8 ITMP3 / 8éme conférence internationale de photoacoustique et photothermique
J. Phys. IV France 04 (1994) C7-107-C7-110
Identification and concentration measurements of atmospheric pollutants through a neural network analysis of photothermal signaturesK. Boccara1 and A.C. Boccara2
1 Process Engineering UMIST, Manchester M601 QD, U.K.
2 Laboratoire d'Optique, ESPCI, UPR A0005 du CNRS, 10 rue Vauquelin, 75005 Paris, France
We propose here to use a neural network as a help in the identification of individual components of a gaz mixture and in the evaluation of the concentration level of each component as they appear in real photoacoustic spectra in presence of background signals and noises. The method is based on the dependance of the activation functions with regard to the input level controled variation and on the desactivation of the first activated outputs.
© EDP Sciences 1994