Learning the Evolution of Correlated Stochastic Power System Dynamics

This addresses uncertainty quantification in power systems, which is a domain-specific problem, and appears incremental as it builds on existing methods for stochastic dynamics.

The authors tackled the problem of quantifying uncertainty in power system dynamics with spatiotemporally correlated stochastic forcing by proposing a machine learning technique that learns one-dimensional linear partial differential equations for probability density functions, which is suitable for high-dimensional systems and helps alleviate the curse of dimensionality.