A general framework integrating exploration, self-assembly and locomotion control for swarm robots

In this paper, we proposed a general control framework for the swarm robots with self-assembly in real environments. There are three tasks: exploration task, self-assembly task and locomotion task. At the start, the swarm robots explore the environment using their sensors and make decision to form some robotic structures to finish the task. Then the swarm robots execute the self-assembly task to construct the corresponding configuration. For the locomotion, a unified CPG controller model as well as unity configuration representation is combined to generate locomotive motions for robotic structure and improve self-adaptation according to any configuration. Based on Sambot simulation platform, we demonstrate simulation experiments that the swarm robots form different robotic structures to adapt to the different environments. Finally, the effectiveness and scalability of the framework are discussed.