摘要
针对传统中心计算模式难以满足大规模计算任务中算力资源合理分配的问题,设计基于多接入边缘计算融合网络的动态算力资源分配系统。在分布式新能源消纳场景中引入多接入边缘计算架构,形成一个包含云侧一体化算力平台与边侧计算节点的动态算力资源分配系统架构。软件设计方面,通过度量多接入边缘计算融合网络下的各节点算力资源,筛选出符合任务计算需求的候选算力节点;将算力资源分配问题转化为动态寻址问题,采用人工蜂群算法根据优先级将任务分为云侧一体化算力平台处理与边侧计算节点处理两部分,实现动态算力资源的分配。测试结果表明,设计系统可以在满足不同任务对处理时延要求的情况下,同时满足算力资源利用率的要求。
Aiming at the problem that the traditional central computing mode is difficult to meet the reasonable allo⁃cation of computing resources in large-scale computing tasks,a dynamic computing resource allocation system based on multi-access edge computing fusion network is designed.The multi-access edge computing architecture is introduced into the distributed new energy consumption scenario to form a dynamic computing resource allocation system architecture including the cloud-side integrated computing platform and edge computing nodes.In terms of software design,by measuring the computing power resources of each node under the multi-access edge computing fusion network,the candidate computing power nodes that meet the requirements of task computing are selected.The problem of computing resource allocation is transformed into a dynamic addressing problem.The artificial bee colony algorithm is used to divide the task into two parts:cloud-side integrated computing platform processing and edge-side computing node processing according to the priority,so as to realize the allocation of dynamic computing resources.The test results show that the design system can meet the requirements of processing delay for differenttasks,and at the same time meet the requirements of computing resource utilization.
作者
徐晓伟
虎啸
徐俊
XU Xiaowei;HU Xiao;XU Jun(State Grid Jiangsu Electric Power Co.,Ltd.Wuxi Power Supply Branch,Wuxi 214031,Jiangsu China)
出处
《粘接》
2025年第8期181-184,共4页
Adhesion
基金
国网江苏省电力有限公司科技项目(项目编号:J2022051)。
关键词
多接入边缘计算
融合网络
动态算力
资源分配
系统设计
multi-access edge computing
converged network
dynamic computing power
resource allocation
sys⁃tem design
