PREDICTIVE CORRELATION MODEL SYSTEM

• US 20070088448A1
• Filed: 10/19/2005
• Published: 04/19/2007
• Est. Priority Date: 10/19/2005
• Status: Abandoned Application

• Current Assignee: Honeywell International Inc.
• Original Assignee: Honeywell International Inc.
• Inventors: Mylaraswamy, Dinkar, Foslien, Wendy
• Application Number:
  US11/163,445
• Publication Number:
  US 20070088448A1
• Time in Patent Office: Days
• US Class Current: 700/44
• 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 having a combination including a predictive controller and a correlation model analysis which may include a principal component analysis (PCA) calculator. The system may encompass a discrete sampler of trajectories of variables from the predictive controller for a correlation model analysis calculation. There may be a compensator for providing initialization and estimation values of other system variables for the correlation model analysis. An uncertainty calculator connected to the predictive controller may output an anomaly impact prediction zone based on upper and lower bounds from the predictive controller.

Claims

1. A predictive system comprising:

a correlation model module; and

a predictive controller connected to the correlation model module; and

wherein the correlation model module comprises;

a sampler connected to the predictive controller;

a principal component analysis calculator connected to the sampler; and

an output interface module connected to the principal component analysis calculator.

2. The system of claim 1, wherein the correlation model module further comprises:

a multiplexer having an input connected to the sampler and an output connected to the principal component analysis calculator;

a compensator having an output connected to the multiplexer; and

a demultiplexer having an input connected to the principal component analysis calculator and an output connected to the compensator.

3. The system of claim 2, wherein:

the compensator receives an initialization signal from a process;

a first input to the predictive controller is a present measurements signal;

a second input to the predictive controller is a predictive controller horizon signal;

a third input to the controller includes setpoints and/or target values;

a first output from the predictive controller to the sampler is a trajectory of manipulated variable moves;

a second output from the predictive controller to the sampler is a trajectory of controlled variables; and

the sampler samples the trajectory of manipulated variable moves and the trajectory of controlled variables into synchronization data vectors as an output to the multiplexer.

4. The system of claim 3, wherein:

an initialization signal and/or a compensating signal goes from the compensator to the multiplexer;

an output of the principal component analysis calculator provides a statistic of a measure of an anomaly to the interface output module and provides estimates to the input of the demultiplexer; and

the interface output module outputs a future trajectory.

5. The system of claim 4, further comprising:

an uncertainty calculator having an input connected to the predictive controller; and

wherein the uncertainty calculator provides an anomaly prediction cone to the output interface module.

6. The system of claim 5, wherein the predictive controller is a horizon based controller.

7. A predictive correlation model system comprising:

a predictive controller; and

a correlation model module connected to the predictive controller.

8. The analyzer of claim 7, wherein the correlation model module comprises:

a sampler connected to the predictive controller;

a multiplexer connected to the sampler; and

a correlation model calculator connected to the multiplexer.

9. The analyzer of claim 8, wherein the correlation module further comprises an output interface mechanism connected to the correlation model calculator.

10. The analyzer of claim 9, wherein the correlation model module further comprises:

a demultiplexer connected to the correlation model calculator; and

a compensator connected to the demultiplexer and the multiplexer.

11. The analyzer of claim 10, further comprising an uncertainty calculator connected to the predictive controller and to the output interface mechanism.

12. A predictor system comprising:

a predictive controller; and

a principal component analyzer connected to the predictive controller.

13. The system of claim 12, wherein the principal component analyzer comprises:

a sampler connected to the predictive controller;

a principal component analysis calculator connected to the sampler; and

an output interface mechanism connected to the principal component analysis calculator.

14. The system of claim 13, wherein:

the predictive controller outputs trajectory information to the sampler; and

the sampler places the trajectory information into data vectors to be sent to the principal component analysis calculator.

15. The system of claim 14, wherein the principal component analysis calculator transforms the data vectors into information about an anomaly.

16. The system of claim 15, wherein the information about an anomaly includes information about a prediction of an impact of the anomaly.

17. The system of claim 16, further comprising a compensator connected to the principal component analysis calculator.

18. The system of claim 17, wherein the compensator may provide information to the principal component analysis calculator not provided by the predictive controller.

19. The system of claim 18, further comprising uncertainty calculator connected to the predictive controller and to the output interface mechanism.

20. The system of claim 19, wherein the uncertainty calculator provides a bound on the prediction of the impact of the anomaly to the output interface mechanism.

21. The system of claim 20, wherein:

a first input signal to the predictive controller includes present measurements of a process;

a second signal to the predictive controller includes a predictive horizon;

a third signal to the predictive controller includes setpoints and/or target values; and

an initialization signal to the compensator includes additional measurements of a process.