マルチエージェントシミュレーションに基づく人間-環境-社会システムのモデル化一試論

抄録

A multi agent simulation model to deal how the Human-Environment-Society System can approach the sustainable society was established. The model expresses dynamic process where three agents, a government, enterprises and consumers play interdependently through a variety of economic activities. Every enterprise and consumer has respective value systems that are both defined by price and environmental load. Each enterprise is imposed to determine how weighty between price and environmental load when she produces goods in order to maximize her profit. While each consumer also determines how weighty between price and environmental load when he buy products in order to maximize his utility. These determination processes are described as learning systems on an individual agent based on Finite State Machine, FSM and Genetic Algorism, GA. As a whole, the model can display a various paths whether the society goes to the sustainable world or the materialism world. The built model was verified qualitatively, and numerical simulations were presented.