The Research and Simulation of a Prey Pursuit Algorithm

Abstract:

General prey pursuit problem used capture strategy between agents in limited grid space. In prey pursuit problem, there are a lot of problems such as occurrence of collision between agents, imperfection capture that can be produced from corner of restricted grid space. But, this causes actuality deficiency problem of experiment environment itself. In this paper, we propose an experiment environment that considers actuality and new capture strategy that using direction vector to solve prey pursuit problem that is typical experiment model of multi agent system. Therefore, we proposed continuous experiment space of grid space of circular type that is environment similar to the actuality world in this paper, and agents in proposed experiment space solved prey pursuit problem through direction vector that uses distance information and direction information between each others. The proposed method could capture prey effectively through using direction vector in new environment, and we solved imperfection capture and agent collision problem that happens in general research.

• Periodical: Advanced Materials Research (Volume 739)
• Pages: 522-526
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.739.522
• Online since: August 2013
• Authors: Wen Chuan Yang, Guang Jie Lin, Jiang Yong Wang
• Keywords: Agent-Based System, Capture Strategy, Direction Vector Heuristics, Prey Pursuit Problem

References

[1] S. Y. Kim, B. C. Kim and B. J. Yoon, Multi-agent Coordination Strategy using Reinforcement Learning. Proceedings of the 22th KIPS Fall Conference (2010) , September 5-8; Tokyo, Japan.

[2] P. Stone, M. Veloso, A Survey from a Machine Learning in Multiagent System, J. Information Processing Reports, 3, 125 (2007).

[3] T. Haynes, S. Sen. Evolving behavioral strategies in prey and predators. Edited G. W. Sandip, Springer Verlag, Berlin (2006).

[4] S. Sen, M. Sekaran and J. Hale, Learning to coordinate without sharing information, Proceedings of the 10th National Conference on Artificial Intelligence (2004); August 12-16, Los Angeles, USA.

[5] L. Stephens, M. Morris, The effect of agent control strategy on the performance of a DAI pursuit problem, Proceeding of the 2000 Distributed AI Workshop (2000); December 1-5, New York, USA.

[6] E. D. Jong, Multi-Agent Coordination by Communication of Evaluation. Proceeding of the 8th Europeon Workshop on MAAMAW 07 (2007) ; May 18-21, Berlin, German.

[7] H. Lee, B. C. Kim, Multiagent Control Strategy using Reinforcement Learning, J. The Database Transactions, 5, 175 (2004).