基于数据驱动的交直流电网分层鲁棒日前优化调度

对于风电并网下基于VSC-MTDC的多区域互联电网日前优化调度问题,提出了基于目标分析级联的分层鲁棒调度模型。为降低风电出力的保守性,采用基于数据驱动的高维椭球(MVEE)多面体集合。为充分发挥VSC-MTDC输电方式的灵活调度能力,以VSC换流站为节点对交直流电网进行分层解构,将跨区域交直流电网的日前调度问题分为了高层直流电网多时段联络线协调优化的主问题和低层区域交流电网两阶段鲁棒优化的子问题,并且各个区域交流电网子问题独立并行求解。通过4区域-6节点互联测试网络的仿真求解,证明所提模型的有助于风电消纳。

Research on Data-driven Hierarchical and Robust Day-ahead optimization scheduling of AC/DC Grid

For the day-ahead optimization scheduling problem of multi-region interconnected power system via VSC-MTDC with wind power, a hierarchical and robust scheduling model based on analytical target cascade was proposed. In order to reduce the conservativeness of wind power output, a data-driven high-dimensional ellipsoid (MVEE) polyhedron set had been adopted. Considering the flexible adjustment capabilities of the VSC-MTDC transmission method, the AC/DC power grids were stratified with the VSC converter station as the node. The day-ahead scheduling problem of cross-regional AC/DC power grids was divided into the main problem of high-level DC grid and the sub-problem of low-level regional AC power grid. The high-level main problem is the coordination and optimization of multi-period DC tie-lines. The low-level sub-problem is a two-stage robust optimal dispatching problem for regional AC power grid, and solved independently and in parallel. The simulation solution of 6-node-4-region test system proves that the proposed model can help improve wind power accommodation ability.