Automatic Appliance Identi cation Investigation of Automatic Appliance Identi cation scenarios and methodologies

Krystalakos, Odysseas

Automatic Appliance Identi cation Investigation of Automatic Appliance Identi cation scenarios and methodologies

Files

S1016895KrystalakosO.pdf (1.37 MB)

Authors

Krystalakos, Odysseas

Issue Date

2020-06-01

Language

en

URI

https://theses.ubn.ru.nl/handle/123456789/10607

Abstract

Automatic Appliance Identi cation refers to the task of identifying household devices given measurements of its power consumption. Solving this problem is crucial for modern energy monitoring applications but, so far, it has been shown to be non-trivial. In addition, there seems to be confusion about the practical scenarios on which Appliance Identi cation can be deployed. In this research project we attempt to untangle the de nition of Appliance Identi cation by proposing a distinction of three di erent scenarios. Among these, we describe the Appliance Load Identi cation scenario that, even though it had been implicitly mentioned in past works, it was never explicitly de ned. With regards to experiments, we initially replicate results of noteable past works using open datasets. Next, we propose a novel set of techniques for Appliance Identi cation that use a mix of VI trajectory data, handpicked features and Multi-Modal Neural Networks. Finally, we propose three classi ers for the newly-de ned Appliance Load Identi cation scenario. Through tests we nd that most existing models are not robust to tests across datasets. We also nd that combining VI trajectory representations with other features leads to increased performance. Last, we provide the results of our Appliance Load Identi cation models as baseline for future research.