配电网潮流特性模拟数据集

在实际配电网复杂运行环境中，精准的配电网络参数往往难以获取，配电系统精确的数学机理模型很难建立且适用能力较为有限，给配电系统的精细化运行调控的实现带来了挑战。随着大数据、人工智能等新一代数字技术的快速发展，配电网数字化、网络化、智能化水平将显著提升。在实际配电网复杂运行环境中，精准的配电网络参数往往难以获取，配电系统精确的数学机理模型很难建立且适用能力较为有限，给配电系统的精细化运行调控的实现带来了挑战。基于边缘侧广泛分布的小微传感器，配电系统可以获得描述网络运行精细状态的大量量测信息。基于图卷积神经网络可构建配电网代理模型，利用配电网历史运行数据构造代理模型训练集并完成模型训练，以数据为基础拟合配电网节点功率与电压分布间的潮流映射关系，实现基于数据驱动的配电网电压特性模拟，为配电网运行控制提供潮流模型支撑。

In the complex operational environment of actual distribution networks, accurate parameters of the distribution network are often difficult to obtain, and it is hard to establish precise mathematical mechanism models for power distribution systems with limited applicability, which poses challenges to the realization of refined operation and regulation of distribution systems. With the rapid development of new-generation digital technologies such as big data and artificial intelligence, the digitization, networking and intelligentization levels of distribution networks will be significantly improved. In the complex operational environment of actual distribution networks, accurate parameters of the distribution network are often difficult to obtain, and it is hard to establish precise mathematical mechanism models for power distribution systems with limited applicability, which poses challenges to the realization of refined operation and regulation of distribution systems. Based on the widely distributed micro-small sensors deployed on the edge side, the power distribution system can acquire a large volume of measurement information describing the fine-grained operational states of the network. A surrogate model for the distribution network can be constructed based on graph convolutional neural networks (GCNs). The training set for the surrogate model is constructed using the historical operational data of the distribution network, and model training is completed. Based on the collected data, the power flow mapping relationship between the node power and voltage distribution of the distribution network is fitted, realizing data-driven simulation of the voltage characteristics of the distribution network, and providing power flow model support for the operation and control of the distribution system.

搜集汇总

背景与挑战

背景概述

该数据集是一个用于配电网潮流特性模拟的训练集，旨在通过图卷积神经网络构建代理模型，利用历史运行数据拟合节点功率与电压分布之间的映射关系，以数据驱动方式支持配电网的精细化运行调控。数据集包含142.88MB的3个文件，由天津大学创建，属于国家重点研发计划项目'数字电网关键技术'的组成部分，适用于电力系统自动化领域的研究和应用。