Xiaobao Song
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Xiaobao Song has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Hao Wang
Guangdong Key Laboratory of Intelligent Information Processing, Shenzhen University, Shenzhen, China
3 shared publications - Wenming Cao
State Key Laboratory of Radio Frequency Heterogeneous Integration, Shenzhen University, Shenzhen, China
3 shared publications - Zeeshan Abbas
Guangdong Key Laboratory of Intelligent Information Processing, Shenzhen University, Shenzhen, China
2 shared publications - Chi-Sing Leung
Department of Electrical Engineering, City University of Hong Kong, Kowloon Tong, Hong Kong
1 shared publication - Dong Wang
Guangdong Provincial Key Laboratory of Intelligent Information Processing, Shenzhen University, Shenzhen, China
1 shared publication - Hongbo Qiu
Guangdong Provincial Key Laboratory of Intelligent Information Processing, Shenzhen University, Shenzhen, China
1 shared publication
Latest Publications
- 15 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. PP (15513203) , 1-11
Aging-Informed Dynamic Graph Network for Aero-Engine RUL Prediction
Remaining useful life (RUL) prediction for safety-critical systems, such as aero-engines, is a fundamental challenge in prognostics and health management. While existing deep learning methods have achieved success in modeling spatiotemporal patterns, they often overlook the physical essence of system degradation: the evolution of intersensor dependencies is intrinsically driven by the aging progression. To address this, we propose aging-informed dynamic graph network for RUL prediction (AIDGN-RUL), which formulates RUL prediction as an aging-guided dynamic graph construction problem. Specifically, the model derives an aging representation from the operational cycle count and integrates it with multivariate sensor features to generate adaptive graph structures that evolve with system aging. A graph convolutional architecture is then employed to propagate information over aging-guided graph structures to obtain refined sensor representations for RUL prediction. Extensive experiments on the CMAPSS and N-CMAPSS datasets demonstrate that AIDGN-RUL significantly outperforms state-of-the-art baselines, achieving a maximum improvement of 27.83% in prognostic score and 15.94% in root mean square error (RMSE). Furthermore, the model exhibits superior robustness to sensor noise and maintains high computational efficiency, highlighting its potential for real-world industrial deployment.
- 15 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. PP (15513203) , 1-11
Aging-Informed Dynamic Graph Network for Aero-Engine RUL Prediction
Remaining useful life (RUL) prediction for safety-critical systems, such as aero-engines, is a fundamental challenge in prognostics and health management. While existing deep learning methods have achieved success in modeling spatiotemporal patterns, they often overlook the physical essence of system degradation: the evolution of intersensor dependencies is intrinsically driven by the aging progression. To address this, we propose aging-informed dynamic graph network for RUL prediction (AIDGN-RUL), which formulates RUL prediction as an aging-guided dynamic graph construction problem. Specifically, the model derives an aging representation from the operational cycle count and integrates it with multivariate sensor features to generate adaptive graph structures that evolve with system aging. A graph convolutional architecture is then employed to propagate information over aging-guided graph structures to obtain refined sensor representations for RUL prediction. Extensive experiments on the CMAPSS and N-CMAPSS datasets demonstrate that AIDGN-RUL significantly outperforms state-of-the-art baselines, achieving a maximum improvement of 27.83% in prognostic score and 15.94% in root mean square error (RMSE). Furthermore, the model exhibits superior robustness to sensor noise and maintains high computational efficiency, highlighting its potential for real-world industrial deployment.
- Open Access
- 26 December 2025
- Published by Elsevier in Results in Engineering
- Vol. 29
Remaining useful life estimation based on improved attention mechanism
- 26 December 2025
- Published by Elsevier in Results in Engineering
- Vol. 29
Open AccessRemaining useful life estimation based on improved attention mechanism
- 4 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 37 (5) , 2223-2234
D2Vformer: A Flexible Time-Series Prediction Model Based on Time-Position Embedding
Existing time-series forecasting methods often struggle to adapt to dynamic scenarios and lack flexibility in prediction. They typically require retraining the model when the prediction length or position changes. Moreover, these methods still face challenges in effectively capturing and utilizing time-position embeddings (PEs). To address these limitations, this article proposes a novel model called D2Vformer. Unlike conventional prediction methods that rely on fixed-length predictors, D2Vformer can directly handle scenarios with arbitrary prediction lengths. In addition, it significantly reduces training resource consumption and proves highly effective in real-world dynamic environments. In D2Vformer, the Date2Vec (D2V) module is devised to leverage timestamp information and feature sequences to generate time PEs. Subsequently, D2Vformer introduces an innovative fusion module that leverages an attention mechanism to capture the mapping between input and target time PEs, thereby enabling flexible prediction. Extensive experiments on six datasets demonstrate that D2V outperforms other time-PE methods, while D2Vformer surpasses state-of-the-art approaches in both fixed-length and arbitrary-length prediction tasks. The code for D2Vformer is available at: https://github.com/TeamofHaoWang/D2Vformer
- 4 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 37 (5) , 2223-2234
D2Vformer: A Flexible Time-Series Prediction Model Based on Time-Position Embedding
Existing time-series forecasting methods often struggle to adapt to dynamic scenarios and lack flexibility in prediction. They typically require retraining the model when the prediction length or position changes. Moreover, these methods still face challenges in effectively capturing and utilizing time-position embeddings (PEs). To address these limitations, this article proposes a novel model called D2Vformer. Unlike conventional prediction methods that rely on fixed-length predictors, D2Vformer can directly handle scenarios with arbitrary prediction lengths. In addition, it significantly reduces training resource consumption and proves highly effective in real-world dynamic environments. In D2Vformer, the Date2Vec (D2V) module is devised to leverage timestamp information and feature sequences to generate time PEs. Subsequently, D2Vformer introduces an innovative fusion module that leverages an attention mechanism to capture the mapping between input and target time PEs, thereby enabling flexible prediction. Extensive experiments on six datasets demonstrate that D2V outperforms other time-PE methods, while D2Vformer surpasses state-of-the-art approaches in both fixed-length and arbitrary-length prediction tasks. The code for D2Vformer is available at: https://github.com/TeamofHaoWang/D2Vformer