Paper Conference

Proceedings of eSim 2020: 11th Conference of IBPSA-Canada


Preparing Weather Data for Real-time Building Energy Simulation

Maryam MeshkinKiya 1, Riccardo Paolini 2
1 Politecnico di Milano, Italy
2 University of New South Wales, Australia

Abstract: The application of actual weather data for building performance simulations has become more popular. Yet, anomalous values defect the results, while missing data lead to an unexpected termination of the simulation process. Traditionally, infilling missing values in weather data is performed through periodic or linear interpolations. However, when missing values exceed many consecutive hours, the accuracy of traditional methods is subject to debate. This study demonstrates how Neural Networks can execute highly accurate data imputation for infilling missing values. Results show that Neural Networks provide a reasonable balance between training time and accuracy compared to popular supervised learning techniques.