System and method for augmenting lumped parameter model diagnosis engine

Patent metadata

Field	Value
Publication number	US-11580267-B2
Application number	US-202016793375-A
Country	US
Kind code	B2
Filing date	Feb 18, 2020
Priority date	Feb 18, 2020
Publication date	Feb 14, 2023
Grant date	Feb 14, 2023

Abstract

A target system is coupled to a diagnosis engine that uses a lumped parameter model of the system for diagnosis. A proximity search in is performed in a computer-aided design model of the system to find groups of components that may be affected by resistive or parasitic interactions between the individual components in the groups. The lumped parameter model is augmented by adding elements that emulate the resistive or parasitic interactions between the individual components in the groups. The augmented lumped model is used by the diagnosis engine to perform diagnosis on the system.

First claim

The invention claimed is:

1. A method comprising:
accessing a computer aided design (CAD) model of a system, the system utilizing a diagnosis engine, the diagnosis engine using a lumped parameter model of the system for diagnosis;
analyzing the CAD model to determine components of the CAD model that are also represented in the lumped parameter model;
performing a proximity search in the CAD model to find groups of the components that may be affected by resistive or parasitic interactions between individual components in the groups due to geometric adjacency between the individual components without direct contact;
augmenting the lumped parameter model by adding new elements to the lumped parameter model that emulate the resistive or parasitic interactions between the individual components without direct contact in the groups; and
using the augmented lumped model by the diagnosis engine to perform the diagnosis on the system.

2. The method of claim 1 , wherein performing the diagnosis comprises at least one of real-time monitoring of the system.

3. The method of claim 1 , wherein the proximity search in the CAD model involves computing a distance field for rigid bodies that represent the individual components, wherein at least one of the rigid bodies moves relative to others of the rigid bodies in the system, the distance field calculated using a workspace of the at least one rigid body.

4. The method of claim 3 , wherein the proximity search comprises determining an exterior medial axis by evaluating non-differentiable points on the distance field at points outside the rigid bodies.

5. The method of claim 1 , further comprising performing a component classification of an abstract syntax tree of the lumped parameter model to determine resistive components, wherein determining the resistive components comprises analyzing the abstract syntax tree to determine known resistive components.

6. The method of claim 5 , wherein determining the resistive components comprises applying a machine learning regression algorithm to components of the abstract syntax tree to solve a constrained optimization problem.

7. The method of claim 5 , wherein determining the resistive components comprises modeling wires based on connect statements, the resistive components comprising parasitic capacitance between the wires.

8. An apparatus comprising: a memory operable to store: a computer aided design (CAD) model of a system, the system utilizing a diagnosis engine; and
a lumped parameter model of the system used by the diagnosis engine for diagnosis; and
a processor coupled to the memory and operable to perform: analyzing the CAD model to determine components of the CAD model that are also represented in the lumped parameter model;
analyzing the CAD model to determine components of the CAD model that are also represented in the lumped parameter model;
performing a proximity search in the CAD model to find groups of the components that may be affected by resistive or parasitic interactions between individual components in the groups due to geometric adjacency between the individual components without direct contact;
augmenting the lumped parameter model by adding new elements to the lumped parameter model that emulate the resistive or parasitic interactions between the individual components without direct contact in the groups;
and sending the augmented lumped model to the diagnosis engine to perform the diagnosis on the system.

9. The apparatus of claim 8 , wherein the proximity search in the CAD model involves computing a distance field for rigid bodies that represent the individual components, wherein the proximity search comprises determining an exterior medial axis by evaluating non-differentiable points on the distance field at points outside the rigid bodies.

10. The apparatus of claim 8 , wherein the processor is further operable to perform a component classification of an abstract syntax tree of the lumped parameter model to determine resistive components.

11. The apparatus of claim 10 , wherein determining the resistive components comprises applying a machine learning regression algorithm to components of the abstract syntax tree to solve a constrained optimization problem.

12. The apparatus of claim 10 , wherein determining the resistive components comprises modeling wires based on connect statements, the resistive components comprising parasitic capacitance between the wires.

13. A system comprising: a target system comprising instrumentation that monitors performance of the target system;
a diagnosis engine coupled to the instrumentation of the target system and configured to perform a diagnosis on the target system using a lumped parameter model;
and an apparatus comprising a processor configured to perform: analyzing a computer aided design (CAD) model of the system to determine components of the CAD model that are also represented in the lumped parameter model;
perform a proximity search in the CAD model to find groups of components that may be affected by resistive or parasitic interactions between individual components in the groups due to geometric adjacency between the individual components without direct contact;
augment the lumped parameter model by adding new elements to the lumped parameter model that emulate the resistive or parasitic interactions between the individual components without direct contact in the groups;
wherein the diagnosis engine is operable to use the augmented lumped model to perform the diagnosis on the system.

14. The system of claim 13 , wherein performing the diagnosis comprises at least one of real-time monitoring of the system and performing a post-hoc diagnosis after a failure of the system.

15. The system of claim 13 , wherein the proximity search in the CAD model involves computing a distance field for rigid bodies that represent the individual components.

16. The method of claim 1 , wherein the new elements comprise at least one of point mass objects, electronic components, simplified mechanical components, and simplified heat transfer components.

17. The method of claim 1 , wherein the new elements that emulate the resistive or parasitic interactions represent unintended interactions between the individual components that are not included in the lumped parameter model.

18. The method of claim 17 , wherein the individual components are not in direct contact or otherwise connected with one another in the CAD model.

19. The apparatus of claim 8 , wherein the individual components are not in direct contact or otherwise connected with one another in the CAD model and wherein the resistive or parasitic interactions represent unintended interactions that are not included in the lumped parameter model.

20. The system of claim 13 , wherein the individual components are not in direct contact or otherwise connected with one another in the CAD model and wherein the resistive or parasitic interactions represent unintended interactions that are not included in the lumped parameter model.

Assignees

• Palo Alto Res Ct Inc

Inventors

• Matei Ion
• Nelaturi Saigopal
• De Kleer Johan
• Morris Clinton
• Feldman Aleksandar B

Classifications

• G01R19/2513

  Arrangements for monitoring electric power systems, e.g. power lines or loads; Logging · CPC title

• G06F30/10 Primary

  Geometric CAD · CPC title

• G06F30/20 Primary

  Design optimisation, verification or simulation (optimisation, verification or simulation of circuit designs G06F30/30) · CPC title