[Back]

A. Mahdavi, F. Tahmasebi:
"An optimization-based approach to recurrent calibration of building performance simulation models";
Talk: ECPPM2012 eWork and eBusiness in Architecture, Engineering and Construction,, Reykjavík, Island; 2012-07-25 - 2012-07-27; in: "eWork and eBusiness in Architecture, Engineering and Construction", G. Gudnason, R. Scherer et al. (ed.); Taylor & Francis, (2012), ISBN: 978-0-415-62128-1; 145 - 150.

The utility of performance simulation applications in building design support is widely known
and documented. However, as our past research has demonstrated, simulation tools and methods can also effectively
support building operation and management processes. Specifically, predictive control methods have
demonstrated how simulation engines can be incorporated in buildings´ automation systems. However, the
effectiveness of such predictive control methods greatly depends on the reliability of the deployed simulation
model, implying the need for systematic and recurrent simulation model calibration. In this context, the present
paper entails a case study of an optimization-based calibration of a thermal simulation model for an office
building. The simulation model is calibrated with the building´s monitoring data in the calibration period and is
subsequently validated using data from a separate validation period. The results display a noticeable improvement
of the simulation model´s predictive performance upon calibration, pointing to the promising potential of
the optimization-based calibration method.

see German Version