DocumentCode
2717544
Title
Voltage and frequency time series analysis improvements in transforming data from quality survey into information
Author
Apetrei, Dan ; Silvas, Ioan ; Chicco, Gianfranco ; Neurohr, Ralf ; Albu, Mihaela ; Postolache, Petru
Author_Institution
DCE, SC Electr. SA, Bucharest, Romania
fYear
2010
fDate
16-19 May 2010
Firstpage
265
Lastpage
268
Abstract
The paper investigates the nature of voltage and frequency time series. The analysis is based on long term half cycle survey fulfilled with dedicated IEC 61000-4-30 class A equipment. The aim of our investigation is to improve the usefulness of half cycle data gathered in measurement campaigns. In order to do this, IEC 61000-4-30 requirements are analyzed with respect to the concept of data, information, knowledge pyramid. We present one simple way to improve the usefulness of data by drawing boxplot charts instead of using aggregated values as described in the standard. Second investigation is basic patterns identification used in order to better describe frequency. In the end, the hypotheses of fractal behaviour of voltage/frequency time series is tested by calculating Hurst exponent for different periods of time.
Keywords
Data mining; Data processing; Frequency measurement; IEC standards; Information analysis; Measurement standards; Monitoring; Power quality; Time series analysis; Voltage; Hurst exponent; IEC 61000-4-30; boxplot; frequency; pattern identification; voltage;
fLanguage
English
Publisher
ieee
Conference_Titel
Environment and Electrical Engineering (EEEIC), 2010 9th International Conference on
Conference_Location
Prague, Czech Republic
Print_ISBN
978-1-4244-5370-2
Electronic_ISBN
978-1-4244-5371-9
Type
conf
DOI
10.1109/EEEIC.2010.5489989
Filename
5489989
Link To Document