Method for improving infrared analysis estimations by automatically compensating for instrument instabilities

A method for estimating a physical property of a sample of a material, comprising:

• (a) obtaining infrared spectra for a representative set of specimens of the material, creating an original calibration set of spectra;
• (aa) duplicating the spectra in the original calibration set a plurality of times to enlarge the original calibration set;
• (b) measuring the physical property of the representative set of specimens using a primary reference method creating a reference set of measured properties;
• (c) introducing transmittance shift and wavelength shift into each spectra in the enlarged calibration set to form an enhanced calibration set;
• (d) repeating step (c) a plurality of times using different amounts of shift to create a plurality of enhanced calibration sets;
• (e) generating a plurality of sets of estimation equations based on correlations of the infrared spectra of the enhanced calibration sets to the reference set of measured properties;
• (f) averaging the sets of estimation equations creating an averaged set;
• and
• (g) estimating the physical property of the sample by applying the averaged set to the infrared spectrum of the sample.

A method for estimating octane numbers of a gasoline sample, comprising:

• (a) creating an original calibration set of spectra comprising the near-infrared (NIR) spectra for a representative set of gasoline specimens;
• (b) measuring the octane numbers of the representative set of specimens using a primary reference method of measurement, whereby a reference set of octane numbers is created;
• (c) introducing transmittance and wavelength shifts into each spectrum of the original calibration set, whereby an enhanced calibration set of spectra is created;
• (f) correlating the NIR spectra of the enhanced calibration set to the reference set of octane numbers;
• and
• (g) estimating the octane numbers of the gasoline sample by applying the correlation to the NIR spectrum for the sample.