Skip to main content
UOIT Home
Please use this identifier to cite or link to this item: http://hdl.handle.net/10155/735

Issue Date: 1-Oct-2016
Title: Machine learning classification techniques for non-intrusive load monitoring
Authors: Chung, Jefferson
Publisher : UOIT
Degree : Master of Applied Science (MASc)
Department : Electrical and Computer Engineering
Supervisor : Ibrahim, Walid Morsi
Keywords: Machine learning
Non-intrusive load monitoring
Co-testing
Abstract: Non-intrusive load monitoring is the concept of determining the operational loads using single-point sensing. The features contained within the electrical load’s signal are used to identify a unique signature which is used by a machine learning classifier to automate the load identification process. In this thesis, existing machine learning classification techniques are reviewed within the context of the non-intrusive load monitoring application. A non-intrusive load monitoring algorithm is developed in this to extract the prominent hidden features contained within the electrical load’s signal which helps identify the operation of different appliances from a single point of an electrical circuit. Decision tree and Naïve Bayes classifiers are used as the machine learning classification technique to automate the load classification process. The co-testing of machine learning classifiers was introduced in this work to improve the classification accuracy of previously seen methods when applying the one-against-the-rest testing approach. When the proposed NILM algorithm was applied to a real test system, a classification accuracy of 99.61% for decision tree and 99.38% for Naïve Bayes was obtained. When compared to previous methods in literature utilizing one-against-the-rest testing approach, a classification accuracy of 76.31% for decision tree and 67.44% for Naïve Bayes was obtained. The results demonstrate the effectiveness of the proposed non-intrusive load monitoring approach through the notable significant increase in the observed classification accuracies.
Appears in Collections:Electronic Theses and Dissertations (Public)
Faculty of Engineering and Applied Science - Master Theses

Files in This Item:

File Description SizeFormat
Chung_Jefferson.pdf15.69 MBAdobe PDFView/Open

Items in e-scholar@UOIT are protected by copyright, with all rights reserved, unless otherwise indicated.