摘要
对WAMS海量数据进行数据优化处理是广域量测系统广泛应用的关键步骤。提出了一种适用于广域量测数据的建模方法和特征提取方法。针对广域量测数据的时空特征构造面板数据模型并进行平稳性和协整性检验;以功角为因变量,通过构造回归方程确定与其它广域量测特征量的权重因子;采用灰色关联聚类进行特征提取,并以权重因子为判据进行聚类中心选择,从而获取最优特征子集。通过与实际量测系统的仿真对比分析,验证了所提方法能够在保存特征子集物理含义的前提下,极大消除冗余,满足在线稳定评估的需要,并具有一定的通用性。
The feature extraction from WAMS mass data and intelligent model parameter estimation is the key steps in the transient stability assessment （TSA）of power system.This paper proposes a data modeling method for the WAMS and a features reduction method.Firstly,the panel data model of the WAMS was established with spatial-temporal property.Then,a regression equation of power angle was proposed after the unit root test and cointegration test with weight factors of other features.Thirdly,grey correlation degree and weight factor were employed in each grey cluster in order to obtain an optimized subset of the features.Finally,in the New England 10-generator 39-bus test system,the simulation results show the validity and universal property of the panel data modeling method and feature selection approach,which could meet the requirement of on-line security assessment.
出处
《陕西电力》
2014年第12期11-15,共5页
Shanxi Electric Power
基金
国家自然科学基金项目资助(51207113)
关键词
特征提取
暂态稳定评估
面板数据
灰色关联
灰色聚类
feature extraction
transient stability assessment
panel data
grey correlation
grey clustering
