Implementation of Ant Colony Algorithm Based-On Multi-agent System

摘要

Ant colony algorithm (ACA) is a simulated evolutionary algorithm which was developed in recent years. ACA has attracted many researchers' attentions for the solving of combinatorial optimization problems. Agent-based simulation (ABS) is one of novel methods for the analysis of complex system. This paper introduces the basic principles of ACA and its method of design and implement in a multi-agent system (MAS). Computer simulation results of MAS based on ACA are introduced and discussed in this thesis. The results show that the reasonable combination of ACA and the simple local rules of agent can effectively improve the colony behaviors of agents.