The objective of this letter is to evaluate the condition of the transformer oil using features extracted from the acoustic and the radio-frequency (RF) partial-discharge (PD) signals. Pulse width, rise time, and frequency components of the measured PD signals were used as features to differentiate between two oil samples (i.e., new and aged samples). The artificial neural network (ANN) was trained and tested using these features. The results have shown that the frequency content of the RF signal is highly correlated with the oil status and, hence, can be used to extract information about the oil condition.
No takes yet. Share an insight, caveat, or question.
El‐Hag et al. (2010) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: