US-20260127876-A1 - SYSTEM AND METHOD FOR SUPPLYING POWER TO CONSUMERS IN AN ELECTRICAL POWER LINE NETWORK

Abstract

The present disclosure provides a system and method for supplying power to consumers in an electrical power line network. The system provides automated integration and networking logic for electrical assets. Further, the system provides consumer indexing in an electrical power-line network and facilitates an integrated drive and drone sensor data collection system for surveying and mapping service providers and consumers. The system generates virtual ground control points (GCP's) to form models of an electrical power-line network or grid extending from power generation sites to the end consumer sites.

Inventors

• Deekshant Saxena
• Senjuti Sen
• Abhishek Pandey
• Soumyadip MAJUMDAR
• Pratik PATIL
• Ashna KUMAR

Assignees

• HERE GLOBAL B.V.

Dates

Publication Date

20260507

Application Date

20230511

Priority Date

20220511

Claims (20)

1. 1 . A method, comprising: processing sensor data to detect an electrical utility object; detecting one or more sub-objects of the electrical utility object; determining a connection between the one or more sub-objects and another electrical utility object; and providing the connection as an output.
2. 2 . The method of claim 1 , further comprising: performing a topology correction of the connection, the electrical utility object, the one or more sub-objects, the another electrical utility object, or a combination thereof.
3. 3 . A method, comprising: processing sensor data using a machine learning model to generate one or more electrical utility asset detection instances; conflating the one or more electrical utility asset detection instances into one or more conflated candidate detections; performing a particle swarm optimization on the one or more conflated candidate detections to determine a detected electrical utility asset; and providing the detected electrical utility asset as an output.
4. 4 . The method of claim 3 , further comprising: initiating an electrical utility asset class detection on the one or more conflated candidate detections, wherein the particle swarm optimization is further based on the electrical utility asset class detection.
5. 5 . The method of claim 3 , further comprising: generating a training data set comprising one or more electrical utility assets used in a developing country, wherein the machine learning model is trained to the detect the one or more electrical utility assets of the developing using the training data set.
6. 6 . A method, comprising: generating a path for a device to capture sensor data depicting one or more objects of an electricity power delivery network; selecting between a drive device, a drone device, or a combination thereof to complete one or more portions of the path to capture the sensor data; merging the sensor data from the drive device, the drone device, or a combination thereof on completion of the path; and providing the merged sensor data as an output.
7. 7 . The method of claim 6 , wherein the path is generated based on digital map data of a geographic database.
8. 8 . The method of claim 6 , further comprising: selecting the drone device for the one or more portions of the path associated with an extra-high-tension power line, a high-tension power line, or a combination thereof.
9. 9 . The method of claim 6 , wherein the output is used for a consumer indexing, a network creation, or a combination of the electricity power delivery network.
10. 10 . The method of claim 6 , further comprising: selecting a known ground control point (GCP) as a base point of a virtual GCP layer; determining a location of the device; calculating an offset of the location based on the base point to generate a virtual GCP of the virtual GCP layer; determining a subsequent location of the device; and calculating a subsequent offset of the subsequent location based on the inputting the base point and the virtual GCP to the machine learning model to generate a subsequent GCP of the virtual GCP layer.
11. 11 . The method of claim 10 , wherein the machine learning model is a spatio-temporal graph convolutional network (ST-GCN).
12. 12 . The method of claim 10 , further comprising: determining the designated time period based on a target level of positioning accuracy.
13. 13 . The method of claim 10 , wherein the subsequent location is determined using a differential positioning.
14. 14 . The method of claim 13 , wherein the differential positioning comprises real time kinematic (RTK), post processing kinematic (PPK), or a combination thereof
15. 15 . The method of claim 6 , further comprising: processing sensor data collected by the device to detect a service line associated with an electricity power delivery network; determining a first geo-position of an endpoint of the service line; determining a second geo-position of an electrical meter; determining a distance between the first geo-position of the endpoint and the second geo-position of the electrical meter; and establishing a connection between the service line and the electrical meter based on the distance.
16. 16 . The method of claim 15 , further comprising: performing a consumer indexing of the electricity power delivery network based on the connection.
17. 17 . The method of claim 15 , further comprising: initiating a detection of a subsequent service line, a subsequent electrical meter, or a combination thereof based on determining that the distance is greater than a threshold distance, wherein the establishing of the connection is based on the subsequent service line, the subsequent electrical meter, or a combination thereof.
18. 18 . The method of claim 15 , further comprising: evaluating a business heuristic with respect to the service line, the electrical meter, a consumer associated with the electrical meter or a combination thereof, wherein the establishing of the connection is further based on the business heuristic.
19. 19 . The method of claim 15 , wherein the processing of the sensor data comprises using a machine learning feature detector to classify the service line into a service line type, and wherein the establishing of the connection is based on the service line type.
20. 20 . The method of claim 19 , wherein the service line type includes a low-tension line, a high-tension line, an extra high-tension line, or a combination thereof.

Description

BACKGROUND Historically, electricity providers have used networks of overhead power lines to deliver electricity to customers/consumers. In many cases, the creation of such electrical networks or power grids, particularly in developing countries where building growth rates can be high, establishing connections to the networks or power grids can be ad hoc and undocumented. Accordingly, electricity providers and/or related mapping service providers face significant technical challenges with respect to automatically mapping electricity delivery networks and/or consumer connections to the networks (e.g., a process referred to as consumer indexing), and documenting assets of the electrical grid. Therefore, there is a need for advances in technologies for mapping power lines and customer connection points (e.g., electrical meters) of an electrical power delivery network. SUMMARY In an aspect, a method may include processing sensor data to detect an electrical utility object. The method may include detecting one or more sub-objects of the electrical utility object. The method may include determining a connection between the one or more sub-objects and another electrical utility object and providing the connection as an output. In an embodiment, the method may include, performing a topology correction of the connection, the electrical utility object, the one or more sub-objects, the another electrical utility object, or a combination thereof. In an aspect, a method may include, processing sensor data using a machine learning model to generate one or more electrical utility asset detection instances. The method may include conflating the one or more electrical utility asset detection instances into one or more conflated candidate detections. The method may include performing a particle swarm optimization on the one or more conflated candidate detections to determine a detected electrical utility asset and may include providing the detected electrical utility asset as an output. In an embodiment, the method may include initiating an electrical utility asset class detection on the one or more conflated candidate detections. The particle swarm optimization may be further based on the electrical utility asset class detection. In an embodiment, the method may include, generating a training data set comprising one or more electrical utility assets used in a developing country. The machine learning model may be trained to the detect the one or more electrical utility assets of the developing using the training data set. In an aspect, a method may include, generating a path for a device to capture sensor data depicting one or more objects of an electricity power delivery network. The method may include selecting between a drive device, a drone device, or a combination thereof to complete one or more portions of the path to capture the sensor data. The method may include merging the sensor data from the drive device, the drone device, or a combination thereof on completion of the path. The method may include providing the merged sensor data as an output. In an embodiment, the path may be generated based on digital map data of a geographic database. In an embodiment, the method may include, selecting the drone device for the one or more portions of the path associated with an extra-high-tension power line, a high-tension power line, or a combination thereof. In an embodiment, the output may be used for a consumer indexing, a network creation, or a combination of the electricity power delivery network. In an embodiment, the method may include, selecting a known ground control point (GCP) as a base point of a virtual GCP layer. The method may include determining a location of the device. The method may include calculating an offset of the location based on the base point to generate a virtual GCP of the virtual GCP layer. The method may include determining a subsequent location of the device. The method may include calculating a subsequent offset of the subsequent location based on the inputting the base point and the virtual GCP to the machine learning model to generate a subsequent GCP of the virtual GCP layer. In an embodiment, the machine learning model may be a spatio-temporal graph convolutional network (ST-GCN). In an embodiment, the method may include, determining the designated time period based on a target level of positioning accuracy. In an embodiment, the subsequent location may be determined using a differential positioning. In an embodiment, the differential positioning may include real time kinematic (RTK), post processing kinematic (PPK), or a combination thereof. In an embodiment, the method may include processing sensor data collected by the device to detect a service line associated with an electricity power delivery network. The method may include determining a first geo-position of an endpoint of the service line. The method may include determining a second geo-position of an electrical meter. The method m