Regulação silvicultural das atividades operacionais na programação de projetos florestais via simulated annealing

• Ano de defesa: 2015
• Autor(a) principal: Cruz, Bruno Rogério
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: UNIVERSIDADE FEDERAL DE LAVRAS; DCF - Departamento de Ciências Florestais; UFLA; BRASIL
• Palavras-chave em Português: Planejamento florestal; Otimização; Metaheurística; Forest planning; Optimization; Metaheuristics; CNPQ_NÃO_INFORMADO
• Link de acesso: http://repositorio.ufla.br/jspui/handle/1/9579

A major difficulty faced by companies is related to the execution of a silvicultural planning capable of meeting planting demand, always respecting the technical determinations as well as operational limitations. This search for balance between planting demand and the capacity of executing all silvicultural activities can be understood as "silvicultural regulation". Due to the complexity of the silvicultural issue, the use of exact methods, most often does not generate solution in viable processing time, with the use of metaheuristics as an option for these cases. Therefore, this work sought to develop and evaluate the solutions obtained by means of the simulated annealing metaheuristics regarding the quality of the obtained solutions for planning silvicultural activities. Chapter two had the main objective of studying the different configuration options for the parametrization of simulated annealing metaheuristics, such as the number of iterations, search percentage and temperature reduction functions, testing their effects over the quality of the obtained solutions, thus seeking an increase in the efficiency and efficacy in the process. Subsequently, in chapter three, we applied the metaheuristics, with the parameters defined as ideal, on the resolution of sivlicultural regulation issues, and compared with the exact method of integer linear programming (ILP), verifying the capacity of metaheuristic in generating silvicultural activity planning, both in the attempt to reduce total costs and in anticipating the final time of the project (Makespan).