Adaptive Routing System by Intelligent Environment with Media Agents
In this paper, we consider a distributed robotic system that includes special agents that convey the information. We address the issue of selecting one course from two;a long one-way detour or a short two-way path on which traffic jams may occur. We consider a system in which the environment, instead of mobile agents, learns feasible parameters for task execution. To correct problems with this system and improve it we introduce media agents that carry data for the learning. They adjust information flow. We formulate the system and evaluate its performance.
KeywordsMobile Robot Media Agent Reinforcement Learning Mobile Agent Average Cost
Unable to display preview. Download preview PDF.
- 1.Jun Ota et al.:Distributed Strategy-Making Method in Multiple Mobile Robot System, DARS, 123–133, 1994.Google Scholar
- 2.Yuji Yoshimura et al: Iterative Transportation Planning of Multiple Objects by Cooperative Mobile Robots, DARS2, 171–182, 1996.Google Scholar
- 3.Luc Steels: Cooperation Between Distributed Agents Through Self-Organization, Decentralized AI, Elsevier Science publishers, pp. 175–196, 1990.Google Scholar
- 4.Alexis Drougoul and Jacques Ferber: From Tom Tumb to the Dockers: Some Experiments with Foraging Robots, Animals to Animats 2, pp. 451–459, 1992.Google Scholar
- 5.Daisuke Kurabayashi et al.:Distributed Guidance Knowledge Management by Intelligent Data Carriers, Int. robotics & automation, Vol. 16, No. 4, pp. 207–216, 2001Google Scholar
- 6.R. Beckers, et al.:From Local Actions to Global Tasks:Stigmergy and Collective Robotics, Artificial Life IV, pp.181–189, 1994.Google Scholar
- 7.Daisuke Kurabayashi, et al.: Performance of Decision Making: Individuals and an Environment, IEEE Int. Conf. on Intelligent Robots and Systems, 2831/2836, 2002.Google Scholar