TU Darmstadt / ULB / TUbiblio

Comparison between support vector regression and grey system theory based models for adaptive load forecasting of building thermal energy consumption

Irmler, Robert and Rüppel, Uwe
Li, Haijiang and de Wilde, Pieter and Rafiq, Yaqub (eds.) (2014):
Comparison between support vector regression and grey system theory based models for adaptive load forecasting of building thermal energy consumption.
In: Proceedings of the 21st International Workshop on Intelligent Computing in Engineering, In: 21st EG-ICE International Workshop, Cardiff, United Kingdom, 16.-18. Juli 2014, ISBN 978-0-9930807-0-8,
[Online-Edition: http://egice2014.engineering.cf.ac.uk/images/BIN/papers/32.p...],
[Conference or Workshop Item]

Abstract

The prediction of building thermal energy consumption plays an important role in building energy management systems. Aside from physical simulations a common approach to predict such energy consumptions is to use machine learning methods like support vector regression (SVR). In addition, there is another group of models that are called ‘grey models’. Grey models are based on the so-called ‘grey system theory’, a relatively new branch of mathematical theory, and are designed to be able to handle systems that are characterized by little and uncertain information. This paper aims on investigating the performance of those models, regarding the adaptive load forecasting of building thermal energy consumption. Therefore, a GM(1,1) grey model is used to forecast building thermal energy consumption using a sliding window technique and synthetic real time measurements. Finally, the results are compared with the results of a ε-SVR-model working under similar conditions.

Item Type: Conference or Workshop Item
Erschienen: 2014
Editors: Li, Haijiang and de Wilde, Pieter and Rafiq, Yaqub
Creators: Irmler, Robert and Rüppel, Uwe
Title: Comparison between support vector regression and grey system theory based models for adaptive load forecasting of building thermal energy consumption
Language: English
Abstract:

The prediction of building thermal energy consumption plays an important role in building energy management systems. Aside from physical simulations a common approach to predict such energy consumptions is to use machine learning methods like support vector regression (SVR). In addition, there is another group of models that are called ‘grey models’. Grey models are based on the so-called ‘grey system theory’, a relatively new branch of mathematical theory, and are designed to be able to handle systems that are characterized by little and uncertain information. This paper aims on investigating the performance of those models, regarding the adaptive load forecasting of building thermal energy consumption. Therefore, a GM(1,1) grey model is used to forecast building thermal energy consumption using a sliding window technique and synthetic real time measurements. Finally, the results are compared with the results of a ε-SVR-model working under similar conditions.

Title of Book: Proceedings of the 21st International Workshop on Intelligent Computing in Engineering
ISBN: 978-0-9930807-0-8
Divisions: 13 Department of Civil and Environmental Engineering Sciences
13 Department of Civil and Environmental Engineering Sciences > Institute of Numerical Methods and Informatics in Civil Engineering
Event Title: 21st EG-ICE International Workshop
Event Location: Cardiff, United Kingdom
Event Dates: 16.-18. Juli 2014
Date Deposited: 20 Jan 2015 15:43
Official URL: http://egice2014.engineering.cf.ac.uk/images/BIN/papers/32.p...
Additional Information:

ISBN: 978-0-9930807-0-8

Related URLs:
Export:

Optionen (nur für Redakteure)

View Item View Item