Title :
Emergent behavior research based on multi-agent coevolution mechanism
Author :
Xue Hongtao ; Yang Guang ; Pan Liang ; Yin Dong
Author_Institution :
Nat. Univ. of Defense Technol., Changsha, China
Abstract :
The process of coevolution exists among multiple species in the ecological system, which is one of the primary features for emergent phenomena and behaviors. In this paper, we analyze the emergent behaviors in the process of multi-agent competitive coevolution, propose the research frame for emergent behaviors based on multi-agent coevolution mechanism, establish the multi-agent coevolution dynamics model and the elect strategy model, and design individual rule set of the agent. At last, we implement the system by multi-agent modeling language, NetLogo. With simulation experiments of the birds-bugs prey problem, we analyze emergent behaviors of aggregation in the process of multi-agent coevolution, and summarize the principle of emergent behaviors and parameters of multi-agent coevolution rules.
Keywords :
multi-agent systems; predator-prey systems; simulation languages; NetLogo multi-agent modeling language; birds-bugs prey problem; coevolution process; ecological system; elect strategy model; emergent behavior research; emergent phenomenon; multi-agent coevolution dynamics model; multi-agent coevolution rules; multi-agent competitive coevolution mechanism; Analytical models; Automation; Biological system modeling; Civil engineering; Educational institutions; Electronic mail; Mechatronics; Aggregation; Coevolution; Emergent Behavior; Multi-Agent System; NetLogo;
Conference_Titel :
Control Conference (CCC), 2013 32nd Chinese
Conference_Location :
Xi´an