Methods and systems for neural network modeling of turbine components

• US 8,065,022 B2
• Filed: 01/08/2008
• Issued: 11/22/2011
• Est. Priority Date: 09/06/2005
• Status: Expired due to Fees

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Zhang, Jianbo, Karaca, Erhan, Minto, Karl Dean
• Primary Examiner(s): Decady; Albert
• Assistant Examiner(s): STEVENS, THOMAS H
• Application Number:
  US11/971,022
• Publication Number:
  US 20100100248A1
• Time in Patent Office: 1,414 Days
• Field of Search: 700/29, 706/31, 706/44, 706/904, 706/920, 703/7, 703/18, 703/1, 703/2, 703/6, 603/50, 607/82
• US Class Current: 700/29
• CPC Class Codes: F01D 11/20 Actively adjusting tip-clearance
• CPC Class Codes: F05D 2270/71 synthesized, i.e. parameter computed by a mathematical model
• CPC Class Codes: G05B 13/027 using neural networks only
• CPC Class Codes: G05B 17/02 electric
• CPC Class Codes: Y10S 706/904 Manufacturing or machine, e.g. agricultural machinery, machine tool
• CPC Class Codes: Y10S 706/92 Simulation

Abstract

Embodiments of the invention can include methods and systems for controlling clearances in a turbine. In one embodiment, a method can include applying at least one operating parameter as an input to at least one neural network model, modeling via the neural network model a thermal expansion of at least one turbine component, and taking a control action based at least in part on the modeled thermal expansion of the one or more turbine components. An example system can include a controller operable to determine and apply the operating parameters as inputs to the neural network model, model thermal expansion via the neural network model, and generate a control action based at least in part on the modeled thermal expansion.

Claims

1. A method for controlling clearance in a turbine, the method comprising:

applying at least one operating parameter as an input to at least one neural network model;

modeling via the at least one neural network model thermal expansion of at least one turbine component; and

implementing a control action based at least in part on the modeled thermal expansion of the at least one turbine component.

2. The method in claim 1, further comprising:

determining an outer shell temperature associated with the turbine; and

determining a rotor temperature associated with the turbine;

wherein the at least one operating parameter comprises the outer shell temperature and the rotor temperature.

3. The method in claim 2, wherein determining the outer shell temperature comprises measuring the outer shell temperature.

4. The method in claim 2, wherein determining the rotor temperature comprises modeling the rotor temperature.

5. The method in claim 4, wherein modeling the rotor temperature comprises:

obtaining steam temperature from at least one point in the turbine;

applying the steam temperature to a rotor thermal model; and

modeling via the rotor thermal model a rotor temperature parameter.

6. The method in claim 1, wherein modeling via the at least one neural network model thermal expansion of the at least one turbine component comprises:

modeling thermal expansion of a shell associated with the turbine; and

modeling thermal expansion of a rotor associated with the turbine.

7. The method in claim 6, further comprising determining a differential expansion based at least in part on the difference between the modeled rotor thermal expansion and the modeled shell thermal expansion.

8. The method in claim 1, further comprising applying the modeled thermal expansion of the at least one turbine component as a feedback input to the at least one neural network model.

9. The method in claim 1, wherein implementing the control action comprises at least one of halting operation of the turbine, adjusting operation of the turbine, triggering an alarm, transmitting a notification message, or altering the clearance of the turbine.

10. The method in claim 9, wherein adjusting the turbine comprises increasing the speed of the turbine if the modeled thermal expansion of the at least one turbine component is within a predefined threshold, and decreasing the speed of the turbine if the modeled thermal expansion of the at least one turbine component is outside of the predefined threshold.

11. The method in claim 1, wherein implementing the control action comprises applying the modeled thermal expansion of the at least one turbine component as an input to a predictive model control for controlling a power plant of which the turbine is a part.

12. A system for controlling a turbine, the system comprising a controller operable to:

determine at least one operating parameter;

apply the at least one operating parameter as an input to at least one neural network model;

model via the at least one neural network model thermal expansion of at least one turbine component; and

generate a control action based at least in part on the modeled thermal expansion of the at least one turbine component.

13. The system in claim 12, wherein the controller is further operable to:

determine an outer shell temperature associated with the turbine; and

determine a rotor temperature associated with the turbine;

wherein the at least one operating parameter comprises the outer shell temperature and the rotor temperature.

14. The system in claim 13, further comprising at least one sensor in communication with the turbine and the controller, wherein:

the at least one sensor comprises at least one shell temperature sensor and at least one steam temperature sensor;

the controller is further operable to determine the outer shell temperature by receiving an outer shell temperature from the at least one shell temperature sensor;

the controller is further operable to obtain a steam temperature from at least one point in the turbine; and

the controller is further operable to determine the rotor temperature by modeling the rotor temperature from a received steam temperature.

15. The system in claim 12, wherein the controller is further operable to:

model via the at least one neural network model a shell thermal expansion associated with the turbine;

model via the at least one neural network model a rotor thermal expansion associated with the turbine; and

determine a differential expansion based at least in part on the difference between the modeled rotor thermal expansion and the modeled shell thermal expansion.

16. The system in claim 12, wherein the controller is further configured to apply the modeled thermal expansion of the at least one turbine component as a feedback input to the at least one neural network model.

17. A method for modeling turbine clearance, the method comprising:

sensing a first and a second operating parameter;

modeling at least one shell temperature parameter based at least in part on the first sensed operating parameter;

modeling at least one rotor temperature parameter based at least in part on the second sensed operating parameter;

determining a shell thermal expansion by applying the at least one shell temperature parameter as an input to a shell expansion neural network model;

determining a rotor thermal expansion by applying the at least one rotor temperature parameter as an input to a rotor expansion neural network model; and

determining a differential expansion based at least in part on the difference between the rotor thermal expansion and the shell thermal expansion.

18. The method in claim 17, wherein:

the first operating parameter comprises exhaust steam temperature; and

the second operating parameter comprises main steam temperature.

19. The method in claim 18, wherein the modeling of the at least one shell temperature parameter comprises:

applying the first and the second operating parameters as inputs to the shell thermal model; and

modeling an outer shell inner surface temperature and an exhaust section shell metal temperature via the shell thermal model, wherein the outer shell inner surface temperature and an exhaust section shell metal temperature comprise the at least one shell temperature parameter; and

wherein the outer shell inner surface temperature and the exhaust section shell metal temperature are applied as inputs to the shell expansion neural network model.

20. The method in claim 19, wherein the modeling of the at least one rotor temperature parameter comprises:

applying the first operating parameter as an input to a rotor thermal model; and

modeling a bowl metal temperature via the rotor thermal model, wherein the bowl metal temperature comprises the at least one rotor temperature parameter; and

wherein the bowl metal temperature and the exhaust section shell metal temperature are applied as inputs to the rotor expansion neural network model.