Prognose e Regulação de Florestas de eucalipto utilizando Modelo de Crescimento e Inteligência Computacional

• Autor(a) principal: Magalhães, Juliana Galvão de Sousa
• Data de Publicação: 2014
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/3162
• topic: Eucalipto - Cultivo; Redes neurais artificiais; Modelagem florestal; Eucalyptus - Cultivation; Artificial neural networks; Forest modeling; CNPQ::CIENCIAS AGRARIAS::RECURSOS FLORESTAIS E ENGENHARIA FLORESTAL::MANEJO FLORESTAL

The purpose of this study is to evaluate the performance of an Artificial Intelligence technique, Support Vector Machines (SVM), on projecting the production of eucalyptus stands. Furthermore, it aimed to analyze the effects of these methodologies on regulating the production in a model farm. This dissertation is organized into three chapters: Chapter 1 consists of a forest modeling review, and a brief overview of these techniques and forest data influence on modeling the growth and yield. It also proposes the use of artificial intelligence techniques as an efficient alternative for forest modeling, covering the concept of Support Vector Machine and Least Square Support Vector Machine. In Chapter 2, different Support Vector Machines were built for eucalyptus production prognosis, and these results were compared to those by growth model and Artificial Neural Networks. In chapter 3, each alternative has been used to develop production tables, which served as input into a linear programming model (LP) formulated to regulate the production of a model farm.