%0 Journal Article %1 SRIVASTAVA200324 %A Srivastava, A.K. %D 2003 %J Sensors and Actuators B: Chemical %K calibration p2map %N 1 %P 24 - 37 %R https://doi.org/10.1016/S0925-4005(03)00477-5 %T Detection of volatile organic compounds (VOCs) using SnO2 gas-sensor array and artificial neural network %U http://www.sciencedirect.com/science/article/pii/S0925400503004775 %V 96 %X This paper presents design and development of an electronic nose system based on tin oxide gas-sensors array and artificial neural network (ANN) for the identification of some of the volatile organic compounds (VOCs) relevant to environmental monitoring such as propane-2-ol, methanol, acetone, ethyl methyl ketone, hexane, benzene and xylene. An array of SnO2-based thick-film gas sensors doped with Pd, Pt and Au is used to generate the response patterns and backpropagation neural network is used for the identification. A new data transformation technique based on mean and variance of individual gas-sensor combination has been applied to improve the classification accuracy of neural network classifier. Effect of data transformation on the classification ability of neural network is studied by varying the size of array and corrupting the data with synthetic noise. Our simulation results demonstrate that the developed system is capable to identify target vapors successfully even in the noisy conditions. It is shown that the neural network processing of transformed data has not only better noise tolerance but also can classify the vapors with the array composed of fewer number of sensors as compared to that for the raw (untransformed) data.

@article{SRIVASTAVA200324, abstract = {This paper presents design and development of an electronic nose system based on tin oxide gas-sensors array and artificial neural network (ANN) for the identification of some of the volatile organic compounds (VOCs) relevant to environmental monitoring such as propane-2-ol, methanol, acetone, ethyl methyl ketone, hexane, benzene and xylene. An array of SnO2-based thick-film gas sensors doped with Pd, Pt and Au is used to generate the response patterns and backpropagation neural network is used for the identification. A new data transformation technique based on mean and variance of individual gas-sensor combination has been applied to improve the classification accuracy of neural network classifier. Effect of data transformation on the classification ability of neural network is studied by varying the size of array and corrupting the data with synthetic noise. Our simulation results demonstrate that the developed system is capable to identify target vapors successfully even in the noisy conditions. It is shown that the neural network processing of transformed data has not only better noise tolerance but also can classify the vapors with the array composed of fewer number of sensors as compared to that for the raw (untransformed) data.}, added-at = {2018-07-19T17:04:03.000+0200}, author = {Srivastava, A.K.}, biburl = {https://www.bibsonomy.org/bibtex/279b6a27b5595a0f068ce0a768a6b9aef/lautenschlager}, description = {Detection of volatile organic compounds (VOCs) using SnO2 gas-sensor array and artificial neural network - ScienceDirect}, doi = {https://doi.org/10.1016/S0925-4005(03)00477-5}, interhash = {cb8396fdb77943c4afd1f8038b3f17b7}, intrahash = {79b6a27b5595a0f068ce0a768a6b9aef}, issn = {0925-4005}, journal = {Sensors and Actuators B: Chemical}, keywords = {calibration p2map}, number = 1, pages = {24 - 37}, timestamp = {2018-07-19T17:04:03.000+0200}, title = {Detection of volatile organic compounds (VOCs) using SnO2 gas-sensor array and artificial neural network}, url = {http://www.sciencedirect.com/science/article/pii/S0925400503004775}, volume = 96, year = 2003 }