Poster Open Access

EMD and Gradient Boosting Regression for NILM (Energy Disaggregation)

Timplalexis, Christos; Krinidis, Stelios; Ioannidis, Dimosthenis; Tzovaras, Dimitrios

Abstract: In this study a novel appliance load estimation in a non-intrusive way is presented. The proposed algorithm includes signal processing techniques such as filtering and Empirical Mode Decomposition (EMD) which is used to decompose random noise from the power consumption data collected from the smart meter. Lag features that capture the variance of the data across time are utilized. Experimental results which showcase the effectiveness of the suggested method are also presented. 

Files (1.4 MB)
Name Size
AI_CON_poster_presentation_CERTH.pdf
md5:6c3dc313ae431de14d418b7ccda032fe
1.4 MB Download
20
17
views
downloads
All versions This version
Views 2020
Downloads 1717
Data volume 24.5 MB24.5 MB
Unique views 1818
Unique downloads 1515

Share

Cite as