Thesis
Machine learning for smart grid event detection
Washington State University
Master of Science (MS), Washington State University
05/2016
Handle:
https://hdl.handle.net/2376/102601
Abstract
Synchrophasor technology with Phasor Measurement Units (PMUs) deployment on electrical transmission lines has enabled real-time wide-area monitoring and a new opportunity to enhance situational awareness on the power grid system. With the high sample rate, the technology has presented a new data stream for grid operators to examine details of the dynamic behaviors of electrical systems. The technology also has brought a new attention to the problem of how to leverage the information from large-scale data streams generated by PMUs to improve situational awareness in control rooms. In this thesis, we introduce an approach to employ machine learning techniques for characterizing and classifying line events from the data streams. These events include Single Line to Ground, Line to Line, and Three Phase faults. Our work examines Bonneville Power Administrations historic synchrophasor data recorded between October 2012 and September 2013 and proposes an approach for wide-area line events detection and eventually constructs a multiple stage cascade classifier for detecting line events for use with its current PMU installation. The performance of our classifier is estimated and compared with seven supervised learning algorithms that typically perform well with large-scale data sets. We show that the learning model generated by our classifier algorithm outperformed the examined algorithms with respect to producing very low false alarms and highly accurate outcome predictions.
Metrics
2 File views/ downloads
15 Record Views
Details
- Title
- Machine learning for smart grid event detection
- Creators
- Duc The Nguyen
- Contributors
- Scott Andrew Wallace (Chair)Xinghui Zhao (Committee Member) - Washington State University, Engineering and Computer Science (VANC), School ofSarah Mocas (Committee Member)
- Awarding Institution
- Washington State University
- Academic Unit
- Electrical Engineering and Computer Science, School of
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University; [Pullman, Washington] :
- Number of pages
- 82
- Identifiers
- 99900525180301842
- Language
- English
- Resource Type
- Thesis