Multivariate Analysis of Wireless Sensor Network Data for Machine Condition Monitoring

• US 20090119243A1
• Filed: 10/15/2008
• Published: 05/07/2009
• Est. Priority Date: 11/02/2007
• Status: Active Grant

• Current Assignee: Siemens Corp. (Siemens AG)
• Original Assignee: Siemens Corporate Research Incorporated (Siemens AG)
• Inventors: Galler, Stefan, Neubauer, Claus, Yuan, Chao, Sastry, Chellury R.
• Granted Patent: US 8,112,381 B2
• Time in Patent Office: Days
• US Class Current: 706/52
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks

Abstract

Machine condition monitoring on a system utilizes a wireless sensor network to gather data from a large number of sensors. The data is processed using a multivariate statistical model to determine whether the system has deviated from a normal condition. The wireless sensor network permits the acquisition of a large number of distributed data points from plural system modalities, which, in turn, yields enhanced prediction accuracy and a reduction in false alarms.

Claims

1. A method for performing maintenance on a system, the method comprising the steps of:

measuring data using a plurality of wireless sensors in a wireless sensor network, the data representing a plurality of modalities of the system;

transmitting the data via wireless radio signals from the wireless sensors to one or more base stations;

comparing the data to a multivariate statistical model representing normal machine conditions, wherein relationships among sensor data representing plural modalities are considered; and

performing maintenance on the system according to results of the comparing step.

2. The method of claim 1, wherein the step of transmitting the data via wireless radio signals from the wireless sensors to one or more base stations includes:

utilizing ZIGBEE wireless protocol.

3. The method of claim 1, wherein the step of transmitting the data via wireless radio signals from the wireless sensors to one or more base stations includes:

transmitting the data through a mesh network.

4. The method of claim 3, wherein at least one of the wireless sensors also functions as a router.

5. The method of claim 1, further comprising the steps of:

processing the data in the base station using middleware implementing a service-oriented architecture (SOA); and

transmitting the processed data to a remote location for performing the comparing step.

6. The method of claim 1, further comprising the step of creating the multivariate statistical model by:

measuring training data during normal operation of the system using the plurality of wireless sensors in the wireless sensor network;

transmitting the training data via the wireless radio signals from the wireless sensors to one or more base stations; and

creating the multivariate statistical model representing normal machine conditions by representing a normal condition of the system as a region in a multidimensional sensor space.

7. The method of claim 6, further comprising the step of:

wherein the comparing step comprises determining whether the measured data falls within the normal region.

8. The method of claim 6, wherein the region representing a normal condition is determined using a regression-based algorithm.

9. The method of claim 1, wherein the plurality of wireless sensors in the wireless sensor network includes redundant wireless sensors measuring a same modality.

10. The method of claim 1, wherein the plurality of wireless sensors in the wireless sensor network includes wireless sensors measuring a same modality at different locations, and wherein the results of the comparing step includes a location of a fault in the system.

11. The method of claim 1, wherein the step of transmitting the data via wireless radio signals from the wireless sensors to one or more base stations further comprises:

creating a routing topology wherein wireless radio signals are routed from wireless sensors through other wireless sensors to the one or more base stations;

identifying at least one wireless sensor having a degraded signal; and

altering the routing topology to avoid routing wireless radio signals through the one identified wireless sensor.

12. A computer-usable medium having computer readable instructions stored thereon for execution by a processor to perform a method for performing maintenance on a system, the method comprising the steps of:

measuring data using a plurality of wireless sensors in a wireless sensor network, the data representing a plurality of modalities of the system;

transmitting the data via wireless radio signals from the wireless sensors to one or more base stations;

comparing the data to a multivariate statistical model representing normal machine conditions, wherein relationships among sensor data representing plural modalities are considered; and

performing maintenance on the system according to results of the comparing step.

13. The computer-usable medium of claim 12, wherein the step of transmitting the data via wireless radio signals from the wireless sensors to one or more base stations includes:

utilizing ZIGBEE wireless protocol.

14. The computer-usable medium of claim 12, wherein the step of transmitting the data via wireless radio signals from the wireless sensors to one or more base stations includes:

transmitting the data through a mesh network.

15. The computer-usable medium of claim 14, wherein at least one of the wireless sensors also functions as a router.

16. The computer-usable medium of claim 12, wherein the method further comprises the steps of:

processing the data in the base station using middleware implementing a service-oriented architecture (SOA); and

transmitting the processed data to a remote location for performing the comparing step.

17. The computer-usable medium of claim 12, wherein the method further comprises the step of creating the multivariate statistical model by:

measuring training data during normal operation of the system using the plurality of wireless sensors in the wireless sensor network;

transmitting the training data via the wireless radio signals from the wireless sensors to one or more base stations; and

representing a normal condition of the system as a region in a multidimensional sensor space.

18. The computer-usable medium of claim 17, wherein the method further comprises the step of:

wherein the comparing step comprises determining whether the measured data falls within the normal region.

19. The computer-usable medium of claim 17, wherein the region representing a normal condition is determined using a regression-based algorithm.

20. The computer-usable medium of claim 12, wherein the plurality of wireless sensors in the wireless sensor network includes redundant wireless sensors measuring a same modality.

21. The computer-usable medium of claim 12, wherein the plurality of wireless sensors in the wireless sensor network includes wireless sensors measuring a same modality at different locations, and wherein the results of the comparing step includes a location of a fault in the system.

22. The computer-usable medium of claim 12, wherein the step of transmitting the data via wireless radio signals from the wireless sensors to one or more base stations further comprises:

creating a routing topology wherein wireless radio signals are routed from wireless sensors through other wireless sensors to the one or more base stations;

identifying at least one wireless sensor having a degraded signal; and

altering the routing topology to avoid routing wireless radio signals through the one identified wireless sensor.