Multi-parameter Uncertainty Analysis in Power System Dynamic Simulation: A New Solution Based on the Stochastic Response Surface Method and Trajectory Sensitivity Method

摘要

Power system simulations are greatly affected by uncertainties arising from models and parameters. The conventional methods of uncertainty analysis involve many thousand rounds of computation, which are both time and effort demanding. Based on the stochastic response surface method and the trajectory sensitivity method, this article proposes a new solution to reduce the number of uncertain inputs and to expedite the calculation of uncertainty. The trajectory sensitivity method is first applied to narrow down a set of parameters with high sensitivity, and the stochastic response surface method is then applied on this set to analyze the uncertainties, which can greatly speed up the computation process. The uncertainties estimated this way are close to those by 10~5 Monte Carlo simulations in case studies of the NE power grid of China.