TY - GEN
T1 - HactarV2
T2 - 9th International Workshop on ProgrammingMulti-Agent Systems, ProMAS 2011
AU - Dekker, Marc
AU - Hameete, Pieter
AU - Hegemans, Michiel
AU - Leysen, Sebastiaan
AU - Van Den Oever, Joris
AU - Smits, Jeff
AU - Hindriks, Koen V.
PY - 2012/8/15
Y1 - 2012/8/15
N2 - In this paper we report on the design and implementation of our multi-agent system, called HactarV2, for the Agent Contest 2011. HactarV2 has been implemented in the agent programming language Goal. One of the main challenges of the Agent Contest is to design a decentralized multi-agent system that is able to strategically compete with other agent teams. To address this challenge, the strategy of HactarV2 is based on implicit coordination between agents and there is no central manager that keeps track of all information. The aim, moreover, has been to minimize the communication between agents. Communication is used by HactarV2 agents to ensure that each of them maintains the same map of the environment. The Mars scenario of this year required agents to explore, locate and occupy high valued zones on the planet Mars. Because initially agents are randomly placed on the map, in the first phase of the game the agents individually explore the map and update each other. Agents have different roles and we describe the strategies used by individual agents per role. In the second phase of the game, which starts when the agents have located high value nodes on the map, the agents group together and act as a swarm to maintain and possibly expand the zone on the map that is occupied by the agents.
AB - In this paper we report on the design and implementation of our multi-agent system, called HactarV2, for the Agent Contest 2011. HactarV2 has been implemented in the agent programming language Goal. One of the main challenges of the Agent Contest is to design a decentralized multi-agent system that is able to strategically compete with other agent teams. To address this challenge, the strategy of HactarV2 is based on implicit coordination between agents and there is no central manager that keeps track of all information. The aim, moreover, has been to minimize the communication between agents. Communication is used by HactarV2 agents to ensure that each of them maintains the same map of the environment. The Mars scenario of this year required agents to explore, locate and occupy high valued zones on the planet Mars. Because initially agents are randomly placed on the map, in the first phase of the game the agents individually explore the map and update each other. Agents have different roles and we describe the strategies used by individual agents per role. In the second phase of the game, which starts when the agents have located high value nodes on the map, the agents group together and act as a swarm to maintain and possibly expand the zone on the map that is occupied by the agents.
UR - http://www.scopus.com/inward/record.url?scp=84864843624&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84864843624&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-31915-0_10
DO - 10.1007/978-3-642-31915-0_10
M3 - Conference contribution
AN - SCOPUS:84864843624
SN - 9783642319143
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 173
EP - 184
BT - Programming Multi-Agent Systems - 9th International Workshop, ProMAS 2011, Revised Selected Papers
Y2 - 3 May 2011 through 3 May 2011
ER -