Yifeng Pan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Yifeng Pan has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
- Peng Zhao
The State Key Laboratory of Fluid Power and Mechatronic Systems, College of Mechanical Engineering Zhejiang University Hangzhou China
8 shared publications Chengqian Zhang
3 shared publicationsChen Zheng
2 shared publicationsDong Yang
2 shared publicationsHao Shi
2 shared publications- Jianzhong Fu
State Key Laboratory of Fluid Power & Mechatronic Systems, College of Mechanical Engineering, Zhejiang University, Hangzhou, People’s Republic of China
2 shared publications
Latest Publications
- Open Access
- 8 July 2026
- Published by Elsevier in SSRN Electronic Journal
Efficient wavenumber method for full-matrix ultrasonic imaging in anisotropic composites
Benefiting from superior resolution, ultrasonic full-matrix imaging is widely adopted for internal defect measurement of CFRP components. However, the pronounced acoustic anisotropy of CFRP degrades the accuracy of total focusing method (TFM), the gold standard in conventional full-matrix imaging. Although several travel-time correction schemes have been proposed for anisotropic TFM in CFRP, their pixel-wise traversal leads to low computational efficiency. To balance imaging accuracy and efficiency in internal detection of CFRP components, this paper proposes an anisotropic phase shift migration imaging method. Based on the dispersion relation and anisotropic characteristics of CFRP, the anisotropic vertical wavenumber is derived without prior material information and incorporated into the anisotropic phase shift operator for wavefield reconstruction at discrete depths in the frequency-wavenumber domain. Combined with an isotropic phase-shift operator, the method also enables imaging of isotropic/CFRP composite structures. The model accuracy is validated using both simulated and experimental wavefields. Furthermore, numerical simulations and experiments are conducted to evaluate its performance on CFRP laminates and composite structures. Experimental results reveal that the proposed method completes defect measurement of CFRP composite structures within only 0.34s, merely 1/147166 of the time cost of anisotropy-corrected TFM. The proposed method possesses great application potential for high-efficiency internal defect quantification of CFRP components and composite structures.
- 8 July 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessEfficient wavenumber method for full-matrix ultrasonic imaging in anisotropic composites
Benefiting from superior resolution, ultrasonic full-matrix imaging is widely adopted for internal defect measurement of CFRP components. However, the pronounced acoustic anisotropy of CFRP degrades the accuracy of total focusing method (TFM), the gold standard in conventional full-matrix imaging. Although several travel-time correction schemes have been proposed for anisotropic TFM in CFRP, their pixel-wise traversal leads to low computational efficiency. To balance imaging accuracy and efficiency in internal detection of CFRP components, this paper proposes an anisotropic phase shift migration imaging method. Based on the dispersion relation and anisotropic characteristics of CFRP, the anisotropic vertical wavenumber is derived without prior material information and incorporated into the anisotropic phase shift operator for wavefield reconstruction at discrete depths in the frequency-wavenumber domain. Combined with an isotropic phase-shift operator, the method also enables imaging of isotropic/CFRP composite structures. The model accuracy is validated using both simulated and experimental wavefields. Furthermore, numerical simulations and experiments are conducted to evaluate its performance on CFRP laminates and composite structures. Experimental results reveal that the proposed method completes defect measurement of CFRP composite structures within only 0.34s, merely 1/147166 of the time cost of anisotropy-corrected TFM. The proposed method possesses great application potential for high-efficiency internal defect quantification of CFRP components and composite structures.
- Open Access
- 9 April 2026
- Published by Elsevier in SSRN Electronic Journal
Efficiency ultrasound full-matrix imaging of heterogeneous-anisotropic CFRP laminate multilayer composite structure internal defects
Carbon Fiber Reinforced Polymer (CFRP) components are widely used in engineering for their high strength, high modulus and lightweight properties to enhance product performance. Ultrasonic full-matrix imaging is an effective method to identify the defects in various components, but the Quasi-Pressure (QP) velocity anisotropy of CFRP severely impairs measurement accuracy and efficiency. The total focusing method (TFM) can deal with the velocity anisotropy through ray-tracing method, but the high computational complexity makes the method very time-consuming. To address this issue, an anisotropic phase shift migration is proposed in this study. The anisotropic vertical wavenumber is calibrated by the QP velocity. Then, the wavefield extrapolation in anisotropic medium is implemented by phase shift migration in the frequency-wavenumber domain. The proposed method is validated by simulations and experiments on diverse CFRP laminates, wherein the imaging quality of the proposed method is comparable with the ray-based homogeneous model TFM while the time cost is several orders of magnitude smaller than the latter (on the order of 10-6). Therefore, it shows great potential in defect detection for industrial CFRP products.
- 9 April 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessEfficiency ultrasound full-matrix imaging of heterogeneous-anisotropic CFRP laminate multilayer composite structure internal defects
Carbon Fiber Reinforced Polymer (CFRP) components are widely used in engineering for their high strength, high modulus and lightweight properties to enhance product performance. Ultrasonic full-matrix imaging is an effective method to identify the defects in various components, but the Quasi-Pressure (QP) velocity anisotropy of CFRP severely impairs measurement accuracy and efficiency. The total focusing method (TFM) can deal with the velocity anisotropy through ray-tracing method, but the high computational complexity makes the method very time-consuming. To address this issue, an anisotropic phase shift migration is proposed in this study. The anisotropic vertical wavenumber is calibrated by the QP velocity. Then, the wavefield extrapolation in anisotropic medium is implemented by phase shift migration in the frequency-wavenumber domain. The proposed method is validated by simulations and experiments on diverse CFRP laminates, wherein the imaging quality of the proposed method is comparable with the ray-based homogeneous model TFM while the time cost is several orders of magnitude smaller than the latter (on the order of 10-6). Therefore, it shows great potential in defect detection for industrial CFRP products.
- 25 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 25 (17) , 32301-32311
A Semi-Physical Simulation System for Evaluation of Cardiopulmonary Resuscitation Mechanical Compression Parameters Based on Fracture Risk and Blood Perfusion
High-quality cardiopulmonary resuscitation (CPR) is a critical determinant of survival following cardiac arrest. In recent years, mechanical compression has become increasingly prevalent in the emergency management of cardiac arrest. The settings of key compression parameters strongly influence the effectiveness of chest compression. This study developed a semi-physical simulation platform and evaluation criteria to assess the optimal parameters for CPR. A multi-spring system was designed to simulate the risk of sternal fractures during chest compression. Additionally, a blood flow model was constructed to simulate blood perfusion. The evaluation criteria, which include quantifying sternal fracture risk and blood perfusion, are used to calculate the compression effect by inputting the compression force and depth data into the evaluation model. (Analysis of Variance)ANOVA analysis demonstrated statistically significant impacts of different compression parameters on compression outcomes. The results demonstrated that the mechanical waveform data more accurately reflected the compression dynamics encountered in real-world CPR circumstances. The trapezoidal compression waveform demonstrated clear superiority over triangle and sine waveforms, enhancing blood circulation. This study’s exploration of the trapezoidal waveform fills a gap in the AHA guidelines. In addition to the waveform, the study confirmed that a compression depth of 50 mm and a frequency of 120 compressions per minute yielded the most effective hemodynamic outcomes. These findings validated and expanded upon the AHA guidelines, offering a novel and comprehensive approach to optimising CPR effectiveness, improving both patient survival rates and the quality of mechanical resuscitation.
- 25 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 25 (17) , 32301-32311
A Semi-Physical Simulation System for Evaluation of Cardiopulmonary Resuscitation Mechanical Compression Parameters Based on Fracture Risk and Blood Perfusion
High-quality cardiopulmonary resuscitation (CPR) is a critical determinant of survival following cardiac arrest. In recent years, mechanical compression has become increasingly prevalent in the emergency management of cardiac arrest. The settings of key compression parameters strongly influence the effectiveness of chest compression. This study developed a semi-physical simulation platform and evaluation criteria to assess the optimal parameters for CPR. A multi-spring system was designed to simulate the risk of sternal fractures during chest compression. Additionally, a blood flow model was constructed to simulate blood perfusion. The evaluation criteria, which include quantifying sternal fracture risk and blood perfusion, are used to calculate the compression effect by inputting the compression force and depth data into the evaluation model. (Analysis of Variance)ANOVA analysis demonstrated statistically significant impacts of different compression parameters on compression outcomes. The results demonstrated that the mechanical waveform data more accurately reflected the compression dynamics encountered in real-world CPR circumstances. The trapezoidal compression waveform demonstrated clear superiority over triangle and sine waveforms, enhancing blood circulation. This study’s exploration of the trapezoidal waveform fills a gap in the AHA guidelines. In addition to the waveform, the study confirmed that a compression depth of 50 mm and a frequency of 120 compressions per minute yielded the most effective hemodynamic outcomes. These findings validated and expanded upon the AHA guidelines, offering a novel and comprehensive approach to optimising CPR effectiveness, improving both patient survival rates and the quality of mechanical resuscitation.
- 2 August 2025
- Published by Elsevier in Journal of Manufacturing Systems
- Vol. 82, 766-785
DAE-SWnet: Unsupervised internal defect segmentation through infrared thermography with scarce samples
- 2 August 2025
- Published by Elsevier in Journal of Manufacturing Systems
- Vol. 82, 766-785
DAE-SWnet: Unsupervised internal defect segmentation through infrared thermography with scarce samples
- 4 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (16) , 33671-33685
Kube-IPM: A Kubernetes-Native Platform for Industrial Process Monitoring With Heterogeneous Sensor Data and Delay-Sensitive Prediction
Large-scale monitoring represents a critical frontier in industrial IoT research, yet real-time tracking of heterogeneous machinery with prognostic capabilities remains a persistent challenge. Unresolved complexities include divergent feature dimensions across machinery, nonlinear interdependencies in multivariate data streams, and dynamic temporal behavior. To overcome these limitations, we introduce Kube-IPM, a Kubernetes-native platform designed for large-scale industrial process monitoring. First, we propose the attention-based padding (ABP) method that generates meaningful feature extensions instead of dummy values, preserving data integrity. Then we apply a temporal attention layer within an encoder-decoder architecture addressing time delays and variable coupling. Finally, we establish a Kubernetes-orchestrated cloud-edge deployment enabling scalable real-time monitoring. Taking the injection molding barrel heating process as an example, we validate our approach using a comprehensive dataset collected from two injection molding machines, one experimental platform, and two simulation models, featuring variable sensor configurations (4-6). We conduct comparisons with state-of-the-art predictive and padding methods. Experiments validate model’s performance, yielding MAE of 0.13± 0.003 (single-step) and 0.26± 0.009 (multi-step). It exceeds benchmarks by 38.1% and 38.0% against top predictive models, with 18.8% and 7.1% improvements over advanced padding techniques. Attention weight analysis validates that the temporal attention module effectively identifies temporal dependencies, enhancing interpretability in forecasting. Furthermore, Kube-IPM maintains 12 concurrent replicas within sampling intervals. Operational costs average $65 per machine, with marginal scaling expenses of $55 per unit, ensuring cost-efficiency for scalability. By integrating cloud-edge and attention-driven prediction technologies, this study constructs an intelligent manufacturing framework with both academic frontier and engineering implementation. Code availability: https://github.com/MangoLab-ZJU/ABP-TA-LSTM.git.
- 4 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (16) , 33671-33685
Kube-IPM: A Kubernetes-Native Platform for Industrial Process Monitoring With Heterogeneous Sensor Data and Delay-Sensitive Prediction
Large-scale monitoring represents a critical frontier in industrial IoT research, yet real-time tracking of heterogeneous machinery with prognostic capabilities remains a persistent challenge. Unresolved complexities include divergent feature dimensions across machinery, nonlinear interdependencies in multivariate data streams, and dynamic temporal behavior. To overcome these limitations, we introduce Kube-IPM, a Kubernetes-native platform designed for large-scale industrial process monitoring. First, we propose the attention-based padding (ABP) method that generates meaningful feature extensions instead of dummy values, preserving data integrity. Then we apply a temporal attention layer within an encoder-decoder architecture addressing time delays and variable coupling. Finally, we establish a Kubernetes-orchestrated cloud-edge deployment enabling scalable real-time monitoring. Taking the injection molding barrel heating process as an example, we validate our approach using a comprehensive dataset collected from two injection molding machines, one experimental platform, and two simulation models, featuring variable sensor configurations (4-6). We conduct comparisons with state-of-the-art predictive and padding methods. Experiments validate model’s performance, yielding MAE of 0.13± 0.003 (single-step) and 0.26± 0.009 (multi-step). It exceeds benchmarks by 38.1% and 38.0% against top predictive models, with 18.8% and 7.1% improvements over advanced padding techniques. Attention weight analysis validates that the temporal attention module effectively identifies temporal dependencies, enhancing interpretability in forecasting. Furthermore, Kube-IPM maintains 12 concurrent replicas within sampling intervals. Operational costs average $65 per machine, with marginal scaling expenses of $55 per unit, ensuring cost-efficiency for scalability. By integrating cloud-edge and attention-driven prediction technologies, this study constructs an intelligent manufacturing framework with both academic frontier and engineering implementation. Code availability: https://github.com/MangoLab-ZJU/ABP-TA-LSTM.git.
- 29 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 21 (5) , 3565-3574
Temporal Re-Attention LSTM for Multivariate Prediction in Industrial Heating Process With Large Time Delays
The heating process contributes to over 60% of the energy consumption in material manufacturing, such as injection molding, metallurgy, die casting, etc. Temperature prediction is the crucial first step to realize energy management. However, the complex heating process poses challenges, such as strong multivariate coupling, large time delay, and frequent changes of heat transfer characteristics. To address these issues, this article introduces a temporal re-attention long short-term memory (TRA-LSTM) method for multivariate time series prediction. We adapt an encoder–decoder framework. A temporal attention module is utilized for extracting key historical features. A re-attention incorporated loss layer is proposed to rectify the temporal attention. TRA-LSTM's performance is compared with five advanced time series prediction methods across six predictions on four real-world datasets collected, including data from the injection molding barrel heating process and the battery thermal runaway process. Results show that TRA-LSTM consistently outperformed the other methods with high accuracy and efficiency. The alignment of re-attention weights with tested time delays in both heating and cooling processes further demonstrates TRA-LSTM's advantage. Ablation experiments underscore the significant contribution of the re-attention module, while hyperparameter sensitivity analysis offers guidance on selection and validates TRA-LSTM's robustness. Overall, the proposed method enables precise multivariate time series prediction and shows substantial potential for optimizing energy management in industrial production.
- 29 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 21 (5) , 3565-3574
Temporal Re-Attention LSTM for Multivariate Prediction in Industrial Heating Process With Large Time Delays
The heating process contributes to over 60% of the energy consumption in material manufacturing, such as injection molding, metallurgy, die casting, etc. Temperature prediction is the crucial first step to realize energy management. However, the complex heating process poses challenges, such as strong multivariate coupling, large time delay, and frequent changes of heat transfer characteristics. To address these issues, this article introduces a temporal re-attention long short-term memory (TRA-LSTM) method for multivariate time series prediction. We adapt an encoder–decoder framework. A temporal attention module is utilized for extracting key historical features. A re-attention incorporated loss layer is proposed to rectify the temporal attention. TRA-LSTM's performance is compared with five advanced time series prediction methods across six predictions on four real-world datasets collected, including data from the injection molding barrel heating process and the battery thermal runaway process. Results show that TRA-LSTM consistently outperformed the other methods with high accuracy and efficiency. The alignment of re-attention weights with tested time delays in both heating and cooling processes further demonstrates TRA-LSTM's advantage. Ablation experiments underscore the significant contribution of the re-attention module, while hyperparameter sensitivity analysis offers guidance on selection and validates TRA-LSTM's robustness. Overall, the proposed method enables precise multivariate time series prediction and shows substantial potential for optimizing energy management in industrial production.
- 15 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 20 (7) , 9298-9307
A Multiphase Dual Attention-Based LSTM Neural Network for Industrial Product Quality Prediction
Quality prediction plays an essential role in industrial process monitoring and optimization. However, the existed methods lack of the ability to extract features in multiphase and nonlinear time series data. To improve the quality prediction accuracy and simultaneously provide some suggestion of the process mechanism, this article proposes a phased dual-attention long short-term memory (PDA-LSTM) neural network, which consists of a phase segmentation module and a prediction module. The whole series data are first divided into phases in the phase segmentation module, and individual models are subsequently constructed for each phase in the prediction module to capture distinct patterns. The LSTM network with input and temporal attention mechanism is utilized to select the most relevant variables and time steps throughout the time sequence. The proposed method is verified with a numerical example and a real injection molding process. Compared with the existed methods, the PDA-LSTM method has the advantages of higher prediction accuracy and the ability to interpret the key input variable and time points of the process data.
- 15 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 20 (7) , 9298-9307
A Multiphase Dual Attention-Based LSTM Neural Network for Industrial Product Quality Prediction
Quality prediction plays an essential role in industrial process monitoring and optimization. However, the existed methods lack of the ability to extract features in multiphase and nonlinear time series data. To improve the quality prediction accuracy and simultaneously provide some suggestion of the process mechanism, this article proposes a phased dual-attention long short-term memory (PDA-LSTM) neural network, which consists of a phase segmentation module and a prediction module. The whole series data are first divided into phases in the phase segmentation module, and individual models are subsequently constructed for each phase in the prediction module to capture distinct patterns. The LSTM network with input and temporal attention mechanism is utilized to select the most relevant variables and time steps throughout the time sequence. The proposed method is verified with a numerical example and a real injection molding process. Compared with the existed methods, the PDA-LSTM method has the advantages of higher prediction accuracy and the ability to interpret the key input variable and time points of the process data.