A Computational Attractive Interval Power Flow Approach With Correlated Uncertain Power Injections

This paper presents a new computational attractive interval power flow (PF) approach with correlated uncertain power injections while maintaining high estimation accuracy. The key idea is to integrate the ellipsoid convex model with the evidence theory. This allows us to effectively account for uncertain power injection correlations and eliminate irrelevant focal elements, yielding significantly improved computational efficiency. Extensive comparisons with other approaches demonstrate its effectiveness under various conditions.