Today, the control of energy consumption is a very important issue for the preservation of the natural resources. Besides transportation domain, substancial energetic gains exist in residential and office buildings. While low energy building is now a reality, the renovation is rarely put into practice. The objective of this thesis is to evaluate the potential energy savings resulting from the renovation of old buildings. In the core of this work are the problematics of the building system and of the household comfort. Based on this knowledge, we propose here a global control architecture of the building, which accomodates thermal and optical confort, and reduction of energy consumption. Specifically, we focus on automatic control and systemic approach, which enable, from modelling to control law synthesis, a global cooperation of all home automation equipments. Our approach is to design a hierarchical control architecture and simulate it in the case of an old building with central heating, lighting and interior Venetian blinds. The control architecture is designed from a systemic analysis of buildings. It is based on a hierarchical approach in order to adapt to user requirements and to different scales of spatial representation (from the building to the room). To experiment this control architecture, a physical multi-scale model of building was developed using a Bond-Graph approach. This model allows to estimate the optical and thermal behavior of the building separating the slow dynamics of the fast ones. The synthesis of actuators control laws was then performed after reducing the simulation model. It helped to implement two control laws : The first is expressed throughout the building by a predictive control of central heating system and considers the future behaviour of the outdoor temperature and solar radiative power. The second takes place at the rooms level as an optimal control of radiators, Venetian blinds and lighting to ensure the operative temperature and illumination desired by the occupants. The study of energy savings offered by the renovation of the control system helped to highlight the interest of the presented control architecture. The results showed that the implementation of predictive control on the central heating leads to reduce more than 20% of the energy consumption. Moreover, the optimal control in rooms increase thermal comfort around 20% and the visual comfort by 75%.
