Paper Conference

Proceedings of Building Simulation 2017: 15th Conference of IBPSA


Development of a Statistical Model for the Prediction of Overheating in UK Homes Using Descriptive Time Series Analysis

Argyris Oraiopoulos 1, Tom Kane 1, Steven K Firth 1, Kevin J Lomas 1
1Building Energy Research Group (BERG), School of Civil and Building Engineering, Loughborough University, Loughborough, UK
2 London-Loughborough (LoLo) Centre for Doctoral Training on Energy Demand, Loughborough University, Loughborough, UK

Abstract: Overheating risk in dwellings is often predicted using modelling techniques based on assumptions of heat gains, heat losses and heat storage. However, a simpler method is to use empirical data to predict internal temperatures in dwellings based on external climate data. The aim of this research is to use classical time series descriptive analysis and construct statistical models that allow the prediction of future internal temperatures based external weather data. Initial results from the analysis of a living room in a house show that the proposed method can successfully predict the risk of overheating based on four different overheating criteria.
Pages: 1939 - 1948