NEURAL NETWORKS AND THEIR ROLE IN PARTIAL DISCHARGE RECOGNITION IN GAS INSULATED SWITCHGEAR AND TRANSFORMERS: A SHORT

Michael G. Danikas - Ning Gao - Martti Aro

   In this paper a short review of the work performed in the field of partial discharge (PD) pattern recognition with the aid of neural networks in gas insulated switchgears and in power transformers is presented. The role of neural networks in shaping modern approaches to PD pattern recognition is discussed. Comments about various types of neural networks are offered as well as about the separation of PD events from noise signals. Possible remaining problem areas are discussed.

Keywords: neural networks, partial discharges, pulse height analysis, phase analysis

[full-paper]