Uma metodologia para priorização de portfólio de projetos em ambientes incertos: aplicação em mineração a céu aberto

Abstract

Operation in an open-pit mining results in a complex system with stochastic nature, in which interactions between many factors influence the production.This work considers optimizing the portfolio chosen projects such a way to maximize the total production. Due to the difficulties inherent from the system uncertanty and the problems factorial complexity, it may not be possible to find good results using conventional optimization methods. Considering these aspects, this work presents a computational method applied to a new formulation based on the knapsack problem, aiming to aid decision making in problems with this nature. This method performs optimization in an environment where each evaluation function is obtained by simulation. Using load-transport cycle characteristics in open-pit mining and measuring evaluations by a limited computational time, this method seeks to identify projects combinations that may generate good results. The method was applied to a simulation model of an open-pit mining and its results were compared with results obtained using conventional approaches also based on the knapsack problem. At last, it was possible to find a good trade-off between the best found solution and computational effort.