基于数字孪生与AI的智慧机房管理应用探索  

Intelligent data center management exploration and applications based on digital twins and AI

作　　者：张珊 刘佳 陈新林 石启亚 ZHANG Shan;LIU Jia;CHEN Xinlin;SHI Qiya(China Telecom Digital Intelligence Technology Co.,Ltd.,Nanjing 210005,China)

机构地区：[1]中电信数智科技有限公司,江苏南京210005

出　　处：《通信与信息技术》2025年第2期112-114,共3页Communication & Information Technology

摘　　要：针对机房管理智能化程度不足、设备故障定位准确度不高、资产价值评估难度大、远程感知差等问题,提出构建基于数字孪生与AI的智慧机房管理系统,通过IOT技术实现机房的全息感知,通过BIM技术构建数字孪生系统,实现孪生模型与物理机房的在线映射,通过AI分析算法实现设备故障预测、资源配置优化和机房的智能化管理。选取某大型数据中心为研究对象进行技术验证,结果表明系统应用大幅减少了机房停机时间和设备故障率、提高了劳动生产率和能源利用率、节约了维护成本,应用成效显著。In response to issues such as inadequate intelligence in data center management,low accuracy in equipment fault local⁃ization,difficulty in asset valuation,and poor remote perception,this paper proposes the construction of a smart data center management system based on digital twins and AI.Through IoT technology,the system achieves comprehensive perception of the data center;using BIM technology,it constructs a digital twin system to map twin models with physical data centers online;and employing AI analysis algo⁃rithms,it achieves equipment fault prediction,resource optimization,and intelligent management of the data center.A large-scale data center is selected as the research subject for technical validation,and the results indicate that the system significantly reduces data center downtime and equipment failure rates,improves labor productivity and energy utilization efficiency,and saves maintenance costs,demon⁃strating significant application effectiveness.

关 键 词：数字孪生 AI分析 智慧机房 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]