Quadrotor Tailsitter Scalability, Design, and Control: Towards Efficient Aerial Mobility

Abstract

This thesis investigates the scalability, design, and control challenges of tailsitter unmanned aerial systems, which uniquely combine the vertical takeoff and landing capabilities of multirotors with the efficient forward flight characteristics of fixed-wing aircraft. In a tailsitter configuration, multiple rotors provide lift and control during hover that also generate thrust for forward flight, while fixed wings supply additional lift. The transition between hover and forward flight is achieved by rotating the entire vehicle about its pitch axis. The mode-switching operation introduces highly nonlinear dynamics, complicating efforts to accurately model, design, and control such vehicles. To address these challenges, this work develops empirical scalability laws derived from historical data, offering practical guidelines for component mass distribution, wing sizing, and performance estimation during the conceptual design phase of tailsitters. These principles are applied to the design of a 5 kg prototype, serving as a case study. Furthermore, a software-in-the-loop simulation is established to concurrently test the aircraft dynamics and the flight controller. Enhancements to the attitude controller are proposed to improve aircraft performance during the transition and forward flight, with a particular focus on mitigating oscillations in forward flight common to tailsitters with a single flight mode. The controller gains are adapted online using a fuzzy logic scheme based on measured error, error derivative, and aircraft pitch, resulting in improved responsiveness and stability as verified through simulation and test flights with the prototype aircraft. In summary, this thesis outlines empirical correlations for estimating preliminary design parameters of tailsitters and proposes a method to develop control algorithms for such vehicles.