An automated method for monitoring a polyethylene recycling process ha
s been developed. The products of catalyzed polyethylene cracking are
alkanes, monounsaturated alkenes (monoene) and dienes, which were anal
yzed by gas chromatography-mass spectrometry. Three expert systems, a
knowledge based expert system, a univariate rule building expert syste
m and a fuzzy multivariate expert system, were developed and evaluated
for their ability to correctly classify products of polyethylene crac
king by their degree of unsaturation. Two test sets were used in the e
valuation. One contained spectra from a mass spectral library. The sec
ond was a collected set of hydrocarbon standards. On both test sets, t
he rule building expert systems performed better than the knowledge ba
sed expert system. The chemometric detectors were effective for mathem
atically resolving chromatograms by chemical functionality.