Lenois Mariotto

An Innovative Loss-of-Excitation Protection Based on the Fuzzy Inference Mechanism

A new loss-of-excitation protection based on fuzzy set theory has been presented. It makes use of conventional concepts of loss-of-excitation protection in synchronous generators (i.e., the behavior of internal voltage and apparent impedance trajectory). Instead of crisp values, a fuzzy inference mechanism is applied. To show the effectiveness of the proposed technique, comparisons are made with traditional protection methods by considering different generators sizes. The protection scheme proposed displays a secure and effective high-speed loss-of-excitation detection during power swings. Furthermore, the performance of the proposed method was not affected by generator parameters.

Passive Method for Distributed-Generation Island Detection Based on Oscillation Frequency

This paper aims to present a passive island detection technique for synchronous distributed generation (DG). The technique is based on frequency oscillation estimation in order to distinguish the islanding from other events that may occur in distribution systems. The island detection uses a small window to estimate the oscillation frequency, obtaining faster responses than the existing methods which use larger windows to estimate the damping and frequency of oscillation. The algorithm performance has been tested considering different generation and load scenarios, including short circuits, load, and capacitor switching, DG outage, and islanding. The technique is reliable since it does not trip for a nonislanding event; the island detection time is less than 40 ms and its nondetection zone is less than 1.6% of the DG nominal power. The proposed method has been compared with one of the most common algorithms used in practice-the rate of change of frequency. The results show that the algorithm based on frequency oscillation detection performs better than the ROCOF, and its mathematical simplicity is adequate for practical relay implementation.

Análise da resposta em regime transitório e permanente de algoritmos para filtragem digital utilizados em relés numéricos: velocidade de convergência, overshoot e sensibilidade em relação a constante de tempo

This paper aims to assess the performance of digital filtering algorithms for numeric relays in terms of: ability to reject the DC offset component, transient response, accuracy, sampling frequency and noise immunity. The filters under investigation are the full cycle Fourier transform filters, cosine and two modified DFT methods. The current waveforms used during the tests were generated by mathematical equations and by computational simulation on a test system. The results show that the modified Fourier filters have the best performance, in most cases. These algorithms require a greater computational effort due to the number of mathematical operations involved. However, this fact has not caused considerable increase in simulation time.