Apparatuses, systems, and methods utilizing adaptive control

• US 8,095,227 B2
• Filed: 06/13/2005
• Issued: 01/10/2012
• Est. Priority Date: 06/13/2005
• Status: Active Grant

• Current Assignee: Carnegie Mellon University
• Original Assignee: Carnegie Mellon University
• Inventors: Ydstie, Birger Erik
• Primary Examiner(s): Kasenge, Charles
• Application Number:
  US11/922,180
• Publication Number:
  US 20090132064A1
• Time in Patent Office: 2,402 Days
• Field of Search: 700/28, 700/29, 700/30, 700/31, 700/32, 700/33, 700/42, 700/44, 700/45
• US Class Current: 700/45
• CPC Class Codes: G05B 13/048 using a predictor

Abstract

A method for controlling a process, wherein the process has an input and an output, and including controlling the input to the process; predicting the output of the process based on a first set of operating parameters; predicting the output of the process based on a second set of operating parameters; updating the first set of operating parameters when a result from predicting the output from the first set of parameters differs from the output of the process by more than a predetermined amount, wherein the predetermined amount is a function the output predicted by the second set of parameters. The invention also includes apparatuses and systems.

Claims

1. A system for controlling a process, wherein the process has an input and an output, comprising:

a computer; and

software, executing on said computer, said software performing the steps of;

controlling the input to the process;

predicting the output of the process based on an adaptive control model having a first set of operating parameters, including the controlled input to the process and one or more process outputs;

predicting the output of the process based on an adaptive audit model having a second set of operating parameters, including the controlled input to the process and one or more process outputs; and

allowing the adaptive control model access to the one or more process outputs only when a result from predicting the output based on the adaptive control model differs from the output of the process by more than a predetermined amount, wherein the predetermined amount is a function of the output predicted by the adaptive audit model.

2. The system of claim 1, wherein:

controlling the input of the process is based on operating parameters which are modified in response to updating the first set of operating parameters.

3. The system of claim 1, further comprising updating the second set of operating parameters in response to the output of the process.

4. The system of claim 1, wherein the process has a state that changes with time, further comprising:

modeling the state of the process based on the input; and

wherein;

predicting the output of the process based on the adaptive control model further includes predicting the output of the process based on results from modeling the state of the process; and

wherein

predicting the output of the process based on the adaptive audit model further includes predicting the output of the process based on results from modeling the state of the process.

5. A system for controlling a process, wherein the process has an input and an output, comprising:

a computer; and

software, executing on said computer, said software performing the steps of;

controlling the input to the process;

predicting the output of the process based on an adaptive control model including a first set of operating parameters;

predicting the output of the process based on an adaptive audit model including a second set of operating parameters;

updating the first set of operating parameters when a result from predicting the output based on the adaptive control model differs from the output of the process by more than a predetermined amount, wherein the predetermined amount is a function of the output predicted by the adaptive audit model; and

updating the second set of operating parameters to reduce an error indicative of a difference between the output of the process and the output predicted based on the adaptive audit model.

6. The system of claim 5, wherein updating the second set of operating parameters includes performing a least squares method.

7. A system for controlling a process, wherein the process has an input and an output, comprising:

a computer; and

software, executing on said computer, said software performing the steps of;

controlling the input to the process;

predicting the output of the process based on an adaptive control model including a first set of operating parameters;

predicting the output of the process based on an adaptive audit model including a second set of operating parameters;

calculating an error between the process output and an output predicted based on the adaptive control model;

calculating an error between the process output and an output predicted based on the adaptive audit model; and

wherein

updating the first set of operating parameters includes updating the first set of operating parameters when the error based on the process output and an output predicted based on the adaptive control model differs by more than a predetermined tolerance from the error between the process output and an output predicted based on the adaptive audit model.

8. The system of claim 7, wherein the predetermined tolerance is greater than zero.

9. The system of claim 7, wherein the predetermined tolerance is constant.

10. A system for controlling a process, wherein the process has an input and an output, comprising:

a controller having an output which affects the input of the process;

an adaptive control model including an adjustable model of the process, wherein the adaptive control model receives a signal indicative of the input to the process, and wherein the adaptive control model produces a signal indicative of a predicted output of the process;

an adaptive audit model including an adjustable model of the process, wherein the adaptive audit model receives a signal indicative of the input to the process, wherein the adaptive audit model receives a signal indicative of the output of the process, and wherein the adaptive audit model produces a signal indicative of a predicted output of the process; and

a switch connecting the output of the process to the adaptive control model, wherein the switch is controlled in response to a model error for the adaptive audit model.

11. The system of claim 10, wherein the adaptive control model includes operating parameters which are modified in response to changes in the switch.

12. The system of claim 11, wherein the switch carries a signal indicative of the output of the process, and wherein the operating parameters of the adaptive control model change in response to the output of the process.

13. The system of claim 11, wherein the controller includes operating parameters, wherein the controller is connected to the adaptive control model, and wherein the operating parameters of the controller are modified in responses to modifications to the operating parameters of the adaptive control model.

14. The system of claim 13, further comprising a control design algorithm connected between the controller and the adjustable model of the adaptive control model.

15. The system of claim 11, wherein the switch carries signal indicative of a model error corresponding to the adaptive control model, and wherein the operating parameters of the adaptive control model change in response to the model error.

16. The system of claim 11, wherein the controller is a PID controller, and wherein the controller includes operating parameters which are modified in response to modifications to the operating parameters of the adaptive control model.

17. The system of claim 10, wherein the process has a state that changes with time, and further comprising a state estimator having an input indicative of the input to the process and having an output.

18. The system of claim 17, wherein the adaptive control model is responsive to the output of the state estimator.

19. The system of claim 17, wherein the adaptive control model includes the state estimator.

20. The system of claim 10, wherein the process has a state that changes with time, further comprising;

a first state estimator having an input indicative of the input to the process, indicative of at least one output from the process, and having an output connected to the adaptive control model; and

a second state estimator having an input indicative of the input to the process, indicative of at least one output from the process, and having an output connected to the adaptive audit model.