Abstract
Hybrid, or multi-modal, aerial vehicles which combine two or more flight configurations have gained popularity due to their increased flight envelope and versatility. Typical Vertical Takeoff and Landing (VTOL) hybrids combine multirotor type aircraft with fixed-wing type aircraft. In this dissertation, the author has proposed a unique and novel UAV platform that takes inspiration from nature, the falling of the maple seed. To achieve flight, the UAV spins its entire body around a central axis to generate lift. By adopting a dual-winged configuration, it is also capable operating in three flight regimes (hover, tailsitter, cruise), with the added advantage of increased flight efficiency during stationary hover compared to many existing hybrids. This is the first such work where a monocopter-type rotary craft achieves and demonstrates full hybrid functionality in a structurally efficient manner. Power usage during this rotary hover mode is found to be at least 30 % less than the traditional tailsitter hover mode. Dynamic models in 6-DOF were also formulated for the purpose of analysis and prediction of flight characteristics and performance as well as optimizations via simulation. Several control strategies for the novel hovering mode were explored, including reinforcement learning where a neural network policy for position control was successfully trained, and the adoption of a nature-inspired cyclical control strategy, the Central Pattern Generator (CPG). Flight transition control strategies and mechanism were also developed and discussed, including open-loop trajectory optimization, trajectory optimization paired with closed-loop angular rate, and control blending via a transition coordinator. Various optimizations using different strategies were also performed on both mechanical design and control parameters which resulted in better flight efficiency and control performance when applied. The prototype with the optimized design and control were shown to consume 18 % less power on average and up to 56 % improved control performance. The optimized blended control for transition showed 20 % less altitude variation and 5 % less throttle usage. Numerous prototypes were built for experimental validation and multiple fight tests were also conducted in real-world outdoor conditions demonstrating full three-mode operation