an electric VTOL (eVTOL) passenger aircraft, incorporating both multirotor and fixed-wing dynamics. Additionally, this work proposes a sampling-based Model Predictive Control (MPC) method using Model Predictive Path Integral (MPPI), that simultaneously controls all 6 degrees of freedom of an aerial manipulator without needing phase-specific controllers. Looking forward, we propose various avenues of exploration. These include generalizing and optimizing our unified controller to operate in real-time and work with over-actuated platforms, incorporating dynamics learning and residual learning, and developing a path planner in full state/control space. With these future works, we aim to further enhance the performance and robustness of over-actuated and fully-actuated aerial vehicles.

Thesis Committee Members: 
Sebastian Scherer, Chair

Zachary Manchester

Wennie Tabib

Junyi Geng, PSU