Document Type

Thesis - University Access Only

Award Date

2011

Degree Name

Master of Science (MS)

Department / School

Electrical Engineering and Computer Science

Abstract

Electronic noses are devices used to identify gas compounds and are useful in many areas. IETS sensors have qualities such as high selectivity and sensitivity that make them desirable for use in electronic nose devices. Work has been done that shows that gas identification from IETS data is feasible, however the variances in IETS responses between samples and devices were not considered. Self organizing map neural networks are pattern analysis techniques that have good generalization performance and have demonstrated ability to classify data from sensors similar to IETS sensors. The objective of this thesis was to determine if a neural network could classify IETS samples from multiple devices with 90% accuracy. Theoretical IETS data were generated and used for training and testing a self organizing map. The results suggest that classification of samples from multiple devices with a self organizing map is feasible. Future work should expand testing with a greater number of gasses including gasses that produce similar IETS responses, using data sets obtained from actual devices.

Library of Congress Subject Headings

Gas detectors

Tunneling spectroscopy

Electron spectroscopy

Format

application/pdf

Number of Pages

106

Publisher

South Dakota State University

Share

COinS