Yudong Cao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Yudong Cao has a total of 50 co-authors. Here are the top co-authors by number of shared publications:
- Minping Jia
School of Mechanical Engineering, Southeast University, Nanjing, China
9 shared publications - Xiaoli Zhao
Nanjing University of Science and Technology
8 shared publications - Yifei Ding
Suzhou University of Technology
8 shared publications - Xiaoan Yan
School of Mechatronics Engineering, Nanjing Forestry University, Nanjing 210037, People’s Republic of China
5 shared publications Xiaoli Zhao
5 shared publicationsMinping Jia
4 shared publications
Latest Publications
- 24 July 2026
- Published by IOP Publishing in Measurement Science and Technology
- Vol. 37 (30) , 305101
LLM-informed stage-aware and physics-guided remaining useful life prediction for rolling bearings
Remaining useful life (RUL) prediction of rolling bearings is a fundamental task in prognostics and health management. Existing data-driven approaches mainly rely on static training objectives and purely signal-level learning, which limits their ability to adapt to different degradation stages and to incorporate high-level degradation knowledge. An large language model (LLM)-informed RUL prediction framework is proposed, integrating LLM-based degradation reasoning, stage-aware adaptive learning, and physics-guided constraint regularization. Instead of directly processing raw vibration signals, the LLM is employed as a degradation reasoning module that infers structured degradation priors, including degradation stage, degradation trend, and physically meaningful constraints, from semantic health representations. These priors are subsequently embedded into a task-specific neural predictor through a stage-aware adaptive loss updating strategy and physics-inspired constraint losses. By explicitly modeling stage-dependent learning objectives and enforcing physically consistent degradation behavior, the proposed framework achieves more reliable lifecycle-wide RUL prediction. Experimental results demonstrate that the proposed method consistently outperforms existing data-driven approaches in terms of prediction accuracy and stability.
- 24 July 2026
- Published by IOP Publishing in Measurement Science and Technology
- Vol. 37 (30) , 305101
LLM-informed stage-aware and physics-guided remaining useful life prediction for rolling bearings
Remaining useful life (RUL) prediction of rolling bearings is a fundamental task in prognostics and health management. Existing data-driven approaches mainly rely on static training objectives and purely signal-level learning, which limits their ability to adapt to different degradation stages and to incorporate high-level degradation knowledge. An large language model (LLM)-informed RUL prediction framework is proposed, integrating LLM-based degradation reasoning, stage-aware adaptive learning, and physics-guided constraint regularization. Instead of directly processing raw vibration signals, the LLM is employed as a degradation reasoning module that infers structured degradation priors, including degradation stage, degradation trend, and physically meaningful constraints, from semantic health representations. These priors are subsequently embedded into a task-specific neural predictor through a stage-aware adaptive loss updating strategy and physics-inspired constraint losses. By explicitly modeling stage-dependent learning objectives and enforcing physically consistent degradation behavior, the proposed framework achieves more reliable lifecycle-wide RUL prediction. Experimental results demonstrate that the proposed method consistently outperforms existing data-driven approaches in terms of prediction accuracy and stability.
- 6 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Reliability
- Vol. 75 (00189529) , 490-503
Investigation of Bonds Between Network Convolution and Time-Frequency Transforms for Ex-Ante Interpretable Machine Health Prognosis
In the era of big data and intelligent sensing, deep neural networks provide new impetus for prognostics and health management (PHM) with their powerful feature extraction capabilities. However, the pursuit of performance through increased network depth and complexity concurrently escalates the number of hyperparameters and model intricacy, thereby exacerbating the inherent opaque nature and restricting their deployment in complex industrial settings. To address this dilemma, this article develops a machine health prognosis framework with ex-ante interpretability based on complex domain time-frequency network (CDTFN). Specifically, this article first investigates the intrinsic bonds between network convolution and time-frequency transforms. Building upon this foundation, four complex observation operators with trainable parameters are designed for extracting fault-related time-frequency information, embedding it into the CDTFN as a preprocessing layer. Simultaneously, by extending the forward and backward propagation mechanisms of real-valued networks to the complex domain, the proposed CDTFN gains the capability to fuse complex-valued time-frequency information and establish end-to-end mapping from feature representation layers to prediction labels. The effectiveness and accuracy of the proposed prognosis framework based on CDTFN are verified by public and self-built run-to-failure rolling bearings datasets. The detailed experimental results further demonstrate its distinct advantages in interpretability and generalization capability.
- 6 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Reliability
- Vol. 75 (00189529) , 490-503
Investigation of Bonds Between Network Convolution and Time-Frequency Transforms for Ex-Ante Interpretable Machine Health Prognosis
In the era of big data and intelligent sensing, deep neural networks provide new impetus for prognostics and health management (PHM) with their powerful feature extraction capabilities. However, the pursuit of performance through increased network depth and complexity concurrently escalates the number of hyperparameters and model intricacy, thereby exacerbating the inherent opaque nature and restricting their deployment in complex industrial settings. To address this dilemma, this article develops a machine health prognosis framework with ex-ante interpretability based on complex domain time-frequency network (CDTFN). Specifically, this article first investigates the intrinsic bonds between network convolution and time-frequency transforms. Building upon this foundation, four complex observation operators with trainable parameters are designed for extracting fault-related time-frequency information, embedding it into the CDTFN as a preprocessing layer. Simultaneously, by extending the forward and backward propagation mechanisms of real-valued networks to the complex domain, the proposed CDTFN gains the capability to fuse complex-valued time-frequency information and establish end-to-end mapping from feature representation layers to prediction labels. The effectiveness and accuracy of the proposed prognosis framework based on CDTFN are verified by public and self-built run-to-failure rolling bearings datasets. The detailed experimental results further demonstrate its distinct advantages in interpretability and generalization capability.
- 22 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (16) , 27828-27839
Online Knowledge Distillation for Machine Health Prognosis Considering Edge Deployment
Complex neural networks with deep structures are beneficial for solving problems, such as fault classification and health prediction of industrial equipment, due to their powerful feature extraction capabilities. Unfortunately, corresponding complex models designed based on deep learning algorithms require huge computational and memory resources, making them difficult to achieve effective edge deployment. In order to solve this difficulty with practical industrial significance, this article proposes an online knowledge distillation framework for machine health prognosis. Within this framework, the learned knowledge of complex networks can be distilled to simple networks that can be deployed on edge devices in sites. Specifically, the response-based knowledge distillation module, feature-based knowledge distillation module, and relation-based knowledge distillation module are, respectively, designed to achieve effective information transmission from different levels. Furthermore, the inherent differences between simple and complex networks have been fully considered for their impact on the efficiency of knowledge distillation, and an adaptive mutual learning strategy has been contrapuntally proposed to address this limitation. Multiple online knowledge distillation experiments were conducted on two different sets of run-to-failure data sets of mechanical key components with different pairs of complex and simple networks to verify the effectiveness of the proposed framework. The experimental results show that the simple student-networks can effectively improve prediction performance after receiving knowledge distillation from the complex teacher-networks, providing a new solution for machine health prognosis under the premise of edge deployment.
- 22 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (16) , 27828-27839
Online Knowledge Distillation for Machine Health Prognosis Considering Edge Deployment
Complex neural networks with deep structures are beneficial for solving problems, such as fault classification and health prediction of industrial equipment, due to their powerful feature extraction capabilities. Unfortunately, corresponding complex models designed based on deep learning algorithms require huge computational and memory resources, making them difficult to achieve effective edge deployment. In order to solve this difficulty with practical industrial significance, this article proposes an online knowledge distillation framework for machine health prognosis. Within this framework, the learned knowledge of complex networks can be distilled to simple networks that can be deployed on edge devices in sites. Specifically, the response-based knowledge distillation module, feature-based knowledge distillation module, and relation-based knowledge distillation module are, respectively, designed to achieve effective information transmission from different levels. Furthermore, the inherent differences between simple and complex networks have been fully considered for their impact on the efficiency of knowledge distillation, and an adaptive mutual learning strategy has been contrapuntally proposed to address this limitation. Multiple online knowledge distillation experiments were conducted on two different sets of run-to-failure data sets of mechanical key components with different pairs of complex and simple networks to verify the effectiveness of the proposed framework. The experimental results show that the simple student-networks can effectively improve prediction performance after receiving knowledge distillation from the complex teacher-networks, providing a new solution for machine health prognosis under the premise of edge deployment.
- 12 May 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 20 (2) , 1313-1323
Unsupervised Fault Detection With Deep One-Class Classification and Manifold Distribution Alignment
Fault detection or anomaly detection relies heavily on learning from datasets where only normal samples are available, resulting in the emergence of numerous one-class classification (OCC) methods. However, learning discriminative deep representatives with good generalization from cross-domain positive samples remains challenging. Therefore, this work proposes an end-to-end framework, deep transfer one-class classification (DTOCC) for unsupervised fault detection, which combines adversarial generative OCC and distribution alignment from the perspective of manifold learning. Specifically, pseudo-negative samples are generated outside the positive manifold, facilitating the model to learn discrimination with respect to normal and anomaly. Further, cross-domain positive samples are aligned in log-Euclidean manifold space to enhance representation learning. Then, we provide the specific implementations for fault detection and validate its superiority through case studies on multiclass and run-to-failure datasets, simulating both offline and online scenarios.
- 12 May 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 20 (2) , 1313-1323
Unsupervised Fault Detection With Deep One-Class Classification and Manifold Distribution Alignment
Fault detection or anomaly detection relies heavily on learning from datasets where only normal samples are available, resulting in the emergence of numerous one-class classification (OCC) methods. However, learning discriminative deep representatives with good generalization from cross-domain positive samples remains challenging. Therefore, this work proposes an end-to-end framework, deep transfer one-class classification (DTOCC) for unsupervised fault detection, which combines adversarial generative OCC and distribution alignment from the perspective of manifold learning. Specifically, pseudo-negative samples are generated outside the positive manifold, facilitating the model to learn discrimination with respect to normal and anomaly. Further, cross-domain positive samples are aligned in log-Euclidean manifold space to enhance representation learning. Then, we provide the specific implementations for fault detection and validate its superiority through case studies on multiclass and run-to-failure datasets, simulating both offline and online scenarios.
- 19 April 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 72 (00189456) , 1-11
Picture-in-Picture Strategy-Based Complex Graph Neural Network for Remaining Useful Life Prediction of Rotating Machinery
Graph neural networks are increasingly explored in the field of Prognostics and Health Management (PHM) due to their excellent performance when dealing with non-Euclidean data. However, current graph neural networks are mostly based on real domain modeling. In addition, existing graph construction methods rely on the prior positional relationship of multiple sensors. In view of the above, this paper proposes complex graph neural network based on the picture-in-picture strategy (CGNN-PIP) to realize the remaining useful life (RUL) prediction of rotating machinery under multi-channel signals. Specifically, the classical graph convolution operation is upgraded to generalized complex graph convolution, and complex graph neural network is further constructed to extract deep degenerate feature representations. Meanwhile, the picture-in-picture strategy is designed to guide graph construction, which takes the single-path graph as a node of the new graph to build a deeper-level graph. We verified the effectiveness and superiority of the proposed method through two case studies on different run-to-failure datasets. The results show that the proposed CGCN-PIP can reasonably construct the topology map of the complex domain data, and extract the temporal and structural information reflecting the equipment degradation. The comparison with state-of-the-art methods also proves that CGCN-PIP has advantages in terms of prediction accuracy and training consumption.
- 19 April 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 72 (00189456) , 1-11
Picture-in-Picture Strategy-Based Complex Graph Neural Network for Remaining Useful Life Prediction of Rotating Machinery
Graph neural networks are increasingly explored in the field of Prognostics and Health Management (PHM) due to their excellent performance when dealing with non-Euclidean data. However, current graph neural networks are mostly based on real domain modeling. In addition, existing graph construction methods rely on the prior positional relationship of multiple sensors. In view of the above, this paper proposes complex graph neural network based on the picture-in-picture strategy (CGNN-PIP) to realize the remaining useful life (RUL) prediction of rotating machinery under multi-channel signals. Specifically, the classical graph convolution operation is upgraded to generalized complex graph convolution, and complex graph neural network is further constructed to extract deep degenerate feature representations. Meanwhile, the picture-in-picture strategy is designed to guide graph construction, which takes the single-path graph as a node of the new graph to build a deeper-level graph. We verified the effectiveness and superiority of the proposed method through two case studies on different run-to-failure datasets. The results show that the proposed CGCN-PIP can reasonably construct the topology map of the complex domain data, and extract the temporal and structural information reflecting the equipment degradation. The comparison with state-of-the-art methods also proves that CGCN-PIP has advantages in terms of prediction accuracy and training consumption.