Experimaental Evaluation of a Distributed Control System for Chain-type Self-reconfigurable Robots

Self-reconfigurable robots are an instance of complex systems that can autonomously rearrange their modules. Recently, there has been a great interest in using these robots in different domains such as reconnaissance, rescue missions, and space applications for their versatility, fault tolerance, low cost of reproducibility, and efficiency in rough terrains. Because of their distributed nature, the control of self-reconfigurable robots is a challenging task. This dissertation contributes a totally distributed control architecture that unifies the control of self-reconfiguration, locomotion, and manipulation tasks in chain-type self-reconfigurable robots. DISTINCT, D-BEST, and FEATURE are three distributed algorithms that provide solutions to the problems of ‘Task Negotiation’, ‘Behavior Collaboration’, ‘Synchronization’ and ‘Autonomous Discovery and Functional Response to Topology Change’ which are fundamental in the field of self-reconfigurable robots control. All modules cooperate in problem solving process, which results a fault tolerant architecture. A special type of content-base messages, called ‘Digital Hormones’, have been used for communication among modules.