Gas chromatograph peak identification using statistically determined fuzzy logic membership functions

This paper presents a fuzzy logic method for identifying peaks in the output from a gas chromatograph (GC), an instrument used in chemical analysis. The fuzzy membership functions are determined based on means and standard deviations of the peaks in a large data set which were identified by traditional methods. The method has been implemented in a C++ computer program. As a test of the method, 30 data sets, including 2400 peaks, were analyzed by the program with a successful identification rate of 99%. The success rate of the normal method (output of the GC laboriously corrected by a human technician) was only 96%