Abstract: In this talk, we will present a model-free screening method to statistically rank the criticality of power line combinations to facilitate the development of an efficient hardening plan within budget to improve power grid weather resilience. The binary status of the power lines, connected/disconnected, in the power grid violates the normal or elliptical contoured distribution assumption in many existing screening methods. The proposed method skillfully uses the symmetric marginal relationship between power loss and power line status, and switches their roles in the marginal model to obtain a well-known statistics as the ranking utility. This role switching in the marginal models also facilitates the regression on a set of B-spline basis functions and consequently enables the proposed method to accommodate flexible marginal association other than linear. The top ranked power line combinations will guide a simulation-based optimization algorithm to efficiently search for the optimal plan with budget constraints to harden a subset of power lines for improved weather resilience. The proposed method is among the first to perform dimension reduction with all binary covariates and shows better numerical performance than existing methods in our preliminary simulation study.

Bio: Jie Xu is an associate professor of systems engineering and operations research at George Mason University. He received his Ph.D. degree from Northwestern. His research interests are the modeling, simulation, and optimization of complex stochastic systems.

Attend via BlueJeans