Photometer system for obtaining reliable data

• US 6,421,614 B1
• Filed: 07/26/1999
• Issued: 07/16/2002
• Est. Priority Date: 07/26/1999
• Status: Expired due to Fees

• Current Assignee: Donald S. Goldman, Nelson Wayne Lytle
• Original Assignee: Donald S. Goldman, Nelson Wayne Lytle
• Inventors: Lytle, Nelson Wayne, Goldman, Donald S.
• Primary Examiner(s): Hoff, Marc S.
• Assistant Examiner(s): Kim, Paul
• Application Number:
  US09/361,090
• Time in Patent Office: 1,086 Days
• Field of Search: 702/72, 702/23, 702/30, 702/32, 250/339.07, 250/339.09, 250/252.1, 356/73
• US Class Current: 702/32
• CPC Class Codes: G01N 21/274 Calibration, base line adjustment, drift correction
• CPC Class Codes: G01N 21/359 using near infrared light

Abstract

A system for obtaining reliable data from a chemical sample for at least one variable environmental condition utilizing a photometer. The photometer obtains sample characteristic data at a plurality of discrete wavelengths. The reliability of the data is determined by obtaining a training set of data, and obtaining a number representing the variation between the mean center of the training set of data and the data representing the sample characteristic obtained by the photometer. Such number represents the basis of determining outliers of the data representative of the reliability of the data.

Claims

1. A system for obtaining reliable chemometric data from a sample including components from a chemical process under at least one variable sample condition, comprising:

a. a fixed filter photometer for obtaining data representing a sample characteristic by interacting at least a first beam of electromagnetic radiation at a first wavelength possessing a narrow band width, and a second beam of electromagnetic radiation at a second wavelength possessing a narrow bandwidth, said first and second wavelengths relating to components in the sample; and

b. means for determining the reliability of said data representing a sample characteristic, said means utilizing a training set of data representing said sample characteristic being repeatedly determined by varying the value of at least one sample condition at said first and second wavelengths, a mean center of said training set of data at said first and second wavelengths, and a number representing the amount of variation between said mean center of said training set of data and said data representing a sample characteristic, to determine outliers of said data representing a sample characteristic.

2. The system of claim 1 in which said number representing said amount of variation comprises a Mahalanobis distance.

3. The system of claim 1 in which said number representing an amount of variation comprises a sum of square error.

4. The system of claim 1 in which said means for determining the reliability of said data representing a sample characteristic further comprises a limit for said number representing an amount of variation.

5. The system of claim 4 in which said number representing said amount of variation comprises a Mahalanobis distance.

6. The system of claim 5 in which said number representing an amount of variation comprises a sum of squares error.

16. The system of claim 5 in which said means for determining the reliability of said data representing a sample characteristic further comprises a limit for said number representing an amount of variation.

7. A method of obtaining reliable chemometric data from a sample including components from a chemical process, under at least one variable sample condition, comprising:

a. obtaining data representing a sample characteristic by utilizing a fixed filter photometer to generate and to interact with the sample at least a first beam of electromagnetic radiation at a first wavelength possessing a narrow bandwidth, and a second beam of electromagnetic radiation at a second wavelength possessing a narrow bandwidth, said first and second wavelengths relating to components in the sample;

b. determining the reliability of said data representing a sample characteristic utilizing a training set of data representing said sample characteristic being repeatedly determined by varying the value of at least one sample condition at said first and second wavelengths, mean centering said training set of data at said first and second wavelengths, and calculating a number representing the amount of variation between said mean center of said training set of data and said data representing a sample characteristic, to determine outliers of said data representing a sample characteristic.

8. The method of claim 7 which further comprises determining a limit for said number representing an amount of variation between said mean center of said training set of data and said data representing a sample characteristic.

9. The method of claim 7 in which said number representing said amount of variation comprises a Mahalanobis distance.

10. The method of claim 7 in said which said number representing said amount of variation comprises a sum of square error.

11. The method of claim 9 which further comprises determining a limit for said number representing an amount of variation between said mean center of said training set of data and said data representing a sample characteristic.

12. The method of claim 10 in which said number representing said amount of variation comprises a Mahalanobis distance.

13. A system for obtaining reliable chemometric data from a sample including components from a chemical process under at least one variable sample condition, comprising:

a. a photometer for obtaining data representing a sample characteristic by interacting electromagnetic radiation with the sample and analyzing said interaction at least at a first wavelength possessing a narrow bandwidth, and a second wavelength possessing a narrow bandwidth, said first and second wavelengths relating to components in the sample; and

b. means for determining the reliability of said data representing a sample characteristic, said means utilizing a training set of data representing said sample characteristic being repeatedly determined by varying the value of at least one sample condition at said first and second wavelengths, a mean center of said training set of data at said first and second wavelengths, and a number representing the amount of variation between said mean center of said training set of data and said data representing a sample characteristic, to determine outliers of said data representing a sample characteristic.

14. The system of claim 13 in which said number representing said amount of variation comprises a Mahalanobis distance.

15. The system of claim 14 in which said number representing an amount of variation comprises a sum of square error.

17. A method of obtaining reliable chemometric data from a sample including components in a chemical process, under at least one variable sample condition, comprising:

a. obtaining data representing a sample characteristic by utilizing a photometer to generate and to interact with the sample at least a first beam of electromagnetic radiation at a first wavelength possessing a narrow bandwidth, and a second beam of electromagnetic radiation at a second wavelength possessing a narrow bandwidth, said first and second wavelengths relating to components in the sample;

b. determining the reliability of said data representing a sample characteristic utilizing a training set of data representing said sample characteristic being repeatedly determined by varying the value of at least one sample condition at said first and second wavelengths, mean centering said training set of data at said first and second wavelengths, and calculating a number representing the amount of variation between said mean center of said training set of data and said data representing a sample characteristic, to determine outliers of said data representing a sample characteristic.

18. The method of claim 17 which further comprises determining a limit for said number representing an amount of variation between said mean center of said training set of data and said data representing a sample characteristic.

19. The method of claim 17 in which said number representing said amount of variation comprises a Mahalanobis distance.

20. The method of claim 17 in said which said number representing said amount of variation comprises a sum of squares error.