Research On Qualitative Evaluation Method Of Data Center Energy Efficiency Based On Multi-metrics Fusion

With the rapid growth of data center scale and energy consumption,formulating a reasonable data center energy efficiency evaluation method has become the primary problem that needs to be solved to improve and manage data center energy efficiency.Facts have proved that a single metric cannot measure the energy efficiency of data centers effectively,and different data center energy efficiency metrics have their own emphasis or even contradict each other.In the qualitative evaluation of energy efficiency of data centers,the selection of evaluation metrics is extremely important.A single evaluation metric is difficult to evaluate the energy efficiency of a data center systematically.However,when using multiple evaluation metrics to evaluate the energy efficiency of a cloud data center,on the one hand,due to the numerous energy efficiency metrics,the measurement and testing of energy efficiency in the cloud environment corresponds to a complex and even contradictory evaluation result.On the other hand,it is difficult to form a qualitative evaluation of the entire cloud data center system.Therefore,how to select multiple evaluation metrics effectively and integrate different energy efficiency metrics for energy efficiency evaluation of data centers is very important.First of all,the paper conducts an in-depth investigation of the current status of energy efficiency evaluation of data centers,and sorts out relevant metrics for energy efficiency evaluation of data centers.On the basis of analyzing the source of energy consumption of the data center,a set of metrics that can reflect the energy efficiency of the data center are selected fully.This group of metrics follows the principle of selecting energy efficiency metrics for data centers,can comprehensively and reflect the energy efficiency of data centers objectively,and has a certain degree of representativeness.Secondly,the different components of the data center have different effects on the data center as a whole,the proportion of energy efficiency metrics that measure the energy efficiency of different components in the energy efficiency evaluation is also different.Consider using the combined weighting method to weight data center metrics after comparing a variety of different energy efficiency evaluation methods to make the energy efficiency evaluation results more reasonable.Finally,a corresponding grading standard is set for the metrics based on the current situation of the industry and relevant knowledge in the data center field.An energy efficiency evaluation model is established for each metric based on the cloud model theory,and the membership degree of each metric for different energy efficiency grades is calculated according to the input data of the metric,and the different energy efficiency metrics are integrated to comprehensively evaluate the data by combining the membership degree and the metric weight.The correctness of the method is proved by using confidence.The method is better compared with other feasible methods.In order to make the evaluation results to guide the energy efficiency improvement of the data center,we propose to use the gray correlation analysis method to analyze the correlation between the calculation results of the cloud model and the energy efficiency metrics of the data center.In this way,the influencing factors of each energy efficiency metric of the data center on the energy efficiency rating of the data center can be obtained.The most relevant energy efficiency metric CUE will guide data center operators to use more clean energy and approach the goal of energy saving and emission reduction,which is more scientific and meaningful than the current single metric energy efficiency evaluation.