Multi-Agent Coordination by Communication of Evaluations

A framework for coordination in multi-agent systems is introduced. The main idea of our framework is that an agent with knowledge about the desired behavior in a certain domain will direct other, domain-independent agents by means of signals which reflect its evaluation of the coordination between its own actions and their actions. Mechanisms for coordination are required to enable construction of open multi-agent systems. The goal of this investigation was to test the feasibility of guiding an agent with coordination evaluation signals, and furthermore to gather experience with instantiating the framework on a testbed domain, the Pursuit Problem. In the testbed system, agents have been created which choose their actions by maximizing the coordination evaluation signals they will receive. The performance of these agents turned out to rank among the best results encountered in literature, and behavior guided by coordination evaluation signals can thus be concluded to be useful in this domain.

[1]  William E. Lorensen,et al.  Object-Oriented Modeling and Design , 1991, TOOLS.

[2]  G. Reeke.  Marvin Minsky, The Society of Mind , 1991, Artif. Intell..

[3]  Walter Van de Velde.  Cognitive Architectures From Knowledge Level To Structural Coupling , 1995 .

[4]  Luís Torgo,et al.  Panel: Learning in Distributed Systems and Multi-Agent Environments , 1991, EWSL.

[5]  Jeffrey S. Rosenschein,et al.  Cooperation without Communication , 1986, AAAI.

[6]  M. Benda,et al.  On Optimal Cooperation of Knowledge Sources , 1985 .

[7]  Sarit Kraus,et al.  The role of representation in interaction (abstract): discovering focal points among alternative solutions , 1992, SIGO.

[8]  Sandip Sen,et al.  Learning to Coordinate without Sharing Information , 1994, AAAI.

[9]  H. Maturana,et al.  The Tree of Knowledge: The Biological Roots of Human Understanding , 2007 .

[10]  Sandip Sen,et al.  Multiagent Coordination with Learning Classifier Systems , 1995, Adaption and Learning in Multi-Agent Systems.

[11]  Yves Kodratoff,et al.  Machine Learning — EWSL-91 , 1991, Lecture Notes in Computer Science.

[12]  Sarit Kraus,et al.  Coordination without Communication: Experimental Validation of Focal Point Techniques , 1997, ICMAS.

[13]  Sandip Sen,et al.  Evolution and learning in multiagent systems , 1998, Int. J. Hum. Comput. Stud..

[14]  Jeffrey S. Rosenschein,et al.  A game theoretic approach to distributed artificial intelligence and the pursuit problem (abstract) , 1992, SIGO.