基于数据驱动的配微协同分布式鲁棒优化调度方法OA

High renewable energy penetration poses significant challenges to coordinated optimization between distribution networks and microgrids,including hierarchical coupling,inadequate protection of stakeholder privacy,and the difficulty of balancing robustness and economic efficiency under uncertain wind and photovoltaic generation.To overcome the limitations of existing approaches,such as centralized models requiring full data sharing,conventional distributed methods with weak robustness,and overly conservative robust optimization techniques,this paper proposes a data-driven distributed robust optimal scheduling method for distribution network-microgrid coordination.First,distribution networks and microgrids are modelled as independent stakeholders.Analytical target cascading and Lagrangian penalty functions are employed to decouple tie-line power and establish a privacy-friendly distributed optimization model.Second,to characterize the uncertainty of wind and photovoltaic generation,a Gaussian mixture model is adopted to fit historical data,construct the joint probability distribution,and perform clustering analysis.A data-driven ambiguity set based on the Wasserstein distance is then developed,and a two-stage distributed robust optimization model is formulated.Finally,the proposed model is decomposed and solved using a column-and-constraint generation method in conjunction with strong duality theory.Simulations conducted on the IEEE141-bus and 281-bus systems verify the feasibility and effectiveness of the proposed data-driven distributed robust optimal scheduling method for coordinated distribution network-microgrid operation.