Detection and Classification of Partial Discharge

In remote detection of PD, a wideband antenna outside the high voltage apparatus will be used to capture the PD signal. Then the PD signal is analyzed to identify the PD source and the type of defects causing it. In addition, pulse level and location will be found with the help of radiometric techniques and analyzing the radio frequency (RF) spectrum [1], [2]. The type of PD source will be recognized by the application of useful statistical pattern classification algorithms. Mainly, nature of PD sources is different based on the mechanism of charge displacement. so automated identification of PD sources based on PD pattern seems to be achievable. Reliable automated classification of PD sources lets the online monitoring of HV insulations in a low cost investment. In this way, capability of fault identification in early stages leads to safety augmentation of HV apparatus, such as transformers, cables, electric machines, and GAS-Insulated Substation (GIS).