- Library Home /
- Search Collections /
- Open Collections /
- Browse Collections /
- UBC Theses and Dissertations /
- Cable health monitoring using power line communications...
Open Collections
UBC Theses and Dissertations
UBC Theses and Dissertations
Cable health monitoring using power line communications signal and machine learning technique Huo, Yinjia
Abstract
Automated cable health monitoring is an indispensable constituent of a smart grid (SG). We employ machine learning (ML) techniques to automatically analyze power line communications (PLC) signals to monitor the health condition of electric cables under changing environments and load conditions. PLC provides an attractive solution for both monitoring and control of an SG. With regards to the former, we reuse the wide-band signals transmitted by PLC modems to also diagnose the cable health conditions. This circumvents the requirement of conventional cable diagnostics solutions to install additional dedicated sensors. In this thesis, we first propose an ML framework using supervised learning to analyze a degradation profile automatically. This includes the type, the severity, the dimension and the location of the degradation. We carry out extensive studies with synthetic channels using a water-treeing degradation model for cross-linked polyethylene cables. We explore various ML algorithms, including neural networks and automated machine learning, for our developed supervised learning cable diagnostics schemes. Then, in the absence of a detailed characterization of a particular type of the cable degradation, we design cable anomaly detection schemes using unsupervised dimension reduction and time-series processing techniques. We incorporate laboratory measurement data and in-field collected experimental data into our design. We illustrate the effectiveness of our proposed schemes using both synthetic studies and measurement data.
Item Metadata
Title |
Cable health monitoring using power line communications signal and machine learning technique
|
Creator | |
Supervisor | |
Publisher |
University of British Columbia
|
Date Issued |
2023
|
Description |
Automated cable health monitoring is an indispensable constituent of a smart grid (SG). We employ machine learning (ML) techniques to automatically analyze power line communications (PLC) signals to monitor the health condition of electric cables under changing environments and load conditions. PLC provides an attractive solution for both monitoring and control of an SG. With regards to the former, we reuse the wide-band signals transmitted by PLC modems to also diagnose the cable health conditions. This circumvents the requirement of conventional cable diagnostics solutions to install additional dedicated sensors.
In this thesis, we first propose an ML framework using supervised learning to analyze a degradation profile automatically. This includes the type, the severity, the dimension and the location of the degradation. We carry out extensive studies with synthetic channels using a water-treeing degradation model for cross-linked polyethylene cables. We explore various ML algorithms, including neural networks and automated machine learning, for our developed supervised learning cable diagnostics schemes.
Then, in the absence of a detailed characterization of a particular type of the cable degradation, we design cable anomaly detection schemes using unsupervised dimension reduction and time-series processing techniques. We incorporate laboratory measurement data and in-field collected experimental data into our design.
We illustrate the effectiveness of our proposed schemes using both synthetic studies and measurement data.
|
Genre | |
Type | |
Language |
eng
|
Date Available |
2023-10-13
|
Provider |
Vancouver : University of British Columbia Library
|
Rights |
Attribution-NonCommercial-NoDerivatives 4.0 International
|
DOI |
10.14288/1.0437183
|
URI | |
Degree | |
Program | |
Affiliation | |
Degree Grantor |
University of British Columbia
|
Graduation Date |
2023-11
|
Campus | |
Scholarly Level |
Graduate
|
Rights URI | |
Aggregated Source Repository |
DSpace
|
Item Media
Item Citations and Data
Rights
Attribution-NonCommercial-NoDerivatives 4.0 International