大规模电网设备拓扑合成数据集

该数据集是由存算一体的电网高性能图计算方法与计算引擎课题合成的大规模电网设备拓扑图数据，数据文件根据统一电网图数据模型与电网专题图数据模型，参照从企业资源业务中台抽取的数据自行生成。整个图结构包含百亿以上的节点和千亿以上的边，可用于图数据库的存储能力、查询能力以及加速能力的测试。在具体的存储模式上，图的点、边数据以csv格式分开存储，对应node.csv和edge.csv两个文件；另有两个json格式的文件，即schema.json和mapping.json，前者为图的模式，后者为图的关系映射。

This dataset is large-scale power grid equipment topology graph data synthesized by the project focused on high-performance graph computing methods and computing engines for power grids with integrated computing and storage. The data files are self-generated based on the unified power grid graph data model and power grid thematic graph data model, with reference to the data extracted from the enterprise resource business platform. The entire graph structure contains more than 10 billion nodes and over 100 billion edges, which can be used to test the storage, query and acceleration capabilities of graph databases. In terms of specific storage modes, the node and edge data of the graph are separately stored in CSV format, corresponding to the two files node.csv and edge.csv. There are also two JSON-format files, namely schema.json and mapping.json, where the former is the graph schema and the latter is the graph relationship mapping.

搜集汇总

背景与挑战

背景概述

该数据集是一个大规模电网设备拓扑合成数据，由北京海致星图科技有限公司和清华大学基于统一电网图数据模型生成，包含百亿以上节点和千亿以上边，主要用于测试图数据库的存储、查询和加速性能。数据以CSV和JSON格式分开存储点、边和模式信息，数据量较小但代表超大规模图结构，适用于电气工程和计算机科学技术领域的应用。