Signal analysis methods to distinguish tracking process using time-frequency analysis

摘要

This paper proposes a method to distinguish the tracking process only in terms of voltage waveforms. Tracking is simulated using the IEC (International Electrotechnical Commission) 60112 method. The tracking process is observed in 5 stages - static state, discharge inception, discharge in progress, carbon formation, and tracking breakdown. The measured voltage waveforms are analyzed by two signal processes. The two kinds of signal processing are: improved time-frequency and timeenergy analysis. The tracking process can be distinguished by using the improved timefrequency analysis visually by stages. It can also be distinguished by using average and standard deviation for the results of the time-energy analysis. Particularly, probability distribution for the result of the time-energy analysis appears a regular pattern in tracking process by stages.