| Summary: | 碩士 === 國立臺灣科技大學 === 電機工程系 === 99 === Underground cables are the main apparatus of electricity transmission in recent urban. The common defects usually existed in the artificial connection part. Under a long-term operation, these defects might become unexpected accident. In recent, a long-term monitoring of underground cable joint is a method to avoid accident. Increasing with time during long-term monitoring, the data size might reach the limit of measuring device.
This research proposed a real-time insulation status assessment method for underground cable joint. This method includes the acquisition of test voltage and PD signal, simplification, transformation, and extraction. In order to solve the problem of limitation of data, the extracted data will be shifted by a fixed width to calculate a feature and assessment value of insulation status. The duration, which feature is over the assessment value, is represented as the transition period of insulation status. When the assessment value is 3 times the standard deviation, two transition periods can be effectively detected. These two transition periods were used to determine the insulation status, including early, middle and final term. Finally, no matter using wavelet filtering or not, this method can assess insulation status effectively. Finally, comparing the use of wavelet filtering, the diagnosis of insulation status without wavelet filtering was feasible and improved the operation speed and program complexity.
|
|---|
