Modelos baseados em agentes e aplicações em engenharia ambiental

Abstract

Recently, computational models (CMs) have gained more notoriety in Engineering. This highlight is mainly due to the growing search for methods that minimize the time, resources, and costs, usually spent on empirical tests. The focus was to explore and present computational simulation methods applied in Environmental Engineering, focused on Cellular Automata (ACs) and Agent-Based Models (MBAs). As a practical motivation, an example of MBA to simulate the nectar foraging of Apis mellifera L. bees in different environments (landscapes) was introduced. The comparative analysis considered aspects such as the flight pattern, the frequency of dances (signaling rate), the rate of recruitment, resting and nectar collection. We built different landscapes through by varying parameters associated with the floral arrangements, such as the number of fragments (NF), the fragment density (D, flowers per m2), the quality of the resource (sweetness) and the maximum quantity of flowers per fragment (QF). The algorithm has been implemented in R language and distributed the software in an open-source package to promote the use of the technology. The program allows the composition of different floral landscapes and the evaluation of the spatiotemporal dynamics in each simulated scenario.