University of the Aegean
Energy monitoring technologies for energy efficiency and device health assessment
Abstract
dc:descriptionThis dissertation focuses on the field of Non-Intrusive Load Monitoring (NILM) and the analysis of electrical energy data to enhance energy efficiency and device health. NILM involves the monitoring of an electrical circuit and the disaggregation of energy to individual appliances/components without the need for individual metering. The field of NILM is interdisciplinary as it utilizes various techniques for feature extraction from electrical data, data analysis for event detection and energy disaggregation, and different systems for data collection. The quality of the collected data can affect the efficiency of a NILM system and the functionality it provides. Collection of data from real-world scenarios is a one of the pillars of the field, as it provides a common benchmark for the evaluation of techniques and approaches, and enables research advances. This thesis considers NILM in various settings. It presents the development of a versatile and low-cost data acquisition system, designed to operate effectively across different environments (residential, commercial and industrial). The system was deployed across all three environments including challenging harsh industrial settings. Furthermore, this dissertation introduces an efficient lightweight algorithm for on-site event detection to offload fog and cloud infrastructures and enable the real-time capabilities of NILM systems. This algorithm was evaluated against other techniques in terms of efficiency and computational cost. The focus of these contributions is on enabling high-frequency sampling, and data analysis and processing without increasing the needs for storage and transmission bandwidth by handling the data on-site. Finally, these contributions aim to help towards the wide adoption of high-frequency NILM by being robust and easily configurable to address different scenarios and achieve high levels of accuracy. Finally, a key contribution of this research is the creation of an industrial NILM system for collecting data from industrial sites. The creation of such a system allowed us to collect data and assist for creating a novel high-frequency industrial dataset intended for NILM applications. This will help addressing the problem of the few available datasets in this field and will be a valuable resource for future research on industrial NILM. Finally, in this dissertation, the development of a simple method to perform device health assessment for predictive maintenance applications, using electrical data (current waveforms), is described.
Degree
thesis:*- Grantor dc:publisher
- University of the Aegean
- Year dc:date
- 2025
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Kotsilitis, Sarantis
- Κοτσιλίτης, Σαράντης
Subjects
dc:subject × 18- Μη Επεμβατική Παρακολούθηση Φορτίου
- Ενεργειακή απόδοση
- Απόκτηση δεδομένων
- Επεξεργασία σήματος
- Αξιολόγηση Υγείας Συσκευών
- Ενέργεια
- Non-Intrusive Load Monitoring
- Energy efficiency technologies
- Data acquisition system (DAQ)
- Signal processing
- Device Health Assessment
- Energy
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Επιστήμες Μηχανικού και Τεχνολογία
- Αυτοματισμοί και Συστήματα ελέγχου
- Electrical Engineering, Electronic Engineering, Information Engineering
- Engineering and Technology
- Automation and control systems
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/59236