# US 2004/0083013 A1 · Method of operating a plant and a control system for controlling the same

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

- Number: US 2004/0083013 A1
- Document ID: US20040083013A1
- Status: Published application
- Kind: A1
- Application number: 10/280,693
- Filed: 2002-10-25
- Published: 2004-04-29
- Earliest claimed: 2002-10-25
- Assignees: GBI Structures, L.L.C.
- Inventors: H. Dennis Tolley
- Web page: https://patentgrep.com/patent/US20040083013A1

## CPC

- G05B 13/028: {using expert systems only}
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- 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)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A novel apparatus and method for controlling a system with multiple observable variables is disclosed. The apparatus and method disclosed use elements of the bottom-up and top-down strategies of artificial intelligence to provide a control system that is able to learn without a training set of information, and that has an learning process that can handle any amount of input data, i.e., cannot become saturated. The control system disclosed is capable of learning and controlling non linear dynamical systems. The control system is also capable of adding additional observable variables or subtracting existing observable variables to determine the state of the plant or system being controlled.

## Claims

### Claim 1

A method of controlling a plant, described by either a linear or nonlinear dynamical system, comprising the steps of:
(a) receiving a set of data signals indicative of a plurality of observable plant variables;
(b) creating a subset of data signals from said set of data signals;
(c) constructing a grade of membership partition from said subset of data signals;
(d) determining the state variables of the plant;
(e) dividing said state variables into a plurality of finite states;
(f) estimating the losses for transitioning from one state to plurality of possible other states;
(g) assigning a probability to each of said losses;
(h) making a decision using a Bayes preference, wherein said decision is based upon said losses and probabilities;
(i) receiving an updated set of data signals from the plant;
(j) updating said table of losses using said updated set of data signals;
(k) updating said probabilities associated with said losses using said update set of data signals; and
(l) making a new decision using a Bayes preference, wherein said new decision is based upon said updated losses and probabilities.

### Claim 2 (depends on 1)

The method of claim 1, wherein the step of updating said estimated loss comprises estimating a subsequent loss that is a marginal value of said previously estimated loss.

### Claim 3 (depends on 2)

The method of claim 2, wherein the step of updating said subsequent loss is a function of the average of all previously (estimated/actual) losses.

### Claim 4 (depends on 2)

The method of claim 2, wherein the step of updating said subsequent loss is a function of an interpolation of all previously (estimated/actual) losses.

### Claim 5 (depends on 2)

The method of claim 2, wherein the step of updating said subsequent loss is a function of an extrapolation of all previously (estimated/actual) losses.

### Claim 6 (depends on 1)

The method of claim 1, wherein said plant comprises harvestable products.

### Claim 7 (depends on 1)

The method of claim 1, wherein said plant comprises transportation equipment.

### Claim 8 (depends on 1)

The method of claim 1, wherein said plant comprises a chemical process.

### Claim 9

A method of controlling a machine, comprising the steps of:
(a) receiving a set of data signals indicative of a plurality of observable machine variables;
(b) creating a subset of data signals from said set of data signals;
(c) constructing a grade of membership partition from said subset of data signals;
(d) determining the state variables of the machine;
(e) dividing said state variables into a plurality of finite states;
(f) estimating the losses for transitioning from one state to a plurality of possible other states;
(g) assigning a probability to each of said losses;
(h) making a decision using a Bayes preference, wherein said decision is based upon said losses and probabilities;
(i) receiving an updated set of data signals from the machine;
(j) updating said table of losses using said updated set of data signals;
(k) updating said probabilities associated with said losses using said updated set of data signals; and
(l) making a new decision using a Bayes preference, wherein said new decision is based upon said updated losses and probabilities.

### Claim 10 (depends on 9)

The method of claim 9, wherein the step of updating said estimated loss comprises estimating a subsequent loss that is a marginal value of said previously estimated loss.

### Claim 11 (depends on 10)

The method of claim 10, wherein the step of updating said subsequent loss is a function of the average of all previously (estimated/actual) losses.

### Claim 12 (depends on 10)

The method of claim 10, wherein the step of updating said subsequent loss is a function of an interpolation of all previously (estimated/actual) losses.

### Claim 13 (depends on 10)

The method of claim 10, wherein the step of updating said subsequent loss is a function of an extrapolation of all previously (estimated/actual) losses.

### Claim 14

In a system comprising a plurality of observable variables, an apparatus for controlling said system, comprising:
(a) means for receiving a set of data signals indicative of at least one of said plurality of observable variables;
(b) means for creating a subset of data signals from said set of data signals;
(c) means for constructing a grade of membership partition from said subset of data signals;
(d) means for determining the state variables of the system;
(e) means for dividing said state variables into a plurality of finite states;
(f) means for estimating the losses for transitioning from a first state to a plurality of possible other states;
(g) means for assigning a probability to each of said losses;
(h) means for making a decision using a Bayes preference, wherein said decision is based upon said losses and probabilities;
(i) means for receiving an updated set of data signals from the system;
(j) means for updating said table of losses using said updated set of data signals;
(k) means for updating said probabilities associated with said losses using said updated set of data signals; and
(l) means for making a new decision using a Bayes preference, wherein said new decision is based upon said updated losses and probabilities.

### Claim 15 (depends on 13)

The apparatus of claim 13, wherein said system is a biosystem.

### Claim 16 (depends on 13)

The apparatus of claim 13, wherein said system is a transportation system.

### Claim 17 (depends on 13)

The apparatus of claim 13, wherein said means for updating said estimated loss comprises means for estimating a subsequent loss that is a marginal value of said previously estimated loss.

### Claim 18 (depends on 16)

The apparatus of claim 16, wherein the step of updating said subsequent loss is a function of the average of all previously (estimated/actual) losses.

### Claim 19 (depends on 16)

The apparatus of claim 16, wherein the step of updating said subsequent loss is a function of an interpolation of all previously (estimated/actual) losses.

### Claim 20 (depends on 16)

The apparatus of claim 16, wherein the step of updating said subsequent loss is a function of an extrapolation of all previously (estimated/actual) losses.
