| US 11,735,912 B2 | ||
| Systems and methods for ai-assisted electrical power grid fault analysis | ||
| Aaron Acton, Pittsburgh, PA (US); Peter Glover, Belfast (GB); Mark Diamond, Bellaghy (GB); Peter Brett, Belfast (GB); David Cole, Ware (GB); and Paul Donegan, Dunoon (GB) | ||
| Assigned to QUALITROL COMPANY LLC, Fairport, NY (US) | ||
| Filed by Qualitrol Company LLC, Fairport, NY (US) | ||
| Filed on Jul. 27, 2022, as Appl. No. 17/875,280. | ||
| Claims priority of provisional application 63/226,123, filed on Jul. 27, 2021. | ||
| Prior Publication US 2023/0035930 A1, Feb. 2, 2023 | ||
| Int. Cl. H02J 3/00 (2006.01); G05B 23/02 (2006.01); G01R 31/08 (2020.01) | ||
| CPC H02J 3/0012 (2020.01) [G01R 31/086 (2013.01); G01R 31/088 (2013.01); G05B 23/0245 (2013.01); G05B 23/0278 (2013.01)] | 21 Claims |
|
1. A system for use in an electrical power grid, the system comprising:
at least one processor; and at least one memory communicatively coupled to the at least one processor, the at least one memory having computer-executable instructions stored thereon that, when executed by the at least one processor, cause the system to:
electronically receive an indication that a first fault has occurred in the electrical power grid, the indication including a first fault record associated with the first fault, the first fault record being generated by a first substation of a plurality of substations;
electronically identify a plurality of additional fault records associated with the first fault, each respective fault record in the plurality of additional fault records being generated by a respective substation in the electrical power grid other than the first substation;
electronically obtain a first prediction of a cause of the first fault by applying the first fault record and the plurality of additional fault records to a trained machine learning model that has been trained to predict the cause of a fault based on historical fault records;
electronically obtain a second prediction of the cause of the first fault by applying the first fault record to a rules-based model;
electronically obtain a final prediction of the cause of the first fault based on the first prediction and the second prediction; and
based on the final prediction of the cause of the first fault, electronically cause the predicted cause of the first fault to be remedied.
|