Agent architecture implementation in models of socio-ecological-economic systems

Abstract

The article presents analysis of deliberative, reactive, and hybrid agent architectures usage in agent-based models of socio-ecological-economic systems. It was found that these three approaches are used by researches, however, there is no objective assessment of their application in relation to the considered tasks. The authors suggest to determine the optimal architecture of each type of agent model basing on classification of environments in which the agents must operate, and on the amount of information needed to make a decision and presence of past events memory that affects current decisions making. The developed approach was applied to identify the agent architecture model of the regional timber complex.