BS2019_210804 as published.pdf (5.19 MB)
Download fileForecasting indoor temperatures during heatwaves: Do more complex models provide better predictions?
conference contribution
posted on 2021-06-21, 11:29 authored by Matej Gustin, Rob McLeod, Kevin LomasKevin LomasA novel application of semi-parametric Generalized Additive Models (GAMs) was developed to forecast elevated indoor temperatures. GAM models were compared to AutoRegressive models with eXogenous inputs (ARX) and validated against monitored data from two case study dwellings, located near to Loughborough in the UK, during the 2013 heatwave. Input variables were selected using backward stepwise regressions based on minimisation of the Akaike Information Criterion (AIC) and Mean Absolute Error (MAE), for the ARX and GAM models respectively. Comparison of the models showed that GAMs are capable of slightly improving the forecasting accuracy, but only at short horizons (3-6 hours ahead).
Funding
EPSRC Centre for Doctoral Training in Energy Demand (LoLo)
Engineering and Physical Sciences Research Council
Find out more...LEEDR: Low Effort Energy Demand Reduction (Part 2 of the Call)
Engineering and Physical Sciences Research Council
Find out more...History
School
- Architecture, Building and Civil Engineering
Published in
Proceedings of Building Simulation 2019: 16th Conference of IBPSAPages
4243 - 4250Source
16th IBPSA International Conference and ExhibitionPublisher
International Building Performance Simulation Association (IBPSA)Version
- VoR (Version of Record)
Rights holder
© International Building Performance Simulation AssociationAcceptance date
2019-04-01Copyright date
2020ISBN
9781775052012ISSN
2522-2708Publisher version
Book series
Building Simulation Conference Proceedings; 6Language
- en