Power transients characterization and classification using higher-order cumulants and neural networks

This paper deals with power-quality (PQ) event characterization using higher-order cumulants. Their maxima and minima are the main features, and classification is based in competitive layers. We concentrate on differentiating two types of transients (short duration and long duration). By measuring the fourth-order cumulants´ maxima and minima, we build the two-dimensional feature measured vector. Cumulants are calculated over high-pass filtered signals, to avoid the 50-Hz signal. We have observed that the minima and maxima produce clusters in the feature space for 4th-order cumulants; third-order cumulants are not capable of differentiate these two very similar PQs. The experience sets the foundations of an automatic procedure.