Distributionally robust optimization of electric–thermal–hydrogen integrated energy system considering source–load uncertainty

With the increasing penetration of renewable energy and the growing energy demand from users, the scheduling of integrated energy system (IES) faces significant challenges. A data-driven distributionally robust optimization (DRO) approach is proposed to solve the scheduling problem under source–load uncertainty. Firstly, conditional generative adversarial networks (CGAN) are utilized to generate scenarios for wind and solar power outputs as well as electrical and thermal loads. The K-medoids clustering algorithm is then used to obtain typical scenarios. Secondly, a comprehensive norm composed of 1-norm and ∞-norm is applied to constrain the typical scenarios to construct an uncertainty set. Finally, a two-stage DRO model of electric–thermal–hydrogen integrated energy system (ETH-IES) is established. The simulation results demonstrate that the proposed method effectively improves system economy, with a 2.1% reduction in operating cost compared to traditional robust optimization, while ensuring efficient model solving.