Modelling Wind Power Generation in Probabilistic Power Systems Analysis: A Case Study with Non-Gaussian Correlated Random Variables

The rapid expansion of wind power generation demands an update to the computational tools used for decision-making in the operation and planning of electrical power systems. This paper presents a methodology for modeling spatial correlations among wind speeds in probabilistic power system analysis. The approach employs the Nataf transformation and Monte Carlo Simulation. Additionally, the stochastic connection between wind speed and wind power is modeled using a conditional quantile copula kernel density estimator. The methodology is demonstrated through case studies involving the modified IEEE 24-Bus Test System and the Brazilian Northeast Power System, a real-world system with 965 buses.