Synthetic Power System Datasets
收藏资源简介:
本研究由麻省理工学院开发,旨在通过隐私保护算法生成电力系统优化和机器学习所需的合成数据集。数据集包含1000条经过处理的电力系统数据,主要用于解决最优电力流和风力发电预测等问题。创建过程中,采用差分隐私技术,通过添加噪声保护原始数据的隐私,同时通过后处理凸优化保持数据准确性。该数据集主要应用于电力系统的隐私保护优化计算和机器学习问题,如分布式控制算法和集中式电网解算器。
This research, developed by the Massachusetts Institute of Technology (MIT), aims to generate synthetic datasets required for power system optimization and machine learning via privacy-preserving algorithms. The dataset contains 1,000 processed power system data entries, primarily used to address problems such as optimal power flow and wind power forecasting. During its creation, differential privacy techniques were adopted: noise is injected to safeguard the privacy of the raw data, while post-processing convex optimization is applied to maintain data accuracy. This dataset is mainly applied to privacy-preserving optimization calculations and machine learning tasks in power systems, such as distributed control algorithms and centralized grid solvers.




