Signal processing technique utilising fourier transform methods and Artificial Neural Network pattern recognition for interpreting complex data from a multipoint optical fibre sensor system

  • Authors:

  • D. King;W. B. Lyons;C. Flanagan;E. Lewis

  • Affiliations:

  • University of Limerick, Castletroy, Co. Limerick, Ireland;University of Limerick, Castletroy, Co. Limerick, Ireland;University of Limerick, Castletroy, Co. Limerick, Ireland;University of Limerick, Castletroy, Co. Limerick, Ireland

  • Venue:
  • WISICT '04 Proceedings of the winter international synposium on Information and communication technologies
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

A multipoint (2 sensing elements) optical fibre based sensor system capable of detecting ethanol in water supplies is presented. The active sensing elements consist of an exposed core U-bend configuration in order to maximise sensitivity and the sensing system is interrogated using Optical Time Domain Reflectometry (OTDR). Artificial Neural Network (ANN) Pattern Recognition techniques have been applied to the optical fibre sensor system data in order to accurately classify each sensor element test condition. In order to reduce the computational resources of the ANN required to accurately classify the sensor system data, novel signal processing techniques utilising Fourier Transform methods are applied to the resulting OTDR data. Results are presented for the interrogation and classification of two sensors on a single fibre loop.