Detection and classification of partial discharge sources under variable frequency and air pressure
Loading...
Date
2018-12-19
Authors
Nasr Esfahani, Ali
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
With the development of electrical apparatus in more-electric aircraft (MEA), demand for
higher electric power is increasing rapidly. In turn, this requires a higher voltage level in
the power generation system of aircraft which increases the electric stress on the insulation systems. At higher voltages, the insulation systems of more-electric aircraft are prone to partial discharge (PDs) initiation under the operating condition.
The objective of this thesis is to perform a comprehensive study on developing diagnostic
methods for the insulation condition monitoring and PD source identification. An algorithm
is developed based on the combination of wavelet and energy techniques to detect the PD
pulses from the measured noisy PD signals. In addition, based on the statistical distributions of PD pulse waveform characteristics, a classification and separation algorithm is developed for the identification of multi-source PDs using kernel support vector machine (KSVM) as the classifier. The experimental results show that the proposed algorithms show a high performance and accuracy for PD source detection and recognition.
Description
Keywords
Partial discharge source identification