Two-stage adjustable robust optimal dispatching model for multi-energy virtual power plant considering multiple uncertainties and carbon trading

In the context of high proportion renewable energy access and multi-energy synergy, low-carbon multi-energy virtual power plants (MEVPP) are gradually getting hot. This paper introduces the carbon trading mechanism into a MEVPP and constructs a novel dynamic robust optimization model considering the source-load multiple uncertainties. Firstly, the Adaptive Neuro-Fuzzy Inference System and spectral clustering method are applied to generate the source-side output scenarios, and then by the strength of the budget uncertainty set, the concept of uncertainty measurement is established to characterize the system's uncertainties. Secondly, a two-stage robust adjustable scheduling model is developed by integrating the carbon trading cost into the objective function, and the model derived from the decomposition is solved using the column and constraint generation algorithm and strong duality theory. Finally, the economics, robustness, and low-carbon performance of the model in addressing the MEVPP scheduling issues are examined by case studies. The proposed model is capable of getting the optimal cost under the worst-case scenario, dynamically balancing the economics and robustness of the scheduling strategy with adjustable uncertainty measurement, as well as contributing to guidance of energy saving and emission reduction in MEVPP.