STATISTICAL PROCESSING METHODS USED IN ABNORMAL SITUATION DETECTION

• US 20120011180A1
• Filed: 09/23/2011
• Published: 01/12/2012
• Est. Priority Date: 04/04/2005
• Status: Active Grant

• Current Assignee: Fisher-Rosemount Systems Incorporated (Emerson Electric Company)
• Original Assignee: Fisher-Rosemount Systems Incorporated (Emerson Electric Company)
• Inventors: Kavaklioglu, Kadir
• Granted Patent: US 8,401,819 B2
• Time in Patent Office: Days
• US Class Current: 708/202
• CPC Class Codes: C10G 11/187 Controlling or regulating
• CPC Class Codes: G05B 23/0221 Preprocessing measurements, e.g. data collection rate adjustment; Standardization of measurements; Time series or signal analysis, e.g. frequency analysis or wavelets; Trustworthiness of measurements; Indexes therefor; Measurements using easily measured parameters to estimate parameters difficult to measure; Virtual sensor creation; De-noising; Sensor fusion; Unconventional preprocessing inherently present in specific fault detection methods like PCA-based methods
• CPC Class Codes: G05B 23/024 Quantitative history assessment, e.g. mathematical relationships between available data; Functions therefor; Principal component analysis [PCA]; Partial least square [PLS]; Statistical classifiers, e.g. Bayesian networks, linear regression or correlation analysis; Neural networks

Abstract

Detection of one or more abnormal situations is performed using various statistical measures, such as a mean, a median, a standard deviation, etc. of one or more process parameters or variable measurements made by statistical process monitoring blocks within a plant. This detection is enhanced in various cases by using specialized data filters and data processing techniques, which are designed to be computationally simple and therefore are able to be applied to data collected at a high sampling rate in a field device having limited processing power. The enhanced data or measurements may be used to provided better or more accurate statistical measures of the data, may be used to trim the data to remove outliers from this data, may be used to fit this data to non-linear functions, or may be use to quickly detect the occurrence of various abnormal situations within specific plant equipment, such as distillation columns and fluid catalytic crackers.

Claims

1. A method of fitting a sine wave to data collected within a process plant, comprising:

determining a first set of parameters of the sine wave based on one or more statistical measures of the process parameter determined from the data collected within the process plant;

storing a variable transformation of a mathematical expression of the sine wave that produces a linear expression having a second set of sine wave parameters associated therewith;

using the variable transformation to produce a set of transformed data points from the data collected within the process plant;

performing a linear regression to fit the transformed data points to the linear expression; and

determining the second set of sine wave parameters based on the linear regression.

2. The method of claim 1, wherein the first set of parameters of the sine wave includes an offset and a gain.

3. The method of claim 2, wherein determining the first set of parameters of the sine wave includes determining the offset as a mean value of the data collected within the process plant and determining the gain based on the difference between a minimum value and a maximum value of the data collected within the process plant.

4. The method of claim 2, wherein the second set of parameters of the sine wave includes a cyclic frequency and a phase.

5. The method of claim 2, wherein the variable transformation is of the form:

6. The method of claim 5, further including applying a variable transformation to produce a further linear expression including the offset and the gain, applying a linear regression to the further linear expression to determine a new set of values for the offset and the gain and determining a new set of values for the periodic frequency and the phase based on the new set of values for the offset and the gain.

7. The method of claim 6, including iteratively determining values for the sine wave offset, gain, periodic frequency and phase until a change in the values for one or more of the sine wave offset, gain, periodic frequency and phase becomes less than one or more threshold values.

8. The method of claim 1, wherein determining the first set of parameters of the sine wave, using the variable transformation, performing the linear regression and determining the second set of sine wave parameters are performed in a device that collects or measures the data collected within the process plant.