For the development of the future smart grid, the detection of power quality events is a key issue for the power system monitoring. Voltage sags, swells, harmonics (variations) and interruptions, which produce large losses for commercial and industrial consumers, are the main events to be considered due to the sensitivity of equipment to these electrical anomalies. The steady-state events are even more frequently accompanied by transients, the discrimination and localization being far more exigent and requiring advanced signal separating tools to be incorporated in the measurement equipment. This paper shows the event detection performance of the spectral kurtosis as a signal separating tool in the frequency domain. The disturbances under test are hybrid signals resulting from the coupling between amplitude defects and non-desired higher frequencies. Being a fourth-order spectrum, the kurtosis is confirmed as a noise-resistant tool that enhances impulsiveness, therefore characterizing the electrical anomalies. In the beginning of the analysis, the voltage sag is established as a reference; then, the disturbances (oscillatory transients and harmonics) are coupled at the starting and ending instants of the sag, resulting in complex hybrid events. The results show that the spectral kurtosis enhances the detection Energies 2015, 8 9778 of both types of events (steady state and transients), which are outlined in a bump shape in the fourth-order frequency pattern and centered in the main carrier frequency. Indeed, while the oscillatory transients are associated with softer and lower-amplitude peaks, the harmonics correspond to crisper and higher ones. As these mixed electrical faults are very common in the actual power grid, the article postulates the higher-order spectra to be implemented in prospective online measurement instruments.

No abstract

No abstract

Abstract. The aim of this work is by using artificial neural networks (ANNs) compare six regression algorithms supported by 14 power-quality features, based on higher-order statistics (HOS). In addition, we have combined time and frequency domain estimators to deal with non-stationary measurement sequences; the final target is to implement the system in a smart grid to guarantee compatibility between all the equipment connected. The main results were based on spectral kurtosis measurements, which easily adapt to the impulsive nature of the power quality events. Through these results we have verified that the developed technique is capable of offering interesting results at classifying power quality (PQ) disturbance. We can conclude that using radial basis networks, generalized regression and multilayer perceptron, we have obtained the best results mainly due to the non-linear nature of data.