Research On Energy Efficiency Evaluation Method Of Cloud Data Center

The increasing scale of data centers has led to the rapid growth of energy consumption in recent years.A scientific and objective evaluation of cloud data center energy efficiency is not only conducive to digital monitoring of data centers,but also instructive for data center management,such as scheduling,optimizing resource allocation,and improving the service quality.This article focuses on the energy efficiency evaluation of cloud data centers.The specific work is as follows:Aiming at the problem of difficult quantitative and accurate evaluation of service quality in data centers,we established a Qo S index evaluation system,which is divided into target layer,index layer,and feature layer;and proposed a Qo S Hierarchical Evaluation Method.(Qo S-HEM).First preprocess the features of the feature layer,then obtain the user's service quality scores under different task types through the user evaluation matrix,standardize the scores,and finally obtain the value of the target layer through the Qo S-HEM model to realize the data center real-time quantitative evaluation of service quality.At the same time,by analyzing the significance and main problems of cloud data center energy efficiency,based on scientific,reasonable,and systematic principles,the energy efficiency evaluation of cloud data center is constructed from four aspects: energy efficiency metrics,thermal performance metrics,green energy metrics,and service quality.Due to the short of samples scored by the experts,this article establishes a SOM-based semi-supervised cloud data center energy efficiency evaluation model.Compared to the models training on the original labeled data,the model can better learn the scoring pattern of experts and is more suitable for the evaluation of cloud data center energy efficiency levels.A cloud data center energy efficiency evaluation system is developed by using Vue to realize the function of real-time energy efficiency evaluation.