Modelos basados en agentes aplicados a estudios urbanosuna aproximación teórica

Abstract

Agent-Based Models (ABM) are relatively new prospective models with characteristics that allow urban simulations at regional scales, although mostly applied at intra-urban scale so far. Their main characteristics, as the ability to link the behavior of the agents (in this case, coincident with actors of urban growth process) to territorial changes, distinguish them from other models commonly applied to urban issues. ABM allows, within (1) a set of agents, (2) an environment in which agent act over, and (3) a set of behavior rules, to simulate urban complex phenomena that generate spatial patterns in an aggregate way derived from individual behaviors. The observed differences for these changes on scale – from an individual to an aggregate – help us to study how urban processes are generated and also to manipulate the inputs in order to generate different scenarios. While other simulation models are limited to study past trends or to design scenarios based on static components or, even beyond, consider transition rules with respect to the neighborhood in the base of mathematics and statistics, ABM are built on the framework of informatics simulation, allowing more realistic interpretations and higher flexibility by considering the agents behavior and their relations. This document presents a theoretical review on Agent-Based Models applied to urban studies, regarding their concept, components, structure, applications, and also identifying their potentialities and limitations.