Extreme Scenario Extraction Method for Power System Flexibility Shortage Based on Feature Selection and Isolated Forest Algorithm

Against the backdrop of the “double carbon” target, large-scale clean power feeds into the receiving system, and the contradiction of insufficient flexibility resources is becoming more and more prominent. Existing schemes are characterized by subjectivization, homogenization of dimensions, and inappropriate response to data demand, which easily lead to a series of problems such as low computational efficiency. In view of this, this paper proposes an extreme scenario extraction method based on feature selection and isolated forest algorithm for power system flexibility shortage, which solves the common subjectivity problem of extreme selection in the existing schemes, and improves the computational efficiency through the feature selection module to ensure that the extreme scenarios are in line with the demand. The extreme nature of the selected scenarios and the fit of the reconstructed scenarios to the original data are also verified through case analysis.