摘要
调控云是电网调度业务的云服务平台,可以实时反映电网运行状态,支持电网在线分析和应用业务。为了减少调控云业务采集层合法节点数据边缘缓存的延迟,防止恶意的、未经授权的用户接入边缘端对系统造成的破坏,提出了一种区块链驱动的调控云缓存系统框架,基于边缘层的许可区块链对用户进行身份验证,对合法授权用户的监控数据进行缓存,基于深度Q网络最小化了合法用户在边缘平面数据缓存的延迟。仿真结果证明了所提算法的有效性。
The dispatching control cloud is a cloud service platform for power grid scheduling services,which can reflect the running status of power grid in real time and support online analysis and application of power grid.In order to reduce the delay of caching edge servers for legitimate nodes in dispatching control cloud acquisition layer and prevent damage to the system caused by malicious and unauthorized users accessing the edge servers,the framework of a blockchain-driven dispatching control cloud caching system is proposed,which authenticates users based on the permissioned blockchainin edge servers,and the monitoring data of legitimate users is cached.The deep Q network is proposed to minimize the time of edge plane for legitimate users.Simulation results prove the effectiveness of the proposed algorithm.
作者
谢江
郭琳
张晶
黄思婕
李发均
唐翰源
XIE Jiang;GUO Ling;ZHANG Jing;HUANG Sijie;LI Fajun;TANG Hanyuan(State Grid Yibin Electric Power Supply Company,Yibin 644000,Sichuan,China;State Grid Sichuan Information and Telecommunication Company,Chengdu 610041,Sichuan,China;State Grid Bazhong Electric Power Supply Company,Bazhong 636000,Sichuan,China;State Grid Panzhihua Electric Power Supply Company,Panzhihua 617000,Sichuan,China)
出处
《四川电力技术》
2024年第5期21-26,97,共7页
Sichuan Electric Power Technology
基金
国网四川省电力公司科技项目(521911230001)。
关键词
调控云
区块链
缓存
深度强化学习
dispatching control cloud
blockchain
caching
deep reinforcement learning
