부산대학교 도서관

Many group of organisms that live in proximity are capable of complicated collective movement, which occurs via periodic oscillation of the individuals. One of the fundamental goal for swarm robotics study is to understand how effective and robust the collective behaviors can emerge from simple principles of interaction. In this context, we characterize the motion behaviors of swimming animals that move through lateral body undulation with multi-link robots and demonstrate that phase adaption can be achieved through the proposed control law. Critically, the ability to achieve phase synchronization is driven through contact interactions between adjacent robots without the requirement for network communication or motion planning. Through bio-inspired experiment and simulation, we demonstrate that collective in-phase and opposite-phase behaviors can arise passively through intermittent collision events, and furthermore we provide principles for control architecture to achieve self-organization in swarms.