Data-based calibration of physics-based thermal models of single-family houses

Abstract

The calibration of building energy simulation models is crucial for addressing the issue of the discrepancy between the simulation output and real-world measurements. The majority of current research studies have used monthly utility bills as calibration data. In this study an automated optimization algorithm was implemented to calibrate an energy model of a singlefamily house using daily and hourly energy consumption data. The performance of the calibrated model was tested on a different dataset and the simulation output matched the measurements with a daily CV(RMSE) of 14%. This study demonstrates that the calibration using currently available district heating data can significantly improve the performance of building energy simulation.