Optimization of building control strategies using dynamic programming for building simulation 2013 conference

Being increasingly insulated, new buildings are more and more sensitive to variations of solar and internal gains. Controlling mechanical or natural ventilation often constitutes an efficient solution to maintain indoor comfort during hot periods. The proposed energy management is a predictive set of optimal commands issued from a dynamic programming optimization knowing in advance the weather, occupancy and internal gains for the next 14 days. This method is tested on a bioclimatic house situated in Chambery, France with an annual heating demand of 26 kWh/m². The results issued from the mechanical and natural ventilation controllers are compared.