Comprehensive Analysis and Visualization Method of Online and Offline Operation Modes of Power System

The AC / DC hybrid power grid has already been built in China using Ultra-High Voltage (UHV) grid as its backbone. Operation modes of power grid become more complex and diverse. With the development of online and offline analysis of the power system, more and more operation mode data have been accumulated. However, it is difficult to study the offline and online data together because of their different distributions. Offline data are distributed on the boundary of security region, while online data are concentrated in some parts for a period of time. Therefore, a comprehensive analysis and visualization method should be proposed to demonstrate the relationship between them. Firstly, a simulation sample database is constructed including both online and offline data. Then, the research of sample distribution is carried out based on an introduced similarity degree index that calculated by designated features. Finally, a new method is proposed to represent the closeness degree of simulation samples, which maps the high-dimensional and complicated relations into 2D images by using t-distribution stochastic neighbor embedding algorithm (t-SNE) to give dispatcher more intuitive impressions. The method is verified by taking the actual data of regional power grid.