Partial-Dimensional Correlation-Aided Convex-Hull Uncertainty Set for Robust Unit Commitment

Correlations help narrow the uncertainty region in robust unit commitment (RUC) of power systems for economic improvement, yet in high-dimensional cases, state-of-the-art full-dimensional correlation (FDC) based uncertainty set methods suffer from either conservativeness or computational burden. This article proposes the novel partial-dimensional correlation (PDC) aided convex-hull uncertainty set (CHUS) for RUC. The PDC-aided framework is established for the first time to utilize the accurate and accessible PDC instead of the assumed but inaccessible FDC, which provides a general formula that covers both the traditional correlation-ignored and the emerging FDC-based methods. The diamond-cut CHUS of correlation data is developed to approach the compact CHUS to reduce conservativeness under an acceptable complexity. The customized scenario-parallel algorithm is proposed for efficient calculation, which combines the extreme scenario-based constraint rebuild and the parallel computing-enabled column-and-constraint generation. Case studies demonstrate the effectiveness of the proposed method in enhancing both economic and computational efficiency.