My da|ra Login

Detailed view

metadata language: English

Thermal images of the server in different running status

Version
V0
Resource Type
Dataset
Creator
  • liu, hang (Dalian University of Technology)
Publication Date
2020-05-26
Description
  • Abstract

    Balancing the uneven temperature distribution of data centers can reduce cooling energy consumption significantly, as the overheated surface of servers is one of the root causes of the uneven distribution. This paper presents a method for intelligently diagnosing server operating status based on the heat distribution of the server surface. Five types of server operation can be diagnosed: normal status, main fan failure, vice-fan failure, air vent blockage and low-load status. The method involves signal processing and pattern recognition techniques such as thermal image enhancement, region segmentation and image classification. First, thermal images of server outlets in running status are captured as data; second, the images are preprocessed for standardization; third, after homomorphic filtering enhancement, the images are subjected to one-dimensional maximum entropy segmentation to obtain server hotspot images; fourth, morphological features, texture features and statistical features are extracted from hotspot images; finally, the server status is diagnosed by a support vector machine.
Availability
Download
Relations
  • Is supplement to
    DOI: 10.1016/j.infrared.2018.08.028 (Text)
Publications
  • Liu, Hang, Chenchen Bao, Ting Xie, Shan Gao, Xianlin Song, and Weina Wang. “Research on the Intelligent Diagnosis Method of the Server Based on Thermal Image Technology.” Infrared Physics & Technology 96 (January 2019): 390–96. https://doi.org/10.1016/j.infrared.2018.08.028.
    • ID: 10.1016/j.infrared.2018.08.028 (DOI)

Update Metadata: 2020-05-27 | Issue Number: 1 | Registration Date: 2020-05-27

liu, hang (2020): Thermal images of the server in different running status. Version: V0. ICPSR - Interuniversity Consortium for Political and Social Research. Dataset. https://doi.org/10.3886/E119603