skip to main content
article

The multilevel hypernetwork dynamics of complex systems of robot soccer agents

Published:01 June 2007Publication History
Skip Abstract Section

Abstract

A mathematical formalism is sketched for representing relational structure between agents. n-ary relations, n > 2, require hypernetworks, which generalize binary relation networks. n-ary relations on sets create structure at higher levels of representation to the elements in multilevel systems. The state of a system is represented by its multilevel relational structure. The dynamics of a system are represented by state changes through time. These can be continuous with no change in the hypernetwork topology, but often they are not. Controlling such systems involves taking actions intended to result in desirable state changes. The concept of multilevel hypernetwork can be applied to multiagent systems in general.

References

  1. Asada, M., Kitano, H., and Veloso, M. 1999. Robocup: Today and tomorrow---what we have learned. AI 110, 1 (Jan.), 193--214. Google ScholarGoogle Scholar
  2. Ashby, W. R. 1956. Cybernetics. Chapman Hall, London, UK.Google ScholarGoogle Scholar
  3. Atkin, R. H. 1974. Mathematical Structure in Human Affairs. Heinemann Educationak Books, London, UK.Google ScholarGoogle Scholar
  4. Atkin, R. H. 1977. Combinatorial Connectivities in Social Systems: An Application of Simplicial Complex Structures to the Study of Large Organizations. Birkhäuser, Basel, Switzerland.Google ScholarGoogle Scholar
  5. Berge, C. 1989. Hypergraphs. North-Holland Mathematical Library, Amsterdam, The Netherlands.Google ScholarGoogle Scholar
  6. Durfee, E. 2001. Scaling up agent coordination strategies. IEEE Comput. 34, 7 (July), 39--46. Google ScholarGoogle Scholar
  7. Fajstrup, L., Raussen, M., and Goubault, E. 2006. Algebraic topology and concurrency. Theoret. Comput. Sci. 357, 1-3, 241--278. Google ScholarGoogle Scholar
  8. Goubault, E. and Raussen, M. 2002. Dihomotopy as a tool in state space analysis: Turorial. Theoretical Informatics: 5th Latin American Symposium (LATIN'02). Google ScholarGoogle Scholar
  9. Hilton, P. and Wylie, S. 1965. Homology Theory: An Introduction to Algebraic Topology. Cambridge University Press, Cambridge, UK.Google ScholarGoogle Scholar
  10. Iravani, P. 2005. Discovering relevant sensor data by q-analysis. Robot Soccer World Cup IX (RoboCup'05). Springer-Verlag, Berlin, Germany. Google ScholarGoogle Scholar
  11. Iravani, P. and Johnson, J. 2005. The emergence of a visual communication language in robotic football. Towards Autonomous Robotics Systems (TAROS-05). Essex University, UK.Google ScholarGoogle Scholar
  12. Iravani, P., Johnson, J., and Rapanotti, L. 2005. Robotics and the Q-analysis of behavior. In Proceedings of the 10th International Conference on Artificial Life and Robotics (AROB 10) M. Sugisaka, Ed. Oita University, Japan.Google ScholarGoogle Scholar
  13. Jelle, K., Matthijs, T., and Nikos, V. 2005. Non-communicative multi-robot coordination in dynamic environments. Robotics Autonom. Syst. 50, 2-3 (Feb.), 99--114.Google ScholarGoogle Scholar
  14. Johnson, J. 2001. Visual communication in swarms of intelligent robot agents. Artific. Life Robotics 5, 1--9.Google ScholarGoogle Scholar
  15. Johnson, J. 2006. The hypernetworks of multilevel complex systems. In European Conference on Complex Systems. Oxford, UK.Google ScholarGoogle Scholar
  16. Johnson, J. and Price, B. 2003. Complexity science and representation in robot soccer. In Robot Soccer World Cup VII (RoboCup'03). Padova, Italy. Springer-Verlag, Berlin, Germany.Google ScholarGoogle Scholar
  17. Johnson, J. and Simon, J. 2001. Fundamental structures for the design of machine vision systems. Mathemat. Geol. 33, 3, 331--352.Google ScholarGoogle Scholar
  18. Johnson, J. H. 2007. Hypernetworks in the Science of Complex Systems. Imperial College Press, London, UK. Google ScholarGoogle Scholar
  19. Law, J. and Johnson, J. H. 2004. Discrete dynamics and the space time possession game. In FIRA World Congress. Busan, South Korea.Google ScholarGoogle Scholar
  20. Von Bertalanffy, L. 1969. General Systems Theory. George Braziller, New York, NY.Google ScholarGoogle Scholar
  21. Wooldridge, M. and Jennings, N. R. 1995. Intelligent agents: Theory and practice. Knowl. Engin. Rev. 10, 2.Google ScholarGoogle Scholar

Index Terms

  1. The multilevel hypernetwork dynamics of complex systems of robot soccer agents

            Recommendations

            Reviews

            Maria L. Gini

            Hypergraphs can be used as a tool to represent relationships among agents. This paper focuses on RoboCup, but includes outlines on how hypergraphs can be used for other applications, such as chess and machine vision. The idea is to capture the n -ary relationships that exist among groups of agents, and to construct a multilevel hypernetwork, where higher levels correspond to a more abstract view of the relationships between the agents. In RoboCup, these relationships are dynamic as players move in the field. At the lowest level, the nodes of the hypernetwork represent pairs of players. As the levels increase, the nodes represent groups of players, and then roles, such as attack and defense. At the highest level, the root node represents the entire team. Sequences of actions of the players produce trajectories in this multidimensional space, and can be analyzed and controlled via actions taken by the players. The hope is that hypernetworks will be tools for the agents to recognize good and bad trajectories, and to act accordingly. The paper relates the use of hypernetworks to the science of complex systems and to emergence, but the connection is only marginally relevant to what is presented. The paper is rich with examples, which makes it easy reading, but it is short on precise definitions, which can be frustrating. For example, I could not find a description of how a hypernetwork differs from a hypergraph. The theory presented is intriguing, and has the potential to change the way interactions among agents in multiagent systems are modeled. Unfortunately, the paper lacks evidence that the theory is capable of producing the desired results. The paper contains detailed examples of situations in RoboCup, but the analysis is done manually after the game, and no algorithm is presented. Online Computing Reviews Service

            Access critical reviews of Computing literature here

            Become a reviewer for Computing Reviews.

            Comments

            Login options

            Check if you have access through your login credentials or your institution to get full access on this article.

            Sign in

            Full Access

            PDF Format

            View or Download as a PDF file.

            PDF

            eReader

            View online with eReader.

            eReader
            About Cookies On This Site

            We use cookies to ensure that we give you the best experience on our website.

            Learn more

            Got it!