Automated control of a power network using metadata and automated creation of predictive process models

• US 8,447,707 B2
• Filed: 06/17/2010
• Issued: 05/21/2013
• Est. Priority Date: 06/19/2009
• Status: Active Grant

• Current Assignee: S&C Electric Company
• Original Assignee: Intelligent Power & Energy Research Corporation (IPERC)
• Inventors: Massie, Darrell D., Miller, Michael A., Curtiss, Peter S.
• Primary Examiner(s): Holmes, Michael B
• Application Number:
  US12/817,651
• Publication Number:
  US 20110071970A1
• Time in Patent Office: 1,069 Days
• Field of Search: 706/12
• US Class Current: 706/12
• CPC Class Codes: H02J 3/00 Circuit arrangements for ac mains or ac distribution networks
• CPC Class Codes: H02J 3/38 Arrangements for parallely feeding a single network by two or more generators, converters or transformers
• CPC Class Codes: H02J 3/388 Islanding, i.e. disconnection of local power supply from the network

Abstract

Automated control of a power network is provided by: providing multiple intelligent power controllers (IPCs) associated with multiple components of the power network, each IPC being associated with a different component; obtaining at least one raw data stream representative of at least one operational aspect of at least one component of the multiple components; and automatically associating, by at least one intelligent power controller associated with at least one component, metadata with the at least one raw data stream to produce at least one self-identifying data stream. The associated metadata describes one or more characteristics of the at least one raw data stream, and the at least one self-identifying data stream facilitates automated creation of predictive process models to assist in automated control of the power network by an IPC manager of the power network.

Claims

1. A method of facilitating automated control of a power network, the method comprising:

providing multiple intelligent power controllers (IPCs) associated with multiple components of the power network, each IPC being associated with a different component of the multiple components of the power network;

obtaining at least one raw data stream from at least one data source of at least one component of the multiple components, the at least one raw data stream representative of at least one operational aspect of the at least one component of the multiple components; and

automatically associating, by the at least one IPC associated with the at least one component, metadata with the at least one raw data stream to produce at least one self-identifying data stream, the metadata describing one or more characteristics of the at least one raw data stream, and the at least one self-identifying data stream facilitating automated control of the power network.

2. The method of claim 1, wherein the obtaining comprises obtaining multiple raw data streams representative of multiple operational aspects of the at least one component of the multiple components and the automatically associating comprises automatically associating, by the at least one IPC associated with the at least one component, different metadata with the multiple raw data streams to produce multiple different self-identifying data streams which facilitate automated control of the power network.

3. The method of claim 1, further comprising referencing the at least one self-identifying data stream of the at least one IPC by another mc, of the multiple IPCs, wherein the at least one self-identifying data stream referenced by the another IPC facilitates automated control of the power network.

4. The method of claim 3, further comprising automatically selecting the another IPC of the multiple IPCs to be an IPC manager which facilitates automated control of the power network.

5. The method of claim 3, wherein the another IPC comprises at least one model template which facilitates generating at least one predictive process model for use in automated control in the power network, the at least one predictive process model facilitating automated evaluation of potential control actions to control operation of one or more components of the multiple components of the power network.

6. The method of claim 5, wherein the generating comprises generating multiple predictive process models, including non-data-based predictive process models and data-based predictive process models, and wherein the at least one self-identifying data stream is employed by the another IPC in generating at least one data-based predictive process model.

7. The method of claim 6, wherein the another IPC;

relates at least two predictive process models of the multiple predictive process models to one another to produce a logical sequence of predictive process models; and

relates the logical sequence of predictive process models to another predictive process model of the multiple predictive process models.

8. The method of claim 5, further comprising employing the at least one predictive process model by one or more IPCs of the power network in controlling one or more control points associated with one or more of the multiple components, and wherein the at least one predictive process model facilitates automated evaluation by providing a prediction of behavior of components of the power network should one or more potential control actions be implemented to control operation of one or more components of the multiple components of the power network, the at least one predictive process model utilizing at least one of a linear regression or a neural network in mapping a predicted component behavior as a function of predictive process model inputs.

9. The method of claim 8, wherein the employing comprises using the at least one predictive process model to produce at least one control signal for controlling one or more control points associated with one or more of the multiple components.

10. The method of claim 1, further comprising automatically selecting one IPC of the multiple IPCs as an IPC manager of the power network and accessing by the IPC manager of the power network the at least one self-identifying data stream to facilitate automated control of the power network.

11. The method of claim 10, further comprising generating, by the IPC manager of the power network, at least one predictive process model for use in automated control of one or more control points associated with one or more components of the multiple components, the at least one predictive process model facilitating an evaluation of potential control actions to control the one or more control points associated with the one or more components of the multiple components, and the generating comprising using at least one model template of the IPC manager of the power network and the at least one self-identifying data stream to generate the at least one predictive process model.

12. The method of claim 1, wherein the metadata further describes one or more characteristics of the at least one component of the multiple components.

13. The method of claim 1, wherein the metadata of the at least one self-identifying data stream comprises for each self-identifying data stream at least a data stream name for the corresponding raw data stream.

14. A method of facilitating automated control of a power network, the method comprising:

providing multiple intelligent power controllers (IPCs), each IPC being coupled to a respective component of multiple components of the power network and at least one IPC of the multiple IPCs being operable to;

access at least one raw data stream representative of at least one operational aspect of the respective component of the multiple components of the power network;

automatically associate metadata with the at least one raw data stream to produce at least one self-identifying data stream accessible by other IPCs of the multiple IPCs, the metadata describing one or more characteristics of the at least one raw data stream;

access self-identifying data streams of other IPCs of the multiple IPCs;

facilitate automated control of the power network, responsive to being selected IPC manager of the power network, using the self-identifying data streams to generate at least one predictive process model from at least one model template, the at least one predictive process model facilitating an evaluation of one or more potential control actions to control one or more components of the multiple components of the power network; and

automatically control the power network, responsive to being selected IPC manager of the power network, by employing one or more predictive process models in evaluating the one or more potential control actions, and producing a set of control signals based on the evaluation of the one or more potential control actions, the set of control signals controlling one or more control points associated with one or more components of the multiple components of the power network.

15. The method of claim 14, further comprising automatically selecting, by the multiple IPCs, one IPC of the multiple IPCs to operate as IPC manager of the power network and accessing by the IPC manager at least one self-identifying data stream of at least one other IPC of the multiple IPCs in generating at least one predictive process model for use in automated control of one or more control points associated with one or more components of the multiple components, the at least one predictive process model facilitating an evaluation of potential control actions to control the one or more components of the multiple components, the generating comprising using at least one model template of the IPC manager of the power network and the at least one self-identifying data stream to generate the at least one predictive process model.

16. A system facilitating automated control of a power network, the system comprising:

multiple intelligent power controllers (IPCs) associated with multiple components of the power network, each IPC being associated with a different component of the multiple components of the power network, wherein at least one IPC of the multiple IPCs;

obtains at least one raw data stream from at least one data source of at least one component of the multiple components, the at least one raw data stream representative of at least one operational aspect of the at least one component of the multiple components; and

automatically associates metadata with the at least one raw data stream to produce at least one self-identifying data stream, the metadata describing one or more characteristics of the at least one raw data stream, and the at least one self-identifying data stream facilitating automated control of the power network.

17. The system of claim 16, wherein the at least one IPC obtains multiple raw data streams representative of multiple operational aspects of the at least one component of the multiple components, and wherein the at least one IPC automatically associates different metadata with the multiple raw data streams to produce multiple different self-identifying data streams which facilitate automated control of the power network.

18. The system of claim 16, wherein another IPC of the multiple IPCs references the at least one self-identifying data stream, and wherein the at least one self-identifying data stream referenced by the another IPC facilitates automated control of the power network.

19. The system of claim 16, wherein another IPC of the multiple IPCs is automatically selected as an IPC manager of the power network, and wherein the IPC manager of the power network accesses the at least one self-identifying data stream to facilitate automated control of the power network.

20. The system of claim 19, wherein the IPC manager of the power network generates at least. one predictive process model for use in automated control of one or more control points associated with one or more components of the multiple components, the at least one predictive process model. facilitating an evaluation of potential control actions to control the one or more control points associated with the one or more components of the multiple components, and wherein the IPC manager of the power network uses at least one model template and the at least one self-identifying data stream in generating the at least one predictive process model.

21. A power network comprising:

multiple power components, the multiple power components comprising at least one source component and at least one load component; and

an automated control system for controlling the power network, the automated control system comprising;

multiple intelligent power controllers (IPCs) associated with the multiple power components, each IPC being associated with a different power component of the multiple power components, wherein at least one IPC of the multiple IPCs;

obtains at least one raw data stream from at least one data source of at least one power componet of the multiple power components, the at least one raw data stream representative of at least one operational aspect of the at least one power component of the multiple power components; and

automatically associates metadata with the at least one raw data stream to produce at least one self-identifying data stream, the metadata describing one or more characteristics of the at least one raw data stream, and the at least one self-identifying data stream facilitating automated control of the power network.