Development of single-zone predictive equations using linear regression for advanced controllers synthesis

Regression equations can be used for predicting temperature and relative humidity in single-zone buildings. This paper presents a methodology for computing the parameters of such kind of models based on weather data and room air hygrotermal conditions. In addition, the computed model can be very effective on the synthesis of advanced controllers applied to HVAC systems. A building hygrothermal simulation tool has been used to generate the data (heating signal, outdoor temperature and relative humidity, total solar radiation, indoor temperature and relative humidity). Validation procedures have shown very good agreement between the regression model and the simulation tool for winter and summer periods in Brazilian cities.