Method for analyzing and classifying process data

• US 20100030521A1
• Filed: 02/13/2008
• Published: 02/04/2010
• Est. Priority Date: 02/14/2007
• Status: Active Grant

• Current Assignee: Isagacity
• Original Assignee: Isagacity
• Inventors: Millett, Peter, Akhrarov, Murad, Sharma, Chanchal, Bates, Joe
• Granted Patent: US 8,036,999 B2
• Time in Patent Office: Days
• US Class Current: 702/182
• CPC Class Codes: G05B 13/0265 the criterion being a learning criterion
• CPC Class Codes: G05B 23/0229 knowledge based, e.g. expert systems; genetic algorithms
• CPC Class Codes: G06F 18/232 Non-hierarchical techniques
• CPC Class Codes: G06N 20/00 Machine learning

Abstract

Process data mining system and method. The system analyzes data from complex process plants or systems and operates in open-book and closed-book modes. In closed-book mode, the system monitors incoming data sets against pre-defined clusters of data values and generates reports, indicating whether incoming data is a match or a no-match with the pre-defined clusters. In open-book mode, the system generates initial clusters, without having a-priori knowledge of the component or process, and also creates clusters “on the fly”, thereby fine-tuning the analysis. A knowledge base encompasses a combination of parameters for a particular component. Clusters are defined within the knowledge base, each cluster representing a particular operating condition. The system expands clusters, within pre-defined limits, or creates new clusters, as needed, in order to accommodate incoming data values. Newly created clusters are then named, so as to indicate the particular operating conditions.

Claims

1-2. -2. (canceled)

3. :

A method of analyzing data relating to parameters of a component in a process plant, a component being a piece of equipment, a location, or a process step, and a process plant being an agglomeration of equipment and/or process steps, said method comprising the steps of;

a) defining a knowledge base that relates to a particular component, said knowledge base comprising a plurality of data channels, each data channel of said plurality transmitting measurement data relating respectively to a particular parameter of said particular component;

b) receiving a discrete transmission of said measurement data from said plurality of data channels;

c) determining whether one or more pre-defined clusters exist within said knowledge base, a pre-defined cluster being a pre-defined data vector having pre-defined boundaries for expected data values from said data channels in said knowledge base;

d) monitoring operation of said particular component by providing an algorithm that determines whether data values of said discrete transmission of said measurement data fall within said pre-defined boundaries of said pre-defined cluster;

e) operating said algorithm in open-book mode, so as to enable generation of an algorithm-defined cluster, wherein said algorithm-defined cluster is a data model that defines a hitherto undefined data vector, based on actual values that are received in said discrete transmission of measurement data;

f) upon determining that no pre-defined cluster exists, automatically generating said algorithm-defined cluster; and

g) upon determining that one or more pre-defined clusters exist, but that said data values of said discrete transmission do not fall within said pre-defined boundaries of any one said pre-defined cluster, automatically generating said algorithm-defined cluster.

4. :

The method of claim 3, wherein said knowledge base includes a plurality of pre-defined clusters, each said pre-defined cluster corresponding to a particular operating state of said particular component, said method comprising the following steps;

d1) identifying each said pre-defined cluster within said knowledge base as corresponding to a particular operating state, wherein a first pre-defined cluster corresponds to a first operating state and a second pre-defined cluster corresponds to a second operating state;

d2) comparing said discrete transmission of measurement data from said data channels with said pre-defined clusters, so as to determine a closest cluster, said closest cluster having cluster boundaries that encompass said data values of said discrete transmission;

d3) determining a residual value of said data values of said discrete transmission relative to an expected value of said closest cluster, said expected value being defined by said cluster boundaries and said residual value defining a distance between said data values of said discrete transmission and said expected value of said closest cluster;

d4) determining a match condition if said residual value is within said cluster boundaries; and

d5) determining a no-match condition, if said residual value is outside said cluster boundaries.

5. :

The method of claim 4, further comprising the steps of;

d6) providing a pre-defined boundary-expansion tolerance for said cluster boundaries of each said pre-defined cluster;

d7) determining that said residual value of said data values of said discrete transmission lie within said pre-defined boundary-expansion tolerance of one of said pre-defined clusters; and

d8) expanding said knowledge base by automatically expanding said cluster boundaries of said one of said pre-defined clusters, so as to obtain an expanded cluster that encompasses said data values of said discrete transmission.

6. :

The method of claim 5, further comprising the steps of;

h) within said knowledge base, defining all said data channels of said knowledge base as base level parameters;

i) within each said pre-defined cluster, defining at least one base level parameter that is particularly indicative of a particular operating state to also be a key level parameter;

j) within said knowledge base, defining a key-level knowledge base that contains a key-level cluster comprising a selection of said key level parameters, said key-level cluster being indicative of a particular operating state of said component;

k) first processing with said algorithm said key level parameters in said key-level knowledge base to determine a match with one said key-level cluster, so as to determine said particular operating state, and then subsequently processing said base level parameters against said pre-defined clusters that correspond to said particular operating state.

7. :

The method of claim 6, said step of processing said base level parameters comprising the steps of;

k1) forming a base-level knowledge base comprising all data channels;

k2) defining within said base-level knowledge base one or more parent clusters and one or more child clusters, wherein a first parent cluster corresponds to a first operating condition and a first child cluster of said first parent cluster corresponds to a particular scenario within said first operating condition;

k3) applying a fingerprinting algorithm to said one or more parent clusters;

k4) searching for a match of said data values of said discrete transmission with said expected value of one parent cluster of said one or more parent clusters;

k5) if a match is found with said expected value of said one parent cluster, associating a cluster ID of said one parent cluster with said data values of said discrete transmission and incrementing a cluster counter for said one parent cluster;

k6) if a no-match with said one or more parent clusters is found, expanding said base-level knowledge base by generating a new algorithm-defined parent cluster that reflects said data values of said discrete transmission; and

k7) creating a report on said expanded knowledge base.

8. :

The method of claim 6, said step of processing said key level parameters further comprising the steps of;

k8) ascertaining whether said process plant includes a plurality of plants;

k9) allocating a collection of said data channels to a corresponding one plant of said plurality plants; and

k10) sequentially processing a first knowledge base of each plant of said plurality of plants and then a second knowledge base of said each plant.

9. :

The method of claim 3, further comprising the steps of;

l) operating said algorithm concurrently in a closed book mode, which disallows generation of said algorithm-defined cluster with said closed book mode;

m) providing pre-existing data for said knowledge base, so as to provide at least one cluster that is a pre-defined cluster that has cluster boundaries that are defined by a minimum-maximum tolerance range for said each data value from said data channels;

n) comparing said data values from said discrete transmission with values in said pre-defined cluster;

o) determining whether said data values of said discrete transmission fall within said cluster boundaries of any said pre-existing cluster;

p) upon determining a match, issuing a match report; and

q) upon determining a no-match, issuing a no-match report.

10. :

The method of claim 3, said step of defining a knowledge base further comprising the step of;

a1) defining a plurality of knowledge bases, each said knowledge base relating to one particular component and containing data channels for said one particular component.

11. :

The method of claim 10 further comprising the steps of;

a2) defining a first knowledge base as comprising a first set of data channels and defining a second knowledge base as comprising a second set of data channels, wherein said first and said second sets of data channels are exclusive of each other.