2019 IEEE International Conference on Prognostics and Health Management (ICPHM) 2019
DOI: 10.1109/icphm.2019.8819403
|View full text |Cite
|
Sign up to set email alerts
|

Fault Detection and Isolation in Industrial Networks using Graph Convolutional Neural Networks

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
9
0
6

Year Published

2019
2019
2022
2022

Publication Types

Select...
6
1
1

Relationship

1
7

Authors

Journals

citations
Cited by 23 publications
(15 citation statements)
references
References 7 publications
0
9
0
6
Order By: Relevance
“…In addition, the traditional methods do not consider the network structure when designing the fault diagnosis, which causes over-fitting problems. To solve these problems, the connected components in power systems can be represented as a weighted undirected graph structure [72]. Then, local relationships between power variables in different components of the distribution network are explored by GCNs to improve fault detection and isolation.…”
Section: ) Fault Detection and Isolationmentioning
confidence: 99%
“…In addition, the traditional methods do not consider the network structure when designing the fault diagnosis, which causes over-fitting problems. To solve these problems, the connected components in power systems can be represented as a weighted undirected graph structure [72]. Then, local relationships between power variables in different components of the distribution network are explored by GCNs to improve fault detection and isolation.…”
Section: ) Fault Detection and Isolationmentioning
confidence: 99%
“…We define a graph with n nodes as, G = { V , Ā}, where V = [v 1 , v 2 , ...v n ] is the matrix of the node embedding in the graph, v i is the node embedding for node i. The edges of a graph are represented by the adjacency matrix, Ā, where each [29] unsupervised 3D shape retrieval [30] Optimization optimize scheduling performance of flexible manufacturing systems [31] find the optimum scheduling policy for job-shop problems [32] Monitoring detect and isolate faulty components in industrial systems [33] improve product failure prediction [34] share multi-level manufacturing knowledge within a system [35] predict the remaining useful life estimation of industrial equipment [36] value a ij corresponds to the relation between nodes i and j.…”
Section: A Preliminaries On Graph Learningmentioning
confidence: 99%
“…In [31], [32], it is shown that graph learning can be used to optimize scheduling performance during operation. In monitoring systems, graph learning has been shown to improve detection and prediction of failures in both products and equipment [33], [34], [36]. While these demonstrations highlight the opportunities that lie for graph learning in manufacturing systems, the graph learning systems that these works develop are restricted to specific applications.…”
Section: B Graph Learning In Manufacturingmentioning
confidence: 99%
“…We use the simulated water tank system dataset (Khorasgani et al, 2019) to demonstrate and validate the performance of our method. This dataset includes a network of water tanks.…”
Section: Water Tank Data Studymentioning
confidence: 99%
“…But the responses in all these papers are continuous multivariate variables rather than multivariate binary variables, which are not applicable to failure prediction. Other researchers employ existing network structure of physical models (Khorasgani, Farahat, Hasanzade, & Gupta, 2019) as one of the inputs for failure prediction. However, in reality we may not be able to obtain the graph structure inside a system or among devices.…”
Section: Introductionmentioning
confidence: 99%