Publications des agents du Cirad

Cirad

Economic theory of renewable resource management : A multi-agent systems approach

Antona M., Bousquet F., Le Page C., Weber J., Karsenty A., Guizol P.. 1998. Lecture Notes in Computer Science. LNAI. Lecture Notes in Artificial Intelligence, 1534 : p. 61-78. MABS'98 International Workshop. 1, 1998-07-04/1998-07-06, Paris (France).

Multi-agent methodology is used for various purposes, such as distributed solving problem, network management, etc. Multi-agent systems are also used for simulation. Usually, the emergence of macroscopic properties is studied by simulating the behavior of agents and their interactions at a microscopic level. It is a bottom-up approach. Recently emphasis has been placed on the links between the macroscopic and the microscopic levels combining the bottom-up and top down directions. We are currently developing a multi-agent simulator to illustrate and discuss the principles of the economic theory of renewable resource management. The management of renewable resources, from an economic point of view, may be based on public interventions at the macroscopic level or on regulation at the microscopic level. The simulator uses the Cormas environment (Bousquet et al., 1998). First, a renewable resource dynamics is simulated on a theoretical grid. Second, economic agents are represented. Different agents, belonging to different stages of an industry, are represented. Agents collect the resources and others use this resource as an input for transformation. The agents' behavior and objectives are heterogeneous. Economic exchanges are simulated. We reproduce the market-oriented approach (Cheng & Wellman, 1996) which assumes that a general equilibrium sets a global price. We also simulate the trade rules proposed by Epstein and Axtell (1996) in which there is no global price. The aim is to evaluate the differences between these two approaches from the ecological and economic points of view. Third, when an industry is established we simulate various actions at the macroscopic level. Economic theory proposes various interventions such as taxes, quotas, trade and market organization. We test these actions and observe their effects on the society of agents. We observe the quantity of resources, the benefits to the agents, the rent obtained from the macroscopic interventions and how it is shared. This simulator provides a didactic framework to explain economic theory for resource management. Later, we will use it for field study by importing field data. From the computer science point of view, this simulator will be used to test various algorithms for agent coordination. From an economic point of view, it introduces the notion of space and a time frame in the scope of resource management. (Résumé d'auteur)

Mots-clés : analyse microéconomique; gestion foncière; analyse macroéconomique; ressource renouvelable; gestion des ressources; modèle de simulation; système multiagents

Thématique : Méthodes mathématiques et statistiques; Conservation de la nature et ressources foncières

Article de revue

Agents Cirad, auteurs de cette publication :