TY - JOUR
T1 - Adaptive Attention-Driven Few-Shot Learning for Robust Fault Diagnosis
AU - Wang, Zhe
AU - Ding, Yi
AU - Han, Te
AU - Xu, Qiang
AU - Yan, Hong
AU - Xie, Min
PY - 2024/8/15
Y1 - 2024/8/15
N2 - The deep learning techniques have propelled significant advancements in intelligent fault diagnosis. However, the limited labeled data due to resource-intensive labeling processes poses the challenges for actual applications. This study proposes an attention-centric model for few-shot fault diagnosis in rotating machinery. The model is informed by few-shot learning and integrates internal and external attention mechanisms, which are leveraged to enhance the feature extraction capability. Performance evaluations under the 5-way 1-shot setting achieve remarkable results. The accuracy reaches 97.147% for the scenario from artificial damage to real damage, and 95.613% for the scenario of different operational conditions. The critical role of the integrated attention modules is further validated through the ablation study. Comparative analysis with state-of-the-art techniques demonstrate the superior performance of the proposed model. In short, this work provides an alternative method for fault diagnosis under the few-shot limitation. © 2024 IEEE.
AB - The deep learning techniques have propelled significant advancements in intelligent fault diagnosis. However, the limited labeled data due to resource-intensive labeling processes poses the challenges for actual applications. This study proposes an attention-centric model for few-shot fault diagnosis in rotating machinery. The model is informed by few-shot learning and integrates internal and external attention mechanisms, which are leveraged to enhance the feature extraction capability. Performance evaluations under the 5-way 1-shot setting achieve remarkable results. The accuracy reaches 97.147% for the scenario from artificial damage to real damage, and 95.613% for the scenario of different operational conditions. The critical role of the integrated attention modules is further validated through the ablation study. Comparative analysis with state-of-the-art techniques demonstrate the superior performance of the proposed model. In short, this work provides an alternative method for fault diagnosis under the few-shot limitation. © 2024 IEEE.
KW - Attention mechanism
KW - deep learning
KW - fault diagnosis
KW - few-shot learning (FSL)
KW - rotating machinery
KW - Sensors
KW - Task analysis
KW - Testing
KW - Training
UR - http://www.scopus.com/inward/record.url?scp=85198236096&partnerID=8YFLogxK
UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85198236096&origin=recordpage
U2 - 10.1109/JSEN.2024.3421242
DO - 10.1109/JSEN.2024.3421242
M3 - RGC 21 - Publication in refereed journal
SN - 1530-437X
VL - 24
SP - 26034
EP - 26043
JO - IEEE Sensors Journal
JF - IEEE Sensors Journal
IS - 16
ER -