Stochastic Uncertainty Propagation in Power System Dynamics using Measure-valued Proximal Recursions

We present a proximal algorithm that performs variational recursion on the space of joint probability measures to propagate stochastic uncertainties in power system dynamics over high dimensional state space. The proposed takes advantage exact nonlinearity structures trajectory-level networked systems, and is nonparametric. Lifting allows us design scalable obviates gridding underlying which computationally prohibitive. implements generalized infinite gradient flow, evolves probability-weighted scattered point clouds. clarify theoretical nuances algorithmic details specific nonlinearities, provide illustrative numerical examples.