Adaptive Flocking of Robot Swarms: Algorithms and Properties

摘要

This paper presents a distributed approach for adaptive flocking of swarms of mobile robots that enables to navigate autonomously in complex environments populated with obstacles. Based on the observation of the swimming behavior of a school of fish, we propose an integrated algorithm that allows a swarm of robots to navigate in a coordinated manner, split into multiple swarms, or merge with other swarms according to the environment conditions. We prove the convergence of the proposed algorithm using Lyapunov stability theory. We also verify the effectiveness of the algorithm through extensive simulations, where a swarm of robots repeats the process of splitting and merging while passing around multiple stationary and moving obstacles. The simulation results show that the proposed algorithm is scalable, and robust to variations in the sensing capability of individual robots.