《计算机应用》唯一官方网站 ›› 2026, Vol. 46 ›› Issue (10): 3191-3198.DOI: 10.11772/j.issn.1001-9081.2025080944

• 数据科学与技术 • 上一篇    

基于LSTM-Transformer的小型风冷数据中心电源使用效率预测

滕雨轩1, 蒋从锋1(), 刘俊明2, 徐传奇2   

  1. 1.杭州电子科技大学 计算机学院,杭州 310018
    2.杭州电子科技大学 信息技术中心,杭州 310018
  • 收稿日期:2025-08-19 修回日期:2025-09-10 接受日期:2025-09-24 发布日期:2026-10-21 出版日期:2026-10-10
  • 通讯作者: 蒋从锋
  • 作者简介:滕雨轩(1999—),男,福建福州人,工程师,硕士,主要研究方向:数据中心电源使用效率
    刘俊明(1994—),男,福建泉州人,助理工程师,主要研究方向:数据治理、大数据、数据库、数据管理、异构算力融合与调度
    徐传奇(1994—),男,河南商丘人,助理工程师,硕士,主要研究方向:数据挖掘、大数据分析、网络与信息安全、云计算、虚拟化。
  • 基金资助:
    国家电网有限公司总部科技项目(5700-202358842A-4-3-WL)

Power usage effectiveness prediction of small air-cooled data centers based on LSTM-Transformer

Yuxuan TENG1, Congfeng JIANG1(), Junming LIU2, Chuanqi XU2   

  1. 1.School of Computer Science,Hangzhou Dianzi University,Hangzhou Zhejiang 310018,China
    2.Information Technology Center,Hangzhou Dianzi University,Hangzhou Zhejiang 310018,China
  • Received:2025-08-19 Revised:2025-09-10 Accepted:2025-09-24 Online:2026-10-21 Published:2026-10-10
  • Contact: Congfeng JIANG
  • About author:TENG Yuxuan, born in 1999, M. S., engineer. His research interests include data center power usage effectiveness.
    LIU Junming, born in 1994, assistant engineer. His research interests include data governance, big data, database, data management, heterogeneous computing power integration and scheduling.
    XU Chuanqi, born in 1994, M. S., assistant engineer. His research interests include data mining, big data analysis, network and information security, cloud computing, virtualization.
  • Supported by:
    Science and Technology Project of State Grid Corporation of China Headquarters(5700-202358842A-4-3-WL)

摘要:

随着数据中心规模的不断扩大,电源使用效率(PUE)已成为衡量数据中心能效水平的重要指标。现有研究大多聚焦于超大规模数据中心,而小型风冷数据中心在扩展能力、负载波动性、网络基础设施以及管理方式等方面具有显著差异,因此相关预测研究相对不足。为此,本文提出一种基于LSTM-Transformer架构的小型风冷数据中心PUE预测模型。该模型结合长短期记忆(LSTM)网络对时间序列特征的提取能力与Transformer的全局依赖建模优势,有效整合多源异构能源数据,提升对PUE变化规律的刻画与预测能力。在真实数据中心数据集上的实验结果表明,该模型的预测均方误差(MSE)、平均绝对误差(MAE)和决定系数(R2)分别为0.000 7、0.021 4和0.958 3;与LSTM模型相比,MSE和MAE分别降低53.33%和28.43%,R2提高0.051 6;与Transformer模型相比,MSE和MAE分别降低78.79%和45.41%,R2提高0.136 3,验证了所提混合架构在PUE预测精度方面的有效性。研究结果表明,该模型能够为小型风冷数据中心的能效优化与运维管理提供可靠支持。

关键词: 数据中心, 能源效率优化, 小型风冷数据中心, 电源使用效率预测, 敏感度分析

Abstract:

With the continuous expansion of data center scale, Power Usage Effectiveness (PUE) has become an important indicator for measuring energy efficiency of data centers. Most existing research focuses on hyperscale data centers, while small air-cooled data centers have significant differences in scalability, load fluctuations, network infrastructure, and management approaches, leading to relatively insufficient related prediction studies. Therefore, a PUE prediction model of small air-cooled data centers based on LSTM-Transformer architecture was proposed. The model was designed to combine the capability of Long Short-Term Memory (LSTM) network in extracting time-series features with the advantage of Transformer in modelling global dependencies, so that multi-source heterogeneous energy data were integrated effectively, thereby enhancing the characterization and prediction capabilities of PUE variation patterns. Experimental results on a real data center dataset show that the Mean Squared Error (MSE), Mean Absolute Error (MAE), and Coefficient of Determination (R2) of the proposed model are 0.000 7, 0.021 4, and 0.958 3, respectively; compared with the LSTM model, the proposed model has the MSE and MAE reduced by 53.33% and 28.43%, respectively, and R2 increased by 0.051 6; compared with the Transformer model, the proposed model has the MSE and MAE reduced by 78.79% and 45.41%, respectively, and R2 increased by 0.136 3, verifying the effectiveness of the proposed hybrid architecture in improving PUE prediction accuracy. It can be seen that the proposed model can provide reliable support for the energy efficiency optimization and operation management of small air-cooled data centers.

Key words: data center, energy efficiency optimization, small air-cooled data center, Power Usage Effectiveness (PUE) prediction, sensitivity analysis

中图分类号: