Article Details

Title · Authors · Abstract · Keywords · DOI · References

Research and Application of Hardware Device Operation and Maintenance Based on Data Reconstruction and AIOps

WANG Yao, CHEN Wanyi, XU Junyuan, NIE Kaixun

(Zhejiang Energy Digital Technology Co., Ltd., Hangzhou 310012, China)

Section:Computer Technology · 2025, Vol. 9, No. 12 · Pages:101-109 · 2096-4706(2025)12-0101-09 · CLC number:TP393;TP311 · Document code:A · DOI:10.19850/j.cnki.2096-4706.2025.12.020

Abstract

As a technology integrating Machine Learning and data analysis, AIOps (Intelligent Operations and Maintenance) demonstrates great potential in improving IT operation and maintenance efficiency and decision-making quality. However, the current industry pays relatively low attention to the operation and maintenance of IT hardware devices, with most practices only staying at the basic monitoring level, which leads to significant room for improvement in the application of AIOps in the field of hardware device operation and maintenance. Aiming at the insufficiency in the operation and maintenance of IT hardware devices, this paper proposes a scheme to enhance the quality and quantity of hardware device data, so as to support the effective implementation of AIOps and ensure the stable operation of business systems. By reconstructing the operation and maintenance data structure of hardware devices, new data structures and automated processes are designed to achieve automatic discovery of hardware devices and data collection. Meanwhile, weak classifier voters are applied for anomaly detection, and predictive models are used to improve the accuracy of data forecasting. In the practice of the Zheneng Group data center, this approach significantly improves the quality of hardware device-related data and operation and maintenance efficiency, thereby enhancing the stability of business systems and the operational sustainability of the data center. The results provide new ideas and solutions for the further development of AIOps in hardware device operation and maintenance.

Keywords

AIOps; asset data reconstruction; anomaly detection; time series data forecasting; data collection; automatic device discovery

References

  1. [1] Han Xiaoguang, Wang Ruosong, Tang Jincai Research and Practice of IT Operation and Maintenance Management System in the New Era [J]. China Media Technology, 2024 (6): 150-155
  2. [2] Li Lecheng, Hu Yafei Exploration of Information System Operation and Maintenance Management System [J]. Information and Communication, 2019 (8): 104-106
  3. [3] Lu Shida, Kang Kai, Liu Shuaihua, et al. Application of deep neural network-based anomaly detection algorithm in AIOps of State Grid Shanghai Company [J]. Microcomputer Applications, 2024, 40 (5): 231-233
  4. [4] Zheng Shihui AIOps Practice of Bank of Communications [J]. Financial Electronics, 2019 (1): 46
  5. [5] Cheng Peng Exploration and Practice of AIOps Intelligent Operations in Industrial and Commercial Bank of China [J]. China Financial Computer, 2021 (5): 68-71
  6. [6] Wang Xindong, Wang Yida, Pang International, et al. Research and Application of Intelligent Operations (AIOps) in China Unicom's Distributed Architecture [J]. Telecommunications Engineering Technology and Standardization, 2021, 34 (1): 48-54
  7. [7] LI Y,JIANG Z M J,LI H,et al. Predicting Node Failures in an Ultra-Large-Scale Cloud Computing Platform [J].ACM Transactions on Software Engineering and Methodology(TOSEM).New York:Association for Computing Machinery,2020:1-24.
  8. [8] BRENNER M,GILLMEISTER M. Designing CMDB Data Models with Good Utility and Limited Complexity [C]//IEEE Network Operations and Management Symposium(NOMS).Krakow:IEEE,2014:1-15.
  9. [9] Gao Yu Design and Implementation of a Cloud Native Topology Service System [D]. Sichuan: Southwest Jiaotong University, 2019
  10. [10] VALLI L N,SUJATHA E,RATHINAM J. A Study on Deep Learning Frameworks to Understand the Real Time Fault Detection and Diagnosis in IT Operations with Aiops [C]//2023 International Conference on Evolutionary Algorithms and Soft Computing Techniques(EASCT).Bengaluru:IEEE,2023:1-6.
  11. [11] Huang Wei Research on AIOps Technology Based on Machine Learning [D]. Beijing: Beijing Jiaotong University, 2019
  12. [12] Cao Wei KPI anomaly detection based on multidimensional feature extraction and XGBoost [D]. Liaoning: Dalian Maritime University, 2020
  13. [13] REN H,XU B,WANG Y,et al. Time-Series Anomaly Detection Service at Microsoft [C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining.New York:ACM,2019:3009-3017.
  14. [14] MEHRA P,AHUJA M S,AERI M. Time Series Anomaly Detection System with Linear Neural Network and Autoencoder [C]//2023 International Conference on Device Intelligence,Computing and Communication Technologies (DICCT).Dehradun:IEEE,2023:659-662.
  15. [15] ROUKERD F R,RAJABI M M. Anomaly Detection in Groundwater Monitoring Data Using LSTM-Autoencoder Neural Networks [J/OL].Environmental Monitoring and Sssessment,2024,196(8):692[2025-06-04].https://link.springer.com/article/10.1007/s10661-024-12848-z.
  16. [16] LI Z,ZHAO Y,BOTTA N,et al. COPOD: Copula-Based Outlier Detection [C]//2020 IEEE International Conference on Data Mining (ICDM).Sorrento:IEEE,2021:1118-1123.
  17. [17] LI Z,ZHAO Y,HU X,et al. Ecod: Unsupervised Outlier Detection Using Empirical Cumulative Distribution Functions [J].IEEE Transactions on Automatic Control,2023,35(12):12181-12193.
  18. [18] LIU F T,TING K M,ZHOU Z H. Isolation-Based Anomaly Detection [J].ACM Transactions on Knowledge Discovery from Data,2012,6(1):1-39.
  19. [19] WINTERS P R. Forecasting Sales by Exponentially Weighted Moving Averages [J].Management Ence,1976,6(3):324-342.

About the authors

Wang Yao (1997-), female, Han ethnicity, from Qingdao, Shandong, is an operations and development engineer with a master's degree. Her research interests include machine learning and intelligent operations; Chen Wanyi (1998-), male, Han ethnicity, from Wenzhou, Zhejiang, is an operations and development engineer with a master's degree. His research interests include intelligent operations and artificial intelligence; Xu Junyuan (1987-), male, Han ethnicity, born in Hangzhou, Zhejiang, is a deputy senior engineer with a bachelor's degree. His research focuses on cloud platforms and network security; Nie Kaixun (1989-), male, Yi ethnic, from Guiyang, Guizhou, is the head of the operation monitoring team with a bachelor's degree. His research interests include AIOPS and intelligent operation and maintenance.