Abstract
Extreme value theory (EVT) deals with extremal behavior at the tails of distributions of random variables (maxima and minima). In this paper the results of applying some basic ideas associated EVT to the statistical analysis of the leakage current peak magnitude (envelope) is presented. The measurement and analysis of leakage current for condition-based monitoring of polluted insulators is well documented. Research has established that, insulator surface arcing causes significant variations in both the magnitude and shape of the leakage current of polluted insulators and several techniques exist for analysis. A technique is presented here which employs EVT to discriminate and characterize the largest and rarest peaks (which occur with highest severity); namely the characteristic largest value and the extreme value risk function.
| Original language | English (US) |
|---|---|
| Title of host publication | Proceedings of the IEEE Power Engineering Society Transmission and Distribution Conference |
| Pages | 2531-2535 |
| Number of pages | 5 |
| Volume | 4 |
| State | Published - 2000 |
| Event | Proceedings of the 2000 Power Engineering Society Summer Meeting - Seattle, WA, United States Duration: Jul 16 2000 → Jul 20 2000 |
Other
| Other | Proceedings of the 2000 Power Engineering Society Summer Meeting |
|---|---|
| Country/Territory | United States |
| City | Seattle, WA |
| Period | 7/16/00 → 7/20/00 |
Keywords
- Envelope
- Exceedance
- Gumbel distribution
- Insulators
- Leakage current
- Pollution
ASJC Scopus subject areas
- General Engineering
- General Energy