Incentive-Based Integrated Demand Response and Distributionally Robust Optimization for Enabling Economic Dispatch of Integrated Energy Systems

Source-load fluctuations have significant impacts on the economic performance of integrated energy system (IES) dispatch. Existing studies mainly focus on alleviating the supply-demand matching pressure using system-side operation optimization and resource configuration. Therefore, two critical issues arise: the first is that the regulation potential of demand-side resources has not been fully exploited, and the second is that coordinated regulation between the system side and the demand side remains insufficient. To address these issues, this paper presents a two-stage optimization framework for IESs considering an incentive-based integrated demand response (IBIDR) model. Firstly, a two-stage optimization framework consisting of day-ahead and intra-day stages is established, which integrates the demand-side IBIDR and the system-side economic dispatch on a global scale. Secondly, in the day-ahead stage, the IBIDR model—considering multi-energy demand coupling and the bidirectional response characteristics of consumers—is presented. A multi-agent proximal policy optimization (MAPPO) algorithm is employed by the model to fully explore the regulation potential of the demand-side resources. Then, the model enables the generation of incentive strategies for the demand-side resources. Finally, in the intra-day stage, a distributionally robust optimization (DRO) model is further presented to address the source-load uncertainties, in which a wind power generation (WPG) ambiguity set constructed using imprecise Dirichlet theory (IDT) is adopted to represent the WPG uncertainty. The experimental results demonstrate that the presented framework can enhance the regulation capability and economic performance of the IES under source-load fluctuations. The IBIDR model can accurately characterize the multi-energy response characteristics of consumers. The MAPPO algorithm generates effective incentive strategies to encourage consumer participation in the IBIDR. The IDT ambiguity set improves WPG uncertainty characterization and reduces dispatch conservativeness.