Modern Smart Grids (SG) represent the new paradigm to plan and manage an efficient, secure and eco-friendly power system with multiple energy carriers. The shortage of traditional resources, the large diffusion of effective technologies based on Renewable Energy Sources (RES) and distributed generation and the liberalization of energy markets are transforming distribution systems into intelligent infrastructures in which every consumer can turn into a producer too. The need for an appropriate framework to forecast and organize RES generation, load demand, storage levels, market prices and many other issues necessary to realize a reliable power network leads to the Energy Hub (EH). In this paper we develop an EH model including electrical and thermal loads and supplies, RES generation and storage devices.. The core of the methodology is based on the formulation and solution of a non-linear discrete optimization problem aimed at optimizing input and output time trajectories for a set of combined power-generation and storage technologies. This model was applied for the demand response forecasting of an area of tertiary buildings but can be also fitted to a vast range of residential, commercial or industrial districts.