A new method for modeling energy performance in buildings

Abstract

This paper presents a new method for modelling building energy performance based on a physical model to predict indoor environment and energy consumption by selecting best match parameters and variables. The innovative aspect is the introduction of an open and closed loop dynamics approach to explicitly presenting the complex behaviour and relationships between building indoor environment and energy consumption over time under uncertainty and disturbances and to simultaneously tracking the evolution of the complex lag-lead relationships between heating system and the building in order to keep the parameters and variables to the minimum. The model system is solved for an approximate analytical solution including physical and generalized parameters calibrated by measurements. Singular value decomposition model reduction technique is also applied to determine the model variables for the best approximation using lower dimensions. The approach is illustrated and validated with a case study of a swimming hall. A two-stage validation shows excellent agreements between the measurements and model (R2>0.9). The developed model is simple, accurate and straightforward that can serve the need in prediction and control in a wide variety of building applications.