Picture of a sphere with binary code

Making Strathclyde research discoverable to the world...

The Strathprints institutional repository is a digital archive of University of Strathclyde research outputs. It exposes Strathclyde's world leading Open Access research to many of the world's leading resource discovery tools, and from there onto the screens of researchers around the world.

Explore Strathclyde Open Access research content

A method for distinguishing between complex transient signals in condition monitoring applications

Pinpart, Tanya and West, Graeme and Galloway, Stuart and Judd, Martin (2008) A method for distinguishing between complex transient signals in condition monitoring applications. In: Fifth International Conference on Condition Monitoring and Machine Failure Prevention Technologies, 2008-07-15 - 2008-07-18.

Full text not available in this repository. (Request a copy from the Strathclyde author)

Abstract

Partial discharges (PDs) are a cause of concern in power systems. Many detection methods and analysis techniques for PDs are available, such as UHF detection, conventional IEC 60270 method, acoustic detection, etc. This paper focuses on analysing signals obtained by the UHF method in order to distinguish between different sources of PD. Two methods based on the frequency and time characteristics are investigated with the aim of testing PD recognition based on the calculation of similarity or difference of two UHF signals. It has previously been shown that PDs can be recognized by a similarity function based on a time-frequency representation constructed using a wavelet transform. However, this is a computationally intensive process that needs to be simplified for practical implementation. Two new approaches are investigated in this paper. The first method uses spectrogram analysis to represent the UHF signal in terms of time, frequency and amplitude, from which a few key features can be extracted. The second method uses a smoothing kernel to capture the envelope of the UHF signal, effectively reducing the quantity of data to be analysed. The results from the two methods of comparing PD signals are discussed to determine their ability to measure similarity, while the results of non-PD signals are compared to show the different behaviours between PD and non-PD signals.