Degree

Doctor of Philosophy (PhD)

Department

Mechanical and Industrial Engineering

Document Type

Dissertation

Abstract

This research focuses on the distance-based formation control of multi-agent systems in two- and three-dimensional spaces. Rigid graph theory is used to describe the interactions between agents. Lyapunov theory, input-to-state stability, and/or switched system theory are used to design and analyze the proposed formation controllers. To deal with the existing issue of convergence to incorrect equilibrium points of distance-based formation controllers, several methods are proposed by introducing additional controlled variables, viz., the signed area or edge angle of a triangle-like structure in two dimensions (2D), and the signed volume of a tetrahedron-like structure in three dimensions (3D). In this research, we seek to generalize these methods to planar and spatial formations of N agents for both first- and second-order agent models while using unidirectional sensing and control. In addition to formation shape control, we also consider the problem of formation motions. We prove that the decentralized controllers ensure the almost-global asymptotic stability of the correct formation and are coordinate frame invariant. Computational and experimental evaluations are presented to support the theoretical results. The experiments were conducted on a 2D ground vehicle testbed. An aerial vehicle testbed for future 3D experimentation is also presented. We conclude this dissertation with a discussion of aerial robotic network and remaining issues in this research.

Committee Chair

de Queiroz, Marcio

Available for download on Monday, March 08, 2027

Share

COinS