A. De Salvia, R. Lollini, I. Meroni
Bibliographic info:
23rd AIVC and EPIC 2002 Conference (in conjunction with 3rd European Conference on Energy Performance and Indoor Climate in Buildings) "Energy efficient and healthy buildings in sustainable cities", Lyon, France, 23-26 October 2002

Together with the definition of innovative plant and envelope technological solutions for buildings, many simulation tools (models) have been developed to make the design choices easier. However the definition of analytical structures able to describe the characteristics of building components installed under real conditions is still difficult. The paper presents some experiences made by ITC. They have been carried out by using System Identification techniques to simulate and predict the performances of various components analysed also through experimental campaigns under real conditions. To this end, parametrical black-box models have been identified and validated. Apart from the physical and dimensional characteristics of the building, they allow to describe it through equations that consist of simple polynomials whose variables are represented by inputs and outputs. The application of the models has allowed to obtain meaningful results, as regards the prediction of the behaviour of the studied components also with a very limited knowledge of their characteristics. They need repeated identifications each time the external environments dynamics change. But as the analytical structure is quite simple and they dont need too long monitoring periods, the identification can be efficaciously automated, like in the presented cases.