Method and apparatus for approximating gains in dynamic and steady-state processes for prediction, control, and optimization
First Claim
1. A method for controlling a nonlinear system, comprising:
- a linear controller receiving inputs representing measured variables of the system, wherein values of the measured variables of the system are in a first operating region, wherein the linear controller is operable to determine predicted control values for manipulable variables that control the system, and wherein the linear controller has a linear gain k;
executing a nonlinear steady-state model of the system, wherein the nonlinear steady-state model has been trained to represent operation of the system over a specified region of the system'"'"'s operating space, and wherein the nonlinear steady state model has a steady-state gain K;
initiating a change in operation of the system, wherein the change operates to move the values of the measured variables of the system from a first operating point to a second operating point, wherein said initiating comprises modifying the nonlinear model in accordance with the change in operation of the system;
determining values of the steady-state gain K at the first and second operating points;
adjusting the linear gain k of the linear controller as a function of the values of the steady-state gain K of the nonlinear steady-state model at the first and second operating points;
the linear controller determining predicted control values for the manipulable variables in response to said adjusting; and
outputting the predicted control values for the manipulable variables, wherein the predicted control values are useable to control the system.
4 Assignments
0 Petitions
Accused Products
Abstract
A method and apparatus for controlling a non-linear mill. A linear controller is provided having a linear gain k that is operable to receive inputs representing measured variables of the plant and predict on an output of the linear controller predicted control values for manipulatible variables that control the plant. A non-linear model of the plant is provided for storing a representation of the plant over a trained region of the operating input space and having a steady-state gain K associated therewith. The gain k of the linear model is adjusted with the gain K of the non-linear model in accordance with a predetermined relationship as the measured variables change the operating region of the input space at which the linear controller is predicting the values for the manipulatible variables. The predicted manipulatible variables are then output after the step of adjusting the gain k.
-
Citations
20 Claims
-
1. A method for controlling a nonlinear system, comprising:
-
a linear controller receiving inputs representing measured variables of the system, wherein values of the measured variables of the system are in a first operating region, wherein the linear controller is operable to determine predicted control values for manipulable variables that control the system, and wherein the linear controller has a linear gain k;
executing a nonlinear steady-state model of the system, wherein the nonlinear steady-state model has been trained to represent operation of the system over a specified region of the system'"'"'s operating space, and wherein the nonlinear steady state model has a steady-state gain K;
initiating a change in operation of the system, wherein the change operates to move the values of the measured variables of the system from a first operating point to a second operating point, wherein said initiating comprises modifying the nonlinear model in accordance with the change in operation of the system;
determining values of the steady-state gain K at the first and second operating points;
adjusting the linear gain k of the linear controller as a function of the values of the steady-state gain K of the nonlinear steady-state model at the first and second operating points;
the linear controller determining predicted control values for the manipulable variables in response to said adjusting; and
outputting the predicted control values for the manipulable variables, wherein the predicted control values are useable to control the system. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
-
18. A system for controlling a nonlinear system, comprising:
-
means for a linear controller receiving inputs representing measured variables of a nonlinear system, wherein values of the measured variables of the nonlinear system are in a first operating region, wherein the linear controller is operable to determine predicted control values for manipulable variables that control the nonlinear system, and wherein the linear controller has a linear gain k;
means for executing a nonlinear steady-state model of the nonlinear system, wherein the nonlinear steady-state model has been trained to represent operation of the nonlinear system over a specified region of the nonlinear system'"'"'s operating space, and wherein the nonlinear steady state model has a steady-state gain K;
means for initiating a change in operation of the nonlinear system, wherein the change operates to move the values of the measured variables of the nonlinear system from a first operating point to a second operating point, wherein said initiating comprises modifying the nonlinear model in accordance with the change in operation of the nonlinear system;
means for determining values of the steady-state gain K at the first and second operating points;
means for adjusting the linear gain k of the linear controller as a function of the values of the steady-state gain K of the nonlinear steady-state model at the first and second operating points;
means for the linear controller determining predicted control values for the manipulable variables in response to said adjusting; and
means for outputting the predicted control values for the manipulable variables, wherein the predicted control values are useable to control the nonlinear system.
-
-
19. A system for operating a nonlinear system, the system comprising:
-
a linear controller, operable to;
receive inputs representing measured variables of a nonlinear system, wherein values of the measured variables of the nonlinear system are in a first operating region; and
determine predicted control values for manipulable variables that control the nonlinear system, and wherein the linear controller has a linear gain k; and
a nonlinear steady-state model of the nonlinear system, coupled to the linear controller, wherein the nonlinear steady-state model has been trained to represent operation of the nonlinear system over a specified region of the nonlinear system'"'"'s operating space, and wherein the nonlinear steady state model has a steady-state gain K;
wherein, in response to an initiated change in operation of the nonlinear system from a first operating point to a second operating point, including a modification to the nonlinear model in accordance with the change in operation of the nonlinear system, the linear controller is further operable to;
adjust the linear gain k as a function of values of the steady-state gain K of the nonlinear steady-state model at first and second operating points;
output predicted control values for the manipulable variables based on the adjusted linear gain k, wherein the predicted control values are useable to control the nonlinear system.
-
-
20. A computer-accessible memory medium that stores program instructions for operating a nonlinear system, wherein the program instructions implement:
-
a linear controller, operable to;
receive inputs representing measured variables of a nonlinear system, wherein values of the measured variables of the nonlinear system are in a first operating region; and
determine predicted control values for manipulable variables that control the nonlinear system, and wherein the linear controller has a linear gain k; and
a nonlinear steady-state model of the nonlinear system, wherein the nonlinear steady-state model has been trained to represent operation of the nonlinear system over a specified region of the nonlinear system'"'"'s operating space, and wherein the nonlinear steady state model has a steady-state gain K, wherein the program instructions are executable by a processor to perform;
executing the nonlinear steady-state model of the system;
initiating a change in operation of the system, wherein the change operates to move the values of the measured variables of the system from a first operating point to a second operating point, wherein said initiating comprises modifying the nonlinear model in accordance with the change in operation of the system;
determining values of the steady-state gain K at the first and second operating points;
adjusting the linear gain k of the linear controller as a function of the values of the steady-state gain K of the nonlinear steady-state model at the first and second operating points;
the linear controller determining predicted control values for the manipulable variables in response to said adjusting; and
outputting the predicted control values for the manipulable variables, wherein the predicted control values are useable to control the system.
-
Specification