Method and apparatus for spectroscopic comparison of compositions
First Claim
Patent Images
1. A method of analysis, comprising the steps of:
- (A) examining at least one characteristic of a sample as a function of a variable parameter to obtain data,(B) normalizing the sample data with respect to a model by;
(1) establishing a plurality of data pairs which are characteristic values for the sample and a reference, both established at a single value of the variable parameter;
(2) performing a regression analysis with the data pairs to obtain normalization parameters by;
(a) performing a first regression, using all data pairs, to obtain a set of regression parameters;
(b) determining acceptability of the regression parameters by comparing variability of the data pairs around the model to predetermined variability limits;
(c) if the regression parameters are not acceptable, excluding data pairs which do not fall within a predetermined statistical limit around the model, and performing a further regression using a diminished set of data pairs to obtain further regression parameters; and
(d) repeating steps (b) and (c) in an iterative manner until acceptable regression parameters are obtained, or until a predetermined number of iterations are completed, whichever first occurs; and
(3) using the normalization parameters to normalize characteristic sample values to reference values,(C) making a comparison of the normalized sample data with reference data, and(D) determining from the comparison quality of the sample.
1 Assignment
0 Petitions
Accused Products
Abstract
A method of analysis of and an apparatus for analysing a sample comprising the steps of: (A) examining at least one characteristic of a sample as a function of a variable parameter to obtain data, (B) normalizing the data with respect to a model, (C) making a comparison of the normalized data with reference data, and (D) determining from the comparison the quality of the sample. As a function of such determination of quality, control functions may be carried out with respect to a sample that was analyzed, such as a blend sample in a chemical processing plant.
-
Citations
29 Claims
-
1. A method of analysis, comprising the steps of:
-
(A) examining at least one characteristic of a sample as a function of a variable parameter to obtain data, (B) normalizing the sample data with respect to a model by; (1) establishing a plurality of data pairs which are characteristic values for the sample and a reference, both established at a single value of the variable parameter; (2) performing a regression analysis with the data pairs to obtain normalization parameters by; (a) performing a first regression, using all data pairs, to obtain a set of regression parameters; (b) determining acceptability of the regression parameters by comparing variability of the data pairs around the model to predetermined variability limits; (c) if the regression parameters are not acceptable, excluding data pairs which do not fall within a predetermined statistical limit around the model, and performing a further regression using a diminished set of data pairs to obtain further regression parameters; and (d) repeating steps (b) and (c) in an iterative manner until acceptable regression parameters are obtained, or until a predetermined number of iterations are completed, whichever first occurs; and (3) using the normalization parameters to normalize characteristic sample values to reference values, (C) making a comparison of the normalized sample data with reference data, and (D) determining from the comparison quality of the sample. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18)
-
-
19. A method of analysis, comprising the steps of:
-
(A) examining the radiation transmission of the sample as a function of wavelength to obtain absorbance data, (B) normalizing the sample absorbance data with respect to a model, by; (1) establishing a plurality of data pairs which are characteristic absorbance values for the sample and a reference, both established at a single value of wavelength; (2) performing a regression analysis with the data pairs to obtain normalization parameters by; (a) performing a first regression, using all data pairs, to obtain a set of regression parameters; (b) determining acceptability of the regression parameters by comparing variability of the data pairs around the model to predetermined variability limits; (c) if the regression parameters are not acceptable, excluding data pairs which do not fall within a predetermined statistical limit around the model, and performing a further regression using a diminished set of data pairs to obtain further regression parameters; and (d) repeating steps (b) and (c) in an iterative manner until acceptable regression parameters are obtained, or until a predetermined number of iterations are completed, whichever first occurs; and (3) using the normalization parameters to normalize characteristic sample values to reference values; (C) subtracting the noramlized sample data from reference data, and (D) determining quality of the sample. - View Dependent Claims (20, 21, 22, 23, 24, 25, 26, 27, 28)
-
-
29. A method of analyzing a sample of a chemical mixture, comprising the steps of:
-
(A) examining infrared light absorption of the sample as a function of wavelength to obtain data, (B) normalizing the sample data with respect to a model by; (1) establishing a plurality of data pairs which are characteristic values for the sample and a reference, both established at a single value of wavelength; (2) performing a regression analysis with the data pairs to obtain normalization parameters by; (a) performing a first linear regression, using all data pairs, to obtain a set of regression parameters; (b) determining acceptability of the regression parameters by comparing variability of the data pairs around the model to predetermined variability limits; (c) if the regression parameters are not acceptable, exclusing data pairs which do not fall within a predetermined statistical limit around the model, and performing a further linear regression using a diminished set of data pairs to obtain further regression parameters; and (d) repeating steps (b) and (c) in an iterative manner until acceptable regression parameters are obtained, or until a predetermined number of iterations are completed, whichever first occurs; and (3) using the normalization parameters to normalize characteristic sample values to reference values, (C) making a comparison of the normalized sample data with reference data, and (D) determining from the comparison quality of the sample.
-
Specification