Two-stage regression model of thermal comfort in office buildings

Models of the predicted mean vote (PMV) play an important role in evaluation of thermal comfort and control design of heating, ventilation, and air conditioning (HVAC) systems. In this article, we present a novel two-stage regression representation of the ASHRAE empirical PMV model that incorporates architectural parameters and control variables as predictors. Extensive measurements from an office building are used to develop and validate the regression model. The resulting model can predict the PMV in different rooms accurately in both short-term and long-term. Over a period of four weeks, for example, the predictions of the PMV have a root mean squared error less than 0.04 with a coefficient of determination larger than 0.96.

► We study a two-stage regression model of thermal comfort.
► Architectural parameters and control variables are explicit model predictors.
► The model provides analytical foundations for parametric study and control design.
► The model demonstrates excellent predictions in both short-term and long-term.