Dynamic formation generating in mobile-agents using algorithm to select function by environmental information

In a multi-agent system, agents must have functions to work with in cooperation with other agents. To produce the cooperative works, we propose Soft DNA (soft computing oriented data driven functional scheduling architecture) to change dynamically agent´s control block based on the fuzzy associative memory organizing units system and conceptual fuzzy set. It is a function to allow each agent´s role in the agents group and select the role dynamically according to the circumstance. We apply this system to intelligent transport systems, and simulation was performed.