This thesis dealt with a study on the algorithm design capable of detecting and analyzing the partial discharge pulses for insulation diagnosis of power facilities. Also, the proposed a TF map(Time frequency map) algorithm which can easily detect noises and discriminate the partial discharge pulse by analyzing the alteration of amplitude by time and frequency of signal.
The measurement system consists of a coupling capacitor of coupling network, and a low noise amplifier. In order to simulate a inner discharge, a point-plane and a plane-plane electrode system were fabricated. The data acquisition module consists of a differential mode of a 2-channel analog input with the horizontal resolution of 100MS/s, vertical resolution of 8bit, and measurement range of ±20V.
The measured partial discharge pulses are input to a data acquisition device and performed through an fabricated TF map, filter application and phase resolved partial discharge analysis(PRPDA) algorithm. The analysis graph is shown on a display unit.
The experiment for evaluating the fabricated algorithm was conducted in accordance with IEC 60270 and the characteristics of discharges such as partial discharge magnitude(q), phase distribution(φ), and the repetition rate(n) were also analyzed.
According to the experiment results, the noise and partial discharge pulse signal were discriminated, and φ-q and φ-q-n were also analyzed.
The phase distribution of partial discharge pulse at each electrode system was 12°～132°, 191°～308° in plane-plane electrode and 7°～90°, 195°～282° in point-plane electrode.
Therefore, it was confirmed that the proposed algorithm was efficient enough in analyzing and detecting the partial discharge pulse.