Regra de operação de reservatório isolado no semiárido: uma aplicação da lógica fuzzy

Abstract

Good reservoir management requires the creation of a set of operational procedures, rules or plans that best meet society's objectives. In general, the operational policy are logical or mathematical equations that take into account system variables to calculate the release of water from a reservoir. The computational techniques, fuzzy logic and evolutionary computing, have been widely used to optimize water resource problems, including the operation of reservoirs. This is a research applied in a quantitative analysis that implements a simulation and optimization model of reservoirs’ operation in order to evaluate a hedging rule with application of fuzzy logic (FHR). The study area is a reservoir inserted in the Brazilian semiarid, the Cruzeta reservoir. The area is located in a region that has intra-annual, inter-annual hydrological variability and severe droughts, that requires better management. The optimization of the operation was carried out via a genetic algorithm associated with the simulation of the reservoir's water balance. It was sought to maximize the efficiency of meeting water demands during dry season, a period with greater water demand. A performance of standard operating policy (SOP), hedging rule (HR) and FHR were assessed by analysis of indicators such as indicators, efficiency, resilience, vulnerability and time on dead volume. Thus, the rules with the best performance in mitigating water scarcity during the dry season, in decreasing order, were FHR, HR and SOP. The improvement is evidenced by the increase in efficiency in meeting priority demands, such as, for example, greater reduction of vulnerability and time with storage in the dead volume. However, the hedging rules induce more time with reduced supply, while rationing. Therefore, the results confirm the desirable advantages of the anticipated rationing of water, since it allows to mitigate future events of continuous and severe scarcity.