Calibration system for spectrographic analyzing instruments

• US 5,798,526 A
• Filed: 01/24/1997
• Issued: 08/25/1998
• Est. Priority Date: 01/24/1997
• Status: Expired due to Term

• Current Assignee: Infrasoft International LLC
• Original Assignee: Infrasoft International LLC
• Inventors: Shenk, John S., Westerhaus, Mark O.
• Primary Examiner(s): EVANS, FANNIE L
• Application Number:
  US08/788,259
• Time in Patent Office: 578 Days
• Field of Search: 356/300, 356/319, 356/326, 356/328, 250/339.09, 250/339.12, 364/498, 364/571.02, 364/571.04, 364/571.01
• US Class Current: 250/339.09
• CPC Class Codes: G01J 2003/2866 Markers; Calibrating of scan
• CPC Class Codes: G01J 3/28 Investigating the spectrum using colour filters G01J3/51
• CPC Class Codes: G01J 3/42 Absorption spectrometry; Double beam spectrometry; Flicker spectrometry; Reflection spectrometry beam switching arrangements G01J3/08

Abstract

In a system for analyzing materials by near infrared analysis, a library of absorbance spectra from sample materials is stored in computer storage. The absorbance spectra of an unknown material is measured and is correlated with the library spectra to select those library spectra which most closely match the spectrum of the unknown material. The selected library spectra are used to determine, by least squares convergence, or by multiple regression, coefficients in equations relating measurable properties of the unknown material to absorbance measurements made on the unknown material.

Claims

1. A method of analyzing an unknown material to measure at least one measurable characteristic of said material comprising storing a library of a multiplicity of spectra obtained from the reflectance or transmittance of a multiplicity of sample materials in each of which the value of said measurable characteristic is known, measuring a target spectrum obtained from the transmittance or reflectance of said unknown material, selecting a subset of spectra from said multiplicity which most closely match said target spectrum, determining coefficients of at least one equation which relates said measurable characteristic to values in the target spectrum, said coefficients being determined from said subset of spectra and the known values of said measurable characteristic in said sample materials corresponding respectively to said spectra of said subset, and calculating said measurable characteristic of said unknown material from said coefficients and said target spectrum in accordance with said equation.

2. A method as recited in claim 1, wherein said measurable characteristic is a concentration of a constituent of the material.

3. A method as recited in claim 2, wherein concentrations of a plurality of constituents of said unknown material are measured, the concentrations of said plurality of constituents in each of said sample materials in said library being known, the step of determining coefficients including determining the coefficients in a plurality of equations each relating a different one of said concentrations to values in the target spectrum, the coefficients of said plurality of equations being determined from said subset of spectra and from the known concentrations of constituents in the sample materials corresponding to the spectra of said subset.

4. A method as recited in claim 1, wherein said measurable characteristic comprises a property of said unknown material.

5. A method as recited in claim 1, wherein said subset of spectra is selected by correlating the target spectrum with said multiplicity of spectra in said library.

6. A method as recited in claim 5, wherein said target spectrum is divided into regions each containing a peak and wherein said regions are correlated with corresponding regions in the spectra of said library.

7. A method as recited in claim 1, wherein said coefficients are determined by partial least squares regression from said subset of spectra and from said known values of said measurable characteristic corresponding respectively to the spectra of said subset.

8. A method as recited in claim 1, wherein said spectra are absorbance spectra.

9. A method as recited in claim 1, wherein said spectra are in the near infrared range.

10. A method as recited in claim 1, wherein said spectra are obtained from reflectance measurements on said sample materials and said unknown material.

11. A method as recited in claim 1, further comprising adding the target spectrum to said library of spectra.