Xiaozhe Wang – Assistant Professor, Department of Electrical and Computer Engineering, McGill University, Canada
The ever-increasing integration of renewable energy sources and new forms of load demand introduces a growing uncertainty level to power systems, which greatly affect various security properties of a system. In this talk, I will present some recent works of my group in utilizing polynomial chaos expansion (PCE)-based methods in power system probabilistic security assessments including probabilistic power flow solutions, available transfer capability assessments, and economic dispatch. In contrast to Monte Carlo-based simulations that require a large number of scenarios and model evaluations, the polynomial chaos expansion method can build a surrogate model for assessing the model response (e.g., probabilistic power flow solution) from a small number of scenarios and model evaluations, which thus saves huge computational efforts. I will also introduce the efforts to relax the assumption of knowing marginal distributions of random variables required in PCE. Insights for decision-making to reduce the negative impacts of uncertainty on power system security will also be discussed.