In this paper, the structure of the highest level of a hierarchical control architecture for microgrids is proposed. Such structure includes two sub-levels: the Energy Management System (EMS) and the tertiary regulation. The first devoted to energy resources allocation in each time slot based on marginal production costs, the latter aiming at finding the match between production and consumption satisfying the constraints set by the EMS level about the energy production in each time slot. Neglecting the efficiency of the different energy generation systems as well as that of the infrastructure for electrical energy distribution, the problem dealt with by the EMS sub-level is linear and can be solved by well known Linear Programming optimization procedures. The tertiary sub-level, below the EMS, optimizes mainly technical objectives and requires the solution of the Optimal Power Flow problem. After a review of the state of the art on the topic, the higher control sub-levels are described and an application is proposed. The application shows the efficiency of Mixed Integer Linear Programming methods for cost minimization of the energy production systems for microgrids.