Print Email Facebook Twitter Autonomous Swarms of Tiny Flying Robots Title Autonomous Swarms of Tiny Flying Robots Author Li, S. (TU Delft Control & Simulation) Contributor de Croon, G.C.H.E. (promotor) Mulder, Max (promotor) Degree granting institution Delft University of Technology Date 2021-11-29 Abstract In the last decade, the research field of aerial swarms has grown at a rapid pace. These multi-robot systems possess desirable abilities including mobility in 3D spaces, efficient task execution in parallel, and redundant characteristics for fault tolerance. Many applications with multiple flying robots have already been demonstrated, such as light shows, search and rescue, area coverage, etc. Most studies for the above applications deal with position estimation, coordinated control, motion planning, or task assignments. However, the fundamental challenge remains to develop autonomous swarm systems that can work together and tackle real-world applications. As a special case of aerial swarms, multiple tiny (pocket-size) flying robots are safer and thus promising for real-world applications. These robots are highly limited in computation power and sensor capability, which makes the system design more challenging. An essential capability required for swarm coordination is that the individual robots are able to localize themselves with respect to others, preferably without the help of external infrastructure. Even though some works address the problem of onboard relative localization, the relative estimation is not accurate or consistent enough for precise swarm behaviors. This thesis investigates how to build a fully autonomous swarm of tiny aerial robots, featuring accurate relative state estimation and distributed control for different multi-robot tasks in unknown 3D environments. Subject Swarm roboticsMicro aerial vehiclesRelative localizationNonlinear predictive controlVisual deep learning To reference this document use: https://doi.org/10.4233/uuid:825f3a6b-3039-4a6e-8f3f-9f0871bd9ce5 ISBN 978-94-6366-472-1 Part of collection Institutional Repository Document type doctoral thesis Rights © 2021 S. Li Files PDF thesis.pdf 17.76 MB Close viewer /islandora/object/uuid:825f3a6b-3039-4a6e-8f3f-9f0871bd9ce5/datastream/OBJ/view