Dual stochastic multi-stage robust optimization of integrated energy system considering long medium and short-term uncertainties

The multi-stage robust optimization of integrated energy system (IES) exhibits strong conservativeness, and the characterization of uncertainties across multiple time scales presents significant challenges. This complexity hinders the achievement of optimal scheduling strategies for “configuration + operation” throughout the entire planning cycle of IES, thereby posing substantial obstacles to the realization of efficient multi-energy supply. To address these challenges, this study proposes a dual stochastic multi-stage robust planning model that incorporates long, medium, and short-term uncertainties in IES. This model essentially consists of a five-level robust planning framework. In the first level, the optimization objective is to derive the optimal configuration of multi-energy devices within the IES. In the second and third levels, to manage long- and medium-term uncertainties, a scenario-probability-driven dual distributionally robust optimization approach is employed to identify the worst-case probabilities for different seasonal-typical day scenarios. In the fourth and fifth levels, multi-stage robust optimization is utilized to determine the worst-case scenarios for source-load-price interactions, addressing short-term uncertainties. Furthermore, to effectively solve this model, a parallelizable AOP-Looped C&CG algorithm is proposed. The efficacy of the proposed model and algorithm is validated through case studies.