Research on Wind Power Probabilistic Characteristics Based on VBEM Weighted Gaussian Mixture Distribution Model

Abstract

In order to quantitatively analyze the probability distribution characteristics of wind power, this paper is based on the Weighted Gaussian Mixture Distribution (WGMD) model fitting the active power probability distribution of wind power field and a single wind turbine at the same time scale, and estimates the parameters of WGMD model with Variational Bayesian Expectation Maximization (VBEM) algorithm, and uses the Sum of Squares due to Error(SSE), the Root Mean Squared Error(RMSE), Adjusted R-Square (ARS) and other indicators to quantify numerically the fitting effect. Finally, the fitting effects of WGMD model, Weibull model and normal distribution are compared by the evaluation indicators, and the validity and feasibility of the proposed method are verified.