Artificial Neural Network Approach for Evaluation of Gas Sensor Array Responses
Neural processing was applied for evaluation of gas sensors responses. Thin layer type gas sensors based on TiO2 sensing film were fabricated using sol-gel technology. Several variants of these devices were obtained by different technology parameters. They were characterized by exposition to air containing various volatile organic compounds. Neural processing was applied for classification of compounds according to the chemical function group and number of carbon atoms in molecules. Different statistics of correct classification obtained for the two variants deliver significant knowledge about sensing mechanisms and quality of information that may be extracted from sensor responses.
KeywordsSensor Array Neural Processing Neural Network Approach Diaphragm Pump Sensor Reaction
Unable to display preview. Download preview PDF.
- 1.Kozłowska, K., Łukowiak, A., Szczurek, A., Dudek, K., Maruszewski, K.: Sol gel coatings for electrical gas sensors. Optica Applicata 35, 783–790 (2005)Google Scholar
- 2.Łukowiak, A., Kozłowska, K., Urbański, K., Szczurek, A., Dudek, K., Maruszewski, K.: The application of an artificial neural network in the processing of output signals from a gas sensor with sol-gel-derived TiO2 film. Materials Science - Poland 25, 861–868 (2007)Google Scholar
- 4.Moseley, P.T., Norris, J., Williams, D.: Techniques and mechanisms of gas sensing. IOP Publishing Ltd., New York (1991)Google Scholar
- 6.Sberveglieri, G.: Gas sensors. Kluwer Academic Publishers, Dordrecht (1992)Google Scholar