Compact aero-thermo model real time linearization based state estimator

Patent metadata

Field	Value
Publication number	US-10539078-B2
Application number	US-201815899800-A
Country	US
Kind code	B2
Filing date	Feb 20, 2018
Priority date	Mar 15, 2013
Publication date	Jan 21, 2020
Grant date	Jan 21, 2020

Abstract

Systems and methods for controlling a fluid based engineering system are disclosed. The systems and methods may include a model processor configured to generate a model output, the model processor including a set state module for setting dynamic states of the model processor, the dynamic states input to an open loop model based on a model operating mode. The model processor may further include an estimate state module configured to determine an estimated state of the model based on at least one of a prior state, current state derivatives, solver state errors, and synthesized parameters, the estimate state module using online linearization and gain calculation to determine estimator gain for determining the estimated state of the model.

First claim

What is claimed is:

1. A control system, comprising:
an actuator configured to position a control surface of a control device;
a control law configured to direct the actuator as a function of a model output; and
a model processor configured to generate the model output, the model processor comprising:
an input object for processing a model input vector and setting a model operating mode;
a set state module for setting dynamic states of the model processor, the dynamic states input to an open loop model based on the model operating mode;
wherein the open loop model generates current state derivatives, solver state errors, and synthesized parameters as a function of the dynamic states and the model input vector, wherein a constraint on the current state derivatives and solver state errors is based on a series of cycle synthesis modules, each member of the series of cycle synthesis modules modeling a component of a cycle of the control device and comprising a series of utilities, the utilities based on mathematical abstractions of physical laws that govern behavior of the component;
an estimate state module configured to determine an estimated state of the model based on at least one of a prior state, the current state derivatives, the solver state errors, and the synthesized parameters, the estimate state module using online linearization and gain calculation to determine estimator gain for determining the estimated state of the model; and
an output object for processing at least the synthesized parameters of the model to determine the model output.

2. The control system of claim 1 , wherein online linearization and gain calculation comprises obtaining partial derivatives based on current states, the model input vector, the model output, and the synthesized parameters.

3. The control system of claim 2 , wherein the partial derivatives are a Jacobian matrix.

4. The control system of claim 2 , wherein online linearization and gain calculation further comprises constructing a linear system for gain design using the partial derivatives.

5. The control system of claim 4 , wherein online linearization and gain calculation further comprises determining the estimator gain based on the linear system for gain design.

6. The control system of claim 1 , wherein the estimator gain is used in minimizing error vectors.

7. The control system of claim 6 , wherein the estimate state module uses numerical integration to determine the estimated state of the model based on the prior state and the current state derivatives.

8. The control system of claim 1 , wherein the control device is a gas turbine engine.

9. The control system of claim 8 , wherein the one or more cycle synthesis modules are based on one or more mathematical abstractions of physical processes associated with components of a thermodynamic cycle of the gas turbine engine.

10. A method for controlling a control device, the method comprising:
generating a model output using a model processor;
processing a model input vector and setting a model operating mode;
setting dynamic states of the model processor, the dynamic states input to an open loop model based on the model operating mode;
generating current state derivatives, solver state errors, and synthesized parameters as a function of the dynamic states and the model input vector, wherein a constraint on the current state derivatives and solver state errors is based on a series of cycle synthesis modules, each member of the series of cycle synthesis modules modeling a component of a cycle of the control device and comprising a series of utilities, the utilities based on mathematical abstractions of physical laws that govern behavior of the component;
determining an estimated state of the model based on at least one of a prior state, the current state derivatives, the solver state errors, and the synthesized parameters using online linearization and gain calculation to determine estimator gain for determining the estimated state of the model;
processing at least the synthesized parameters of the model to determine the model output;
directing an actuator associated with the control device as a function of the model output using a control law; and
positioning the control device comprising a control surface using the actuator, wherein the actuator positions the control surface.

11. The method of claim 10 , wherein online linearization and gain calculation comprises obtaining partial derivatives based on current states, the model input vector, the model output, and the synthesized parameters.

12. The method of claim 11 , wherein online linearization and gain calculation further comprises constructing a linear system for gain design using the partial derivatives.

13. The method of claim 12 , wherein online linearization and gain calculation further comprises determining the estimator gain based on the linear system for gain design.

14. The method of claim 10 , wherein the estimator gain is used in minimizing error vectors.

15. The method of claim 14 , further comprising using numerical integration to determine the estimated state of the model based on the prior state and the current state derivatives.

16. A gas turbine engine comprising:
an actuator for positioning the gas turbine engine, wherein the actuator positions a control surface of an element of the gas turbine engine;
a control law configured to direct the actuator as a function of a model output; and
a model processor configured to generate the model output, the model processor comprising:
an input object for processing a model input vector and setting a model operating mode;
a set state module for setting dynamic states of the model processor, the dynamic states input to an open loop model based on the model operating mode;
wherein the open loop model generates current state derivatives, solver state errors, and synthesized parameters as a function of the dynamic states and the model input vector, wherein a constraint on the current state derivatives and solver state errors is based on a series of cycle synthesis modules, each member of the series of cycle synthesis modules modeling a component of a cycle of the gas turbine engine and comprising a series of utilities, the utilities based on mathematical abstractions of physical laws that govern behavior of the component;
an estimate state module configured to determine an estimated state of the model based on at least one of a prior state, the current state derivatives, the solver state errors, and the synthesized parameters, the estimate state module using online linearization and gain calculation to determine estimator gain for determining the estimated state of the model; and
an output object for processing at least the synthesized parameters of the model to determine the model output.

17. The gas turbine engine of claim 16 , wherein online linearization and gain calculation comprises obtaining partial derivatives based on current states, the model input vector, the model output, and the synthesized parameters.

18. The gas turbine engine of claim 17 , wherein online linearization and gain calculation further comprises constructing a linear system for gain design using the partial derivatives.

19. The gas turbine engine of claim 18 , wherein online linearization and gain calculation further comprises determining the estimator gain based on the linear system for gain design.

20. The gas turbine engine of claim 16 , wherein the estimator gain is used in minimizing error vectors.

Assignees

• United Technologies Corp

Inventors

• Karpman Boris
• Meisner Richard P
• Dhingra Manuj

Classifications

• F04D29/321

  for axial flow compressors · CPC title

• F05D2260/85

  Starting · CPC title

• F01D21/003

  Arrangements for testing or measuring (for measuring vibrations G01H) · CPC title

• F05D2240/35

  Combustors or associated equipment · CPC title

• F05D2260/81

  Modelling or simulation · CPC title