为了丰富电力物联网的计算资源,减少数据传输的时延,提出电网二次设备的云边协同体系,结合云数据计算与边缘计算,以解决边缘集群负载不均衡的问题.提出差分人工蜂群算法并应用到云边协同的资源调度,引入差分修正的思想并利用已知的最优解为蜂群搜索提供参考,增加原有搜索空间中的搜索密度,解决资源调度的多目标等问题.对电力终端计算任务卸载执行时延进行优化,应用基于生物地理学优化(BBO)算法的任务调度进化算法,合理分配时隙长度,减少传输时延.实验结果显示,所提云边协同体系在任务数达到500时,负载均衡度低至0.07,任务调度平均时延不超过1 s.
Research on Cloud-side Collaborative System of IoT for Power Grid Secondary Equipment
In order to enrich the computing resources of the power IoT and reduce the data transmission delay,this paper propo-ses a cloud-side collaborative system for the secondary equipment of the power grid,which combines cloud data computing and edge computing to solve the problem of unbalanced load in the edge cluster.The differential artificial bee colony algorithm is proposed and applied to the resource scheduling of cloud-side collaboration.The idea of differential correction is introduced and the known optimal solution is used to provide reference for bee colony search.The search density is increased in the original search space to solve the multi-objective problem of resource scheduling.The unload execution delay of power terminal compu-ting task is optimized,and the biogeography-based optimization(BBO)algorithm based task scheduling evolutionary algorithm is applied to rationally allocate the time slot length and reduce transmission delay.The experimental results show that when the task number of the cloud-side collaboration system in this paper reach 500,the unload balancing degree is as low as 0.07,and the average delay of task scheduling is less than 1 s.
万方数据
维普
