摘要
近年来,人工智能技术在智能电网的调控和数据分析中发挥着日益重要的作用,但是在智能电网数据分析中的“数据壁垒”和隐私泄露是当前亟待解决的问题。为此,文章引入了人工智能领域新兴的联邦学习概念,分析了联邦学习在电力数据分析中的研究现状,并探索了联邦学习在电力数据分析中的应用场景。联邦学习方法能够在最大程度提高模型精确度的同时保证机器学习算法的收敛性与优良性能;将联邦学习与电力数据分析相结合,既能最大化地发挥利益相关者的作用,又能满足各利益相关者的隐私保护需求。联邦学习将为智能电网的信息化和智能化发展开辟全新的路径。
In recent years,artificial intelligence technology has played an increasingly important role in the dispatching control and data analysis of smart grids,but the problems of"data barriers"and privacy leakage in smart grid data analysis are urgent problems that need to be solved.Aiming at this,this paper introduces the emerging concept of federated learning in the field of artificial intelligence,analyzes the research status of federated learning in power data analysis,and explores the application scenarios of federated learning in power data analysis.Federated learning methods can maximize model accuracy while ensuring convergence and excellent performance of machine learning algorithms.Combining federated learning with power data analysis can not only maximize the role of stakeholders,but also meet the privacy protection needs of various stakeholders.Federated learning will open up a whole new path for the informatization and intelligent development of smart grids.
作者
戴理朋
杨鑫
徐茹枝
DAI Lipeng;YANG Xin;XU Ruzhi(School of Control and Computer Engineering,North China Electric Power University,Beijing 102206,China)
出处
《电力信息与通信技术》
2022年第11期47-56,共10页
Electric Power Information and Communication Technology
基金
国家自然科学基金项目(61972148)。
关键词
联邦学习
隐私保护
电力数据分析
机器学习
深度学习
federated learning
privacy protection
power data analysis
machine learning
deep learning
