From Grid Environment to Geographic Vector Agents, Modeling with the GAMA simulation platform

Patrick Taillandier (1) , Alexis Drogoul (1, 2)

1 UMMISCO - Unité de modélisation mathématique et informatique des systèmes complexes [Bondy]
2 MSI - Modélisation et Simulation Informatique de systèmes complexes

These last years has seen an important spreading of the agent-based simulations. If the early models proposed a very basic representation of the spatial environment --usually a grid-- the newest models tend to propose a very precise one that integrates geographic vector data. Unfortunately, if all of the existing agent-based simulation platforms allow to use a grid as environment, rare of the ones that allow to integrate geographic vector data. Moreover, these platforms are often complex to use and their understanding can require a time investment from the modeler that can be similar to the one needed to develop a model from scratch. Thus, today, most of the models that integrate complex vector data are developed from scratch. The GAMA agent-based simulation platform was developed to address these issues. This platform, which is under GPL license, provides a complete modeling and simulation development environment for building spatially explicit multi-agent simulations. Its main advantages come from its versatility (domain independent) and the simplicity to define a model with it. Indeed, GAMA provides a rich, yet accessible, modeling language, the GAML, that allows to define complex models integrating grids and geographic vector data. In this paper, we present the spatial capabilities of the new version of GAMA (v1.3) and we illustrate its utilization by case study concerning the classic Shelling model. In this case study, we show how to build different kinds of models, from the simplest (grid environment) to the more complex (geographic vector agents) with the GAMA platform.

Domaines: Modélisation et simulation

Patrick Taillandier, Alexis Drogoul. From Grid Environment to Geographic Vector Agents, Modeling with the GAMA simulation platform. Conference of the International Cartographic Association, 2011, Paris, France. ⟨hal-00688421⟩