Systems and methods for adaptive non-linear control of process systems
OwnerKyle Hunteinventor
Life of this patent
Estimating the term…
Abstract
The invention provides systems and methods for generating an adaptive nonlinear controller and utilizing the adaptive nonlinear controller to regulate the operation of nonlinear process systems. In particular, a method is provided for generating a control model by defining an objective function utilizing a target function that specifies the desired response of the system and a state-space model representing the dynamics of the non-linear system. When executed by a controller the control model causes the regulated system to operate as specified by the target function and thereby produce a product that is consistent with various prescribed quality metrics.
Claims14
A method for configuring a controller of a non-linear process system for producing a product, comprising:
- providing a state-space model of the system, wherein the state space model comprises one or more state-space equations representing a relationship between at least one system variable x, and at least one system output y and is a function of at least one control variable u, wherein u is an input to the system and wherein y relates to a quality measure of the product;
- transforming, with one or more processors configured by executing instructions in the form of code therein, the state-space model into one or more controllable canonical equations, wherein at least one of the one or more controllable canonical equations includes the control variable u;
- generating, with the one or more processors, an objective function J, wherein generating the objective function includes:
- selecting, with the one or more processors, the objective function J from among a plurality of objective functions stored in a non-transitory computer readable storage medium, wherein J is selected based on at least an order of the state space model, and wherein J is a function of the system output y, the derivatives of y and a control input v, a gain λ and system constants a and b, wherein control input v defines one or more operational set-points of the system, and
- inputting, with one or more processors, the one or more controllable canonical equations into the selected objective function;
- minimizing, with one or more processors, the objective function with respect to the control variable u;
- selecting, with one or more processors, parameters of the control system including the gain λ and a and b according to constraints of the non-linear process system;
- programmatically configuring the controller to implement a control algorithm defined by the minimized objective function in view of the selected parameters.
A method for regulating a non-linear process system for producing a product, comprising:
- providing, to a computer-based controller, a state-space model of the non-linear process system (“the system”), wherein the state-space model represents a relationship between at least one system variable and at least one system output as a function of at least one system control variable, wherein the at least one system control variable is an input to the process system and is controlled by the controller, and wherein the at least one system output relates to a quality measure of the product produced by the system, and
- providing, to the controller, a reference model, wherein the reference model defines a target response of the at least one system output in relation to at least one controller input to produce a product having a prescribed quality measure, wherein the controller input defines one or more operational set-points of the system;
- providing, to the controller, one or more objective functions and a set of operational constraints relating to the non-linear process system;
- generating, with the controller, an adaptive non-linear control model according to the state-space model, the reference model and one or more objective functions and the set of operational constraints;
- receive, with the controller from one or more sensors coupled to the system during operation of the process system, process information including values of the at least one system variable and the at least one system output;
- and
- dynamically controlling, with the processor executing the adaptive non-linear control model based on the received process information, a value of the at least one system control variable to manipulate the least one system output in accordance with the target response thereby producing a product having the prescribed quality measure.
Dependencies are inferred from the claim text.