Open AccessPrediction of DHW energy use in a hotel in Norway
Author(s) -
Dmytro Ivanko,
Nataša Nord,
Åse Lekang Sørensen,
Igor Sartori,
Plesser Thale Sofie Wester,
Harald Taxt Walnum
Publication year - 2019
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/609/5/052018
Subject(s) - energy (signal processing) , computer science , energy modeling , regression analysis , variable (mathematics) , operations research , statistics , engineering , machine learning , mathematics , mathematical analysis
Domestic hot water (DHW) systems are significant consumers of energy in buildings. This article is dedicated to hourly and daily DHW energy use modeling, with the aim of achieving energy savings in buildings. The methods investigated in the article were tested using statistical data obtained from a hotel located in Oslo, Norway. For better modeling, the influence of various factors on DHW energy use in the hotel was studied. For this purpose, the wrapper approach was used. The analysis indicates that the most important variable that should be used in the model is number of guests. There are also other factors that can be taken in account, even though they do not have such strong influence. Traditionally, only daily data about number of guests are available in the hotels. These data do not allow us to develop accurate hourly model of DHW energy. The article therefore proposes a method which, based on introduction of artificial variables, improve accuracy of the hourly DHW model. Eight models are compered, based on criteria of their adequacy. The Support vector machine model shows the best results for daily modeling and the Partial least squares (PLS) regression for hourly modeling.