Method and apparatus for implementing competing control models

• US 7,067,333 B1
• Filed: 06/28/2002
• Issued: 06/27/2006
• Est. Priority Date: 06/28/2002
• Status: Expired due to Fees

• Current Assignee: GlobalFoundries, Inc.
• Original Assignee: Advanced Micro Devices, Inc.
• Inventors: Pasadyn, Alexander J., Purdy, Matthew A.
• Primary Examiner(s): NGUYEN, THANH T
• Application Number:
  US10/186,145
• Time in Patent Office: 1,460 Days
• Field of Search: 438 14- 18, 438 5- 13, 257/48
• US Class Current: 438/5
• CPC Class Codes: G05B 17/02 electric

Abstract

A method for controlling a process includes determining incoming state information associated with the process. A plurality of control models associated with the process is provided. A confidence metric is determined for each of the control models based on the incoming state information. The one of the plurality of control models having the highest associated confidence metric is selected. A control action for determining at least one parameter in an operating recipe used to implement the process is generated using the selected control model. A system includes a process tool and a process controller. The process tool is configured to process a workpiece in accordance with an operating recipe. The process controller is configured to determine incoming state information associated with processing the workpiece, determine a confidence metric for each of a plurality of control models based on the incoming state information, select the one of the plurality of control models having the highest associated confidence metric, and generate a control action for determining at least one parameter in the operating recipe using the selected control model.

Claims

1. A method for controlling a process, comprising:

determining incoming state information associated with the process;

providing a plurality of control models associated with the process;

determining a confidence metric for each of the control models based on the incoming state information;

selecting the one of the plurality of control models having the highest associated confidence metric; and

generating a control action for determining at least one parameter in an operating recipe used to implement the process using the selected control model.

2. The method of claim 1, wherein generating the control action comprises generating a feedback control action.

3. The method of claim 1, wherein generating the control action comprises generating a feedforward control action.

4. The method of claim 1, wherein determining the incoming state information further comprises determining a characteristic of a current workpiece processed using the process.

5. The method of claim 4, further comprising processing the current workpiece in accordance with the generated parameter of the operating recipe.

6. The method of claim 4, further comprising processing a subsequent workpiece in accordance with the generated parameter of the operating recipe.

7. The method of claim 1, wherein determining the incoming state information further comprises determining a characteristic of a process tool used to implement the process.

8. The method of claim 1, wherein determining the confidence metric for each of the control models further comprises:

assigning a range to each of the control models; and

generating the confidence metric for each control model based on the whether the incoming state information is encompassed by the associated range.

9. The method of claim 8, wherein selecting the one of the plurality of control models having the highest associated confidence metric further comprises selecting the one of the plurality of control models having an associated range encompassing the incoming state information.

10. The method of claim 1, wherein determining the confidence metric for each of the control models further comprises:

analyzing historical data collected for the process to identify previous process runs having state information falling within a range about the incoming state information;

determining the performance of each of the control models based on the historical data; and

generating the confidence metric for each of the control modes based on the determined performance.

11. The method of claim 1, wherein determining the confidence metric for each of the control models further comprises:

assigning a center point to each of the control models; and

generating the confidence metric for each control model based on a comparison between the incoming state information and the center point.

12. A system, comprising:

a process tool configured to process a workpiece in accordance with an operating recipe; and

a process controller configured to determine incoming state information associated with processing the workpiece, determine a confidence metric for each of a plurality of control models based on the incoming state information, select the one of the plurality of control models having the highest associated confidence metric, and generate a control action for determining at least one parameter in the operating recipe using the selected control model.

13. The system of claim 12, wherein the control action comprises a feedback control action.

14. The system of claim 12, wherein the control action comprises a feedforward control action.

15. The system of claim 12, wherein the incoming state information further comprises a characteristic of a current workpiece processed by the process tool.

16. The system of claim 15, wherein the process tool is further configured to process the current workpiece in accordance with the generated parameter of the operating recipe.

17. The system of claim 15, wherein the process tool is further configured to process a subsequent workpiece in accordance with the generated parameter of the operating recipe.

18. The system of claim 12, wherein the incoming state information further comprises a characteristic of a process tool used to implement the process.

19. The system of claim 12, wherein the process controller is further configured to assign a range to each of the control models and generate the confidence metric for each control model based on the whether the incoming state information is encompassed by the associated range.

20. The system of claim 19, wherein the process controller is further configured to select the one of the plurality of control models having an associated range encompassing the incoming state information.

21. The system of claim 12, wherein the process controller is further configured to analyze historical data collected for the process to identify previous process runs having state information falling within a range about the incoming state information, determine the performance of each of the control models based on the historical data, and generate the confidence metric for each of the control modes based on the determined performance.

22. The system of claim 12, wherein the process controller is further configured to assign a center point to each of the control models and generate the confidence metric for each control model based on a comparison between the incoming state information and the center point.

23. A system, comprising:

means for determining incoming state information associated with a process;

means for providing a plurality of control models associated with the process;

means for determining a confidence metric for each of the control models based on the incoming state information;

means for selecting the one of the plurality of control models having the highest associated confidence metric; and

means for generating a control action for determining at least one parameter in an operating recipe used to implement the process using the selected control model.