Transformer inrush causes interference harmonics at frequencies used by railway signal systems and leads to failures in train operations. For safety reasons dynamic detection and prediction of such inrush signals is an important issue. Conventional analysis techniques like Fourier transform or filters often fail to work well due to the poor time-localisation. In this paper two different approaches using modern signal analysis methods are described in detail which provide precise trend information of inrush current for operation decision within a reasonable short time. At first an analysis procedure and simulation results using Morlet wavelet transform are presented which clearly show the effectiveness of the time-frequency analysis technique. In the second part a time-domain approach using least-squares estimation algorithm is introduced and successfully applied to detect and predict the inrush current based on a suitable signal model. The advantages and disadvantages of both methods are shown by a comparison.
Detection and prediction of transformer inrush signal in railway applications
2001
8 Seiten, 7 Quellen
Conference paper
English
Wavelet-based fast discrimination of transformer magnetizing inrush current
Online Contents | 2007
|Transformer Inrush Currents Harmonic analysis in interconnected power systems.
Online Contents | 2010
|Superconducting Railway Transformer
British Library Conference Proceedings | 1999
|