Systems and methods for parameter dependent riccati equation approaches to adaptive control

• US 9,058,028 B2
• Filed: 04/30/2012
• Issued: 06/16/2015
• Est. Priority Date: 04/29/2011
• Status: Active Grant

• Current Assignee: Georgia Tech Research Corporation (University System of Georgia)
• Original Assignee: Georgia Tech Research Corporation (University System of Georgia)
• Inventors: Kim, Kilsoo, Yucelen, Tansel, Calise, Anthony J.
• Primary Examiner(s): BARNES-BULLOCK, CRYSTAL JOY
• Application Number:
  US13/460,663
• Publication Number:
  US 20120277888A1
• Time in Patent Office: 1,142 Days
• Field of Search: 700 28- 30, 700/32, 700/33, 700/37, 700/47, 700/48, 700/34, 703/2, 703/6, 706 14- 16, 706/22, 706/23, 706/21, 706/26, 706/38, 706/39, 701 1- 3
• US Class Current: 1/1
• CPC Class Codes: G05B 13/0205 not using a model or a simulator of the controlled system

Abstract

Systems and methods for adaptive control are disclosed. The systems and methods can control uncertain dynamic systems. The control system can comprise a controller that employs a parameter dependent Riccati equation. The controller can produce a response that causes the state of the system to remain bounded. The control system can control both minimum phase and non-minimum phase systems. The control system can augment an existing, non-adaptive control design without modifying the gains employed in that design. The control system can also avoid the use of high gains in both the observer design and the adaptive control law.

Claims

1. An adaptive control system for controlling a plant, the adaptive control system comprising:

a state observer configured to receive a sensed quantity from the plant, the state observer further configured to output a state estimate based at least in part on the sensed quantity; and

a controller configured to receive the state estimate and employ the state estimate in an adaptive control law to calculate an adaptive control, the adaptive control law based at least in part on a weight estimate, the controller further configured to produce a control command that is an input to the plant;

wherein at least a component of the control command is the adaptive control; and

wherein the weight estimate is based at least in part on a parameter of a parameter dependent Riccati equation, the value of the parameter being selectable, and the parameter dependent Riccati equation having a positive definite solution for the selected value of the parameter.

2. The adaptive control system of claim 1, wherein there is a maximum value of the parameter of the parameter dependent Riccati equation for which the parameter dependent Riccati equation has a positive definite solution.

3. The adaptive control system of claim 1, wherein the state of the plant remains bounded when the value of the parameter of the parameter dependent Riccati equation is greater than zero.

4. The adaptive control system of claim 3, wherein the plant comprises a non-minimum phase system.

5. The adaptive control system of claim 1, wherein the parameter dependent Riccati equation is based at least in part on a state observer gain matrix.

6. The adaptive control system of claim 1, wherein the adaptive control system does not comprise a reference model.

7. The adaptive control system of claim 1, wherein the adaptive control law is based at least in part on a vector of basis functions.

8. The adaptive control system of claim 1,

wherein the control command further comprises a nominal control from a control system of the plant, the nominal control comprising a gain;

wherein the adaptive control augments the nominal control; and

wherein the gain of the nominal control is not modified to accommodate the adaptive control.

9. A method for adaptive control of an uncertain system, the method comprising:

providing a first control command as an input to the uncertain system, the first control command based at least in part on a first adaptive control;

receiving a signal by a state observer, the signal based at least in part on a sensed quantity from the uncertain system;

outputting a state estimate from the state observer;

employing the state estimate in an adaptive control law to calculate a second adaptive control, the adaptive control law based at least in part on a weight estimate; and

providing a second control command as an input to the uncertain system, the second control command based at least in part on the second adaptive control;

wherein the weight estimate is based at least in part on a parameter of a parameter dependent Riccati equation, the value of the parameter being selectable, and the parameter dependent Riccati equation having a positive definite solution for the selected value of the parameter.

10. The method for adaptive control of claim 9 further comprising bounding the state of the system.

11. The method for adaptive control of claim 9, wherein the uncertain system comprises a non-minimum phase system.

12. The method for adaptive control of claim 9, wherein a reference model is not used to define an ideal response of the uncertain system.

13. The method for adaptive control of claim 9, wherein the weight estimate is a function of continuous time.

14. The method for adaptive control of claim 9, wherein the weight estimate is a function of discrete time.

15. The method for adaptive control of claim 9,

wherein the second control command comprises the second adaptive control added to a nominal control,

wherein the nominal control comprises a gain;

and

wherein the gain of the nominal control is not modified to accommodate the adaptive control.

16. An adaptive control system for controlling an uncertain system, the adaptive control system comprising:

a reference model configured to output a state estimate; and

a controller configured to receive the state estimate and employ the state estimate in an adaptive control law to calculate an adaptive control, the adaptive control law based at least in part on a weight estimate, the controller further configured to produce a control command that is an input to the uncertain system;

wherein at least a component of the control command is the adaptive control; and

wherein the weight estimate is based at least in part on a parameter of a parameter dependent Riccati equation, the parameter being selected from a set of values that generate a positive definite solution for the parameter dependent Riccati equation.

17. The adaptive control system of claim 16, wherein the state of the uncertain system remains bounded when the value of the parameter of the parameter dependent Riccati equation is greater than zero.