This paper deals with a global model based anticipative energy management system adapted for the CANOPEA building prototype proposed by the team Rhone-Alpes for the Solar Decathlon Europe contest. It presents a practical application of theoretical studies originally developed in Grenoble Research labs (G-SCOP, G2Elab and LIG) and improved by the Vesta-System company. The energy manager system can receive and takes into account signals that came from a smart grid. The different configurations of the building system including the envelop with the HVAC system, a battery and domestic appliances are presented: shutters, water storage, a temperature phase shifter, an air/air heat pump, a water/water heat pump, hybrid panels. The system is able to adapt to many different contexts to harvest the maximum energy. Nevertheless, it is so complex for a human to define the best configuration according to weather context and occupant activities. The proposed energy management system embeds a CANOPEA model connected to actual measurements, forecasts, realtime energy costs and occupant demands collected thanks to an Android tablet. It is able to support decision regarding the day ahead best CANOPEA configurations with a one-hour resolution.