计算机科学 ›› 2025, Vol. 52 ›› Issue (9): 178-185.doi: 10.11896/jsjkx.241100174
李思琪, 俞琨, 陈宇皓
LI Siqi, YU Kun, CHEN Yuhao
摘要： 随着科学研究和工程模拟中数据规模和实验复杂度的不断提升,对高性能计算资源的需求日益增长。然而由于资源有限,如何高效利用现有计算资源成为亟待解决的问题。基于2022年1月至2023年11月期间华东师范大学高性能计算中心集群收集的40万条作业数据,整理得到集群每日运行作业数和CPU资源利用率,以表征计算资源的使用情况。采用ARIMA模型、由LSTM改进的2DLSTM模型和ARIMA-2DLSTM组合模型对历史数据进行拟合,实现了对集群计算资源使用情况的长短期预测。通过平均绝对误差(MAE)和均方误差(MSE)指标评估模型预测效果,实验结果表明,ARIMA-2DLSTM组合模型在预测准确性上优于单独使用ARIMA模型和2DLSTM模型,且能够精确预测趋势变化以及波峰与低谷出现的时间,为高性能计算中心的资源分配提供了有效支持。
中图分类号:
| [1]LU P J,XIONG Z Y,LAI M C.Survey on High-performanceComputing Technology and Standards[J].Computer Science,2023,50(11):1-7. [2]Editorial Department of this Journal.Top 10 Frontiers ofScience and Technology in 2023[J].Science and Technology Think Tank,2023(1):1-4. [3]WANG J Y,ZHOU B Y,ZHANG F,et al.Overview of Energy Consumption Models and Energy Efficiency Algorithms in Data Centers[J].Journal of Computer Research and Development,2019,56(8):1587-1603. [4]CHEN Y J.Energy Consumption Analysis and Power Consump-tion Prediction of High-Performance Computing Clusters[D].Jinan:Qilu University of Technology,2023. [5]SUN J W.Data-Driven Research on Execution Time Predictionand Optimization for High-Performance Computing Programs[D].Hefei:University of Science and Technology of China,2020. [6]ZHANG J P.Research on Active Learning Methods for Empirical Performance Modeling of High-Performance Computing Programs[D].Hefei:University of Science and Technology of China,2020. [7]KUMAR A S,MAZUMDAR S.Forecasting HPC workloadusing ARMA models and SSA[C]//2016 International Confe-rence on Information Technology(ICIT).IEEE,2016:294-297. [8]BI J,ZHANG L,YUAN H,et al.Hybrid task prediction based on wavelet decomposition and ARIMA model in cloud data center[C]//2018 IEEE 15th International Conference on Networking,Sensing and Control(ICNSC).IEEE,2018:1-6. [9]IPEK E,DE SUPINSKI B R,SCHULZ M,et al.An approach to performance prediction for parallel applications[C]//Euro-Par 2005 Parallel Processing:11th International Euro-Par Confe-rence,Lisbon,Portugal,August 30-September 2,2005.Procee-dings 11.Berlin:Springer,2005:196-205. [10]BORGHESI A,BARTOLINI A,LOMBARDI M,et al.Predictive modeling for job power consumption in HPC systems[C]//High Performance Computing:31st International Conference,ISC High Performance 2016,Frankfurt,Germany,June 19-23,2016,Proceedings.Springer,2016:181-199. [11]KONTOPOULOU V I,PANAGOPOULOS A D,KAKKOS I,et al.A review of ARIMA vs.machine learning approaches for time series forecasting in data driven networks[J].Future Internet,2023,15(8):255. [12]SHARIFF S M.Autoregressive Integrated Moving Average(ARIMA) and Long Short-Term Memory(LSTM) network models for forecasting energy consumptions[J].European Journal of Electrical Engineering and Computer Science,2022,6(3):7-10. [13]SIRISHA U M,BELAVAGI M C,ATTIGERI G.Profit prediction using ARIMA,SARIMA and LSTM models in time series forecasting:A comparison[J].IEEE Access,2022,10:124715-124727. [14]FAN D,SUN H,YAO J,et al.Well production forecastingbased on ARIMA-LSTM model considering manual operations[J].Energy,2021,220:119708. [15]SELVIN S,VINAYAKUMAR R,GOPALAKRISHNAN E A,et al.Stock price prediction using LSTM,RNN and CNN-sliding window model[C]//2017 International Conference on Advances in Computing,Communications and Informatics(ICACCI).IEEE,2017:1643-1647. [16]SCHAFFER A L,DOBBINS T A,PEARSON S A.Interrupted time series analysis using autoregressive integrated moving average(ARIMA) models:a guide for evaluating large-scale health interventions[J].BMC Medical Research Methodology,2021,21:1-12. [17]BOX G E P,JENKINS G M,REINSEL G C,et al.Time series analysis:forecasting and control[M].John Wiley & Sons,2015:88-126. [18]ARUNKUMAR K E,KALAGA D V,KUMAR C M S,et al.Forecasting the dynamics of cumulative COVID-19 cases(confirmed,recovered and deaths) for top-16 countries using statistical machine learning models:Auto-Regressive Integrated Mo-ving Average(ARIMA) and Seasonal Auto-Regressive Integrated Moving Average(SARIMA)[J].Applied Soft Computing,2021,103:107161. [19]HOCHREITER S,SCHMIDHUBER J.Long Short-term Me-mory[J].Neural Computation,1997,9(8):1735-1780. [20]CHOUDAKKANAVAR G,MANGAI J A.A hybrid 1D-CNN-Bi-LSTM based model with spatial dropout for multiple fault diagnosis of roller bearing[J].International Journal of Advanced Computer Science and Applications,2022,13(8):0130873. |
|
||
首页