Diagnosis of Device Exception Based on Causality of Device Indicators
Existing intelligent exception diagnosis methods face challenges in unstable feature extraction and difficulty detecting anomalies in high-dimensional spaces. To address these limitations, we propose CGNN—a novel anomaly detection framework integrating causal reasoning with graph neural n...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11048560/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850094278773047296 |
|---|---|
| author | Zhaohui Wang Yan Wei Longhua Shang Shiwei Zhang Shixiong Bao Zhengren Li |
| author_facet | Zhaohui Wang Yan Wei Longhua Shang Shiwei Zhang Shixiong Bao Zhengren Li |
| author_sort | Zhaohui Wang |
| collection | DOAJ |
| description | Existing intelligent exception diagnosis methods face challenges in unstable feature extraction and difficulty detecting anomalies in high-dimensional spaces. To address these limitations, we propose CGNN—a novel anomaly detection framework integrating causal reasoning with graph neural networks. The framework operates in three stages: 1) Causality Detection (CD) employs the LASAR algorithm to construct sparse causal graphs from monitoring variables. 2) Prediction of Device Status (PODS) leverages GraphSAGE to extract spatio-temporal features from causal graphs for state prediction. 3) Diagnosis of Exception (DOE) utilizes kernel density estimation (KDE) for distribution-agnostic anomaly scoring. Experimental validation on the extended-TE dataset demonstrates that our CGNN framework outperforms baseline algorithms, achieving faster fault identification and higher prediction accuracy. Crucially, CGNN requires minimal prior causal knowledge and adapts to arbitrary data distributions, significantly enhancing interpretability and reliability in complex industrial systems. |
| format | Article |
| id | doaj-art-1992612f56ee4c6095aab28f87dbb091 |
| institution | DOAJ |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-1992612f56ee4c6095aab28f87dbb0912025-08-20T02:41:42ZengIEEEIEEE Access2169-35362025-01-011311030711032110.1109/ACCESS.2025.358234611048560Diagnosis of Device Exception Based on Causality of Device IndicatorsZhaohui Wang0Yan Wei1Longhua Shang2Shiwei Zhang3Shixiong Bao4Zhengren Li5https://orcid.org/0000-0003-2908-2233China Satellite Network Digital Technology Company Ltd., Xiong’an, ChinaChina Satellite Network Digital Technology Company Ltd., Xiong’an, ChinaChina Satellite Network Digital Technology Company Ltd., Xiong’an, ChinaChina Satellite Network Digital Technology Company Ltd., Xiong’an, ChinaChina Satellite Network Digital Technology Company Ltd., Xiong’an, ChinaSchool of Economics and Management, Beijing University of Posts and Telecommunications, Beijing, ChinaExisting intelligent exception diagnosis methods face challenges in unstable feature extraction and difficulty detecting anomalies in high-dimensional spaces. To address these limitations, we propose CGNN—a novel anomaly detection framework integrating causal reasoning with graph neural networks. The framework operates in three stages: 1) Causality Detection (CD) employs the LASAR algorithm to construct sparse causal graphs from monitoring variables. 2) Prediction of Device Status (PODS) leverages GraphSAGE to extract spatio-temporal features from causal graphs for state prediction. 3) Diagnosis of Exception (DOE) utilizes kernel density estimation (KDE) for distribution-agnostic anomaly scoring. Experimental validation on the extended-TE dataset demonstrates that our CGNN framework outperforms baseline algorithms, achieving faster fault identification and higher prediction accuracy. Crucially, CGNN requires minimal prior causal knowledge and adapts to arbitrary data distributions, significantly enhancing interpretability and reliability in complex industrial systems.https://ieeexplore.ieee.org/document/11048560/Exception diagnosiscausal reasoninggraph neural networks |
| spellingShingle | Zhaohui Wang Yan Wei Longhua Shang Shiwei Zhang Shixiong Bao Zhengren Li Diagnosis of Device Exception Based on Causality of Device Indicators IEEE Access Exception diagnosis causal reasoning graph neural networks |
| title | Diagnosis of Device Exception Based on Causality of Device Indicators |
| title_full | Diagnosis of Device Exception Based on Causality of Device Indicators |
| title_fullStr | Diagnosis of Device Exception Based on Causality of Device Indicators |
| title_full_unstemmed | Diagnosis of Device Exception Based on Causality of Device Indicators |
| title_short | Diagnosis of Device Exception Based on Causality of Device Indicators |
| title_sort | diagnosis of device exception based on causality of device indicators |
| topic | Exception diagnosis causal reasoning graph neural networks |
| url | https://ieeexplore.ieee.org/document/11048560/ |
| work_keys_str_mv | AT zhaohuiwang diagnosisofdeviceexceptionbasedoncausalityofdeviceindicators AT yanwei diagnosisofdeviceexceptionbasedoncausalityofdeviceindicators AT longhuashang diagnosisofdeviceexceptionbasedoncausalityofdeviceindicators AT shiweizhang diagnosisofdeviceexceptionbasedoncausalityofdeviceindicators AT shixiongbao diagnosisofdeviceexceptionbasedoncausalityofdeviceindicators AT zhengrenli diagnosisofdeviceexceptionbasedoncausalityofdeviceindicators |