Method, apparatus and computer program product for constructing a diagnostic network model

• US 8,510,083 B2
• Filed: 07/31/2003
• Issued: 08/13/2013
• Est. Priority Date: 07/31/2003
• Status: Active Grant

• Current Assignee: The Boeing Co.
• Original Assignee: The Boeing Co.
• Inventors: Kipersztok, Oscar, Wang, Haiqin, Xue, Ping
• Primary Examiner(s): ALHIJA, SAIF A
• Application Number:
  US10/631,697
• Publication Number:
  US 20050028033A1
• Time in Patent Office: 3,666 Days
• Field of Search: 703/2
• US Class Current: 703/2
• CPC Class Codes: G05B 23/0283 Predictive maintenance, e.g. involving the monitoring of a system and, based on the monitoring results, taking decisions on the maintenance schedule of the monitored system; Estimating remaining useful life [RUL] preventive maintenance, i.e. planning maintenance according to the available resources without monitoring the system G06Q10/06
• CPC Class Codes: G06F 11/2257 using expert systems

Abstract

A method, apparatus and computer program product are provided for constructing a diagnostic network model in a more efficient and timely manner and with improved consistency. A structured electronic representation of a document is automatically reviewed to identify the symptoms experienced by the complex system, the plurality of components of the complex system and the causal relationships between a failure of the respective components and the occurrence of the various symptoms. Respective probabilities can be associated with the causal relationships between the symptoms and the failure of the various components. A network model, such as a Bayesian network model, may then be automatically constructed to represent the symptoms, the components, the causal relationships between failure of the components and exhibition of the symptoms, and the probabilities of the respective causal relationships.

Claims

1. A method for constructing a diagnostic network model comprising:

automatically reviewing, with a processor, a structured electronic representation of each of a plurality of different types of documents, wherein automatically reviewing comprises automatically identifying at least one symptom, a plurality of components and causal relationships between a failure of the respective components and the occurrence of at least one symptom, wherein automatically identifying comprises mapping the failure of a respective component to at least one symptom to establish a respective causal relationship;

associating respective probabilities, with the processor, with the causal relationships between the at least one symptom and the plurality of components whose failure contributes to the at least one symptom, wherein associating respective probabilities comprises obtaining respective probabilities from an individual participating in construction of the network model, and wherein obtaining respective probabilities from an individual comprises eliciting the probability associated with the causal relationship between at least one symptom and the failure of a respective component and receiving input from the individual of the elicited probability;

automatically constructing, with the processor, the network model of the at least one symptom, the plurality of components, the causal relationships therebetween and the probabilities of the respective causal relationships; and

displaying, with the processor, a representation of at least a portion of the network model.

2. A method according to claim 1 wherein the structured electronic representation of the document comprises a plurality of tags identifying respective fields including the at least one symptom and the plurality of components that are causally related to the at least one symptom, and wherein automatically reviewing the structured electronic representation of the document comprises automatically identifying the at least one symptom and the plurality of components that are causally related to the at least one symptom based, at least in part, upon the plurality of tags.

3. A method according to claim 1 wherein the structured electronic representation of the document comprises a markup language representation of the document, and wherein automatically reviewing the structured electronic representation of the document comprises automatically reviewing the markup language representation of the document.

4. A method according to claim 1 wherein associating the respective probabilities comprises at least one of receiving input defining at least some of the respective probabilities and retrieving at least some of the respective probabilities from a database.

5. A method according to claim 1 wherein automatically constructing the network model comprises automatically constructing a Bayesian network model of the at least one symptom, the plurality of components, the causal relationships therebetween and the probabilities of the respective causal relationships.

6. An apparatus for constructing a diagnostic network model comprising:

a processor for automatically reviewing a structured electronic representation of each of a plurality of different types of documents including automatically identifying at least one symptom, a plurality of components and causal relationships between a failure of respective components and occurrence of the at least one symptom by mapping the failure of a respective component to at least one symptom to establish a respective causal relationship, wherein said processor also associates respective probabilities with the causal relationships between the at least one symptom and the plurality of components whose failure contributes to the at least one symptom, wherein the processor is configured to obtain respective probabilities from an individual participating in construction of the network model by eliciting the probability associated with the causal relationship between at least one symptom and the failure of a respective component and receiving input from the individual of the elicited probability, wherein said processor automatically constructs the network model of the at least one symptom, the plurality of components, the causal relationships therebetween and the probabilities of the respective causal relationships, and wherein the processor is configured to display a representation of at least a portion of the network model.

7. An apparatus according to claim 6 wherein the structured electronic representation of the document comprises a plurality of tags identifying respective fields including the at least one symptom and the plurality of the components that are causally related to the at least one symptom, and wherein said processor automatically identifies the at least one symptom and the plurality of components that are causally related to the at least one symptom based, at least in part, upon the plurality of tags.

8. An apparatus according to claim 6 further comprising at least one of an interface for permitting input of at least some of the respective probabilities and a memory device for storing some of the respective probabilities.

9. An apparatus according to claim 6 wherein said processor automatically constructs a Bayesian network model of the at least one symptom, the plurality of components, the causal relationships therebetween and the probabilities of the respective causal relationships.

10. A computer program product comprising a computer-readable storage medium having computer-readable program code portions stored therein that, when executed, cause a processor to perform respective functions, the computer-readable program portions comprising:

a first processor-executable portion capable of reviewing a structured electronic representation of each of a plurality of different types of documents and of automatically identifying at least one symptom, a plurality of components and causal relationships between a failure of the respective components and the occurrence of the at least one symptom by mapping the failure of a respective component to at least one symptom to establish a respective causal relationship;

a second processor-executable portion capable of receiving respective probabilities associated with the causal relationships between the at least one symptom and the plurality of components whose failure contributes to the at least one symptom, wherein the second executable portion is configured to receive respective probabilities from an individual participating in construction of the network model, and wherein the second executable portion is further configured to obtain respective probabilities from an individual by eliciting the probability associated with the causal relationship between at least one symptom and the failure of a respective component and receiving input from the individual of the elicited probability;

a third processor-executable portion capable of constructing a diagnostic network model of the at least one symptom, the plurality of components, the causal relationships therebetween and the probabilities of the respective causal relationships; and

a fourth processor-executable portion configured to display a representation of at least a portion of the network model.

11. A computer program product according to claim 10 wherein the structured electronic representation of the document comprises a plurality of tags identifying respective fields including the at least one symptom and the plurality of components that are causally related to the at least one symptom, and wherein said first executable portion is capable of automatically identifying the at least one symptom and the plurality of components that are causally related to the at least one symptom based, at least in part, upon the plurality of tags.

12. A computer program product according to claim 10 wherein said second executable portion is capable of at least one of receiving input defining at least some of the respective probabilities and retrieving at least some of the respective probabilities from a database.

13. A computer program product according to claim 10 wherein said third executable portion is capable of constructing a Bayesian network model of the at least one symptom, the plurality of components, the causal relationships therebetween and the probabilities of the respective causal relationships.

14. A method comprising:

accessing, with a processor, at least a portion of a diagnostic network model that associates a plurality of components and symptoms based upon causal relationships therebetween, wherein the causal relationships associate a failure of respective components and occurrence of the at least one symptom;

associating, with the processor, a probability with the causal relationship between a symptom and a respective component whose failure contributes to the symptom, wherein associating a probability comprises providing the probability which was obtained from an individual participating in construction of the diagnostic network model;

displaying, with the processor, a representation of the plurality of components and symptoms including a plurality of parent nodes representative of the components, a plurality of child nodes representative of the symptoms and a plurality of influence arcs extending between respective parent and child nodes and associated with respective probabilities;

receiving a selection, with the processor, of at least one of the plurality of components and symptoms from among the plurality of components and symptoms that are displayed; and

in response to the selection and with the processor, obtaining and displaying additional information related to the at least one of the components and symptoms that is selected.

15. A method according to claim 14 further comprising constructing the diagnostic network model comprising:

automatically reviewing a structured electronic representation of a document, wherein automatically reviewing comprises automatically identifying the plurality of symptoms and components and the causal relationships therebetween; and

automatically constructing the network model of the plurality of symptoms and components and the causal relationships therebetween.

16. A method according to claim 15 wherein displaying additional information comprises displaying at least a portion of the document.

17. A method according to claim 14 further comprising defining an additional causal relationship between a symptom and a respective component that have been selected.

18. A method according to claim 14 further comprising deleting a causal relationship between a symptom and a respective component that have been selected.

19. A method according to claim 14 further comprising displaying each symptom that is causally related to the failure of the at least one component that is selected.

20. A method according to claim 14 further comprising displaying each component whose failure is capable of causing the at least one symptom that is selected.

21. A computer program product comprising a non-transitory computer-readable storage medium having computer-readable program code portions stored therein that, when executed, cause a processor to perform respective functions, the computer-readable program portions comprising:

a first processor-executable portion capable of accessing at least a portion of a diagnostic network model that associates a plurality of components and symptoms based upon causal relationships there between, wherein the causal relationships associate a failure of respective components and occurrence of the at least one symptom, and wherein the diagnostic model also associates a probability obtained from an individual participating in construction of the diagnostic network model with the causal relationship between a symptom and a respective component whose failure contributes to the symptom;

a second processor-executable portion capable of displaying a representation of the plurality of components and symptoms including a plurality of parent nodes representative of the components, a plurality of child nodes representative of the symptoms and a plurality of influence arcs extending between respective parent and child nodes and associated with respective probabilities;

a third processor-executable portion capable of receiving a selection of at least one of the plurality of components and symptoms from among the plurality of components and symptoms that are displayed; and

a fourth processor-executable portion, responsive to the selection, capable of obtaining and displaying additional information related to the at least one of the components and symptoms that is selected.

22. A computer-program product according to claim 21 further comprising a fifth executable portion capable of constructing the diagnostic network model, wherein said fifth executable portion is further capable of automatically reviewing a structured electronic representation of a document to identify the plurality of symptoms and components and the causal relationships therebetween, and of automatically constructing the network model of the plurality of symptoms and components and the causal relationships therebetween.

23. A computer program product according to claim 22 wherein said fourth executable portion is capable of displaying at least a portion of the document.