Picture of neon light reading 'Open'

Discover open research at Strathprints as part of International Open Access Week!

23-29 October 2017 is International Open Access Week. The Strathprints institutional repository is a digital archive of Open Access research outputs, all produced by University of Strathclyde researchers.

Explore recent world leading Open Access research content this Open Access Week from across Strathclyde's many research active faculties: Engineering, Science, Humanities, Arts & Social Sciences and Strathclyde Business School.

Explore all Strathclyde Open Access research outputs...

Localisation of partial discharge sources using radio fingerprinting technique

Iorkyase, E. T. and Tachtatzis, C. and Atkinson, R. C. and Glover, I. A. (2015) Localisation of partial discharge sources using radio fingerprinting technique. In: 2015 Loughborough Antennas & Propagation Conference LAPC. IEEE. ISBN 9781479989430

[img]
Preview
Text (Iorkyase-etal-APC2015-localisation-partial-discharge-sources-using-radio-fingerprinting-technique)
Iorkyase_etal_APC2015_localisation_partial_discharge_sources_using_radio_fingerprinting_technique.pdf - Accepted Author Manuscript

Download (1MB) | Preview

Abstract

Partial discharge (PD) is a well-known indicator of the failure of insulators in electrical plant. Operators are pushing toward lower operating cost and higher reliability and this stimulates a demand for a diagnostic system capable of accurately locating PD sources especially in ageing electricity substations. Existing techniques used for PD source localisation can be prohibitively expensive. In this paper, a cost-effective radio fingerprinting technique is proposed. This technique uses the Received Signal Strength (RSS) extracted from PD measurements gathered using RF sensors. The proposed technique models the complex spatial characteristics of the radio environment, and uses this model for accurate PD localisation. Two models were developed and compared: k-nearest neighbour and a feed-forward neural network which uses regression as a form of function approximation. The results demonstrate that the neural network produced superior performance as a result of its robustness against noise.