基于云边协同的配电物联网管理模型研究

Research on the management model of distribution IoT based on cloud-edge collaboration

针对配电云服务器中计算任务繁重导致推理延时过大的问题,提出配电物联网多节点协同计算管理模型。首先,基于物联网可控智能的特点以及新型配电网对功能和传输时延的需求,讨论了边缘计算在配电网中的应用。然后,结合PTN(分组传送网)网络结构,构建配电物联网技术架构。最后,基于边缘多节点协同推理算法和协同计算架构模型构建了配电物联网管理模型。仿真结果表明,当配电物联网中大量智能设备终端接入网络时,该模型降低了云计算的推理延迟。

Given the excessive inference delay due to heavy computing tasks in cloud servers for distribution net⁃works,a multi-node collaborative computing management model for distribution IoT is proposed.First,the applica⁃tion of edge computing in distribution networks is discussed since IoT is controllable and intelligent as well as the de⁃mand for function and transmission delay in new distribution networks.Then,the technical architecture of distribu⁃tion IoT is constructed by combining the packet transport network(PTN)structure.Finally,the distribution IoT management model is constructed using the edge multi-node collaborative inference algorithm and a collaborative computing architecture model.Simulation results show that the model reduces the inference delay of cloud comput⁃ing when a large number of smart device terminals in distribution IoT are connected to the network.