考虑风电不确定性的交直流配电网低碳分布鲁棒优化调度

为增加配电网风电的消纳能力，减少碳排放，建立了一种交直流配电网低碳分布鲁棒优化调度模型。分析风电预测误差和预测出力历史数据之间的正相关性，采用混合Copula函数，建立它们之间的联合概率分布，得到风电预测误差的条件概率分布。将交直流配电网解耦为交流和直流子网，以各自综合运行成本最小为优化目标，在交流子网优化模型中引入碳交易机制，建立交直流配电网分散协调优化模型。以得到的风电预测误差的条件概率分布为参考，构建了基于K-L散度的分布鲁棒模糊集。利用拉格朗日对偶理论，将优化模型转化为单层优化目标模型，并利用交替方向乘子法进行分散协调优化求解。基于修改后33节点交直流配电网模型的仿真结果表明所提模型能有效减少配电网侧碳排放量，显著提高风电消纳能力。

Low-carbon distributionally robust optimal scheduling for AC/DC distribution network considering wind power uncertainty

In order to increase the wind power absorption capacity of the distribution network and reduce the carbon emission, a low-carbon distributionally robust optimal scheduling model for AC/DC distribution network is established. The positive correlation between historical data of wind power forecast error and wind power forecast output is analyzed. The mixed Copula function is used to establish the joint probability distribution between them, and the conditional probability distribution of wind power forecast error is obtained. The AC/DC distribution network is decoupled as AC and DC subnets, taking the minimum comprehensive operating cost of each subnet as the optimization objective, and the carbon trading mechanism is introduced in the AC subnet optimization model, so that the decentralized coordinated optimization model of AC/DC distribution network is established. Taking the obtained conditional probability distribution of wind power forecast error as the reference, the distributionally robust ambiguous set based on K-L divergence is constructed. The proposed optimization model is converted into a single-layer optimization objective model by using Lagrange dualism theory and the alternating direction method of multipliers is used for decentralized coordination optimal solution. Simulative results based on the modified 33-bus AC/DC distribution network model show that the proposed model can effectively reduce the carbon emission at distribution side and significantly improve the consumption capacity of wind power.