Collaborative Fault Prediction and Intelligent Localization of Distributed Operation and Maintenance Platform Based on Spatiotemporal Graph Neural Network

Authors

  • Jianwei Ma Power Dispatching and Control Center of Guizhou Power Grid Co., Ltd., Guiyang 550002, Guizhou, China
  • Tongfu Fu Liupanshui Power Supply Bureau of Guizhou Power Grid Co., Ltd., Liupanshui 553001, Guizhou, China
  • Ming Lei Liupanshui Power Supply Bureau of Guizhou Power Grid Co., Ltd., Liupanshui 553001, Guizhou, China
  • Zhongqiang Zhou Power Dispatching and Control Center of Guizhou Power Grid Co., Ltd., Guiyang 550002, Guizhou, China
  • Ling Liang Power Dispatching and Control Center of Guizhou Power Grid Co., Ltd., Guiyang 550002, Guizhou, China

Abstract

Traditional approaches to Distributed Systems Operations and Maintenance cannot adequately predict faults or locate root causes. To solve these problems, this paper presents a Spatio-Temporal Graph Neural Network (STGNN) as an intelligent framework. The STGNN framework leverages dynamic graphs for both the topological and the state evolution models of the distributed system. The prediction component utilizes deep learning algorithms. The localization component applies cross-node attention and graph reasoning to automatically reconstruct the chain of fault propagation and identify the node that caused the fault. Experimental results indicate that the proposed approach achieves a fault prediction accuracy of 90.2%, and reduces the average root cause localization time for write latency in databases by 35.6%. This paper provides a more effective and reliable means of enabling intelligent operations and maintenance in distributed systems.

Keywords: distributed operations management; spatio temporal graph neural network; dynamic graph structure; cross-node attention; fault propagation chain

Cite As

J. Ma, T. Fu, M. Lei, Z. Zhou, L. Liang, "Collaborative Fault Prediction and Intelligent Localization of Distributed Operation and
Maintenance Platform Based on Spatiotemporal Graph Neural Network", Engineering Intelligent Systems, vol. 34 no. 5,
pp. 607-620, 2026.





Downloads

Published

2026-09-01

Issue

Vol. 34 No. 5 (2026): International Journal of Engineering Intelligent Systems

Section

Article

License

The submission of a paper implies that, if accepted for publication, it will not be published elsewhere in the same form, in any language, without the prior consent of the publisher. Before publication, authors are requested to assign copyright to CRL Publishing Ltd. This allows CRL to sanction photocopying, and to authorize the reprinting of issues or volumes according to demand. Authors' traditional rights will not be jeopardized by assigning Copyright in this way, as they retain the right to reuse the material following publication, and to veto third-party publication.