# US 2023/0273575 A1 · Empirical modeling with globally enforced general constraints

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2023/0273575 A1
- Document ID: US20230273575A1
- Status: Published application
- Kind: A1
- Application number: 18/312,797
- Filed: 2023-05-05
- Published: 2023-08-31
- Earliest claimed: 2012-11-06
- Assignees: Rockwell Automation Technologies, Inc.
- Applicants: Rockwell Automation Technologies, Inc.
- Inventors: Bijan Sayyar-Rodsari; Eric Jon Hartman; Carl Anthony Schweiger
- Web page: https://patentgrep.com/patent/US20230273575A1

## CPC

- G05B 13/041: {in which a variable is automatically adjusted to optimise the performance}
- G06F 17/10: Complex mathematical operations {(function generation by table look-up G06F1/03; evaluation of elementary functions by calculation G06F7/544)}
- G06N 3/02: Neural networks
- G06N 3/048: Activation functions
- G06N 20/00: Machine learning
- G06N 20/10: using kernel methods, e.g. support vector machines [SVM]
- G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

In certain embodiments, a method includes formulating an optimization problem to determine a plurality of model parameters of a system to be modeled. The method also includes solving the optimization problem to define an empirical model of the system. The method further includes training the empirical model using training data. The empirical model is constrained via general constraints relating to first-principles information and process knowledge of the system.

## Claims

### Claim at position 1

1-20. (canceled)

### Claim 21

A system, comprising:
a processor;
a tangible, non-transitory, computer-readable medium, comprising instructions configured to be executed by the processor, the instructions comprising:
receiving a plurality of operational parameters associated with inputs and outputs of an industrial automation process for an industrial system;
formulating an optimization problem to determine a plurality of model parameters of an empirical model based on the plurality of operational parameters;
determining a plurality of dimensions of a model space of the empirical model;
selecting a subset of the plurality of the model parameters based on the plurality of dimensions;
solving the optimization problem to define the empirical model subject to general constraints based on the selected subset of the plurality of model parameters such that the general constraints are enforced over the plurality of dimensions; and
outputting a control signal to adjust an operation of an industrial automation component of the industrial system based on the empirical model.

### Claim 22 (depends on 21)

The system of claim 21, wherein the instructions comprise selecting the subset of the plurality of the model parameters based on a relative arrangement of the plurality of model parameters within the plurality of dimensions.

### Claim 23 (depends on 21)

The system of claim 21, wherein the subset of the plurality of the model parameters comprises a vertex, an edge, a face, or a combination thereof, of the model space.

### Claim 24 (depends on 21)

The system of claim 21, wherein the instructions comprise randomly selecting the subset of the plurality of the model parameters based on the plurality of dimensions.

### Claim 25 (depends on 21)

The system of claim 21, wherein the instructions comprise selecting the subset of the plurality of the model parameters such that at least one of the model parameters of the subset of the plurality of the model parameters comprises a boundary of the model space along one or more of the plurality of dimensions.

### Claim 26 (depends on 21)

The system of claim 21, wherein the instructions comprise:
selecting probability values based on Lagrangian information associated with the plurality of model parameters.

### Claim 27 (depends on 21)

The system of claim 21, wherein the instructions comprise:
determining a first asymptotic behavior of the industrial system based on the plurality of operational parameters;
determining a basis/kernel function of the empirical model based at least in part on the first asymptotic behavior of the industrial system; and
solving the optimization problem to define the empirical model such that empirical model comprises a second asymptotic behavior that conforms with the first asymptotic behavior of the industrial system, and wherein the empirical model comprises the basis/kernel function.

### Claim 28

A method, comprising:
receiving a plurality of operational parameters associated with inputs and outputs of an industrial automation process for an industrial system;
formulating an optimization problem to determine a plurality of model parameters of an empirical model based on the plurality of operational parameters;
determining a plurality of dimensions of a model space of the empirical model;
selecting a subset of the plurality of the model parameters based on the plurality of dimensions;
solving the optimization problem to define the empirical model subject to general constraints based on the selected subset of the plurality of model parameters such that the general constraints are enforced over the plurality of dimensions; and
outputting a control signal to adjust an operation of an industrial automation component of the industrial system based on the empirical model.

### Claim 29 (depends on 28)

The method of claim 28, comprising selecting the subset of the plurality of the model parameters based on a lowest-dimensional boundary type of the model space.

### Claim 30 (depends on 28)

The method of claim 28, wherein the subset of the plurality of the model parameters comprises a vertex, an edge, a face, or a combination thereof, of the model space.

### Claim 31 (depends on 28)

The method of claim 28, wherein selecting the subset of the plurality of the model parameters based on a relative arrangement of the plurality of model parameters within the plurality of dimensions.

### Claim 32 (depends on 28)

The method of claim 28, wherein selecting the subset of the plurality of the model parameters based on the plurality of dimensions comprises:
assigning a probability value to each model parameter of the plurality of model parameters based on the plurality of dimensions of the model space; and
selecting the subset of the plurality of the model parameters based on the probability values.

### Claim 33 (depends on 32)

The method of claim 32, comprising tuning the empirical model to generate an updated empirical model based on the probability value associated with each model parameter of the subset of the plurality of the model parameters.

### Claim 34 (depends on 28)

The method of claim 28, wherein the operational parameters comprise sensor data sensed by a sensor during operation of the industrial system, actuator data related to operation of an actuator in the industrial system, or both.

### Claim 35

A tangible, non-transitory, computer-readable medium configured to store instructions executable by processing circuitry, wherein the instructions comprise instructions to cause the processing circuitry to perform operations comprising:
receiving a plurality of operational parameters associated with inputs and outputs of an industrial automation process for an industrial system;
formulating an optimization problem to determine a plurality of model parameters of an empirical model based on the plurality of operational parameters;
determining a plurality of dimensions of a model space of the empirical model;
selecting a subset of the plurality of the model parameters based on the plurality of dimensions;
solving the optimization problem to define the empirical model subject to general constraints based on the selected subset of the plurality of model parameters such that the general constraints are enforced over the plurality of dimensions; and
outputting a control signal to adjust an operation of an industrial automation component of the industrial system based on the empirical model.

### Claim 36 (depends on 35)

The computer-readable medium of claim 35, wherein the subset of the plurality of the model parameters comprises a vertex, an edge, a face, or a combination thereof, of the model space.

### Claim 37 (depends on 35)

The computer-readable medium of claim 35, wherein the instructions comprise selecting the subset of the plurality of the model parameters based on a relative arrangement of the plurality of model parameters within the plurality of dimensions.

### Claim 38 (depends on 35)

The computer-readable medium of claim 35, wherein the instructions comprise:
determining a first asymptotic behavior of the industrial system based on the plurality of operational parameters;
determining a basis/kernel function of the empirical model based at least in part on the first asymptotic behavior of the industrial system; and
solving the optimization problem to define the empirical model such that empirical model comprises a second asymptotic behavior that conforms with the first asymptotic behavior of the industrial system, and wherein the empirical model comprises the basis/kernel function.

### Claim 39 (depends on 35)

The computer-readable medium of claim 35, wherein adjusting an operation of the industrial automation component comprises causing an actuator to perform a control action.

### Claim 40 (depends on 35)

The computer-readable medium of claim 35, wherein the instructions to cause the processing circuitry to perform the operations comprising solving the optimization problem comprise instructions to cause the processing circuitry to perform the operations comprising automatically adapting the empirical model during operation of the industrial system.
