Probabilistic Power Flow of AC/DC Hybrid Grids With Addressing Boundary Issue of Correlated Uncertainty Sources
Sui Peng, Xingyu Lin, Junjie Tang, Kaigui Xie, Ferdinanda Ponci, Antonello Monti, Wenyuan Li
Abstract
In practical AC/DC hybrid grids, the uncertainty sources such as wind speeds and loads are massive and bounded intrinsically, the latter feature of which is neglected in most probabilistic power flow (PPF) analyses unfortunately. In this paper, a scaled unscented transformation (SUT)-based PPF on the large AC/DC hybrid grids, in particular for addressing the boundary issue of these high-dimensional random input variables under Pearson correlation, is proposed and investigated. The empirical formula used for determining the scaling parameter of the SUT method is designed to obtain the sample points, with a purpose of capturing sufficient probability information from the preliminary standard normal distributions. Boundary inverse transformation (BIT) is developed to transform these samples into the original probability space via handling each dimensional boundary value, and to guarantee that all the sample points of SUT fall into a reasonable, physical interval. The effectiveness and advantage of the proposed method are validated by using a set of test results on the modified IEEE 1354-bus (PEGASE) system.