AUTOMATIC SIGNAL PROCESSING-BASED LEARNING IN A PROCESS PLANT

• US 20160098647A1
• Filed: 10/06/2014
• Published: 04/07/2016
• Est. Priority Date: 10/06/2014
• Status: Active Grant

• Current Assignee: Fisher-Rosemount Systems Incorporated (Emerson Electric Company)
• Original Assignee: Fisher-Rosemount Systems Incorporated (Emerson Electric Company)
• Inventors: Nixon, Mark J., Zornio, Peter, Wojsznis, Wilhelm K., Lucas, J. Michael, Muston, Paul R., Rotvold, Eric D., Blevins, Terrence L.
• Granted Patent: US 10,282,676 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G05B 13/0265 the criterion being a learning criterion
• CPC Class Codes: G05B 13/028 using expert systems only
• CPC Class Codes: G06N 20/00 Machine learning

Abstract

Techniques for automatically or autonomously performing signal processing-based learning in a process plant are disclosed. Generally, said techniques automatically or autonomously perform signal processing on a real-time signal that is generated based on the process plant controlling a process. Typically, the signal corresponds to a parameter value that varies over time, and the signal is processed as it is generated in real-time during on-line plant operations. Results of the signal processing may indicate characteristics of the signal, and one or more analytics functions may determine the sources of the characteristics, which may include a process element or device, a piece of equipment, and/or an asset of the process plant that is upstream, within the process, of the source of the signal. An autonomous signal processor may be integrated with or included in a process control device and/or a big data node of the process plant.

Claims

1. A system for providing big data based learning in a process plant controlling a process, the system comprising:

a signal processing module including;

an input to receive a signal generated by a signal source associated with the process plant as the signal source generates the signal in real-time, the signal indicative of a parameter value that varies over time based on the process plant controlling the process by using at least one field device that performs a physical function; and

an output to provide, to a recipient application, an indication of at least one characteristic of the signal,

the signal processing module configured to perform signal processing on the signal to determine the at least one characteristic of the signal.

2. The system of claim 1, wherein the signal source is at least one of:

a process control device controlling at least a portion of the process, a big data provider node included in a process control big data network of the process plant, a piece of equipment within the process plant, or an asset of the process plant.

3. The system of claim 1, wherein the signal processing module and the signal source are included in an integral device.

4. The system of claim 3, wherein the integral device is a process control device, and the process control device is one of a controller, a field device performing a physical function to control the at least the portion of the process, or an input/output (I/O) device communicatively connecting the controller and the field device.

5. The system of claim 3, wherein the integral device is a big data node included in a process control big data network of the process plant, and wherein the big data node is one of:

a local big data node, a regional big data node, a centralized big data node, or another big data node.

6. The system of claim 1, wherein the signal processing module is included in a first device coupled to a second device in which the signal source is included.

7. The system of claim 6, wherein the second device is a process control device controlling at least a portion of the process, the process control device being one of a controller, a field device performing a physical function to control the at least the portion of the process, or an input/output (I/O) device communicatively coupling the controller and the field device.

8. The system of claim 1,

further comprising an analytics module, the analytics module including:

an input to receive the indication of the at least one characteristic of the signal from the signal processing module; and

an output to provide an indication of a source of the at least one characteristic of the signal to a recipient application, the recipient application being a user interface application or another application; and

wherein;

the analytics module is configured to determine that at least one member of a set of upstream elements is the source of the at least one characteristic of the signal, the set of upstream elements being elements that are associated with the process plant and that are upstream, within the process, of the signal, and the determination that the at least one member is the source of the at least one characteristic of the signal is based on respective strengths of impact of members of the set of upstream elements on the at least one characteristic of the signal; and

the source of the at least one characteristic of the signal includes at least one of a process control device, a process variable, a piece of equipment, or an asset of the process plant that is upstream of the signal source.

9. The system of claim 8, wherein:

the recipient application is a first recipient application;

the system further comprises a presentation module, the presentation module including;

an input to receive learned output from the analytics module; and

an output to provide the learned output to one or more recipient applications, the one or more recipient applications including the first recipient application; and

the learned output includes at least one of the indication of the source of the at least one characteristic of the signal, or other information learned by the analytics module.

10. The system of claim 9, wherein the presentation module is configured to determine the one or more recipient applications.

11. The system of claim 8, wherein the analytics module is a first analytics module, and the recipient application is the another application included in a second analytics module.

12. The system of claim 8, wherein the analytics module is a first analytics module, and the first analytics module determines the source of the at least one characteristic of the signal based on (i) the respective strengths of impact of the members of the set of upstream elements on the at least one characteristic of the signal, and (ii) information provided by a second analytics module.

13. The system of claim 12, wherein the second analytics module comprises a process element alignment module.

14. The system of claim 1, wherein the signal processing module and the analytics module are both included in a single big data node that is communicatively connected to a network of the process plant.

15. The system of claim 1, wherein the signal processing module is included in a first node corresponding to the process plant, the analytics module is included in a second node corresponding to the process plant, and the second node is communicatively connected to the first node.

16. The system of claim 15, wherein the first node further includes the signal source, and the second node is a big data node included in a process control big data network of the process plant.

17. The system of claim 1, wherein the parameter value is indicative of a value of one of an output signal of a process control device, a process variable, a measurement, a balance of energy, a balance of mass, or an output of another analytics module.

18. The system of claim 1, wherein the variations over time of the parameter value comprise one or more oscillations.

19. The system of claim 1, wherein the signal processing performed on the signal includes at least one of:

filtering, magnitude or amplitude analysis, power analysis, intensity analysis, phase analysis, frequency analysis, spectrum or spectral analysis, correlation analysis, convolution, smoothing, Hilbert transformation, level detection, linear signal processing, non-linear signal processing, or another signal processing technique.

20. The system of claim 1, wherein the at least one characteristic of the signal comprises at least one of:

a dominant frequency, an n-th order frequency wherein n is an integer greater than one, a harmonic, a subharmonic, or a bifurcation of the signal.

21. The system of claim 1, wherein the signal processing module operates in real-time without supervision.

22. A system for automatically performing big data based learning in a process plant controlling a process, the system comprising:

a signal processing module configured to;

receive a signal generated by a signal source as the signal source generates the signal in real-time, the signal indicative of a value of a parameter that varies over time based on the process plant controlling the process by using at least one field device performing a respective physical function; and

perform signal processing on the signal to determine one or more characteristics of the signal; and

an analytics module configured to;

receive an indication of the one or more characteristics determined by the signal processing module;

determine one or more sources of the one or more characteristics of the signal, the one or more sources being one or more members of a set of elements that are upstream, within the process, of the signal, and each of the one or more members of the set of upstream elements has a respective impact on the one or more characteristics of the signal that is greater than a respective impact of at least one other member of the set of upstream elements that is not included in the one or more members; and

cause an indication of the one or more sources of the one or more characteristics of the signal to be provided to a recipient application, the recipient application being a user interface application or another application.

23. The system of claim 22, wherein:

the recipient application is a first recipient application;

the system further comprises a presentation module configured to receive an output of the analytics module and cause the output to be delivered to a determined set of recipient applications;

the first recipient application is included in the determined set of recipient applications; and

the indication of the one or more sources of the one or more characteristics of the signal is included in the output of the analytics module.

24. The system of claim 23, wherein the output of the analytics module further includes additional learned knowledge determined by the analytic module over various stages of learning.

25. The system of claim 22, wherein at least one of an indication of the signal, an indication of the signal source, or an indication of the one or more characteristics of the signal is provided to the recipient application.

26. The system of claim 22, wherein the signal processing module is included in a first node, and the analytics module is included in a second node communicatively coupled to the first node.

27. The system of claim 26, wherein the signal source is included in the first node.

28. The system of claim 26, wherein:

the first node is at least one of;

a process control device operating to control at least a portion of the process, the process control device being one of a field device performing a physical function to control at least a portion of the process, a controller, or an input/output (I/O) device communicatively connecting the field device and the controller;

or

a first big data provider node included in a process control big data network of the process plant; and

the second node is a second big data node included in the process control big data network.

29. The system of claim 22, wherein the signal processing module and the analytics module are included in an integral device or node.

30. The system of claim 29, wherein the integral device or node is at least one of:

a process control device operating to control at least a portion of the process, the process control device being one of a field device performing a physical function to control at least a portion of the process, a controller, or an input/output (I/O) device communicatively connecting the field device and the controller;

or

a big data node included in a process control big data network of the process plant.

31. The system of claim 22, wherein the value of the parameter corresponds to one of a value of an output signal of a process control device, a value of a process variable, a measurement, a value indicative of a balance of energy, a value indicative of a balance of mass, or a value of an output generated by a big data analysis function.

32. The system of claim 22, wherein the signal processing performed on the signal comprises causing at least one of:

filtering, magnitude or amplitude analysis, power analysis, intensity analysis, phase analysis, frequency analysis, spectrum or spectral analysis, correlation analysis, convolution, smoothing, Hilbert transformation, level detection, linear signal processing, non-linear signal processing, or another signal processing technique.

33. The system of claim 22, wherein the one or more characteristics of the signal comprises at least one of:

a dominant frequency, an n-th order frequency wherein n is an integer greater than one, a harmonic, a subharmonic, or a bifurcation of the signal.

34. The system of claim 22, wherein the one or more sources of the one or more characteristics of the signal include at least one of a process control device, a process variable, a piece of equipment, or an asset corresponding to the process plant and upstream of the signal source.

35. The system of claim 22, wherein the analytics module is a first analytics module, the system further includes a second analytics module, and the recipient application is included in the second analytics module.

36. The system of claim 22, wherein the system operates autonomously in real-time without any user input.

37. A method for providing big data based-learning in a process plant controlling a process, the method comprising:

obtaining, at a signal processing-based learning system, indications of variations over time of a value of a parameter, the variations over time of the parameter value being a signal generated by a signal source, the variations over time of the parameter value generated based on controlling the process in the process plant, and the signal being obtained as the signal source generates the signal in real-time;

causing, by the signal processing-based learning system, one or more signal processing functions to be applied to the signal;

determining, by the signal processing-based learning system, at least one characteristic of the signal based on the application of the one or more signal processing functions;

determining, by the signal processing-based learning system, at least a subset of a set of elements that are upstream, within the process, of the signal to be one or more sources of the at least one characteristic of the signal, the determination of the at least the subset based on respective strengths of impact of the set of upstream elements on the at least one characteristic of the signal; and

causing, by the signal processing-based learning system, an indication of the one or more sources of the at least one characteristic of the signal to be provided to a recipient application, the recipient application being a user interface application or another application.

38. The method of claim 37, wherein obtaining the indications of the variations over time of the parameter value comprises obtaining indications of variations over time of a value of one of:

an output signal of a process control device, a process variable, a measurement, a parameter indicative of a balance of energy, a parameter indicative of a balance of mass, or an output of an analytics function performed on big data generated by the process plant.

39. The method of claim 37, wherein causing signal processing to be performed on the signal comprises causing at least one of:

filtering, magnitude or amplitude analysis, power analysis, intensity analysis, phase analysis, frequency analysis, spectrum or spectral analysis, correlation analysis, convolution, smoothing, a Hilbert transformation, level detection, linear signal processing, non-linear signal processing, or another signal processing technique to be performed on the signal.

40. The method of claim 37, wherein causing the signal processing to be performed on the signal to determine the at least one characteristic of the signal comprises causing the signal processing to be performed on the signal to determine at least one of:

a dominant frequency, an n-th order frequency wherein n is an integer greater than one, a harmonic, a subharmonic, or a bifurcation.

41. The method of claim 37, wherein determining the at least one characteristic of the signal based on the signal processing comprises determining the at least one characteristic of the signal based on the signal processing and based on at least one of:

measurements obtained while at least a portion of the process is on-line, measurements obtained while the at least the at least the portion of the process is off-line, or data obtained from a model of the at least the portion of the process.

42. The method of claim 37, wherein determining the at least the subset of the set of upstream elements to be the one or more sources of the at least one characteristic of the signal comprises determining at least one of a process control device, a piece of equipment, a process variable, or an asset to be the one or more sources of the at least one characteristic of the signal.

43. The method of claim 37, wherein at least one of:

the method is performed over a duration of time of one second or less, or

the method is an unsupervised method.

44. The method of claim 37, wherein at least one of:

obtaining the indications of the variations over time of the parameter value comprises detecting the variations over time of the parameter value;

or

obtaining the variations over time of the signal comprises obtaining an indication of one or more oscillations included in the signal.

45. The method of claim 37, wherein:

the signal processing-based learning system includes a signal processing module coupled to an analytics module;

obtaining the signal at the signal processing-based learning system comprises obtaining the signal at the signal processing module of the signal processing-based learning system;

causing, by the signal processing-based learning system, the one or more signal processing functions to be applied to the signal comprises applying, by the signal processing module of the signal processing-based learning system, the one or more signal processing functions to the signal; and

determining, by the signal processing-based learning system, the one or more sources of the at least one characteristic of the signal comprises determining, by the analytics module of the signal processing-based learning system, the one or more sources of the at least one characteristic of the signal.

46. The method of claim 45, wherein the analytics module is a first analytics module, and at least one of:

obtaining the signal comprises obtaining a signal from a second analytics module;

or

causing the indication of the one or more sources of the at least one characteristic of the signal to be provided to the recipient application comprises causing the indication of the one or more sources of the at least one characteristic of the signal to be provided to the second analytics module or to a third analytics module.