基于聚类分析法的电力系统负荷建模  

The Power System Load Model Based on Clustering Analysis

作　　者：郭永鑫[1] 庞松岭 闫珺[3] 刘耀年[1] 龚卫国 

机构地区：[1]东北电力大学电气工程学院,吉林吉林132012 [2]海南电力技术研究院,海口570000 [3]河南省电力公司洛阳供电公司,河南洛阳470123 [4]河南许昌许继变压器有限公司,河南许昌461000

出　　处：《电测与仪表》2013年第1期10-14,共5页Electrical Measurement & Instrumentation

摘　　要：提出一种基于聚类分析的负荷建模方法,通过求取变电站实测数据的PSQ、PSF、SPRSQ三个指标,确定数据分类的最优分类数,对数据进行聚类分析,分别建立不同的负荷模型。仿真结果表明,本文方法不仅能够描述短路时电动机负荷向系统倒送无功的现象,还能反映故障清除后负荷节点中感应电动机对电压恢复的影响。A modeling method based on statistical cluster analysis is presented to create a mathematical model. Through the calculation of the three indicators such as PSQ,PSF and SPRSQ of the substation, the method determines the optimal number of categories of data classification, does cluster analysis of the data, and creates different load models. Simula- tion results Show that this method is not only able to describe the phenomenon of motor load transmission of reactive pow- er to the system, but also reflects the induction motor impact on voltage restoring after the fault cleared.

关 键 词：聚类分析法 负荷模型 ZIP模型 动态负荷模型 综合模型 R2统计量 伪F统计量 半偏统计量 

分 类 号：TM711[电气工程—电力系统及自动化]