Co-optimized Unit Commitment Based On Scenario Generation And Electric Vehicles

Abstract

The high penetration of renewable energy brings unprecedented randomness and uncertainty to the power system, which makes the traditional unit commitment (UC) model uneconomical and unreliable. In this paper, we propose a novel wind power scenario generation method based on bidirectional conditional generative adversarial network (BCGAN) which is an improvement based on generative adversarial networks (GAN). Furthermore, we propose a unit commitment model which collaboratively takes the wind power and charging electric vehicles (EVs) into consideration. Finally, several experiments were carried out on a modified IEEE-RTS 96 system to verify the effectiveness of this research.