基于AP聚类和鲁棒优化的电网规划灵活性评估  

Evaluation on Grid Planning Flexibility Based on Affinity Propagation Clustering and Robust Optimization

作　　者：魏联滨[1] 王伟臣[1] 李慧[1] 宣文博[1] 刘忠义 WEI Lianbin;WANG Weichen;LI Hui;XUAN Wenbo;LIU Zhongyi(State Grid Tianjin Electric Power Company,Tianjin 300010,China;Economic and Technological Research Institute,State Grid Tianjin Electric Power Company,Tianjin 300171,China)

机构地区：[1]国网天津市电力公司,天津300010 [2]国网天津市电力公司经济技术研究院,天津300171

出　　处：《电力系统及其自动化学报》2020年第3期99-106,150,共9页Proceedings of the CSU-EPSA

基　　金：国网天津市电力公司科技项目(KJ19-1-24特大型城市电网远景年电力需求预测及主网架规划技术研究)。

摘　　要：高比例可再生能源的并网,给电力系统规划的灵活性带来了挑战。本文考虑源-荷双端的不确定性,提出了一种基于AP(affinity propagation)聚类和两阶段鲁棒优化的电网规划灵活性评估方法。首先,本文提出了基于AP聚类的远景年典型日负荷建模方法。其次,以区间的形式考虑可再生能源出力的不确定性,构建基于鲁棒优化的电力系统规划方案灵活性评估模型。最后,在整个周期的角度和各时间断面的角度提出了一组电网规划灵活性评估指标。本文利用IEEE RTS-24算例验证所提出方法的合理性。The high-proportion integration of renewable energy into grid poses a challenge for the flexibility of power sys tem planning.In this paper,an evaluation method for grid planning flexibility is proposed based on affinity propagation(AP)clustering and two-stage robust optimization,which considers the uncertainties on both source and load sides.First,a modeling method for typical daily load in a future year is put forward based on AP clustering.Then,the uncer tainties in renewable energy output are considered in the form of interval,and an evaluation model of power system plan ning flexibility is constructed based on robust optimization.Finally,a set of evaluation indexes for grid planning flexibil ity is formulated from perspectives of the whole cycle and each time section.The rationality of the proposed method is verified by an IEEE RTS-24 test system.

关 键 词：AP聚类 灵活性评估 电网规划 鲁棒优化 可再生能源 

分 类 号：TM715[电气工程—电力系统及自动化]