elib
DLR-Header
DLR-Logo -> http://www.dlr.de
DLR Portal Home | Imprint | Privacy Policy | Contact | Deutsch
Fontsize: [-] Text [+]

Benchmarking Reservoir Computing for Residential Energy Demand Forecasting

Brucke, Karoline and Schmitz, Simon and Köglmayr, Daniel and Baur, Sebastian and Räth, Christoph and Ansari, Esmail and Klement, Peter (2024) Benchmarking Reservoir Computing for Residential Energy Demand Forecasting. Energy and Buildings, 314 (114236). Elsevier. doi: 10.1016/j.enbuild.2024.114236. ISSN 0378-7788.

[img] PDF - Published version
1MB

Official URL: https://www.sciencedirect.com/science/article/pii/S0378778824003529?via%3Dihub

Abstract

In the energy sector, accurate demand forecasts are vital but often limited by the available computational power. Reservoir computing (RC) or echo-state networks excel in chaotic time series prediction, with lower computational requirements compared to other recurrent network based methods like LSTMs. Next-generation reservoir computing (NG-RC) is a newer, more efficient variant of classical RC originating from nonlinear vector autoregression and therefore missing the randomness of classical RC. In our study, we evaluate RC and NG-RC for day-ahead energy demand predictions on four data sets and compare it to LSTMs and a naive persistence approach. We find that NG-RC outperforms all other methods when considering the root mean squared error on all data sets but struggles with very small or zero demands. Additionally, it offers a very computationally effective hyperparameter optimization and excels in replicating the inherent volatility and the erratic behavior of energy demands.

Item URL in elib:https://elib.dlr.de/204259/
Document Type:Article
Title:Benchmarking Reservoir Computing for Residential Energy Demand Forecasting
Authors:
AuthorsInstitution or Email of AuthorsAuthor's ORCID iDORCID Put Code
Brucke, KarolineUNSPECIFIEDhttps://orcid.org/0000-0002-4510-8969UNSPECIFIED
Schmitz, SimonUNSPECIFIEDUNSPECIFIEDUNSPECIFIED
Köglmayr, DanielUNSPECIFIEDUNSPECIFIEDUNSPECIFIED
Baur, SebastianUNSPECIFIEDhttps://orcid.org/0000-0003-1924-8009UNSPECIFIED
Räth, ChristophUNSPECIFIEDUNSPECIFIEDUNSPECIFIED
Ansari, EsmailUNSPECIFIEDUNSPECIFIEDUNSPECIFIED
Klement, PeterUNSPECIFIEDhttps://orcid.org/0000-0001-7175-6145UNSPECIFIED
Date:1 July 2024
Journal or Publication Title:Energy and Buildings
Refereed publication:Yes
Open Access:Yes
Gold Open Access:No
In SCOPUS:Yes
In ISI Web of Science:Yes
Volume:314
DOI:10.1016/j.enbuild.2024.114236
Publisher:Elsevier
ISSN:0378-7788
Status:Published
Keywords:Reservoir computing, Next generation reservoir computing, Recurrent network architectures, Energy demand forecasting, LSTM
HGF - Research field:Energy
HGF - Program:Energy System Design
HGF - Program Themes:Digitalization and System Technology
DLR - Research area:Energy
DLR - Program:E SY - Energy System Technology and Analysis
DLR - Research theme (Project):E - Energy System Technology, R - Artificial Intelligence, R - Machine Learning
Location: Oldenburg
Institutes and Institutions:Institute for AI Safety and Security
Institute of Networked Energy Systems > Energy System Technology
Institute for Software Technology
Deposited By: Brucke, Karoline
Deposited On:20 May 2024 11:03
Last Modified:12 Jun 2024 07:03

Repository Staff Only: item control page

Browse
Search
Help & Contact
Information
electronic library is running on EPrints 3.3.12
Website and database design: Copyright © German Aerospace Center (DLR). All rights reserved.