Clustering multiple robots with serialisation

Kambayashi, Yahiko; Satta, Keisuke; Ko, Shibata; Sugiyama, Shota; Takimoto, Munehiro; Yamachi, Hidemi

doi:10.1504/ijiids.2014.066641

Cited by 1 publication

(1 citation statement)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We have successfully instructed a group of robot carts to self-organize into multiple lines. [18][19][20] Oikawa et al have taken advantages of the pheromone communication of ants and proposed a multi-robot system to compose formations. [21][22][23] In their method as well as Mizutani and Shintani's proposals, they have implemented ants and pheromones as mobile software agents and demonstrated that it is feasible to use them to compose two-dimensional formations.…”

Section: Introductionmentioning

confidence: 99%

Formation Control of Swarm Robots Using Mobile Agents

Kambayashi

Yajima

Shyoji

et al. 2019

Vietnam J. Comp. Sci.

Self Cite

View full text Add to dashboard Cite

In this paper, we propose an algorithm for controlling a fleet of swarm robots that construct three-dimensional forms. The swarm robots coordinate with each other through network communication, and compose formations such as polyhedrons presented as spherical coordinates. Our control algorithm achieves communication through mobile software agents, which introduce control programs to robots that initially have no information about the formation. Mobile software agents are autonomous objects that can migrate from one robot to other robots through a communication network and can deliver control programs as they are needed. We have made our swarm robot system to mimic the behaviors of the leafcutter ants. A leafcutter ant is a typical social insect and uses pheromone for communication. In our robot control system, we have implemented ants and pheromones as mobile software agents. We call the mobile software agents that drive the mobile robots as ant agents, and call the other agents that provide communication as pheromone agents. The ant agents drive the swarm robots to locations identified by the pheromone agents. Each ant agent has only partial information. There is no need for either a central control or an agent that has the entire design of the formation. In order to diffuse the partial information among the neighboring robots, each ant agent generates pheromone agents and dispatches them to the surrounding robots. Dispatched pheromone agent looks for a proper ant agent to influence towards a desired relative location. It is the ant agent that actually drives the robot by following the guidance of the pheromone agent, and the collective actions of ant agents and pheromone agents achieve the composition of the objective formation. We have implemented a simulator based on our algorithm and conducted numerical experiments. The results demonstrate that our mobile robot control system is feasible and efficient in practice in practical situations.

show abstract