Kalman filter state estimation for a manufacturing system

• US 6,757,579 B1
• Filed: 07/31/2002
• Issued: 06/29/2004
• Est. Priority Date: 09/13/2001
• Status: Expired due to Term

• Current Assignee: Advanced Micro Devices, Inc.
• Original Assignee: Advanced Micro Devices, Inc.
• Inventors: Pasadyn, Alexander J.
• Primary Examiner(s): RODRIGUEZ, PAUL L
• Application Number:
  US10/209,758
• Time in Patent Office: 699 Days
• Field of Search: 700/28-30, 700/38, 700/44, 700/45, 700/47, 700/51, 700/52, 700/54, 700/55, 700/95, 700/96, 700/108, 702/188
• US Class Current: 700/108
• CPC Class Codes: G05B 13/04 involving the use of models or simulators
• 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 method for monitoring a manufacturing system includes defining a plurality of observed states associated with the manufacturing system. State estimates are generated for the observed states. An uncertainty value is generated for each of the state estimates. Measurement data associated with an entity in the manufacturing system is received. The state estimates are updated based on the measurement data and the uncertainty values associated with the state estimates. A system for monitoring a manufacturing system includes a controller configured to define a plurality of observed states associated with the manufacturing system, generate state estimates for the observed states, generate an uncertainty value for each of the state estimates, receive measurement data associated with an entity in the manufacturing system, and update the state estimates based on the measurement data and the uncertainty values associated with the state estimates.

Claims

1. A method for monitoring a manufacturing system, comprising:

defining a plurality of observed states associated with the manufacturing system;

generating state estimates for the observed states;

generating an uncertainty value for each of the state estimates;

receiving measurement data including associated with an entity in the manufacturing system, including a first measurement associated with a first processing event;

generating a first output mapping based on the measurement data;

updating the state estimates based on the first output mapping, the measurement data and the uncertainty values associated with the state estimates;

receiving a second measurement associated with a second processing event occurring after the first processing event;

generating a second output mapping based on the second measurement; and

updating the state estimates based on the second output mapping, the second measurement, and the associated uncertainty values.

2. The method of claim 1, further comprising generating a state estimate matrix including the state estimates.

3. The method of claim 2, further comprising generating a covariance matrix associated with the state estimate matrix, the covariance matrix including the uncertainty values.

4. The method of claim of claim 3, wherein generating the covariance matrix further comprises generating the covariance matrix including the uncertainty values as diagonal terms.

5. The method of claim of claim 4, wherein generating the covariance matrix further comprises generating the covariance matrix including off-diagonal terms defining interdependency relationships amongst the observed states.

6. The method of claim 1, further comprising:

performing a processing step in the manufacturing system; and

updating each of state estimates based on the processing step and the associated uncertainty values.

7. The method of claim 1, wherein the manufacturing system includes a plurality of tools, and defining the plurality of observed states further comprises defining at least one tool state.

8. The method of claim 1, wherein the manufacturing system is configured to process a plurality of products and defining the plurality of observed states further comprises defining at least one product state.

9. The method of claim 1, wherein the manufacturing system includes a plurality of tools configured to process a plurality of products and defining the plurality of observed states further comprises defining at least one tool state and at least one product state.

10. The method of claim 1, wherein the manufacturing system includes a plurality of tools configured to perform a plurality of processes associated with a plurality of products and defining the plurality of observed states further comprises defining an observed state for each product/process combination.

11. The method of claim 1, further comprising updating the state estimates using a Kalman filter.

12. A method for monitoring a manufacturing system, comprising:

defining a plurality of observed states associated with the manufacturing system;

generating state estimates for the observed states;

generating a state estimate matrix including the state estimates;

generating an uncertainty value for each of the state estimates;

generating a covariance matrix associated with the state estimate matrix, the covariance matrix including the uncertainty values;

receiving measurement data associated with an entity in the manufacturing system;

updating the state estimates based on the measurement data and the uncertainty values associated with the state estimates;

identifying an additional observed state associated with the manufacturing system;

generating an initial state estimate for the additional observed state; and

merging the initial state estimate with the state estimate matrix.

13. The method of claim 12, further comprising:

generating an initial covariance estimate for the additional observed state; and

merging the initial covariance estimate with the covariance matrix.

14. The method of claim 13, wherein the covariance matrix includes a diagonal term associated with the additional observed state and generating the initial covariance estimate includes setting the initial covariance estimate higher than corresponding diagonal terms for other observed states.

15. The method of claim 13, wherein the covariance matrix includes a plurality of off-diagonal terms associated with the additional observed state and generating the initial covariance estimate includes setting the off-diagonal terms to zero.

16. A method for monitoring a manufacturing system, comprising:

defining a plurality of observed states associated with the manufacturing system;

generating state estimates for the observed states;

generating an uncertainty value for each of the state estimates;

receiving measurement data associated with an entity in the manufacturing system;

updating the state estimates based on the measurement data and the uncertainty values associated with the state estimates;

identifying a set of initial state estimates at a first time;

maintaining a queue of events associated with the manufacturing system occurring after the first time, the events including process events and measurement events;

receiving a first measurement event;

identifying one of the process events associated with the first measurement event;

selecting events occurring after the identified process event; and

iteratively updating the initial state estimates based on the first measurement event and the selected set of events.

17. The method of claim 16, further comprising retiring events from the queue after a predetermined time interval.

18. The method of claim 16, further comprising retiring a selected process event responsive to not receiving a measurement event associated with the selected process event after a predetermined time interval.

19. A system for monitoring a manufacturing system, comprising a controller configured to define a plurality of observed states associated with the manufacturing system, generate state estimates for the observed states, generate a state estimate matrix including the state estimates, generate an uncertainty value for each of the state estimates, generate a covariance matrix associated with the state estimate matrix, the covariance matrix including the uncertainty value receive measurement data associated with an entity in the manufacturing system, update the state estimates based on the measurement data and the uncertainty values associated with the state estimates, identify an additional observed state associated with the manufacturing system, generate an initial state estimate for the additional observed state, and merge the initial state estimate with the state estimate matrix.

20. The system of claim 19, wherein the controller is further configured to generate an initial covariance estimate for the additional observed state and merge the initial covariance estimate with the covariance matrix.

21. The system of claim 20, wherein the covariance matrix includes a diagonal term associated with the additional observed state and the controller is further configured to set the initial covariance estimate higher than corresponding diagonal terms for other observed states.

22. The system of claim 20, wherein the covariance matrix includes a plurality of off-diagonal terms associated with the additional observed state and the controller is further configured to set the off-diagonal terms to zero.

23. A system for monitoring a manufacturing system, comprising a controller configured to define a plurality of observed states associated with the manufacturing system, generate state estimates for the observed states, generate an uncertainty value for each of the state estimates, receive measurement data associated with an entity in the manufacturing system, update the state estimates based on the measurement data and the uncertainty values associated with the state estimates, identify a set of initial state estimates at a first time, maintain a queue of events associated with the manufacturing system occurring after the first time, the events including process events and measurement events, receive a first measurement event, identify one of the process events associated with the first measurement event, select events occurring after the identified process event, and iteratively update the initial state estimates based on the first measurement event and the selected events occurring after the identified process event.

24. The system of claim 23, the controller is further configured to retire events from the queue after a predetermined time interval.

25. The system of claim 23, the controller is further configured to retire a selected process event responsive to not receiving a measurement event associated with the selected process event after a predetermined time interval.

26. A system for monitoring a manufacturing system, comprising a controller configured to define a plurality of observed states associated with the manufacturing system, generate state estimates for the observed states, generate an uncertainty value for each of the state estimates, receive measurement data associated with an entity in the manufacturing system including a first measurement associated with a first processing event, generate a first output mapping based on first measurement, update the state estimates based on the first output mapping, first measurement, and the uncertainty values associated with the state estimates, receive a second measurement associated with a second processing event occurring after the first processing event, generate a second output mapping based on the second measurement, and update the state estimates based on the second output mapping, the second measurement, and the associated uncertainty values.

27. The system of claim 26, wherein the controller is further configured to generate a state estimate matrix including the state estimates.

28. The system of claim 27, wherein the controller is further configured to generate a covariance matrix associated with the state estimate matrix, the covariance matrix including the uncertainty values.

29. The system of claim of claim 28, wherein the controller is further configured to generate the covariance matrix including the uncertainty values as diagonal terms.

30. The system of claim of claim 29, wherein the controller is further configured to generate the covariance matrix including off-diagonal terms defining interdependency relationships amongst the observed states.

31. The system of claim 26, further comprising:

a process tool configured to perform a processing step in the manufacturing system,

wherein the controller is further configured to update each of state estimates based on the processing step and the associated uncertainty values.

32. The system of claim 26, wherein the manufacturing system includes a plurality of tools, and the controller is further configured to define at least one tool state.

33. The system of claim 26, wherein the manufacturing system is configured to process a plurality of products and the controller is further configured to define at least one product state.

34. The system of claim 26, wherein the manufacturing system includes a plurality of tools configured to process a plurality of products and the controller is further configured to define at least one tool state and at least one product state.

35. The system of claim 26, wherein the manufacturing system includes a plurality of tools configured to perform a plurality of processes associated with a plurality of products and the controller is further configured to define an observed state for each product/process combination.

36. The system of claim 26, the controller is further configured to update the state estimates using a Kalman filter.

37. A system for monitoring a manufacturing system, comprising:

means for defining a plurality of observed states associated with the manufacturing system;

means for generating state estimates for the observed states;

means for generating an uncertainty value for each of the state estimates;

means for receiving measurement data associated with an entity in the manufacturing system including a first measurement associated with a first processing event;

means for generating a first output mapping based on the measurement data;

means for updating the state estimates based on the first output mapping, the measurement data, and the uncertainty values associated with the state estimates;

means for receiving a second measurement associated with a second processing event occurring after the first processing event;

means for generating a second output mapping based on the second measurement; and

means for updating the state estimates based on the send output mapping, the second measurement, and the associated uncertainty values.