Comparing unscented transformation and point estimate method for probabilistic power flow computation

Abstract

PurposeUnscented transformation (UT) and point estimate method (PEM) are two efficient algorithms for probabilistic power flow (PPF) computation. This paper aims to show the relevance between UT and PEM and to derive a rule to determine the accuracy controlling parameters for UT method.Design/methodology/approachThe authors derive the underlying sampling strategies of UT and PEM and check them in different probability spaces, where quadrature nodes are selected.FindingsGauss-type quadrature rule can be used to determine the accuracy controlling parameters of UT. If UT method and PEM select quadrature nodes in two probability spaces related by a linear transform, these two algorithms are equivalent.Originality/valueIt shows that UT method can be conveniently extended to (km + 1) scheme (k = 4; 6; : : : ) by Gauss-type quadrature rule.