Automated detection and prediction of electrical disturbances in a power transmission system

As power quality becomes a higher priority in the electric utility industry, utilities simply do not have the required personnel to analyze the ever-growing amount of data by hand. This thesis presents an automated approach for the analysis of power quality phenomena within a power transmission system by leveraging rule-based analytics as well as machine learning to analyze the characteristics of the recorded data. Waveform signatures analyzed within this thesis include: various faults, motor starting, and incipient instrument transformer failure. The developed analytics were tested on 160 waveform files and yielded an average accuracy of 99%. Machine learning techniques are also used to predict voltage unbalance on the transmission system above a certain threshold, which yielded an accuracy of over 91%. This work will result in time savings for engineers as well as increased reliability of the transmission system by providing near real-time detection, identification, and prevention of disturbances.

Grantor dc:publisher

University of Tennessee at Chattanooga

Year dc:date.available

2024

Author dc:creator

• Boyd, Jonathan

Contributors dc:contributor

• Reising, Donald R.
• Disfani, Vahid R.; Karrar, Abdelrahman
• College of Engineering and Computer Science

• Electric power system stability
• Power transmission
• Machine learning

Licence

• http://rightsstatements.org/vocab/InC/1.0/

Language dc:language

English, eng

Repository record dc:identifier

https://scholar.utc.edu/theses/792

OAI identifier oai:identifier

oai:scholar.utc.edu:theses-1971