Summary
The characterization of partial discharges (PD) in equipment and electrical machines, allows obtaining relevant information on the state of the insulation system. In some cases, measurement processes may be affected by the presence of unwanted noise sources or by the presence of multiple PD sources acting simultaneously. In this sense, power ratio (PR) maps allow to successfully separate the different sources of PD and electrical noise present in a measurement. This separation technique arises by comparing the relative spectral power in two frequency bands, a low-frequency band, and a high-frequency band. Both relative spectral powers construct a two-dimensional map that, through data mining techniques, is separated into "clusters," each one associated with the sources found. This approach, based on the spectral power of PD, may additionally reveal important information in the event of any change in the insulation system, due to the appearance or growth of vacuoles or cracks, since a variation in the equivalent capacity of the same insulation would occur, and therefore, the frequency behavior of the PD pulses would also be affected. In the present study, the sensitivity of PR maps to variations in the size of different vacuoles within the insulation system will be evaluated. The behavior of the clusters regarding mean value and dispersion will also be evaluated.
Additional informations
Publication type | ISH Collection |
---|---|
Reference | ISH2017_519 |
Publication year | |
Publisher | ISH |
File size | 761 KB |
Pages number | 5 |
Price for non member | Free |
Price for member | Free |
Authors
PEREIRA DA SILVA, PRESBÍTERO XAVIER