Research on Intelligent Detection and Early Warning of Thermal Hazard in Data Center Based on Deep Time-series Learning Models
The equipments in data center generate enormous heat.Weak cooling system may cause thermal accumulation,leading to thermal hazard in data center.It proposes a thermal risk management method in data center based on Bi-LSTM deep learning network.By preprocessing the temperature field data of the computing nodes in the data center and identifying the thermal risks,historical data is used to warn the thermal risks of the computing nodes in the data center.Aiming at the complex thermal environment in the computer room,statistical spatial-temporal analysis is employed to determine thermal risk labels from two dimensions of time and space for network training.The proposed method is compared to the traditional SVM model,and the results show that the proposed method provides thermal hazard prediction of 99.07%accuracy,with 6.6%improvement compared to SVM,which presents reliable prediction ability of thermal hazard.
万方数据
维普
