Inverse Model Identification of the Thermal Dynamics of a Norwegian Zero Emission House

Abstract

Dynamical model identification is an essential element in the implementation of a model predictive controller. In this work, a control-oriented first order model was identified in a dedicated experiment on a super-insulated single-family house. First, parameters resulting from CTSM and the MATLAB System Identification toolbox were compared. Then, a comparison of model predictions and measurements showed that this simple model captures well the main dynamics of the building-averaged indoor temperature, after one week of training on rich data with sample time below 15 min. It was also observed that this prediction performance was not affected by the configuration of internal doors.