Systems and methods for classifying in-situ sensor response data patterns representative of grid pathology severity

• US 10,698,371 B2
• Filed: 07/14/2015
• Issued: 06/30/2020
• Est. Priority Date: 07/17/2014
• Status: Active Grant

• Current Assignee: 3M Innovative Properties Company (3M Company)
• Original Assignee: 3M Innovative Properties Company (3M Company)
• Inventors: Brooks, Brian E., Lu, Yang, Tio, Andrew T., Benoit, Gilles J.
• Primary Examiner(s): Shah, Kamini S
• Assistant Examiner(s): Ayoub, Faraj
• Application Number:
  US15/324,763
• Publication Number:
  US 20170205781A1
• Time in Patent Office: 1,813 Days
• Field of Search: 703 18
• CPC Class Codes: G05B 13/045 using a perturbation signal
• CPC Class Codes: G06Q 50/06 Energy or water supply

Abstract

The present invention is directed towards methods and systems for characterizing sensors and developing classifiers for sensor responses on a utility grid. Experiments are conducted by selectively varying utility grid parameters and observing the responses of utility grid to the variation. Methods and systems of this invention then associate the particular responses of the utility grid sensors with specific variations in the grid parameters, based on knowledge of the areas of space and periods of time where the variation in grid parameters may affect the sensor response. This associated data is then used to updating a model of grid response.

Claims

1. A computer-implemented method for determining the effects of signal injections on sensor responses, comprising the following steps executed by a processor:

implementing a first signal injection on a utility grid, wherein the first signal injection is a first change in a state of a grid control;

implementing a second signal injection on the utility grid, wherein the second signal injection is a second change in the state of the grid control;

receiving first sensor data from a first sensor;

receiving second sensor data from a second sensor;

receiving a first spatial reach for the first signal injection, wherein the first spatial reach is a first area over which sensors will show a first response to the first signal injection;

receiving a first temporal reach for the first signal injection, wherein the first temporal reach is a first time period over which the first response will be observed;

receiving a second spatial reach for the second signal injection, wherein the second spatial reach is a second area over which sensors will show a second response to the second signal injection;

receiving a second temporal reach for the second signal injection, wherein the second temporal reach is a second time period over which the second response will be observed;

coordinating the first and second signal injections such that the first spatial reach overlaps the second spatial reach, and the first temporal reach overlaps the second temporal reach;

associating, using the processor, the first sensor data within the first spatial reach and within the first temporal reach with the first signal injection;

associating, using the processor, the second sensor data within the second spatial reach and within the second temporal reach with the second signal injection; and

updating a model of the first and second sensor response based on the associated first and second sensor data and the first and second signal injections, wherein the model maps the first and second sensor data to particular states, conditions, or events occurring at the first and second sensors in order to interpret outputs from the first and second sensors during operation of the utility grid.

2. The method of claim 1, wherein updating the model of the first and second sensor response comprises:

predicting the first and second sensor response to the first and second signal injections using models of sensor response;

comparing the associated first and second sensor data to the predicted first and second sensor response; and

falsifying the models of first and second sensor response where the predicted first and second sensor response deviates from the associated first and second sensor data.

3. The method of claim 1, wherein updating the model of the first and second sensor response comprises:

updating a database of sensor response to the first and second signal injections with the associated first and second sensor data; and

computing, based on the updated database of sensor response to the first and second signal injections, relationship between the first and second sensor response and a grid parameter affected by the first and second signal injections.

4. The method of claim 1, wherein updating the model of the first and second sensor response includes removing data from the model of the first and second sensor response based on age of the data and an inclusion period.

5. The method of claim 1, wherein the model of the first and second response is an alert threshold.

6. The method of claim 1, wherein the model of the first and second response is a classifier.

7. The method of claim 1, wherein the first or second sensor data are electrical measurements.

8. The method of claim 1, wherein the first or second sensor data are methane concentration measurements.

9. The method of claim 1, wherein the first or second spatial reach is computed based on a database of grid response to previous signal injections.

10. The method of claim 1, wherein the first or second temporal reach is computed based on a database of grid response to previous signal injections.

11. A system for associating sensor data with particular signal injections into a utility grid, comprising:

a plurality of utility grid controls;

a plurality of sensors;

a sensor data memory configured to receive and store sensor data;

a spatial reach memory configured to receive and store spatial reaches for the signal injections, wherein the spatial reaches are areas over which the plurality of sensors will show responses to the signal injections;

a temporal reach memory configured to receive and store temporal reaches for the signal injections, wherein the temporal reaches are time periods over which the responses will be observed;

an association processor configured to associate data with the signal injections based on the spatial and temporal reaches of the signal injections, wherein the signal injections are changes in states of the plurality of utility grid controls;

a memory configured to store models of the plurality of sensors response; and

a processor configured to update models of the plurality of sensors response using associated sensor data, wherein the models map the sensor data to particular states, conditions, or events occurring at the plurality of sensors in order to interpret outputs from the plurality of sensors during operation of the utility grid,

wherein the signal injections are coordinated such that the spatial reaches overlap and the temporal reaches overlap.

12. The system of claim 11, further comprising:

the memory configured to store a database of previous grid responses to the signal injections;

a spatial reach processor configured to compute the spatial reaches of the signal injections based on the database of previous grid responses to the signal injections; and

a temporal reach processor configured to compute the temporal reaches of the signal injections based on the database of previous grid responses to the signal injections.

13. The system of claim 11, wherein the processor is configured to update models of the plurality of sensors response by comparing the outputs of models of the plurality of sensors response to data associated with the signal injections and falsifying at least some of the models of the plurality of sensors response.

14. The system of claim 11, wherein the plurality of sensors are located along the utility grid.

15. The system of claim 11, wherein the association processor is configured to compare time and location metadata of the sensor data with the spatial and temporal reaches of the signal injections.