Methods for Optimal Microgrid Management

Abstract During the last years, the number of distributed generators has grown significantly and it is expected to become higher in the future. Several new technologies are being de-veloped for this type of generation (including microturbines, photovoltaic plants, wind turbines and electrical storage systems) and have to be integrated in the electrical grid. In this framework, active loads (i.e., shiftable demands like electrical vehicles, intelligent buildings, etc.) and storage systems are crucial to make more flexible and smart the dis-tribution system. This thesis deals with the development and application of system engi-neering methods to solve real-world problems within the specific framework of microgrid control and management. The typical kind of problems that is considered when dealing with the manage-ment and control of Microgrids is generally related to optimal scheduling of the flows of energy among the various components in the systems, within a limited area. The general objective is to schedule the energy consumptions to maximize the expected system utility under energy consumption and energy generation constraints. Three different issues related to microgrid management will be considered in detail in this thesis: 1. The problem of Nowcasting and Forecasting of the photovoltaic power production (PV). This problem has been approached by means of several data-driven techniques. 2. The integration of stations to charge electric vehicles in the smart grids. The impact of this integration on the grid processes and on the demand satisfaction costs have been analysed. In particular, two different models have been developed for the optimal integration of microgrids with renewable sources, smart buildings, and the electrical vehicles (EVs), taking into account two different technologies. The first model is based on a discrete-time representation of the dynamics of the system, whereas the second one adopts a discrete-event representation. 3. The problem of the energy optimization for a set of interconnencted buildings. In ths connection, an architecture, structured as a two-level control scheme has been developed. More precisely, an upper decision maker solves an optimization problem to minimize its own costs and power losses, and provides references (as 3 regars the power flows) to local controllers, associated to buildings. Then, lower level (local) controllers, on the basis of a more detailed representation of each specific subsystem (the building associated to the controller), have the objective of managing local storage systems and devices in order to follow the reference values (provided by the upper level), to contain costs, and to achieve comfort requirements.