Self-learning simulation environments

Patent metadata

Field	Value
Publication number	US-2016259869-A1
Application number	US-201514636133-A
Country	US
Kind code	A1
Filing date	Mar 2, 2015
Priority date	Mar 2, 2015
Publication date	Sep 8, 2016
Grant date	—

Abstract

Systems and methods may include receiving performance data of a system. The systems and methods may include generating a virtual representation of the system based on the performance data. The systems and methods may include simulating interactions between the virtual representation and an environment based on the performance data. The systems and methods may include generating an alert indicating an occurrence of an anomalous event based on the simulated interactions. The systems and methods may include determining that the alert was a false positive indicator of the occurrence of the anomalous event. The systems and methods may include, in response to determining that the alert was a false positive indicator, determining differences between a behavior of the system and the simulated interactions. The systems and methods may include determining an adjustment to the virtual representation to mitigate the differences between the behavior of the system and the simulated interactions.

First claim

What is claimed is:

1 . A method comprising:
receiving performance data of a physical system; generating a virtual representation of the physical system based on the performance data of the physical system; simulating interactions between the virtual representation and an environment of the physical system based on the performance data of the physical system; generating an alert indicating an occurrence of an anomalous event in the physical system based on the simulated interactions; determining that the alert was a false positive indicator of the occurrence of the anomalous event in the physical system; in response to determining that the alert was a false positive indicator of the occurrence of the anomalous event in the physical system, determining differences between a behavior of the physical system and the simulated interactions between the virtual representation and the environment; determining an adjustment to the virtual representation of the physical system to mitigate the differences between the behavior of the physical system and the simulated interactions; and modifying the virtual representation of the physical system in accordance with the determined adjustment.

2 . The method of claim 1 , wherein generating the alert indicating the occurrence of the anomalous event in the physical system comprises:
determining whether the simulated interactions include a simulation of the anomalous event; and in response to determining that the simulated interactions include the simulation of the anomalous event, generating the alert.

3 . The method of claim 2 , wherein determining the adjustment to the virtual representation of the physical system comprises:
identifying parameters of the performance data of the physical system that were not used to generate the virtual representation; generating another virtual representation of the physical system based on one or more of the parameters of the performance data of the physical system that were not used to generate the virtual representation; simulating other interactions between the other virtual representation and the environment; determining whether the other simulated interactions include the simulation of the anomalous event; and in response to determining that the other simulated interactions do not include the simulation of the anomalous event, determining that the adjustment to the virtual representation of the physical system includes modifying the virtual representation to be based on the one or more of the parameters.

4 . The method of claim 1 , further comprising:
receiving performance data of a plurality of physical systems, the plurality of physical systems comprising the physical system; and recording a behavior history of the plurality of physical systems by storing the performance data of the plurality of physical systems, wherein generating the virtual representation of the physical system comprises generating the virtual representation of the physical system based on the behavior history of the plurality of physical systems, and wherein simulating the interactions between the virtual representation and the environment comprises simulating the interactions between the virtual representation and the environment based on the behavior history of the plurality of physical systems.

5 . The method of claim 1 , wherein the differences between the behavior of the physical system and the simulated interactions include different responses by the physical system and the virtual representation to the same performance data of the physical system.

6 . The method of claim 1 ,
wherein the physical system comprises a plurality of physical components, wherein the performance data of the physical system comprises performance data for each physical component of the plurality of physical components, wherein the virtual representation of the physical system comprises a plurality of virtual components, each virtual component of the plurality of virtual components corresponding to a respective physical component of the plurality of physical components, and wherein simulating the interactions between the virtual representation and the environment comprises:
simulating future interactions between each virtual component of the plurality of virtual components and other virtual components of the plurality of virtual components; and
simulating future interactions between each virtual component of the plurality of virtual components and a virtual representation of the environment.

7 . The method of claim 6 ,
wherein the anomalous event includes receiving anomalous performance data of a physical component of the plurality of physical components that indicates an anomalous behavior, and wherein generating the alert comprises generating the alert in response to simulating a future interaction between a virtual component corresponding to the physical component that matches the anomalous behavior.

8 . A system comprising:
a processing system configured to:
receive performance data of a physical system;
generate a virtual representation of the physical system based on the performance data of the physical system;
simulate interactions between the virtual representation and an environment of the physical system based on the performance data of the physical system;
generate an alert indicating an occurrence of an anomalous event in the physical system based on the simulated interactions;
determine that the alert was a false positive indicator of the occurrence of the anomalous event in the physical system;
in response to determining that the alert was a false positive indicator of the occurrence of the anomalous event in the physical system, determine differences between a behavior of the physical system and the simulated interactions between the virtual representation and the environment;
determine an adjustment to the virtual representation of the physical system to mitigate the differences between the behavior of the physical system and the simulated interactions; and
modify the virtual representation of the physical system in accordance with the determined adjustment.

9 . The system according to claim 8 , wherein, when generating the alert indicating the occurrence of the anomalous event in the physical system, the processing system is configured to:
determine whether the simulated interactions include a simulation of the anomalous event; and in response to determining that the simulated interactions include the simulation of the anomalous event, generate the alert.

10 . The system according to claim 9 , wherein, when determining the adjustment to the virtual representation of the physical system, the processing system is configured to:
identify parameters of the performance data of the physical system that were not used to generate the virtual representation; generate another virtual representation of the physical system based on one or more of the parameters of the performance data of the physical system that were not used to generate the virtual representation; simulate other interactions between the other virtual representation and the environment; determine whether the other simulated interactions include the simulation of the anomalous event; and in response to determining that the other simulated interactions do not include the simulation of the anomalous event, determine that the adjustment to the virtual representation of the physical system includes modifying the virtual representation to be based on the one or more of the parameters.

11 . The system according to claim 8 , wherein the processing system is further configured to:
receive performance data of a plurality of p

Assignees

• Parikh Prashant
• Greenspan Steven L
• Danielson Debra J
• Mankovskii Serge
• Velez-Rojas Maria C
• Reno James D
• Ca Inc

Inventors

• Parikh Prashant
• Greenspan Steven L
• Danielson Debra J
• Mankovskii Serge
• Velez-Rojas Maria C
• Reno James D

Classifications

• G05B23/0245

  based on a qualitative model, e.g. rule based; if-then decisions · CPC title

• G06F30/20 Primary

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

• G05B23/0243

  model based detection method, e.g. first-principles knowledge model · CPC title

• G05B23/0262

  Confirmation of fault detection, e.g. extra checks to confirm that a failure has indeed occurred · CPC title

• G05B17/02 Primary

  electric · CPC title