Improving functionalities in a multi-agent architecture for ocean monitoring

Abstract

This paper presents an improved version of a multiagent architecture aimed at providing solutions for monitoring the interaction between the atmosphere and the ocean. The ocean surface and the atmosphere exchange carbon dioxide. This process is can be modeled by a multiagent system with advanced learning and adaption capabilities. The proposed multiagent architecture incorporates CBR-agents. The CBR-agents proposed in this paper integrate novel strategies that both monitor the parameters that affect the interaction, and facilitate the creation of models. The system was tested and this paper presents the results obtained.