Dynamic thermal models in building simulations

Paper in proceedings (conference paper)

en

Mathematical description of a linear system can be obtained in various ways. Theoretical models are desirable when basic physical laws of nature are known that describe the relationship between input and output signals, mostly in terms of differential equations. If this is not the case, approximate dynamic characterization may be received from statistical analysis of experimental data. Both of methods are jointly called system identification. The paper provides the compendium of analytical modelling techniques that are pertinent to space-load calculations. Predominantly, the conduction transfer functions and state space formulation are of interest, whereas they are more efficient than solution of finite-difference and response factors methods. If excited by meteorological conditions, the year-long simulations of the hourly heating or cooling loads for the building could be done.

Dynamic linear system, identification, conduction transfer function, state-space method

2012-06-26

Vysoké učení technické v Brně

Brno

978-80-214-4529-1

Sborník příspěvků 31. mezinárodní konference Setkání kateder mechaniky tekutin a termomechaniky

257–260

4