摘要
大数据时代背景下智能电网各类监测数据量激增,为了提升处理效率、避免数据拥堵或丢失、增强系统运行可靠性,设计了基于云计算技术的“中枢层-次级层-区域层”分层集群架构,采用改进的蜂群算法及爬山算法寻求最优区域的最佳计算节点,引入处理任务价值特征与时间特征综合判断调度优先级,根据优先级进行负载均衡调度。经过实例数据验证,与经典调度算法相比,随着任务数的增加本研究设计的调度算法重点任务完成率更加平稳可靠且处理时间更短,为智能电网的数据流分发与调度提供了可参考方法。
With the monitoring data of smart grid increasing rapidly in the big data era,a hierarchical cluster architecture of"Center layer-sub layer-region layer"is designed based on cloud computing technology,so as to improve the processing efficiency,avoid data jam or loss,and enhance the reliability of the system.The improved bee colony algorithm and hill-climbing algorithm are used to find the best computing nodes in the optimal region,and the value and time features of the processing task are introduced to judge the scheduling priority synthetically,and then the load balance is made according to the priority.Results show that compared with the classical scheduling algorithm,with the increase of the number of tasks,the key task completion rate of the scheduling algorithm designed is more stable and reliable,and the processing time is shorter,which provides a reference method for data flow distribution and dispatching in smart grid.
作者
钟云南
陈行滨
占彤平
林旭军
陈珺
谢妙红
ZHONG Yunnan;CHEN Xingbin;ZHAN Tongping;LIN Xujun;CHEN Jun;XIE Miaohong(State Grid Fujian Electric Power Co.,Ltd.,Fuzhou 350000,China;State Grid Xintongyili Technology Co.,Ltd.,Fuzhou 350003,China)
出处
《粘接》
CAS
2022年第12期182-185,共4页
Adhesion
关键词
智能电网
云计算集群
任务调度
价值特征
时间特征
smart grid
cloud computing cluster
task scheduling
value characteristics
time characteristics
