Soft Sensors For Complex Systems

Résumé

Recently, there has been growing interest in soft sensing techniques for monitoring complex systems in both academic and industrial sectors. Of particular note is the analysis of residential electricity consumption, which has attracted the attention of several researchers. However, the application of non-intrusive load monitoring (NILM) techniques to residential buildings is complicated by the unique characteristics of the data collected at these sites. In addition, publicly available datasets for NILM studies of residential structures are scarce and part of the privacy of the occupants of these structures. In this thesis, we addressed these problems by proposing three solutions: a novel unsupervised machine learning approach to detect abnormal energy consumption in homes, an event detection algorithm that involves the Tukey closure, and a NILM solution based on the integrated deep clustering (DEC) approach.

Citer ce document

Accès au document

Ce document est hébergé sur son dépôt institutionnel d'origine.

Auteur(s)

Statistiques

Consultations : 3

Téléchargements : 0