Research On Task Scheduling Strategy Of Cloud Data Center In Electric Power Corporation

Establishing an integrated information platform is the development trend of powerenterprise information construction, which is the foundation support platform through sixaspects of the smart grid. With the construction and development of integratedmonitoring platform of intelligent substation and communication and informationplatform, the electric power data showing an explosive growth. But the existing datacenters in electric power corporation have a low efficiency and cannot meet therequirements of multi-QoS of electric power users. The core technology of cloudcomputing include massive cheap server clusters, virtualization, massive distributed dataprocessing and parallel programming models, etc. They can meet the need ofconstruction about the smart grid data center platform better.Firstly, analyzes the application situation of the data center and cloud computingtask scheduling. Then, start with the demand analysis of cloud computing data centerscheduling, explicit the load balancing and quality of service is the technical goal of taskscheduling. Research the process of task execution refer to the MapReduce parallelmodel framework. In case of there are many problems in power data center, it should beimproving the existing power data center task scheduling strategy with cloud computingtechnology. By analyzing the design scheme of cloud data center in electric powercorporation, combining its own characteristics, defines an evaluative model of multi-QoSfor power users. And a greedy algorithm and genetic algorithm based on load balancing,which targets are the QoS utility value and the completion time, are put forward.Finally, establish a substation equipment condition assessment center test platform,simulation scheduling tests of diagnosis, repair and maintenance with Cloudsim. Theresults showed that the algorithms can satisfy the needs of multi-QoS of electric powerusers, improve the operating efficiency of data center in electric power corporation, andachieve a more favorable effect of load balancing compared with traditional solutions.