Implementing Heterogeneous Agents in Dynamic Environments, a Case Study in RoboCupRescue
Design and construction of multi-agent systems is a challenging but an intriguing problem. It is because of the intrinsic distribution of the intelligent components. In such environments the interaction and communication between the constituent parts extends the complexity since appropriate coordination methods need to be designated and employed. In this paper a successful experiment in designing and implementing such an environment is presented. The test bed for this research is the rescue simulation environment. The architecture of the implemented heterogeneous agents takes advantage of various algorithms. These algorithms make the agents act intelligently by themselves albeit they happen to act quite in coordination with each other. The implemented algorithms for the sake of cooperation between the heterogeneous agents enhance the overall pay off of the system. The autonomy of the agents is guaranteed by means of some methods such as reinforcement learning, decision trees and some sort of heuristic functions. In order to settle the agents in coordination with each other and make them act cooperatively, some other methods have been applied. Among these methods, combinatorial auctions, coalition formation, function approximation for evaluating the value of cooperation, and some probabilistic and heuristic methods can be named.
KeywordsPath Planning Coalition Formation Coalition Structure Heuristic Function Combinatorial Auction
Unable to display preview. Download preview PDF.
- 1.Ahmadi, M., Sayyadian, M., Habibi, J.: A Learning Method for Evaluating Messages in Multi Agent Systems. In: Proceedings of the Agent Communication Languages (ACL 2002), Italy, Bolognia (2002)Google Scholar
- 2.Deshpande, A., Godbole, D.N., A.G., Varaiya, P.: Design and Evaluation Tools for Automated Highway Systems. Hybrid Systems, 138–148 (1995)Google Scholar
- 3.Hunsberger, L., Grosz, B.J.: A Combinatorial Auction for Collaborative Planning. In: Proceedings of the Fourth International Conference on Multi-Agent Systems (ICMAS-2000), pp. 151–158 (2000)Google Scholar
- 4.Kitano, H., Tadokor, S., Noda, H., Matsubara, I., Takhasi, T., Shinjou, A., Shimada, S.: Robocup-rescue: Search and rescue for large scale disasters as a domain for multi-agent research. In: Proc. of the IEEE Conference on Systems, Men, and Cybernetics (1999)Google Scholar
- 5.Noda, I., Stone, P.: The RoboCup Soccer Server and CMUnited Clients: Implemented Infrastructure for MAS Research. Journal of Autonomous Agents and Multi-Agent Systems (2002)Google Scholar
- 7.Van Roy, B.: Learning and Value Function Approximation in Complex Decision Processes, Ph.D. Thesis, Massachusetts Institute of Technology, 30 (May 1998)Google Scholar
- 8.Sandholm, T., Lesser, V.: Issues in Automated Negotiation and Electronic Commerce: Extending the Contract Net Framework. In: Proceedings of the First International Conference on Multi-Agent Systems (ICMAS 1995), pp. 328–335. The MIT Press, Cambridge (1995)Google Scholar
- 9.Simon, D., Espiau, B., Castillo, E., Kapellos, K.: Computer-aided design of a generic robot controller handling reactivity and real-time control issues. RR-1801 (1995)Google Scholar
- 11.Zurel, E., Nisan, N.: An Efficient Approximate Allocation Algorithm for Combinatorial Auctions. In: Proceedings of ACM Conference on Electronic Commerce (EC 2001) (2001)Google Scholar