Advertisement
Published in:
2017 | OriginalPaper | Chapter
Short Term Load Forecasting for Residential Buildings
An Evaluation Based on Publicly Available Datasets
Author : Carola Gerwig
Published in: Smart Grid Inspired Future Technologies
Publisher: Springer International Publishing
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
Abstract
Short Term Load Forecasting is an essential component for optimizing the energy management of individual houses or small micro grids. By learning consumption patterns on smart metering data, smart grid applications such as Demand-Side-Management can be applied. However, most of the research done in this field is based on data which is not publicly available. Moreover, the evaluations also vary in the evaluation settings and the error measurements. In this work, five state-of-the-art approaches are compared on three publicly available datasets in the most common scenarios. By doing this, the most promising methods and model settings are pointed out. Furthermore, it can be seen that forecasting the consumption 24 h ahead achieves about the same accuracy as doing it four hours ahead. Still, the best results for individual households are rather inaccurate. By aggregating ten households, the results enhance by a factor of about 60%.