ABSTRACT The rapid growth of renewable energy has brought stochasticity into the operation of integrated power‐gas systems (IPGSs). However, current data‐driven methods disregard data quality when estimating probability distributions based on datasets, despite its crucial role in uncertainty modelling. This paper bridges this gap by proposing a multi‐source distributionally robust optimization (DRO) model for IPGSs. Specifically, we utilize the Wasserstein metric to measure the distance between distributions, and then introduce a multi‐source DRO framework to quantify the quality of datasets collected from multiple wind farms. Furthermore, the gas network is extended into a hydrogen‐enriched compressed natural gas (HCNG) network to provide flexibility in renewable energy integration, and a risk‐aware model is then built for HCNG networks to balance flexibility support capability and network state control under uncertain gas loads. To ensure tractability, the entire model is reformulated as a second‐order cone programming problem through the linear decision rule and conditional value‐at‐risk approximation. Finally, we adopt a double‐level distributed algorithm to achieve a decentralized solution to enhance computational efficiency and preserve privacy. Numerical experiments demonstrate that the proposed method effectively quantifies the impact of data quality on dispatch results and reduces wind power spillage. Moreover, the distributed algorithm achieves faster convergence and reduces the solution time.
Li et al. (Thu,) studied this question.