The feasibility of fluctuation-enhanced sensing (FES) for the detection and classification of different gases by using a single microsensor has been experimentally demonstrated. A nanostructured semiconducting metal oxide film has been used as a chemiresistive transducer in the experiments. Similarity fingerprints have been introduced to be used with the FES signatures and the use of similarity fingerprints together with the slope fingerprints have been observed to result in a considerable increase in the classification accuracy attained by both Linear Discriminant Analysis (LDA) and Support Vector Machine (SVM) classification techniques.
Citation: Sensors and Actuators B-Chemical
Pub Type: Journals
fluctuation enhanced sensing, gas detection, linear discriminant analysis, microsensor, oxide, support vector machine