Role of stochasticity in self-organization of robotic swarms

摘要

This paper investigates the effectiveness of designed random behavior in self-organization of swarm of robotic agents. Inspired by the self-organization observed in biological cells and the role played by random forces in providing robustness in cell self-organization, we investigate the possibility of designing a decentralized controller for a swarm of agents in which the stochastic process is included. This paper considers flocking as a self-organizing behavior example to validate our findings. The controller is designed in the framework of Lyapunov function, and it is based on the artificial potential due to interactions among agents. Our analysis shows that the flocking behavior of the swarm is improved and is more robust when the stochastic process is included in the agent controller.