论文部分内容阅读
基于电气接线图并按特定图形设计的时序态势图,其图形特征是网络结构、运行方式、接入源荷及各类电气元件行为、量测数据等动态潮流的一种高维融合映射。通过基于刻画态势图片的关键但数量有限的图形特征量的聚类分析,从中获取态势图中隐藏的规律和有价值的知识。文中提出变电站为中心的配电网电压等高线时序态势图片的特征向量及其计算算法;探索了基于图片特征向量的态势时序图片聚类分析方法,提出了聚类中心图片(典型图片)的图片相似度、时间离散度及其计算算法。以典型日96幅电压态势时序图为案例,介绍了所提方法的效果以及分析结果,并进行了近两个月的长时序图片聚类的初步应用分析。
Based on the electrical wiring diagram and the timing diagram of a specific graphic design, the graphic features are a kind of high-dimensional fusion map of the dynamic power flow such as the network structure, operation mode, access source charge and various electrical component behaviors and measurement data. Through the clustering analysis based on the key but limited amount of graphic features characterizing the situation picture, the hidden rules and valuable knowledge in the situation diagram are obtained. In this paper, the eigenvectors and calculation algorithms of temporal contour picture of voltage contour in substations are put forward. The method of situation sequential picture clustering analysis based on image eigenvectors is proposed and the clustering center picture (typical picture) is proposed Image similarity, time dispersion and its calculation algorithm. A typical example of the time series of 96 voltage regimes is presented. The effects of the proposed method and the results of the analysis are introduced. The preliminary application analysis of the long-time picture clustering in the past two months is also given.