Coordination and control of multi-robot systems

Abstract

Modern technology facilitates the versatile use of robotic systems in multiple facets of life. The increasing need for efficient automation propels research in multi-robot systems (MRSs), essential for tasks like surveillance and search and rescue missions. Optimally covering an environment requires a framework that drives the robotic team to cover multiple distinct areas in the environment, be capable of reacting to time varying features in the environment and in the team itself, and optimally exploit the heterogeneity in the robot team at all times. The work presented in this thesis is a step in this direction. More specifically, this thesis presents the design, implementation and validation of a holistic framework tailored for the coverage control application using an MRS. Current literature on this research area focuses on solving one singular aspect that affects the coverage control of the MRS, such as introducing energy awareness in the control algorithm, or tracking the time varying features of the environment. However, to the best of our knowledge, to date no work in literature simultaneously addresses multiple factors that typify most real-life coverage control applications. To this end, the novel modular framework proposed in this thesis exploits the heterogeneity in the robot team by optimally allocating robots to particular regions of relatively higher importance in the environment, according to the requirements of each region and the capabilities of the robots. Furthermore, the environment is smartly segmented into cells according to the energy reserves of the robot, its sensors’ performance and its maximum velocity, and each robot is made to cover one cell. Finally, a novel energy-aware controller that accurately and efficiently drives the robots to their respective cells and keeps tracking them is designed. The proposed control law is shown to converge the MRS to a centroidal voronoi tessellation (CVT). All these algorithms are combined into a non-trivial modular framework that addresses various practical aspects for the coverage control of a given environment. Finally, comprehensive and realistic simulations are employed to test and validate each module and the complete framework.