Application of Reconstructability Analysis to the NW Power Grid
This talk will focus on preliminary results from Reconstructability Analysis (RA) models, Bayesian Network (BN) models, and standard linear regression to predict dynamics on the bulk electric grid. The best possible prediction models were identified using exploratory RA and BN search algorithms that search the lattice of possible graphs structures to find the best model fit. Preliminary results show that RA and BN both outperform linear regression in overall prediction and prediction in the tails of the distribution, whereas BN marginally outperforms RA overall. In addition to preliminary results, the talk will offer explanations for differences in prediction performance as well as opportunities for extensions of the research.
Marcus Harris is a PSU Systems Science PhD ABD. His research focuses on the theoretical differences between Reconstructability Analysis and Bayesian Networks, both probabilistic graphical modeling methods, working towards unifying the methods into a joint analytical framework. Additionally his research applies these methods to build better prediction models of certain features of the electric power grid. Marcus works at the Bonneville Power Administration and in prior roles applied these methods to predict outcomes on the NW power grid.
Reconstructability analysis, Bayesian statistical decision theory, Electric power systems -- Control, Electric power transmission, System theory, Electric power system stability
Power and Energy | Systems Engineering
Harris, Marcus, "Application of Reconstructability Analysis to the NW Power Grid" (2020). Systems Science Friday Noon Seminar Series. 67.