基于機(jī)器學(xué)習(xí)的數(shù)據(jù)中心分布式制冷系統(tǒng)控制優(yōu)化及節(jié)能策略

打開文本圖片集

[摘    要]通過(guò)機(jī)器學(xué)習(xí)技術(shù)對(duì)數(shù)據(jù)中心分布式制冷系統(tǒng)進(jìn)行大數(shù)據(jù)分析，進(jìn)行室外溫度濕度預(yù)測(cè)及數(shù)據(jù)中心熱負(fù)荷預(yù)測(cè)，建立機(jī)器學(xué)習(xí)模型，從而優(yōu)化分布式AHU系統(tǒng)的節(jié)能運(yùn)行模式、制冷量調(diào)節(jié)，及基于IT負(fù)載SLA的聯(lián)合控制策略。并以某數(shù)據(jù)中心為例，說(shuō)明其節(jié)能效果。

[關(guān)鍵詞]數(shù)據(jù)中心;分布式;制冷系統(tǒng);控制策略;節(jié)能優(yōu)化;機(jī)器學(xué)習(xí);負(fù)荷預(yù)測(cè)

[中圖分類號(hào)]TU83;TP308;TM61 [文獻(xiàn)標(biāo)志碼]A [文章編號(hào)]2095–6487（2022）06–00–04

Control Optimization and Energy Saving Strategy of Data Center

Distributed Refrigeration System Based on Machine Learning

Sun Jing-jing，Jing Tang-bo

[Abstract]Through the machine learning technology， the big data analysis of the distributed refrigeration system in the data center is carried out， the outdoor temperature and humidity prediction and the heat load prediction of the data center are carried out， and the machine learning model is established， so as to optimize the energy-saving operation mode， refrigeration capacity regulation and the joint control strategy based on it load SLA of the distributed AHU system. Take a data center as an example to illustrate its energy-saving effect.

[Keywords]data Center; distributed; refrigeration system; control strategy; energy saving optimization; machine learning; load forecasting

近年來(lái)，隨著大數(shù)據(jù)、云計(jì)算和人工智能的迅猛發(fā)展，數(shù)據(jù)中心尤其是超大集群互聯(lián)網(wǎng)數(shù)據(jù)中心迅猛發(fā)展，成為下一代七大“新基建”之一。（剩余4437字）