Quantitative Evaluations of Uncertainties in Multivariate Operations of Microgrids

Abstract: Microgrid (MG) provides a promising solution to managing various distributed energy resources and loads within defined electrical boundaries. However, variable renewable energy generation and loads as MG inputs would inject complex uncertainties into MG operations, which have significant impacts on MG outputs (e.g., system frequency, bus voltages, and power flows). It is imperative to evaluate the impacts of pertinent uncertainties on MG multivariate outputs for securing MG operations. In this paper, a multivariate global sensitivity analysis (MV-GSA) with three MV-indices is proposed for quantifying the impacts of correlated renewable energy generation on MG multivariate outputs. The proposed MV-GSA overcomes the shortcomings of conventional univariate GSA (UV-GSA) in which the correlations among multivariate outputs are neglected. Also, analytical representations of indices for the proposed MV-GSA are derived by which the computation burdens are significantly reduced as compared with those of using the Monte Carlo simulation. The effectiveness of the proposed method is verified by using the modified IEEE 33-bus and IEEE 123-bus MG systems.