Share Email Print
cover

Proceedings Paper

Artificial neural network approach for humidity-influenced methane sensor response processing
Author(s): Guido Huyberechts; Przemyslaw M. Szecowka; J. Roggen; Benedykt W. Licznerski
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

The problem of methane sensor application in the domestic environment is presented. Tin dioxide thick film technology sensor conductance changes depending on methane concentration. Simple calibration cannot be approached because the conductance is also strongly influenced by humidity. Nevertheless, despite changing humidity accurate methane concentration determination is possible using a two-sensor system. It includes methane sensor and very selective humidity sensor. Artificial neural network is used for reasoning about methane concentration based on responses of these sensors. Feedforward type network is simulated, trained with backpropagation method and tested. The network's accuracy is compared with simple calibration of the sensor.

Paper Details

Date Published: 8 April 1996
PDF: 5 pages
Proc. SPIE 2780, Metal/Nonmetal Microsystems: Physics, Technology, and Applications, (8 April 1996); doi: 10.1117/12.238160
Show Author Affiliations
Guido Huyberechts, IMEC (Belgium)
Przemyslaw M. Szecowka, Technical Univ. of Wroclaw (Poland)
J. Roggen, IMEC (Belgium)
Benedykt W. Licznerski, Technical Univ. of Wroclaw (Poland)


Published in SPIE Proceedings Vol. 2780:
Metal/Nonmetal Microsystems: Physics, Technology, and Applications
Benedykt W. Licznerski; Andrzej Dziedzic, Editor(s)

© SPIE. Terms of Use
Back to Top