Method and system for detecting abnormal operation in a process plant

• US 7,912,676 B2
• Filed: 07/25/2006
• Issued: 03/22/2011
• Est. Priority Date: 07/25/2006
• Status: Active Grant

• Current Assignee: Fisher-Rosemount Systems Incorporated (Emerson Electric Company)
• Original Assignee: Fisher-Rosemount Systems Incorporated (Emerson Electric Company)
• Inventors: Miller, John P.
• Primary Examiner(s): Wachsman; Hal D
• Assistant Examiner(s): Suarez; Felix E
• Application Number:
  US11/492,467
• Publication Number:
  US 20080027678A1
• Time in Patent Office: 1,701 Days
• Field of Search: 702/118, 702/119, 702/122, 702/179, 702183-185, 702/186, 700/108, 703/2, 706/16
• US Class Current: 702/185
• CPC Class Codes: G05B 23/021 adopting a different treatment of each operating region or a different mode of the monitored system, e.g. transient modes; different operating configurations of monitored system
• CPC Class Codes: G05B 23/0243 model based detection method, e.g. first-principles knowledge model
• CPC Class Codes: G05B 23/0297 Reconfiguration of monitoring system, e.g. use of virtual sensors; change monitoring method as a response to monitoring results

Abstract

A system for detecting abnormal operation of at least a portion of a process plant includes a model to model at least the portion of the process plant. The model may be configurable to include multiple regression models corresponding to multiple different operating regions of the portion of the process plant. The system may also include a deviation detector configured to determine if the actual operation of the portion of the process plant deviates significantly from the operation predicted by the model. If there is a significant deviation, this may indicate an abnormal operation.

Claims

1. A method for facilitating detection of abnormal operation of a process in a process plant, the method comprising:

collecting first data sets for the process is producing a product by processing physical materials using one or more field devices communicatively coupled to one or more process controllers and while the device is in a first normal operating region, the first data sets generated from process variables of the process in the first normal operating region;

generating a first regression model of the process in the first normal operating region using the first data sets, the first regression model a function of an independent process variable;

determining a first range for the independent process variable in which the first regression model is valid;

generating a process model to include the first regression model, the process model a function of the independent process variable;

collecting second data sets for the process while the process is in a second normal operating region, the second data sets generated from process variables of the process in the second normal operating region;

generating a second regression model of the process in the second operating region using the second data sets, the second regression model a function of the independent process variable;

determining a second range for the independent process variable in which the second regression model is valid;

revising the process model to include simultaneously the first regression model for the first range of the independent process variable and the second regression model for the second range of the independent process variable; and

using the process model to detect and provide an indication of abnormal operation of the process,

wherein abnormal operation is indicated based on one or more difference values, each difference value the difference between a measured value of a dependent process variable at a corresponding value of the independent process variable and a modeled value of the dependent process variable at the corresponding value of the independent process variable.

2. A method according to claim 1, further comprising revising at least one of the first range and the second range, if necessary, so that the first range and the second range do not overlap.

3. A method according to claim 1, further comprising:

modeling the process using the process model after generating the process model to include the first regression model and prior to revising the process model to include the second regression model; and

modeling the process using the process model after generating the process model to include the first regression model and after revising the process model to include the second regression model.

4. A method according to claim 3, further comprising:

prior to collecting second data sets, receiving a data set outside of the first range;

generating an indicator that the received data set is outside of a validity range of the process model;

wherein collecting second data sets comprises collecting second data sets in response to receiving an authorization.

5. A method according to claim 4, wherein the authorization comprises an authorization to train the process model for the second operating region.

6. A method according to claim 1, further comprising:

collecting at least third data sets for the process while the process is in at least a third normal operating region, the at least third data sets generated from process variables of the process in the at least the third normal operating region;

generating at least a third regression model of the process in the at least the third operating region using the at least third data sets, the third regression model a function of the independent process variable;

determining at least a third range for the independent process variable in which the at least the third regression model is valid; and

revising the process model to include simultaneously the first regression model for the first range of the independent process variable, the second regression model for the second range of the independent process variable and the at least the third regression model for at least the third range of the independent process variable.

7. A method according to claim 6, further comprising revising at least one of the first range, the second range, and the at least the third range, if necessary, so that the first range, the second range and the at least the third range do not overlap.

8. A method according to claim 6, further comprising:

modeling the process using the process model after generating the process model to include the first regression model and prior to revising the process model to include the second regression model;

modeling the process using the process model after generating the process model to include the first regression model, after revising the process model to include the second regression model, and prior to revising the process model to include the at least the third regression model; and

modeling the process using the process model after generating the process model to include the first regression model, after revising the process model to include the second regression model, and after revising the process model to include the at least the third regression model.

9. A method according to claim 1, further comprising:

generating an interpolated model for a third range of the independent process variable between the first range of the independent process variable and the second range of the independent process variable;

revising the process model to include simultaneously the first regression model for the first range of the independent process variable, the second regression model for the second range of the independent process variable and the interpolated model for the third range of the independent process variable between the first range and the second range.

10. A method according to claim 9, further comprising:

modeling the process using the process model after generating the process model to include the first regression model and prior to revising the process model to include the second regression model and the interpolated model;

modeling the process using the process model after generating the process model to include the first regression model and after revising the process model to include the second regression model and the interpolated model.

11. A method according to claim 1, wherein revising the process model comprises revising the process model in response to receiving a user authorization.

12. A method according to claim 11, wherein the user authorization comprises an authorization to train the process model;

wherein collecting second data sets comprises collecting second data sets in response to receiving the user authorization.

13. A method according to claim 12, wherein the user authorization comprises an authorization to begin monitoring the process;

wherein generating the second regression model comprises generating the second regression model in response to receiving the authorization to begin monitoring the process after receiving the authorization to train the process model.

14. A method according to claim 1, wherein generating the first regression model comprises generating the first regression model if at least a minimum number of first data sets are received.

15. A method according to claim 1, wherein collecting first data sets comprises at least one of collecting samples of the process variables, collecting processed samples of the process variables, and collecting statistical data generated from samples of the process variables.

16. A method according to claim 1, wherein the first regression model is configured to generate a prediction of first data generated from a first process variable as a function of second data generated from a second process variable.

17. A method according to claim 16, wherein the first regression model is configured to generate a prediction of a sample of a first process variable as a function of a sample of a second process variable.

18. A method according to claim 16, wherein the first regression model is configured to generate a prediction of a mean of the first process variable as a function of a mean of a second process variable.

19. A method according to claim 1, wherein the first regression model is configured to generate a prediction of first data generated from a first plurality of process variables as a function of second data generated from a second plurality of process variables.

20. A method according to claim 1, wherein the first regression model comprises a first linear regression model, and wherein the second regression model comprises a second linear regression model.

21. A method according to claim 1, further comprising:

determining if the corresponding signal generated from the process variable significantly deviates from the output of the process model.

22. A method according to claim 21, further comprising generating an indicator of significant deviation if the corresponding signal generated from the process variable significantly deviates from the output of the process model.

23. A method according to claim 22, wherein the indicator comprises an alert.

24. A method according to claim 21, wherein modeling the process comprises:

generating an interpolated model for a third range of the independent process variable between the first range and the second range;

revising the process model of the process to include simultaneously the first regression model for the first range of the independent process variable, the second regression model for the second range of the independent process variable and the interpolated model for the third range of the independent process variable between the first range and the second range;

receiving a data set including an X value generated from the independent process variable and a Y value generated from the dependent process variable;

if the X value is in the first range, generating a predicted Y value using the first regression model;

if the X value is in the second range, generating the predicted Y value using the second regression model;

if the X value is between the first range and the second range, generating the predicted Y value using the interpolated model;

wherein determining if the corresponding signal generated from the process variable significantly deviates from the output of the process model comprises determining if the Y value significantly differs from the predicted Y value.

25. A tangible, non-transitory medium storing machine readable instructions, the machine readable instructions capable of causing one or more processors to:

collect first data sets for a process operating in a process plant while the process is in a first normal operating region, the first data sets generated from process variables of the process in the first normal operating region, the process producing a product by processing physical materials using one or more field devices communicatively coupled to one or more process controllers;

generate a first regression model of the process in the first operating region using the first data sets, the first regression model a function of an independent process variable;

determine a first range for the independent process variable in which the first regression model is valid;

generate a process model to include the first regression model, the process model a function of the independent variable;

collect second data sets for the process while the process is in a second normal operating region, the second data sets generated from process variables of the process in the second normal operating region;

generate a second regression model of the process in the second operating region using the second data sets, the second regression model a function of the independent variable;

determine a second range for the independent variable in which the second regression model is valid, the second range different than the first range; and

revise the process model to include simultaneously the first regression model for the first range of the independent process variable and the second regression model for the second range of the independent process variable.

26. A system for detecting an abnormal operation in a process plant, comprising:

a configurable process model of the process in the process plant, the configurable process model a function of an independent process variable and including a first regression model in a first range of the independent variable corresponding to a first normal operating region of the process, the first regression model a function of the independent process variable, the configurable process model capable of being subsequently configured to include simultaneously with the first regression model a second regression model in a second range of the independent variable corresponding to a second normal operating region different than the first normal operating region, the second regression model a function of the independent process variable;

wherein, after the configurable process model is configured to include the second regression model simultaneously with the first regression model, the configurable process model is capable of modeling the process in the first normal operating region using the first regression model, and modeling the process in the second normal operating region using the second regression model; and

the system further comprising a deviation detector coupled to the configurable process model, the deviation detector configured to determine if the process significantly deviates from an output of the process model by comparing one or more measured values of a dependent process variable at a corresponding value of the independent process variable to one or more modeled values of the dependent process variable at the corresponding value of the independent process variable.

27. A system according to claim 26, wherein the configurable process model is capable of being configured, after being configured to include the second regression model simultaneously with the first regression model, to include simultaneously with the first and second regression models at least a third regression model in at least a third range of the independent process variable corresponding to at least a third normal operating region different than the first normal operating region and the second normal operating region;

wherein, after the configurable process model is configured to include the third regression model simultaneously with the first and second regression models, the configurable process model is capable of modeling the process in the first normal operating region using the first regression model, modeling the process in the second normal operating region using the second regression model, and modeling the process in the third normal operating region using the third regression model.

28. A system according to claim 26, wherein the configurable process model is capable of generating an interpolated model for a third range of the independent process variable between the first range of the independent process variable and the second range of the independent process variable.

29. A system according to claim 28, wherein the configurable process model is capable of generating a prediction of a Y value as a function of an X value, wherein the Y value is generated from a dependent process variable and the X value is generated from an independent process variable;

wherein, after the configurable process model is configured to include the second regression model simultaneously with the first regression model and after generating the interpolated model, the configurable process model is capable of;

generating the prediction of the Y value using the first regression model if the X value is in the first range of the independent process variable,

generating the prediction of the Y value using the second regression model if the X value is in the second range of the independent process variable,

generating the prediction of the Y value using the interpolated model if the X value is in the third range of the independent process variable between the first range of the independent process variable and the second range of the independent process variable.

30. A system according to claim 29, wherein the deviation detector is configured to determine if the Y value significantly differs from the predicted Y value.

31. A system according to claim 30, wherein the deviation detector is configured to determine if the Y value significantly differs from the predicted Y value by comparing a difference between the Y value and the predicted Y value to a threshold.

32. A system according to claim 31, wherein the threshold varies as a function of the X value.

33. A system according to claim 30, further comprising:

a first statistical generator coupled to the configurable process model, the first statistical generator to generate the X values as a mean of the independent process variable; and

a second statistical generator coupled to the configurable process model, the second statistical generator to generate the Y values as a mean of the dependent process variable.

34. A system according to claim 33, wherein the second statistical generator additionally generates a standard deviation values of the dependent process variable;

the system further comprising an additional model coupled to the first statistical generator, the second statistical generator, and the deviation detector, the additional model to generate a prediction of the standard deviation of the dependent process variable as a function of the mean of the independent process variable;

wherein the deviation detector is configured to determine if the Y value significantly differs from the predicted Y value by comparing a difference between the Y value and the predicted Y value to a threshold, wherein the threshold is a function of the prediction of the standard deviation.