Processing data base information having nonwhite noise

• US 5,410,492 A
• Filed: 05/27/1993
• Issued: 04/25/1995
• Est. Priority Date: 01/29/1992
• Status: Expired due to Fees

• Current Assignee: ARCH Development Corporation (University of Chicago)
• Original Assignee: ARCH Development Corporation (University of Chicago)
• Inventors: Morreale, Patricia, Gross, Kenneth C.
• Primary Examiner(s): Walsh, Donald P.
• Assistant Examiner(s): CARROLL, CHRISMAN DION
• Application Number:
  US08/068,712
• Time in Patent Office: 698 Days
• Field of Search: 376/215, 376/216, 376/217, 395/914, 395/915, 364/492, 364/576, 364/554, 364/600
• US Class Current: 702/185
• CPC Class Codes: G05B 9/03 with multiple-channel loop, i.e. redundant control systems
• CPC Class Codes: G21C 17/00 Monitoring; Testing measuring in general G01; Maintaining
• CPC Class Codes: Y02E 30/30 Nuclear fission reactors
• CPC Class Codes: Y10S 706/914 Process plant

Abstract

A method and system for processing a set of data from an industrial process and/or a sensor. The method and system can include processing data from either real or calculated data related to an industrial process variable. One of the data sets can be an artificial signal data set generated by an autoregressive moving average technique. After obtaining two data sets associated with one physical variable, a difference function data set is obtained by determining the arithmetic difference between the two pairs of data sets over time. A frequency domain transformation is made of the difference function data set to obtain Fourier modes describing a composite function data set. A residual function data set is obtained by subtracting the composite function data set from the difference function data set and the residual function data set (free of nonwhite noise) is analyzed by a statistical probability ratio test to provide a validated data base.

Claims

1. A method of processing at least one data base obtained from at least one of an industrial process and at least a first and second sensor for determining fault conditions of data characteristic of a process, comprising the steps of:

collecting data from at least a first and second sensor to redundantly accumulate data from at least one physical variable of the industrial process to provide a first data signal from said first sensor and a second data signal from said second sensor, each said data signal being characteristic of the one physical variable;

obtaining a difference function of the data characteristic of the arithmetic difference pairwise between said first data signal and said second data signal at each of a plurality of different times of sensing the one physical variable;

obtaining a frequency domain transformation of said first difference function to procure Fourier coefficients corresponding to Fourier frequencies;

generating a composite function over a time domain using the Fourier coefficients;

obtaining a residual function over time by determining the arithmetic difference between the difference function and the composite function, the residual function having reduced serially correlated noise;

operating on the residual function using computer means for performing a statistical analysis technique to determine whether alarm condition data is present in at least one of the data base from the industrial process and from the at least first and second sensor, the residual function including white noise characteristics of an uncorrelated function of reduced skewness relative to the difference function and input to the statistical analysis technique; and

producing said at least one data base having data with identified alarm conditions associated therewith, allowing separation of said data with identified alarm conditions.

2. The method described is claim 1 wherein said computer means comprises an artificial intelligence system.

3. The method as defined in claim 1 wherein the residual function further comprises reduced Markov dependent noise.

4. The method as defined in claim 1 wherein the industrial process data base comprises at least one of a chemical process data base, a mechanical process data base and an electrical process data base.

5. The method as defined in claim 1 wherein the step of obtaining Fourier coefficients comprise iteratively determining the minimum number of Fourier harmonics able to generate the composite function.

6. The method as defined in claim 1 further including at least one of the steps of identifying or removing data having alarm conditions associated therewith.

7. A method of accumulating a data base with validated information from at least one of an industrial process and a sensor for determining fault conditions therein, comprising the steps of:

collecting data from at least one sensor to detect at least one physical variable of the industrial process to provide a real signal from said one sensor;

generating an artificial data signal characteristic of the one physical variable;

obtaining a difference function characteristic of the difference pairwise between said real signal and said artificial signal at each of a plurality of different times of sensing the one physical variable;

obtaining a frequency domain transformation of said difference function;

generating a composite function data base over a time domain;

obtaining a residual function over time by determining the difference between the difference function and the composite function data base, the residual function including white noise characteristics of an uncorrelated signal of reduced skewness compared to the difference function; and

operating on the residual function using a computer means for performing a statistical analysis technique to determine whether alarm condition data is present in at least one of the industrial process and the at least one sensor, the residual function having said white noise characteristics input to the statistical analysis technique for providing the data base with validated information.

8. The method as defined in claim 7 wherein the step of obtaining a frequency domain transformation comprises performing a Fourier transformation.

9. The method as deemed in claim 7 wherein the steps of obtaining a composite function over time comprises performing an autoregressive moving average analysis.

10. The method as deemed in claim 7 further including the step of determining a difference function for both the artificial signal and the real signal, as well as a separate pair of real signals.

11. The method as defined in claim 7 wherein the residual function further comprises reduced Markov dependent noise.

12. The method as defined in claim 8 wherein the step of obtaining a frequency domain transformation comprises obtaining Fourier coefficients iteratively to determine the minimum number of Fourier harmonics able to generate the composite function data base.

13. A system for accumulating a data base with validated signals from data obtained from at least one of an industrial process and a sensor for determining a fault condition therein, comprising:

first means for providing a first set of data characteristic of at least one physical variable of the industrial process;

second means for providing a second set of data calculationally determined and related to the one physical variable of the industrial process;

third means for determining a difference function of data characteristic of the arithmetic difference pairwise between said first set of data and said second set of data at each of a plurality of different times of the one physical variable being sensed;

fourth means for obtaining a residual function of data over time by means for determining the arithmetic difference between the difference function data and the composite function data, the residual function data having reduced serially correlated noise; and

fifth means for operating on the residual function data including a computer means for performing a statistical analysis technique and for determining whether corrupted data is present in at least one of the data obtained from the industrial process and the sensor and with said third means, said fourth means, and said fifth means cooperatively providing the residual function data with white noise characteristics of an uncorrelated signal of reduced skewness relative to the difference function as an input to the statistical analysis technique to provide the data base with validated signals.

14. The system as defined in claim 13 further including means for obtaining a frequency domain transformation of said difference function data.

15. The system as defined in claim 13 wherein said computer means comprises an artificial intelligence system.

16. The system as defined in claim 13 wherein said means for calculationally generating a second set of data comprises computer means for executing a computer program.

17. The system as defined in claim 16 wherein the computer program includes an autoregressive moving average procedure.

18. The system as defined in claim 13 wherein the system includes at least one pair of sensor data sets.

19. The system as defined in claim 13 wherein said computer means executes a computer program including a statistical probability ratio test on the residual function data.

20. The system as defined in claim 13 further including means for at least one of identifying or removing alarm data from the industrial process data and the sensor data.

21. A method of processing at least one data base obtained from at least one of an industrial process and at least a first and second sensor for determining fault conditions of data characteristic of a process, comprising the steps of:

collecting data from the at least first and second sensor to accumulate comparative data for at least one physical variable of the industrial process to provide a first data signal from said first sensor and a second data signal from said second sensor, each said data signal being characteristic of the one physical variable;

obtaining a difference function of the data characteristic of the arithmetic difference pairwise between said first data signal and said second data signal at each of a plurality of different times of sensing the one physical variable;

obtaining a frequency domain transformation of said first difference function to procure Fourier coefficients corresponding to Fourier frequencies;

generating a composite function over a time domain using the Fourier coefficients;

obtaining a residual function over time by determining the arithmetic difference between the difference function and the composite function, the residual function having reduced serially correlated noise;

operating on the residual function using computer means for performing a statistical analysis technique to determine whether alarm condition data is present in the at least one data base from the industrial process and the first and second sensor, the residual function including white noise characteristics of an uncorrelated function of reduced skewness relative to the difference function and input to the statistical analysis technique; and

producing said at least one data base having data with identified alarm conditions associated therewith, allowing separation of said data with identified alarm conditions.

22. The method as defined in claim 21 wherein said first and second sensor measure wear out conditions in machinery.

23. A system for accumulating a data base with validated signals from data obtained from at least one of an industrial process and a sensor for determining a fault condition therein, comprising:

first means for providing a first set of data characteristic of at least one physical variable of the industrial process;

second means for providing a second set of data related to the one physical variable of the industrial process;

third means for determining a difference function of data characteristic of the arithmetic difference pairwise between said first set of data and said second set of data at each of a plurality of different times of the one physical variable being sensed;

means for generating composite function data over a time domain;

fourth means for obtaining a residual function of data over time by means for determining the arithmetic difference between the difference function data and the composite function data, the residual function data having reduced serially correlated noise; and

fifth means for operating on the residual function data including a computer means for performing a statistical analysis technique and for determining whether degraded data is present in at least one of the data from the industrial process and from the sensor and with said third means, said fourth means, and said fifth means cooperatively providing the residual function data with white noise characteristics of an uncorrelated signal of reduced skewness relative to the difference function as an input to the statistical analysis technique to provide the data base with validated signals.