期刊论文

中文

电网与清洁能源

1674-3814

2018

34

2

53-59

10.3969/j.issn.1674-3814.2018.02.008

国家自然科学基金项目（51277015） 国网河南省电网公司（52170216001F）

本校为第一机构

负荷特性分析的研究对电力系统稳定运行和调度具有重要意义。以历史负荷数据为基础,采用改进的k-means聚类方法得到具有代表性的负荷特性曲线。在考虑了检验聚类结果有效性模型的基础上加入了最大最小算法,让初始聚类中心稳定;采用改进的k-means聚类典型场景缩减算法,考虑到气候因素的影响,按照春、夏、秋、冬4个季节来对负荷数据进行分类,对历史负荷数据进行处理,得到具有代表性的负荷特性曲线。以某市的历史数据为例进行仿真分析,结果表明,该算法可以将原始的大规模场景缩减合并后得到可以表述该市的几个更少的负荷特性曲线,具有代表性和准确性,更具实际应用价值。

The research on the power load characteristics analysis is of great significance to the stable operation and dispatching of the power system. Based on the historical load data,this paper uses the improved k-means clustering method to obtain the representative load characteristic curve. First,based on the model of validity of test clustering results,the maximumminimum algorithm is added to stabilize the initial cluster center. Second,the improved k-means clustering algorithm is used to reduce the typical scene. Taking into account the influence of climatic factors,the load data is classified in...