成本考量下之散熱系統最佳化問題

Abstract

雲端運算(Cloud computing)是近年來相當熱門的話題之一，而雲端數據機房中的伺服器在運作時會產生大量的熱能，如散熱不當容易造成當機、無法有效運作等問題。因此雲端數據機房的散熱問題是個非常重要的議題。然而，在講求環保與永續發展的二十一世紀，「節能減碳」是許多企業與國家追求的目標，固此本研究針對雲端數據機房之耗電量、耗水量、碳排放量及成本等四個反應變數使用Box & Behnken設計(BBD)來規劃實驗，發展一套同時最佳化之演算法，同時也將成本納入考量。最後在散熱量必須滿足發熱量的限制下，求得散熱系統最佳之參數設定值。

Cloud Computing is one of the most popular topic in recent years. The server in the cloud data center will produce a lot of heat during operation, therefore, improper dissipation of heat may cause the computer breakdown or work ineffectively. Cooling problem of Cloud data center, become an important issue. In recent years, environmental protection is also an important issue for government policy makers of many countries. Hence the main objective of this study is under the consideration of energy saving and carbon reduction , to optimize the four response variables including the cloud data center power consumption, water consumption, carbon emissions and cost simultaneously. Box & Behnken design(BBD) is used to design the experiment and the Dual response surface methodology (DRSM) is employed to determine the optional parameter settings for the cooling system to meet the minimum requirements of heat dissipation.