Coordination of mobile multi-robot systems
This thesis is concerned with the coordination problems in multi-robot systems. A methodology for building a multi-robot system is introduced. General coordination tasks are presented as well as existing architectures. In addition general taxonomies of multi-robot systems are reviewed. It is shown that environment and task set the requirements to the system in respect to different aspects of these classifications. In the end a line of action considering all reviewed factors is presented. A coordination problem in form of a path finding problem is stated. To solve this problem an algorithm that can be classified as swarm architecture is presented. This algorithm is simulated by a modified cellular automaton. General information about cellular automata is reviewed and the modifications in respect to classical cellular automata are discussed. By analyzing simulations of several cases involving paths with obstacles and without obstacles as well as varying numbers of agents, the strengths and weaknesses of this algorithm are discussed. The work done has shown that the modified cellular automaton approach is efficient in solving path finding tasks. Cooperation by following simple rules without any explicit communication was achieved in simulations. Finally a mobile robot kit is introduced. A kinematics analysis is done as well as a derivation to determine its position based on the assumptions of the developed algorithm. Required changes to the robot kit are presented that enable the robot to perform the path finding problem stated out.
"Coordination of mobile multi-robot systems"
Dissertations and Master's Theses (Campus Access).