Show simple item record

dc.contributor.authorVan Zaen, Jérôme.
dc.contributor.authorMoufawad El Achkar, Christopher.
dc.contributor.authorCarrillo, Rafael E.
dc.contributor.authorHutter, Andreas.
dc.date.accessioned2021-12-09T14:02:02Z
dc.date.available2021-12-09T14:02:02Z
dc.date.issued2018
dc.identifier.citationNILM2018 Proceedings, March 7-8, 2018 | Austin, Texas
dc.identifier.urihttps://yoda.csem.ch/handle/20.500.12839/280
dc.description.abstractNon-intrusive load monitoring aims to estimate the power consumption of individual appliances from a single meter. Several methods have been proposed to solve this blind source separation problem, such as clustering, hidden Markov models, or neural networks. We present two approaches for detecting and classifying refrigeration units in a commercial environment. The first one is based on unsupervised event detection and the second one on neural networks with convolutional layers. We show that both approaches can accurately recognize power cycles of refrigeration units. The extracted cycles can then be used to reduce total energy consumption or for predictive maintenance by identifying units with an increased risk of failure.
dc.subjectNILM, unsupervised event detection, convolutional neuronal networks, deep learning, artificial intelligence, energy management, predictive maintenance
dc.titleDetection and Classification of Refrigeration Units in a Commercial Environment: Comparing Neural Networks to Unsupervised Clustering
dc.typeProceedings Article
dc.type.csemdivisionsDiv-V
dc.type.csemresearchareasDigital Energy
dc.identifier.urlhttp://nilmworkshop.org/2018/proceedings/Poster_ID20.pdf


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

  • Research Publications
    The “Research Publications” collection provides bibliographic information for scientific papers including conference proceedings and presentations.

Show simple item record