Dynamic Reliability Equivalence for Smart Distribution Grids Considering Wind Power Uncertainties

Reliability evaluation of smart distribution grids is computationally demanding when wind power uncertainty and numerous component states are considered. This paper proposes a dynamic reliability‐equivalence method whose equivalent network varies with wind power uncertainty, load point, and reliability index. A modified Sobol’ sensitivity analysis quantifies each component’s contribution and retains only influential components. Probability integral transformation handles nonuniform wind‐power variables, while sampled reliability indices represent the implicit reliability function. These treatments allow variance‐based component screening when conventional Sobol’ analysis cannot be applied directly. Case studies on a modified IEEE 123‐node system show that the equivalent regions change with wind power uncertainty and evaluation objectives. The proposed method reduces reliability‐evaluation time by more than 80% while limiting the maximum error to 3.11% under the baseline setting. These results support the proposed method as an efficient alternative to full‐network reliability evaluation for distribution grids with wind power uncertainty.