Characteristics of and Approaches to Flocking in Swarm Robotics

摘要

One of the basic collective behaviors of swarm robotic systems is f locking , which has been actively studied for more than a decade and mimics a natural phenomenon where a group of animals move together as a single entity. In flocking, each robot in the flock tries to adjust its velocity and align with other robots in the flock while maintaining a predetermined pattern (formation) and avoiding collisions with obstacle and other members of the flock. This paper presents an up-to-date review on the characteristics of flocking problems and tasks, as well as solution approaches to flocking problems. We have addressed flock characteristics from the formation type, robustness, leader-follower, information, and communication aspects. Also, various flocking tasks of exploration, motion planning and navigation, shepherding, covering, object transportation, and simultaneous object collection and shepherding are investigated. Also, a new categorization of approaches to flocking is presented, which contains Leader-Follower, Behavior-based, Control-based, Fault-Tolerant, and Hybrid approaches. Finally, a comparative table on the characteristics of flocking problems appeared in various works of the literature is presented.