The reservoir management is an important task that aims at the optimization of oil reservoir exploitation. To support this challenging mission, the oil and gas industry has been developing new technologies such as intelligent wells. The purpose of these wells is to reduce costs of the most common restoring operations by control of their actuators. Thus, this work deals with intelligent fields development and presents a decision support system able to optimize, by using evolutionary algorithms, the intelligent wells technology control process considering geological and technical uncertainties. In addition, the system gives support for the decision of rather to use or not intelligent wells, given a reservoir ready to be explored or to receive expansion investments. The control of Intelligent Wells Technology (IWT), as applied in this study, refers to the opening and closing operations of valves in these types of wells. An optimization based on genetic algorithms is used to produce a pro-active control strategy, that is, one that anticipates the actions to be taken in present time in order to achieve better results in the future. Such a strategy proposes a valve configuration that will be able to: delay the water cut on producer wells, advance the oil production or benefit the oil recuperation. As a result, the obtained configuration leads to a maximization of the NPV (Net Present Value). The usage of control strategies that aim to benefit completion identifies the oil field as intelligent. Other works also deal with valve control optimization problems in intelligent wells. However, they use classical optimization methods; these methods limit the number of valves or optimize strategies without considering time. The evolutionary model, based on genetic algorithm, applied in this study, can formulate a control strategy for all valves in a certain production configuration, for any desired time interval, according to the economical criteria of NPV maximization. In order to support the decision making for the use or not of intelligent wells, technical and geological uncertainties are considered. The proposed model was evaluated in three oil reservoirs. The first one is a synthetic reservoir, simple and not real; the other two are more complex with close to real characteristics. The results obtained indicate that the proposed model allows good control strategies that increase the NPV. The main contribution of this work is the conception and implementation of a system based on intelligent techniques that is able to support the development and management of intelligent oil reservoirs considering uncertainties.