135. Using Markovian Stochastic Modelling to Predict. Energy Performances and Thermal Comfort of Passive. Solar-Systems. J.-L. SCARTEZZINI and A. FAIST.
Energy and Buildings, 10 (1987) 135 - 150
135
Using Markovian StochasticModelling to Predict Energy Performances and Thermal Comfort of Passive Solar-Systems J.-L. SCARTEZZINI and A. FAIST Department of Physics, Swiss Federal Institute of Technology EPFL, 1015 Lausanne (Switzerland) Th. LIEBLING Department of Mathematics, Swiss Federal Institute of Technology EPFL, 1015 Lausanne (Switzerland) (Received January 27, 1987; in revised form July 22, 1987; in final form September 4, 1987) I
SUMMARY A Markovian stochastic approach to the simulation o f passive and hybrid solar devices has been developed. The driving variables o f the modelled device, as well as the temperatures which characterize its thermal state, are discretized. These quantities are used to set up a random state vector, which defines a discrete homogeneous Markov chain. The transition probabilities o f the chain are calculated using the stochastic matrices which are obtained by reduction o f weather and users' dependent real sequences o f data into the appropriate form. Determination o f the long-run distribution o f probability o f the chain provides for the evaluation o f energy performance and thermal c o m f o r t indicators. A n intermodel comparison, using several internationally well-known deterministic computer programs (D YWON, HELIOS, PASSIM, S E R I R E S ) , has been carried out by way o f the simulation o f a direct-gain office room. This analysis has shown close correspondence between the stochastic and deterministic modelling results. A m o n g others, the main advantage o f this approach is the ability o f the model to account for the random fluctuations o f the driving variables, which affect the thermal performances o f the solar device. The development o f microcomputer programs based on this approach is regarded as the final goal o f this research topic.
1. INTRODUCTION Energy performance and thermal comfort indicators of passive and hybrid solar devices 0378-7788/87/$3.50
are strongly dependent on meteorological conditions and on the influence of the behaviour of the users of the solar system (e.g., shutting of blinds, opening of windows) [1]. From the point of view of computer simulation, weather and user
