Methods and systems for detecting deviation of a process variable from expected values

• US 20080027677A1
• Filed: 07/25/2006
• Published: 01/31/2008
• 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., Kant, Ravi
• Granted Patent: US 7,657,399 B2
• Time in Patent Office: Days
• US Class Current: 702/182
• CPC Class Codes: G05B 19/4184 characterised by fault tolerance, reliability of production system
• CPC Class Codes: G05B 2219/32058 Execute program as function of deviation from predicted state, result
• CPC Class Codes: G05B 2219/33314 Failure reason analysis, simple strategy or multiple outcome analysis
• CPC Class Codes: G05B 2219/34465 Safety, control of correct operation, abnormal states
• CPC Class Codes: G05B 2219/34477 Fault prediction, analyzing signal trends
• CPC Class Codes: Y02P 90/02 Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
• CPC Class Codes: Y02P 90/80 Management or planning

Abstract

In methods and systems that may facilitate detecting abnormal operation in a process plant, values of a process variable are analyzed to determine whether they significantly deviate from expected values. If there is a significant deviation, an indicator may be generated. Analyzing the process variable may include, for example, determining whether a number of values of the process variable are increasingly deviating from expected values, or determining whether the process variable is cycling. Analyzing the process variable may also include, for example, processing and analyzing the process variable to determine whether the process is in a normal state, one or more abnormal states, or an indeterminate state.

Claims

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

receiving process variable data associated with the process plant;

generating predicted process variable data using a model;

generating a difference signal based on the received process variable data and the predicted process variable data;

detecting whether the difference signal is increasingly deviating from zero; and

generating an indicator of significant deviation based on whether the difference signal is increasingly deviating from zero.

2. A method according to claim 1, wherein detecting whether the difference signal is increasingly deviating from zero comprises at least one of:

detecting whether a rate of change of the difference signal is greater than a first threshold;

detecting whether the rate of change of the difference signal is less than a second threshold; and

detecting whether a first derivative of the difference signal is greater than a third threshold;

detecting whether the first derivative of the difference signal is less than a fourth threshold;

detecting whether a second derivative of the difference signal is greater than a fifth threshold; and

detecting whether the second derivative of the difference signal is less than a sixth threshold.

3. A method according to claim 1, wherein detecting whether the difference signal is increasingly deviating from zero comprises detecting whether at least a subset of values in a set of values of the difference signal are increasingly deviating from zero.

4. A method according to claim 3, wherein detecting whether the difference signal is increasingly deviating from zero further comprises at least one of:

detecting whether a rate of change of the difference signal is greater than a first threshold;

detecting whether the rate of change of the difference signal is less than a second threshold; and

detecting whether a first derivative of the difference signal is greater than a third threshold;

detecting whether the first derivative of the difference signal is less than a fourth threshold;

detecting whether a second derivative of the difference signal is greater than a fifth threshold; and

detecting whether the second derivative of the difference signal is less than a sixth threshold.

5. A method according to claim 3, wherein detecting whether at least the subset of values in the set of values of the difference signal are increasingly deviating from zero comprises at least one of:

detecting whether the difference signal includes a first number of consecutively increasing values; and

detecting whether the difference signal includes a second number of consecutively decreasing values.

6. A method according to claim 3, wherein detecting whether at least the subset of values in the set of values of the difference signal are increasingly deviating from zero comprises:

detecting whether at least the subset of values in the set of values of the difference signal are increasingly deviating from zero in a positive direction, and

detecting whether at least the subset of values in the set of values of the difference signal are increasingly deviating from zero in a negative direction;

the method further comprising;

generating an indicator of an upward trend based on whether the at least the subset of values in the set of values of the difference signal are increasingly deviating from zero in the positive direction;

generating an indicator of a downward trend based on whether the at least the subset of values in the set of values of the difference signal are increasingly deviating from zero in the negative direction.

7. A method according to claim 3, further comprising comparing the difference signal to a threshold;

wherein generating the indicator of significant deviation is further based on whether the at least the subset of values in the set of values exceeds the threshold.

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

receive process variable data associated with a process plant;

generate predicted process variable data using a model;

generate a difference signal based on the received process variable data and the predicted process variable data;

detect whether the difference signal is increasingly deviating from zero; and

generate an indicator of significant deviation based on whether the difference signal is increasingly deviating from zero.

9. A system for facilitating detection of abnormal operation of a process in a process plant, comprising:

at least one computer readable medium;

at least one processor coupled to the at least one computer readable medium, the processor configured according to executable instructions stored on the at least one computer readable medium to;

receive process variable data associated with the process plant,

receive predicted process variable data generated using a model,

generate a difference signal based on the received process variable data and the predicted process variable data,

detect whether the difference signal is increasingly deviating from zero, and

generate an indicator of significant deviation based on whether the difference signal is increasingly deviating from zero.

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

receiving process variable data, the process variable data generated by a device in a process plant;

generating predicted process variable data using a model;

generating a difference signal based on the received process variable data and the predicted process variable data;

analyzing the difference signal to determine if the difference signal is cycling; and

generating an indicator of significant deviation based on whether the difference signal is cycling.

11. A method according to claim 10, wherein analyzing the difference signal comprises:

generating a frequency domain representation of the difference signal; and

analyzing the frequency domain representation to determine if the difference signal is cycling.

12. A method according to claim 10, wherein analyzing the difference signal comprises:

generating a plurality of bandpass filtered signals from the difference signal; and

analyzing the plurality of bandpass filtered signals to determine if the difference signal is cycling.

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

receive process variable data, the process variable data generated by a device in a process plant;

generate predicted process variable data using a model;

generate a difference signal based on the received process variable data and the predicted process variable data;

analyze the difference signal to determine if the difference signal is cycling; and

generate an indicator of significant deviation based on whether the difference signal is cycling.

14. A system for facilitating detection of abnormal operation of a process in a process plant, comprising:

a difference signal generator to generate a difference signal based on process variable data generated by a device in a process plant and predicted process variable data generated using a model;

a frequency domain analyzer coupled to the difference signal generator, the frequency domain analyzer to generate a plurality of signals indicative of frequency components of the difference signal; and

a peak detector coupled to frequency domain analyzer to generate an indicator of significant deviation based on the plurality of signals indicative of frequency components of the difference signal.

15. A system according to claim 14, wherein the frequency domain analyzer comprises a power spectral density estimate generator.

16. A system according to claim 14, wherein the frequency domain analyzer comprises a plurality of bandpass filters tuned to different frequencies.

17. A system according to claim 14, wherein the frequency domain analyzer comprises an envelope generator.

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

receiving process variable data generated in a process plant;

processing the received process variable data to generate processed data;

determining, based on the processed data, whether the process is in a normal state;

determining, based on the processed data, whether the process is in at least one abnormal state;

if it cannot be determined that the process is in the normal state or the at least one abnormal state, determining that the process is in an indeterminate state; and

generating an indicator of significant deviation if it is determined that the process is in an abnormal state of the at least one abnormal state.

19. A method according to claim 18, wherein processing the received process variable data comprises determining a difference between a value of the received process variable data and a corresponding expected value;

wherein determining whether the process is in the normal state is based at least on the difference; and

wherein determining whether the process is in at least one abnormal state is based at least on the difference.

20. A method according to claim 19, wherein processing the received process variable data further comprises generating a current processed value based on the difference and a previous processed value;

wherein determining whether the process is in the normal state is based at least on the current processed value; and

wherein determining whether the process is in at least one abnormal state is based at least on the current processed value.

21. A method according to claim 20, wherein determining whether the process is in the normal state is based at least on whether the current process value is within a first region;

wherein determining whether the process is in at least one abnormal state is based at least on whether the current process value is within a second region different than the second region; and

wherein determining that the process is in the indeterminate state is based at least on whether the current process value is not in the first region and not in the second region.

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

receive process variable data generated in a process plant;

process the received process variable data to generate processed data;

determine, based on the processed data, whether the process is in a normal state;

determine, based on the processed data, whether the process is in at least one abnormal state;

if it cannot be determined that the process is in the normal state or the at least one abnormal state, determine that the process is in an indeterminate state; and

generate an indicator of significant deviation if it is determined that the process is in an abnormal state of the at least one abnormal state.

23. A system for facilitating detection of abnormal operation of a process in a process plant, comprising:

a difference generator that generates a difference signal based on process variable data generated by a device in a process plant and model process variable data generated from a model;

a signal processor coupled to the difference generator to process the difference signal;

a comparator coupled to the signal processor, the comparator configured to

determine, based on the processed difference signal, whether a process of the process plant is in a normal state,

determine, based on the processed difference signal, whether the process is in at least one abnormal state,

if it cannot be determined that the process is in the normal state or the at least one abnormal state, determine that the process is in an indeterminate state, and

generate an indicator of significant deviation if it is determined that the process is in an abnormal state of the at least one abnormal state.

24. A system according to claim 23, wherein the signal processor is further configured to:

generate a current processed value based on a value of the difference signal and a previous processed value;

determine whether the process is in the normal state based on the current processed value; and

determine whether the process is in at least one abnormal state based on the current processed value.

25. A system according to claim 24, wherein the signal processor is further configured to:

determine whether the process is in the normal state based on whether the current process value is within a first region;

determine whether the process is in at least one abnormal state based on whether the current process value is within a second region different than the second region; and

determine that the process is in the indeterminate state based on whether the current process value is not in the first region and not in the second region.