| Under the background of energy transformation and the Energy Internet,the new generation of power system presents the characteristics of multi-energy coordination,flexible and controllable loads,flexible energy storage and market-oriented operation.The edge computing terminal of the Distribution Power Internet of Things(DPIo T)undertakes important functions such as local processing and optimal control of data and services,but its resources are limited and cannot deal with massive information and new services of DPIo T.Therefore,this paper focuses on the cloud-edge collaborative optimization scheduling of distribution service workload under the premise of limited computing resources of edge computing terminal,so as to adapt the workload and computing resources and improve the overall performance of the system.Firstly,this paper studies the schedulable modeling method of microservice workload for monitoring services,optimizes the deployment of service flow according to the characteristics of workload,and accordingly deploys computing resources of edge computing terminal;Next,in order to reduce the delay response of distribution services,computing tasks are offloaded and optimized based on the service characteristics.Finally,based on the global load balancing optimization,the service workload and edge computing resources are spatiotemporally scheduled.This paper studies and analyzes the resource optimization of edge computing terminal from two aspects of resource allocation and operation optimization.The main work is as follows:(1)Combing the overall architecture,key resources and operation mechanism of the DPIo T;Based on the cloud edge collaboration mechanism of DPIo T,the relationship between resource elements of edge computing terminal and cloud center is established;This paper also analyzes the basic functions of monitoring services of the DPIo T,classifies the microservices of monitoring services,analyzes the workload of different microservices,and the information source model and directed acyclic graph model of services are established.(2)Studying the optimization of service information flow under cloud-edge collaboration mechanism,and used to guide the resource allocation optimization of edge computing terminal.The information flow model of power distribution monitoring services is established,and the cloud-edge collaborative optimization of the service information flow is carried out with the minimum resource demand as the goal.Based on the optimization results,the hardware resource configuration of the edge computing terminal is optimized under the resource operation constraints of the edge computing terminal,so as to reduce the redundant configuration of the computing resource of the edge computing terminal and the communication pressure of the core network.(3)Studying the optimization method of cloud-edge collaborative microservice deployment strategy during operation,sorting out the operation mechanism of edge computing terminal container resources,transforming the service DAG model into the temporal logic model of container execution,and carries out service cloud-edge collaboration for the temporal logic to reduce service response delay.The cloud-edge collaborative optimization strategy can effectively reduce the response delay under the condition of limited resources,and improve the performance of the system.(4)Studying the global computing task offloading strategy to optimize computing resources of edge computing terminal,analyzing the mismatch of service computing load and computing resources in time and space zone.A terminal container queue model is established to describe the resource load,and the overall container load is balanced by task offloading method.An optimization strategy model of computing task offloading considering delay penalty is proposed,and the effectiveness of computing task offloading on computing resource optimization is proved by simulation results. |
PDF Full Text Request