基于数据驱动的多微电网互联系统分布鲁棒运行优化

针对多微电网互联系统调度中分布式电源出力不确定性以及运行效率低下等问题，提出基于数据驱动的多微电网互联系统分布鲁棒运行优化策略。首先基于多元正态Copula理论，计及风电功率预测误差的时间相关性及其与预测值的条件相关性，运用聚类方法对处理后的分布式电源数据进行聚类分析，通过场景削减方法生成具有代表性的多变量典型场景，得出分布式电源出力的概率分布模糊集。然后，建立多微电网两阶段滚动调度优化模型，考虑微电网内新能源实时出力的不确定性，提出一种基于列约束生成算法（C&CG）与交替方向乘子法（ADMM）的分布式求解算法，以实现两阶段分布式迭代求解。仿真结果表明，该方法可有效降低源荷预测不确定性对多微电网系统安全运行的影响，提高多微电网区域的经济效益。

Distributed Robust Operation Optimization of Multi-microgrid Interconnection System Based on Data Driven

Aiming at the problems of distributed power output uncertainty and low operation efficiency in multi-microgrid interconnection system scheduling，optimization strategy for multi-microgrid interconnection system based on data-driven is proposed. First of all， based on multivariate normal Copulas theory， considering the wind power prediction error of time correlation and its correlation with the conditions of the predicted value， using the SVM clustering method to deal with distributed power data clustering analysis， through the scene cuts to generate the typical multi-variable scenario，distributed power output probability distribution of fuzzy sets is concluded. Then，two-stage rolling scheduling optimization model of multi-microgrid is established，considering the uncertainty of new energy real-time output in microgrid，a distributed solution algorithm based on column constraint generation algorithm （C&CG） and alternate direction multiplier method （ADMM） is proposed to realize two-stage distributed iterative solution. The simulation results show that the influence of the uncertainty of source-load prediction on the safe operation of multi-micro grid market is effectively reduced，and the economic benefits of multi-micro grid region are improved.