Modelo para gestão de recursos renováveis

Abstract

There are several studies applying Dynamic Programming (DP) to the management of renewable resources, where the main objective is to generate optimal control policies in order to produce greater benefits or reduce costs. This work deals with a model for managing renewable resources, with a critical population level (smin) below which no exploitation of the resource should be done, in order to avoid its extinction. The objective of this work, through the application of the dynamic programming technique, is to describe the solution structure of a dynamic program. For this work, we consider a control model with independent and identically distributed random disturbances, obeying the uniform distribution, and the space of joint states and actions in IR. The first part of the paper is concerned with the introduction of the basic concepts of PD, as well as some mathematical and PD results on the structural properties of PD models, where the monotonicity of the value functions Vn(s) and the continuity of the maximizers fn(s) are emphasized. As a final part of the paper, the final state of the resource is simulated and it is verified that the probabilistic distribution of the final state depends on the critical level smin. From the study done, we conclude that, applying the basic control theorem, the value functions, that is the gain obtained after N periods, are increasing, which means that with the population level initially fixed, the gain grows with the number of decision periods. (TRADUÇÃO NOSSA)