Data-driven probabilistic small signal stability analysis for grid-connected PV systems

Abstract

Probabilistic small signal stability analysis is useful to quantify the effect of uncertain parameters, such as the solar and wind generations, on system stability margin. In the literature, a semi-analytic method based on general polynomial chaos expansion (gPCE) has been proposed for probabilistic small signal stability analysis. The gPCE method can achieve high order accuracy with lower computation burden, but it needs a predefined probability distribution of random parameters, which may not be available in practical engineering applications. To reduce such dependency on the full information of probability distribution, this paper presents a data-driven polynomial chaos expansion (DD-PCE) method for assessing the impact of uncertain solar irradiance on the stability margin of a grid-connected photovoltaic (PV) system. In the proposed method, the system stability margin is characterized in terms of its critical mode damping ratio, and the functional relationship between the critical mode damping ratio and the uncertain solar irradiance is estimated by orthonormal polynomial expansion. The orthonormal polynomial basis is constructed using the moments of solar irradiance, and the coefficients of the orthonormal polynomial expansion are calculated by the collocation points of polynomial basis. The efficiency of the proposed method is validated by comparisons with gPCE method and Monte Carlo method.