PI: Arman Ahmed

Co-PI(s): A. Srivastava

Sponsor: Battelle – Pacific Northwest National Laboratory – DGRP

Award Amount: $103,562

Project Period: 1/2021 – 9/2021

Summary: This research will develop a decentralized scalable multi agent federated prediction/inference event analysis approach, answering following questions:

  1. Given different power grid data sets and predictive models, how to enable profiling of raw data sets and predictive models?
  2. How to enable and improve/optimize event analysis by making use of individual prediction models for event detection, localization and classification?
  3. How to map observed data to decide what’s the best prediction model for event analysis?
  4. How to develop a federated inference/prediction approach for decentralized power grid analysis in scenarios involving cascading failures etc.?