电网技术2017，Vol.41Issue(4)：1132-1138,7.DOI:10.13335/j.1000-3673.pst.2016.2792
基于面板数据有序聚类的主动配电网规划场景降维技术
Scene Dimensionality Reduction Technology in Active Power Distribution Network Planning Based on Orderly Clustering of Panel Data
摘要
Abstract
Scene dimensionality reduction technology in active power distribution network operation based on orderly clustering of panel data is investigated in this paper.Firstly,timing simulation of medium-and long-term planning of active power distribution is performed,and different types of load growth sequence models with different load factors are built.Considering natural capacity growth,PV and wind power timing model is set up using auto regression-moving average model (ARMA).Change point theory in statistics is introduced to partition DG and load scenes in chronological order respectively.Change points of DG and load are merged together to get comprehensive change points,and temporal scenarios with these comprehensive change points are segmented.Results show that orderly clustering can show operation conditions of loads and DGs effectively for different time scale.Millions of sequential scenarios can be orderly clustered into 14 continuous scenarios,greatly increasing efficiency.关键词
主动配电网/时序/面板数据/有序聚类/场景降维Key words
active power distribution network/timing/panel data/orderly clustering/scene dimensionality reduction分类
信息技术与安全科学引用本文复制引用
刘洁颖,刘俊勇,黄媛,刘友波,高红均,庄丹,马铁丰..基于面板数据有序聚类的主动配电网规划场景降维技术[J].电网技术,2017,41(4):1132-1138,7.基金项目
国家科技部863计划项目(2014AA051901) （2014AA051901）
国家自然科学基金资助项目(51437003).The National High Technology Research and Development of China 863 Program (2014AA051901) （51437003）
Project Supported by National Natural Science Foundation of China (51437003). （51437003）
