BAYESIAN METHOD FOR ESTIMATING AIRTIGHTNESS COEFFICIENTS FROM PRESSURIZATION MEASUREMENTS

A new method for processing pressurization data of a building is prese nted. The inverse problem theory is applied to the estimation of airti ghtness coefficients. The method, based on the conjunction of theoreti cal and experimental information, provides better confidence intervals than a classical least mean square fit method and allows one to take into account more sources of error. The airtightness coefficients are major parameters for predicting air renewal from weather data and they usually have very large confidence intervals whose reduction merits s ome numerical efforts.