基于机器学习的数据中心分布式制冷系统控制优化及节能策略

打开文本图片集

[摘    要]通过机器学习技术对数据中心分布式制冷系统进行大数据分析，进行室外温度湿度预测及数据中心热负荷预测，建立机器学习模型，从而优化分布式AHU系统的节能运行模式、制冷量调节，及基于IT负载SLA的联合控制策略。并以某数据中心为例，说明其节能效果。

[关键词]数据中心;分布式;制冷系统;控制策略;节能优化;机器学习;负荷预测

[中图分类号]TU83;TP308;TM61 [文献标志码]A [文章编号]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

近年来，随着大数据、云计算和人工智能的迅猛发展，数据中心尤其是超大集群互联网数据中心迅猛发展，成为下一代七大“新基建”之一。（剩余4437字）