Tangbin Xia
Papers
3
Total Citations
25
H-Index
3
About
Dr. Tangbin Xia is pioneering the integration of physics-informed deep learning with industrial fault diagnosis, addressing critical challenges in modern manufacturing. His research centers on developing interpretable, data-efficient diagnostic frameworks for industrial robots operating under complex, real-world conditions. Dr. Xia’s most impactful work introduces a hybrid physics-embedded recurrent neural network that fuses physical models with deep learning to achieve robust fault diagnosis from multivariate proprioceptive signals, even under time-varying operational states (15 citations). He further advances the field with an unsupervised motion-based anomaly detection system using graph attention networks, enabling automated fault labeling without costly manual annotation (7 citations). His recent contribution of a deep complex wavelet denoising network tackles the dual challenges of noise interference and imbalanced data, providing both high diagnostic accuracy and physical interpretability—a critical step toward trustworthy AI in manufacturing (3 citations). By bridging the gap between data-driven methods and physical understanding, Dr. Xia’s work is shaping the next generation of intelligent, resilient industrial automation systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3