An International Research Journal

AJP

SSN : 0971 - 3093

Vol  12, Nos 2&3, April - September, 2003

Asian Journal of Physics                                                                                                  Vol. 12, Nos 2&3 (2003) 183-204


Novel signal processing for advanced optical sensors


M N Taib1 and R Narayanaswamy2

1Faculty of Electrical Engineering, Universiti Teknologi MARA (UiTM)

40450 Shah Alam, Selangor D E Malaysia

2 Department of Instrumentation, and Analytical Science,UMIST, P O Box 88

Manchester M60 1QD, U K

 ___________________________________________________________________________________________________________________________________

A novel technique for processing signals of an optical fibre sensor has been demonstrated  in this paper. A multilayer artificial neural network (ANN), with only one hidden layer, was trained via the recursive prediction error (PRE) algorithm to adapt to the input-output signals from a Light Emitting Diode  used in conjunction with a reflectance fibre optic sensor. Three LEDs (red, Orange and yellow) operating in the visible region at peak wavelengths of 660nm, 620nm and 590nm,respectively were employed as light source, the optical signals reflected  from the tip of a fibre optic sensor based on immobilized thymol blue were correlated directly to pH measurements. A network architecture of 3-input neuros, 13-hidden neurons and 1-output neuron was found most suitable for this application and a linear calibration output  was produced for a pH measurement range of 0.90 to 10.28.The application of (ANN) to the optical fibre sensor also exhibited a very robust system where fluctuations in the sensor signals of upto 7.0% can be tolerated before the prediction error magnitude becomes larger than 1.0pH.

Total Refs: 23

___________________________________________________________________________________________________________________________________


Asian Journal of Physics                                                                                                  Vol. 12, Nos 2&3 (2003) 205-211

 

Long Period Grating Sensor to monitor Fuel Quality

 

Rosane Falate, José Luís Fabris, Márcia Müller and Hypolito José Kalinowski

Centro Federal de Educação Tecnológica do Paraná

Av. Sete de Setembro, 3165 – 80230-901 Curitiba, Brazil

e-mail:rfalate@cefetpr.br

 ___________________________________________________________________________________________________________________________________

We describe the production and characterisation of  Long Period Gratings in fibre optics. The gratings are written in standard single-mode fibre using the electrical arc technique while their transmission spectrum is used to control the grating length. Optical characterisation  and  wavelength dependence  with temperature  and  refractive  index of the external medium are  shown. An  application of  the LPG as a  sensor to monitor the fuel quality (Brazilian standard mixture of ethanol and gasoline) is presented.

Total refs: 16

Cited: 7 times

___________________________________________________________________________________________________________________________________

© ANITA PUBLICATIONS

All rights reserved