학술논문
Data-Driven Sparsification and Multi-Resolution Analysis-Based Framework for Load Identification
Document Type
Periodical
Author
Source
IEEE Transactions on Smart Grid IEEE Trans. Smart Grid Smart Grid, IEEE Transactions on. 15(2):2387-2390 Mar, 2024
Subject
Language
ISSN
1949-3053
1949-3061
1949-3061
Abstract
The present work focuses on certain pivotal aspects regarding load identification for non-intrusive load monitoring (NILM) using convolutional neural networks (CNN). Primarily, image sets of load signatures are generated via multi-resolution analysis leading to improved image sets compared to the recent relevant works. Secondly, to avoid manual settings for deep learning optimization algorithm hyper-parameters, Bayesian optimization is performed. Lastly, a data-driven strategy based on Taylor’s score is considered to significantly compress the CNN architecture. The proposed overall strategy leads to high classification performance and reduced memory footprint.