[39]

Electronic Nose Data Enhancement through Algorithmic Modeling    T. McKenzie1,  B.R. Linnell2.  1Diné College, Shiprock NM, and 2ASRC Aerospace Coorporation, Kennedy Space Center, FL.

   With the small enclosed environment of the Space Shuttle and International Space Station the detection of toxic vapors is essential.  Research previously done at Kennedy Space Center has shown that the AirSense i-PEN 3 electronic nose can identify and quantify vapors in about 90 seconds with little or no false positives.  To accomplish an increase of performance, a model of the vapors of interest, across a wide range of concentrations and relative humidities was done, however, it was suspected that when the humidity is significantly different from the model values tested, the error in concentration would be very large.  Therefore, it was found that the suspected modeling error is correct, and alternative modeling methods tested did not improve the results.