Abstract
Gas chromatography and pattern recognition techniques have been used to develop a potential method to authenticate fuel spill identification standards. The test data consisted of 65 gas chromatograms of JP-4, Jet-A, JP-7, JPTS, JP-5, AVGAS, and Diesel fuels, which were obtained from Mukilteo and Wright Patterson Energy Management Laboratories. A genetic algorithm (GA) was used to identify features in the gas chromatograms of neat jet fuels characteristic of fuel type. Authentication of the fuel recognition standards was achieved by projecting each standard onto a principal component map developed from 64 neat jet fuel samples and the 14 chromatographic peaks selected by the GA. The proposed method has been validated using nine fuel recognition standards purchased from an unnamed commercial supplier.