Method and system for detecting abnormal operation of a level regulatory control loop

• US 20080125879A1
• Filed: 07/25/2006
• Published: 05/29/2008
• 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.
• Granted Patent: US 8,145,358 B2
• Time in Patent Office: Days
• US Class Current: 700/30
• CPC Class Codes: G05B 23/0254 based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks

Abstract

A system facilitates detecting an abnormal operation associated with a level regulatory control loop in a process plant. A model for modeling at least a portion of the level regulatory control loop may be utilized with respect to first and second signals associated with regulatory control of a level of material in a tank. The model may include a first regression model in a first range corresponding to a first operating region of the level regulatory control loop. The model may be capable of being subsequently configured to include at least a second regression model in at least a second respective range corresponding to at least a second respective operating region different than the first operating region. The model may generate a prediction of the second signal as a function of first signal. It may be determined whether the second signal significantly deviates from the prediction of the second signal generated by the model. If there is a significant deviation, this may indicate an abnormal operation associated with the level regulatory control loop.

Claims

1. A method for detecting abnormal operation associated with a level regulatory control loop, comprising:

generating a first signal associated with regulatory control of a level of material in a tank;

generating a second signal associated with regulatory control of the level of material in the tank;

using a first model to generate a prediction of the second signal as a function of the first signal, the first model including a first regression model in a first range corresponding to a first operating region of the level regulatory control loop, the first model capable of being subsequently configured to include at least a second regression model in at least a second respective range corresponding to at least a second respective operating region different than the first operating region;

determining whether the second signal significantly deviates from the prediction of the second signal; and

generating a first indicator of an abnormal operation associated with the level regulatory control loop based on whether the second signal significantly deviates from the prediction of the second signal.

2. A method according to claim 1, wherein generating the first signal comprises generating a mean of a first process variable associated with regulatory control of the level of material in the tank.

3. A method according to claim 2, wherein generating the second signal comprises generating a mean of a second process variable associated with regulatory control of the level of material in the tank.

4. A method according to claim 3, wherein the first process variable is a signal indicative of the level of material in the tank;

wherein the second process variable is a control demand signal;

wherein generating the first indicator comprises;

generating the first indicator to indicate that the mean of the control demand signal is too low if the mean of the control demand signal is significantly lower than the prediction generated by the first model; and

generating the first indicator to indicate that the mean of the control demand signal is too high if the mean of the control demand signal is significantly higher than the prediction generated by the first model.

5. A method according to claim 3, wherein the first process variable is a signal indicative of the level of material in the tank;

wherein the second process variable is a difference between a valve position signal and a control demand signal;

wherein generating the first indicator comprises generating the first indicator to indicate a valve problem if the mean of the difference between the valve position signal and the control demand signal significantly deviates from the prediction generated by the first model.

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

generating a third signal associated with regulatory control of the level of material in a tank;

using a second model to generate a prediction of the third signal as a function of the first signal; and

determining whether the third signal significantly deviates from the prediction of the third signal;

wherein generating the first indicator comprises generating the first indicator further based on whether the third signal significantly deviates from the prediction of the third signal.

7. A method according to claim 6, wherein generating the third signal comprises generating a difference between a fourth signal associated with the level of material in the tank and the second signal.

8. A method according to claim 7, wherein generating the first signal comprises generating a mean of a signal indicative of the level of material in the tank;

wherein generating the second signal comprises generating a mean of a control demand signal;

wherein generating the third signal comprises generating a mean of a difference between a valve position signal and the control demand signal.

9. A method according to claim 8, wherein generating the first indicator comprises:

generating the first indicator to indicate a valve problem if the mean of the difference between the valve position signal and the control demand signal significantly deviates from the prediction generated by the second model; and

generating the first indicator to indicate a measurement drift problem if the mean of the difference between the valve position signal and the control demand signal does not significantly deviate from the prediction generated by the second model and if the mean of the control demand significantly deviates from the prediction generated by the first model.

10. A method according to claim 9, further comprising generating a second indicator of an abnormal operation, wherein generating the second indicator comprises:

generating the second indicator to indicate that the mean of the control demand signal is too low if the mean of the control demand signal is significantly lower than the prediction generated by the first model;

generating the second indicator to indicate that the mean of the control demand signal is too high if the mean of the control demand signal is significantly higher than the prediction generated by the first model;

generating the second indicator to indicate that the mean of the control demand signal is not too high or too low if the mean of the control demand signal does not significantly deviate from the prediction generated by the first model;

the method further comprising generating a third indicator of an abnormal operation, wherein generating the third indicator comprises;

generating the third indicator to indicate that the mean of the difference between the valve position signal and the control demand signal is too low if the mean of the difference between the valve position signal and the control demand signal is significantly lower than the prediction generated by the second model;

generating the third indicator to indicate that the mean of the difference between the valve position signal and the control demand signal is too high if the mean of the difference between the valve position signal and the control demand signal is significantly higher than the prediction generated by the second model; and

generating the third indicator to indicate that the mean of the difference between the valve position signal and the control demand signal is not too high or too low if the mean of the difference between the valve position signal and the control demand signal does not significantly deviate from the prediction generated by the second model.

11. A method according to claim 1, wherein generating the first indicator comprises generating an indicator of a liquid leak condition.

12. A method according to claim 1, wherein generating the first indicator comprises generating an indicator of a head loss condition.

13. A tangible medium storing machine readable instructions, the machine readable instructions capable of causing one or more machines to:

generate a first signal associated with regulatory control of a level of material in a tank;

generate a second signal associated with regulatory control of the level of material in the tank;

use a first model to generate a prediction of the second signal as a function of the first signal, the first model including a first regression model in a first range corresponding to a first operating region of the level regulatory control loop, the first model capable of being subsequently configured to include at least a second regression model in at least a second respective range corresponding to at least a second respective operating region different than the first operating region;

determine whether the second signal significantly deviates from the prediction of the second signal; and

generate an indicator of an abnormal operation associated with the level regulatory control loop based on whether the second signal significantly deviates from the prediction of the second signal.

14. A system for detecting an abnormal operation of a level regulatory control loop in a process plant, comprising:

a first model to generate a prediction of a second signal associated with regulatory control of a level of material in a tank as a function of a first signal associated with regulatory control of the level of material in the tank, the first model including a first regression model in a first range corresponding to a first operating region of the level regulatory control loop, the first model capable of being subsequently configured to include at least a second regression model in at least a second respective range corresponding to at least a second respective operating region different than the first operating region;

a first deviation detector coupled to the first model, the first deviation detector to generate an output indicative of whether the second signal significantly deviates from the prediction of the second signal; and

an indicator generator coupled to the first deviation detector, the indicator generator to generate a indicator of an abnormal operation associated with the level regulatory control loop based on whether the second signal significantly deviates from the prediction of the second signal.

15. A system according to claim 14, further comprising:

a first mean generator coupled to the first model to generate the first signal as a mean of a first process variable associated with the level of material in the tank; and

a second mean generator coupled to the first model to generate the second signal as a mean of a second process variable associated with the level of material in the tank.

16. A system according to claim 15, wherein the first process variable is a signal indicative of the level of material in the tank;

wherein the second process variable is a control demand signal;

wherein the first deviation detector is configured to;

generate the first deviation detector output to indicate whether the mean of the control demand signal is significantly lower than the prediction generated by the first model, and

generate the first deviation detector output to indicate whether the mean of the control demand signal is significantly higher than the prediction generated by the first model;

wherein the indicator generator is configured to;

generate the indicator to indicate that the mean of the control demand signal is too low if the mean of the control demand signal is significantly lower than the prediction generated by the first model, and

generate the indicator to indicate that the mean of the control demand signal is too high if the mean of the control demand signal is significantly higher than the prediction generated by the first model.

17. A system according to claim 15, wherein the first process variable is a signal indicative of the level of material in the tank;

wherein the second process variable is a difference between a valve position signal and a control demand signal;

wherein the indicator generator is configured to generate the indicator to indicate a valve problem if the output of the first deviation detector indicates that the mean of the difference between the valve position signal and the control demand signal significantly deviates from the prediction generated by the first model.

18. A system according to claim 14, further comprising:

a second model to generate a prediction of a third signal associated with regulatory control of the level of material in the tank as a function of the first signal associated with regulatory control of the level of material in the tank;

a second deviation detector coupled to the second model, the second deviation detector to generate an output indicative of whether the third signal significantly deviates from the prediction of the third signal; and

wherein the indicator generator is coupled-to the second deviation detector, and wherein the indicator generator is configured to generate the indicator further based on whether the third signal significantly deviates from the prediction of the third signal.

19. A system according to claim 18, further comprising:

a first mean generator coupled to the first model and the second model to generate the first signal as a mean of a signal indicative of the level of material in the tank;

a second mean generator coupled to the first model to generate the second signal as a mean of a control demand signal; and

a third mean generator coupled to the second model to generate the third signal as a mean of a difference a difference between a valve position signal and the control demand signal.

20. A system according to claim 19, wherein the indicator generator is configured to:

generate the indicator to indicate a valve problem if the mean of the difference between the valve position signal and the control demand signal significantly deviates from the prediction generated by the second model, and

generate the indicator to indicate a measurement drift problem if the mean of the difference between the valve position signal and the control demand signal does not significantly deviate from the prediction generated by the second model and if the mean of the control demand signal significantly deviates from the prediction generated by the first model.

21. A system according to claim 20, wherein the first deviation detector is configured to generate the output of the first deviation detector to indicate whether the mean of the control demand signal is significantly lower than the prediction generated by the first model, and to indicate whether the mean of the control demand signal is significantly higher than the prediction generated by the first model;

and

wherein the second deviation detector is configured to generate the output of the second deviation detector to indicate whether the mean of the difference between the valve position signal and the control demand signal is significantly lower than the prediction generated by the second model, and to indicate whether the mean of the difference between the valve position signal and the control-demand signal is significantly higher than the prediction generated by the second model.

22. A system according to claim 21, wherein the first deviation detector is configured to generate the output of the first deviation detector to indicate whether the mean of the control demand signal does not significantly deviate from the prediction generated by the first model;

and

wherein the second deviation detector is configured to generate the output of the second deviation detector to indicate whether the mean of the difference between the valve position signal and the control demand signal does not significantly deviate from the prediction generated by the second model.

23. A system according to claim 14, wherein the indicator generator is configured to generate an indicator of a liquid leak condition.

24. A system according to claim 14, wherein the indicator generator is configured to generate an indicator of a head loss condition.