Back to results
US 7,949,415 B2Granted
Apparatus and method for model-based control for integrating processes
AssigneeControlSoft, Inc.
InventorsTien-Li Chia, Irving Lefkowitz, Coleman B Brosilow
Life of this patent
Estimating the term…
Abstract
Various methods and systems for the parametric control of a process include representing the process with a process model used to generate future predictions of a process variable. In one embodiment, the process exhibits integrating behavior that is represented by a non-integrating process model. In another embodiment, an inverse of the model is filtered using a filter that includes a lead time constant that is selected to minimize a steady state error of the predicted process variable. In yet another embodiment, an array of output model values is revised or reindexed in response to a change in a time-varying parameter related to the process.
Claims8
A control system for controlling an integrating process, comprising:
- an input through which a set point for the process is input;
- an output through which a control output is output to the process for controlling the process;
- and
- a controller in which the integrating process is represented using a non-integrating process model that approximates the behavior of the integrating process, and the controller configured to:
- in response to a set point change, predict a process variable of the integrating process using the non-integrating model as part of a parametric model-based control technique;
- and
- generate the control output based on the predicted process variable;
- wherein the control technique inverts at least an invertible portion of the non-integrating process model and applies a filter to the inverted portion of the model;
- and
- wherein the filter is represented by:
- (
- γ
-
-
- s
- +
- 1
- )
- m
- (
- ɛ
-
-
- s
- +
- 1
- )
- n
- where n is greater than m and m is greater than or equal to zero, ε is a filter time constant, γ is lead time constant of the filter and S is a Laplace variable, and where γ is selected to minimize a steady state error for a step input disturbance to the process by representing γ as
- 1
- m
-
- (
- n
-
-
- ɛ
- +
- D
- ^
- )
- where {circumflex over (D)} is an estimated dead time of the process.
39 more elements
A control system for controlling an integrating process, comprising:
- an input through which a set point for the process is input;
- an output through which a control output is output to the process for controlling the process;
- and
- a controller in which the integrating process is represented using a non-integrating process model that approximates the behavior of the integrating process, and the controller configured to:
- in response to a set point change, predict a process variable of the integrating process using the non-integrating model as part of a parametric model-based control technique;
- and
- generate the control output based on the predicted process variable;
- wherein the non-integrating process model is a first order model;
- wherein the first order model is represented by:
- α
-
-
- k
- ^
- α
-
-
- s
- +
- 1
-
- ⅇ
- -
- D
- ^
-
- S
- where α is an adjustment parameter, {circumflex over (k)} is an estimated steady state gain of the process, {circumflex over (D)} is an estimated dead time of the process, and S is a Laplace variable;
- wherein α is selected to tune a process variable response;
- and
- wherein α is about 100 to about 1,000.
28 more elements
A control system for controlling an integrating process, comprising:
- an input through which a set point for the process is input;
- an output through which a control output is output to the process for controlling the process;
- and
- a controller in which the integrating process is represented using a non-integrating process model that approximates the behavior of the integrating process, and the controller configured to:
- in response to a set point change, predict a process variable of the integrating process using the non-integrating model as part of a parametric model-based control technique;
- and
- generate the control output based on the predicted process variable;
- and
- wherein the integrating process is a second or higher order process and the model is represented by
- α
-
-
- k
- ^
- α
-
-
- s
- +
- 1
-
- ⅇ
- -
- D
- ~
-
- S
- where α is an adjustment parameter, {circumflex over (k)} is an estimated steady state gain of the process, {tilde over (D)} is a sum of an estimated dead time of the process and a lag, and S is a Laplace variable.
25 more elements
A control system for controlling an integrating process, comprising:
- an input through which a set point for the process is input;
- an output through which a control output is output to the process for controlling the process;
- and
- a controller in which the integrating process is represented using a non-integrating process model that approximates the behavior of the integrating process, and the controller configured to:
- in response to a set point change, predict a process variable of the integrating process using the non-integrating model as part of a parametric model-based control technique;
- and
- generate the control output based on the predicted process variable;
- wherein the non-integrating process model is a second or higher order model;
- and
- wherein the model is represented by:
- G
- m
-
-
- 1
-
- (
- s
- )
- =
- α
-
-
- k
- ^
- α
-
-
- s
- +
- 1
-
- G
- m
-
-
- 1
- 0
-
- (
- s
- )
-
- ⅇ
- -
- D
- ^
-
- S
- where
- s
-
- →
- 0
- lim
-
- G
- m
-
-
- 1
- 0
-
- (
- s
- )
- =
- 1
- where α is an adjustment parameter, {circumflex over (k)} is an estimated steady state gain of the process, {circumflex over (D)} is an estimated dead time of the process, S is a Laplace variable, and Gm10 (s) represents second or higher order terms of the model.
66 more elements
A control system for controlling an integrating process using a parametric model-based control technique, comprising:
- an input through which a set point for the process is input;
- an output through which a control output is output to the process for controlling the process;
- and
- a controller in which the integrating process is represented using a process model, and the controller configured to:
- apply a filter to an inverse of the process model and predict a process variable of the process in response to a set point change, the filter represented by:
- (
- γ
-
-
- s
- +
- 1
- )
- m
- (
- ɛ
-
-
- s
- +
- 1
- )
- n
- where n is greater than m and m is greater than or equal to zero, ε is a filter time constant and y is lead time constant of the filter, and γ is selected to minimize a steady state error for a step input disturbance to the process by representing γ by:
- 1
- m
-
- (
- n
-
-
- ɛ
- +
- D
- ^
- )
- where {circumflex over (D)} is an estimated dead time of the process;
- and generate the control output.
35 more elements
Dependencies are inferred from the claim text.