Comparison of different methods of incorporation of stochastic factors into deterministic models of indoor air quality.

The paper will discuss problems connected with incorporation of stochastic factors into deterministic models of indoor air quality. Three different methods: a quasi dynamic multi-zone modelling with generating of input data time series, multi-zone modelling based on the theory of stochastic differential equations, Monte-Carlo simulation with independent random generating of stochastic parameters, will be shortly presented. Described methods are compared on the base of computer simulation of CO2 concentration in simple two compartment office. The comparison of simulation outcomes shows that the way of stochastic disturbances incorporation into indoor air quality models does not have an important influence on mean value of predicted concentration. At the same time the analysis of standard deviations indicates that the method of disturbances generation and its later incorporation have a great influence on probability distribution of estimated concentrations. The author would like to highlight that differences. The conclusions include also subjective opinion of the author on main advantages and disadvantages of each of proposed methods.