A Collaborative Multi-Agent Model with Knowledge-Based Communication for the RoboCupRescue Simulation

An effective cooperation model in a Multi-Agent system is required to enable agents to interact and achieve their task proficiently. To address this problem, we propose a collaborative hierarchical Multi-Agent model based on two layers (Administrant layer and Autonomous layer), which makes use of the knowledge-based communication to actualize interaction in this model. Task planning and allocation are implemented in the adminktrant layer by using U-Tree algorithm, where reward function and evaluation value are introduced to control selection factors. Agents are allowed to work together using architectures like swarm and information update value in the autonomous layer, so that centralized and decentralized control can be integrated effectively. Moreover, this collaboration model has been applied successfully in CSU'Yunlu RoboCupRescue simulation team to show its advantages over other approaches. which will be about to participate in internationa1 RoboCupRescue 2006 in June.

[1]  Gerry Dozier,et al.  Adapting Particle Swarm Optimizationto Dynamic Environments , 2001 .

[2]  Russell C. Eberhart,et al.  A new optimizer using particle swarm theory , 1995, MHS'95. Proceedings of the Sixth International Symposium on Micro Machine and Human Science.

[3]  Andrew McCallum,et al.  Reinforcement learning with selective perception and hidden state , 1996 .

[4]  Milind Tambe,et al.  Team Formation for Reformation in Multiagent Domains Like RoboCupRescue , 2002, RoboCup.

[5]  Russell C. Eberhart,et al.  Comparison between Genetic Algorithms and Particle Swarm Optimization , 1998, Evolutionary Programming.

[6]  James Kennedy,et al.  The particle swarm: social adaptation of knowledge , 1997, Proceedings of 1997 IEEE International Conference on Evolutionary Computation (ICEC '97).