Research on a Multi-agent System That Mimics Ant Foraging Behavior

Abstract

Artificial life is a research field that uses artificial systems, such as computer simulations, to reproduce life phenomena to expand our understanding of life. Artificial life research focuses primarily on emergent behavior using multi-agent systems, which is a part of artificial intelligence research. In conventional research on artificial life, simulation studies have been conducted using models of organisms that behave in groups, such as fish, birds and insects. Studying pheromone communication in ants focuses on the organized behavior of ants that use pheromones to communicate with their fellow ants and efficiently transport food during foraging. Ants have low individual abilities and behave according to simple behavioral principles. In this study, we focus on the efficient activities of ants and apply the ant foraging behavior to establish an algorithm that mimics this behavior in a multi-agent system using their organized behavior. The purpose of this study is to confirm that pheromone trail formation and foraging behavior are emergent in multiple ants and pheromone volatilization rate affects pheromone trails.