Multiagent systems and the dynamics of a settlement system

Abstract

In order to simulate the evolution of a system of settlements with a dynamic model, many processes must be integrated: the spatial aggregation of population, the complexification of urban activities, and the increasing hierarchical differentiation of settlements. The model must also simulate the progressive structuration of the settlement system through a growing variety and enlarged range of interactions between its elements. "Multiagent systems" provides a flexible modeling method for dealing with the multiple spatial interactions of cooperation and competition and relations that generate and regulate the evolution of a settlement system. Its principles are described and applied to building an evolutionary model, including a simulation tool. The model combines economic and spatial rules to produce birth, growth, decline, and functional diversification of the towns. The "urban transition" from an agrarian settlement system toward a hierarchical system of trade- and manufacturing-oriented towns and cities can be simulated.