Journal Paper: Cross-correlation heat-maps synthesis with Walsh-Fourier transformation in articulated robotic systems diagnostics
Admin 
Feb. 14, 2025 
Comments
Our lab has presented research introducing a novel diagnostic framework for articulated robotic systems in the journal Measurement (SCI, JCR Q1, 中科院 2区, IF(2025) = 5.6). The paper, titled “Cross-correlation heat-maps synthesis with Walsh-Fourier transformation in articulated robotic systems diagnostics,” (DOI: 10.1016/j.measurement.2026.120862) addresses the critical challenges of monitoring harmonic drives under nonstationary dynamics and variable speed/load conditions. Real-world robotic vibration signals are notoriously difficult to decipher because practical deployments often lack a priori characteristic fault frequencies and complete operating context, which typically undermines the robustness of traditional vibration-based diagnostics. Existing methods often become unstable across speed variations, making the accurate isolation of mechanical wear a significant hurdle for automated systems.
To overcome these obstacles, we proposed a comprehensive framework known as Cross-correlation Heatmap Synthesis via Walsh–Fourier Transform (CHS-WFT). This method hinges on speed-aware partitioning with frame-level decomposition, together with a joint Walsh–Fourier cross-correlation representation that fuses Fourier’s sensitivity to harmonic content with Walsh’s responsiveness to abrupt, piecewise transients. By revealing inter-signal dependencies obscured in Fourier-only analyses, the framework employs multi-scale energy entropy to quantify irregularity and complexity across scales. Experimental results confirm that this approach achieves higher diagnostic discriminability and stabilized detection with a reduced dependence on labeled data, enabling reliable early-warning fault detection to support condition-based maintenance in articulated robotic systems.
Journal Paper: Vibration-weighted maximum correlated kurtosis deconvolution and latent cyclic pattern discovery for fault diagnosis of high-speed rail bogies
Admin Jan. 08, 2025 Comments
Our lab has presented research introducing a novel multi-source vibration demodulation framework in Journal of Sound and Vibration (SCI, JCR Q1, 中科院(工程技术) 2区, IF(2025) = 4.9). The paper, titled “Vibration-weighted maximum correlated kurtosis deconvolution and latent cyclic pattern discovery for fault diagnosis of high-speed rail bogies,” (DOI: 10.1016/j.jsv.2026.119657), addresses critical challenges in ensuring the safety of high-speed rail systems. Real-world bogie vibration signals are notoriously difficult to decipher due to a complex mix of mechanical oscillations, deterministic periodic components, and heavy environmental noise, which often obscure fault features. Traditional diagnosis methods are further hindered by the need for manual hyperparameter tuning, poor performance at low Signal-to-Noise Ratios (SNR), and high computational costs, making efficient and accurate fault isolation a significant hurdle.
To overcome these obstacles, we proposed a comprehensive framework combining an optimized Maximum Correlated Kurtosis Deconvolution (MCKD) with Latent Cyclic Pattern Discovery (LCPD). This method introduces a Vibration Amplitude-based Grading and Weighting Distribution (VAGWD) and spectral-negentropy-driven adaptivity, allowing for automatic filter length determination and enhanced extraction of weak impulsive signatures without empirical tuning. By exploiting envelope-cepstral cues and cyclic spectral coherence, the LCPD module effectively recovers hidden or time-warped periodicities even under compound-fault conditions. Experimental results confirm that this approach achieves higher diagnostic accuracy, greater noise robustness, and improved computational efficiency compared to standard methods like MED and MOMEDA.
Journal Paper: Semi-supervised transfer graph neural representation learning with few-shot adaptation for gearbox diagnostics under extraneous transient noise
Admin Dec. 24, 2025 Comments
Our lab has presented research introducing a novel diagnostic framework for industrial gearboxes in Structural Health Monitoring (SCI, JCR Q1, 中科院(工程技术) 2区, IF(2025) = 5.7). The paper, titled “Semi-supervised transfer graph representation learning with few-shot adaptation for gearbox diagnostics under extraneous transient noise,” (DOI: 10.1177/14759217251414344) addresses the pressing challenge of maintaining high diagnostic accuracy in critical mechanical components when faced with extremely sparse labeled datasets and intense extraneous transient noise interference.
The proposed SSTGRL-FSA method integrates a pseudo-label reliability enhancement mechanism based on systematic knowledge transfer with an advanced label transmission strategy and an integrated first-order Markov state probability transition matrix utilizing amplitude-constrained scaling. By exploiting homologous signal patterns and modeling sophisticated temporal dependencies to maintain stable feature scales, the framework significantly enhances diagnostic robustness and generalization capabilities in data-scarce environments, marking an important step forward for reliable industrial gearbox fault diagnosis under challenging operational conditions.
Journal Paper: Markovian spectral transition modeling with temporal dependencies for railway bogie axle bearing diagnostics in non-stationary transient environments
Admin Dec. 01, 2025 Comments
Our lab has presented research introducing a novel diagnostic framework for railway bogie axle bearings in Nonlinear Dynamics (SCI, JCR 1区, 中科院(工程技术) 2区, IF(2025) = 6, Top Journal). The paper, titled “Markovian spectral transition modeling with temporal dependencies for railway bogie axle bearing diagnostics in non-stationary transient environments,”(DOI: 10.1007/s11071-025-12114-y) addresses the pressing challenge of accurately diagnosing bearing faults under complex, non-stationary operating conditions with strong, multivariate interference noise typical of real railway systems.
The proposed method integrates Markovian spectral transition modeling with multi-resolution wavelet analysis and an amplitude-adaptive interference suppression mechanism based on statistical signal modeling. By embedding wavelet coefficients into Markovian state representations and enabling dynamic thresholding, the framework significantly enhances demodulation robustness and fault detection reliability in challenging transient environments, marking an important step forward for railway bearing diagnostics.
Journal Paper: SSKD: Self-Supervised Knowledge Distillation for reliable remote surface defect detection under wireless transmission attenuation
📝 Admin Nov. 25, 2025 Comments
Our Lab has presented research introducing SSKD, a Self-Supervised Knowledge Distillation framework for reliable remote surface defect detection under wireless transmission attenuation, in Engineering Research Express. The paper, titled “SSKD: Self-Supervised Knowledge Distillation for Reliable Remote Surface Defect Detection under Wireless Transmission Attenuation,” (DOI: 10.1088/2631-8695/ae25ba) addresses critical challenges in robust defect inspection when wireless sensor networks (WSNs) degrade image quality through noise, compression artifacts, and packet loss.
SSKD reconstructs attenuated images on the fly without requiring historical paired datasets, enabling flexible deployment in real-world smart manufacturing environments. The framework integrates three key innovations: self-supervised learning decoupled from rigid paired data, cross-domain knowledge transfer that preserves diagnostic accuracy under severe transmission-induced degradation, and autonomous reconstruction that restores defect-salient cues to stabilize downstream detection. Evaluated under diverse wireless transmission conditions, SSKD achieves superior robustness and diagnostic fidelity compared to strong baseline methods, paving the way for reliable, cost-effective remote surface defect inspection using commodity WSN deployments.
喜报频传！S3DP-Lab多名学子荣获2025年度国家奖学金及学业一等奖学金
Admin Oct. 28, 2025 Comments
2025年10月28日 —— 近日，2025年度研究生国家奖学金评审结果正式公布。凭借扎实的学术功底与突出的科研创新能力，S3DP-Lab学子再创佳绩，斩获多项荣誉。
其中，吴宇豪、张锐进两位同学凭借优异的综合表现，荣获“2025年硕士研究生国家奖学金”。该奖项旨在奖励在学业成绩、科研创新和社会实践等方面全面发展的优秀研究生，是国家设立的研究生最高荣誉之一。同时，高嘉、吴宇豪、张锐进、王雅正、张春五位同学荣获“研究生学业一等奖学金”，充分展现了S3DP-Lab在人才培养方面的成效突出。
此次多项荣誉的获得，不仅是对获奖同学个人努力的充分肯定，也体现了S3DP-Lab在科研指导、学术氛围营造和学生全面发展方面的坚实基础与显著成果。实验室始终坚持以创新为导向、以育人为根本，持续助力青年学子在科研道路上砥砺前行。
谨此向所有获奖同学表示热烈祝贺！期待他们在未来的学习与科研中再接再厉，勇攀高峰！
陈鹏受邀作国家自然科学基金优秀结题口头报告
📝 Admin Oct. 24, 2025 Comments
鉴于在2024年度国家自然科学基金结题评估中的优异表现，陈鹏老师经国家自然科学基金委机械设计与制造学科遴选，受邀就2024年度结题基金项目作优秀结题口头报告。2025年10月24日，他以“风电齿轮箱耦合信号时频特征表征增强及融合金字塔模型研究”为题作专题汇报，系统交流了研究思路与阶段性成果。
此次报告展示了团队在风电装备关键部件信号处理与模型构建方面的最新进展，促进了同行专家的学术交流与经验共享，体现了项目在机械设计与制造领域的创新与示范作用。本次基金结题项目报告会从2024年度结题评价较高的项目中精选61项进行口头展示，涵盖面上项目、青年项目（C类）及地区项目，集中展现机械学科前沿研究的阶段性成果与科研活力。
2025 Global Reliability & Prognostics and Health Management (GlobalRel&PHM) Conference
📝 Admin Oct. 12, 2024 Comments
Xi’an, China — October 10–12, 2025 — Dr. Peng Chen participated in the 2025 Global Reliability & Prognostics and Health Management (GlobalRel&PHM) Conference, co-sponsored by the IEEE Reliability Society. The event gathered international experts to discuss advances in reliability engineering and health management technologies. Dr. Chen presented “Neurons-State-Attention-Driven Trustworthy Framework: An Interpretable Network for Wind Turbine Gearbox Fault Diagnosis.” The conference reaffirmed its role as a premier platform for cutting-edge research and professional collaboration.
Journal Paper: CyclicalNet: A dense-connected architecture for multi-scale cyclic feature learning in planetary gearbox fault diagnosis
📝 Admin Aug. 16, 2025 Comments
Our Lab has presented research introducing CyclicalNet, a dense-connected deep learning architecture for multi-scale cyclic feature learning in planetary gearbox fault diagnosis, in ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems Part B-Mechanical Engineering. The paper, titled “CyclicalNet: A dense-connected architecture for multi-scale cyclic feature learning in planetary gearbox fault diagnosis.” (DOI: 10.1115/1.4069653) This research addresses critical challenges in accurate, reliable fault diagnosis for planetary gearboxes operating under complex vibrations and harsh conditions.
CyclicalNet integrates Cyclical blocks, Encoder blocks, and dense connections to emphasize high amplitude-frequency components often overlooked by conventional models. A strategic reshape operation converts 1D time series into 2D representations, enabling structured, simultaneous modeling of temporal and periodic features, while dense connectivity promotes multi-scale feature learning and robust information flow. Evaluated on planetary gearbox datasets, CyclicalNet achieves superior fault diagnosis accuracy compared to general-purpose deep models (e.g., Transformers, ResNets, DenseNets, GANs), establishing a new benchmark for predictive maintenance.
Journal Paper: Semi-Supervised Transfer Learning Preserving Spatial Homogeneity for Gearbox Diagnostics in Extraneous Transient Noise
📝 Admin July 26, 2025 Comments
Our Lab has presented research introducing a approach for steel wire rope inspection in Nondestructive Testing and Evaluation (SCI, JCR 1区, 中科院(材料科学) 2区, IF(2025) = 4.2) . The paper, titled “Semi-supervised transfer learning preserving spatial homogeneity for gearbox diagnostics in extraneous transient noise.” (DOI: 10.1080/10589759.2025.2544893) This research addresses critical challenges in planetary gearbox fault diagnosis, particularly relevant for applications in electric motors, automotive systems, and wind turbines.
The innovative Semi-Supervised Transfer Learning (SSTL) framework introduces a unique combination of semi-supervised learning with transfer learning, effectively addressing the challenges of unlabeled data and transient interference in industrial fault diagnosis. Through comprehensive case studies, the approach has demonstrated superior fault detection accuracy compared to traditional methods. This advancement represents a significant step forward in automated fault diagnosis technology, offering practical solutions for real-world industrial applications where equipment reliability is crucial, particularly in scenarios with limited labeled data and frequent transient interference.
Journal Paper: Multi-Channel Fusion Scale Transformed Signals with Magnetic Leakage for Damage Detection in Steel Wire Ropes
📝 Admin July 26, 2025 Comments
Our Lab has presented research introducing a approach for steel wire rope inspection in Nondestructive Testing and Evaluation (SCI, JCR 1区, 中科院(材料科学) 2区, IF(2025) = 4.2) . The paper, titled “Multi-channel Fusion Scale Transformed Signals with Magnetic Leakage for Damage Detection in Steel Wire Ropes,”(DOI: 10.1080/10589759.2025.2543513) addresses critical challenges in structural safety monitoring across diverse industrial applications including elevators, mining equipment, and cable-supported structures.
The innovative Multi-channel Fusion Scale Transformation (MCFST) framework effectively tackles the persistent challenge of detecting low-amplitude faults in steel wire ropes, particularly those affected by lift-off effects and complex noise environments. Through experimental validation, this method has achieved a remarkable 97.65% detection accuracy, significantly outperforming traditional approaches like Canny Edge Detection and Adaptive Threshold methods. This breakthrough in detection capability, combined with improved computational efficiency, marks a significant advancement in steel wire rope inspection technology, particularly crucial for real-time industrial applications where structural integrity is paramount.
Journal Paper: EKBD-MK: Entropy-Kurtosis bilateral discernment with maximum kurtosis blind deconvolution for fault diagnosis in wind turbine systems
📝 Admin July 21, 2025 Comments
Our Lab has presented research introducing a novel diagnostic framework for wind turbine fault detection in ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems Part B-Mechanical Engineering. The paper, titled “EKBD-MK: Entropy-Kurtosis bilateral discernment with maximum kurtosis blind deconvolution for fault diagnosis in wind turbine systems,” (DOI: 10.1115/1.4069528) addresses critical challenges in wind turbine bearing system maintenance and reliability.
The innovative EKBD-MK framework effectively suppresses non-stationary transient disturbances and harmonic noise in vibration signals, utilizing Shannon entropy-based noise quantification and Variational Mode Decomposition (VMD) with an entropy-frequency dual-constraint system. Through experimental validation, this method has demonstrated superior diagnostic precision compared to existing approaches, marking a significant advancement in wind turbine fault detection technology.
Journal Paper: Physics-aware dynamic spectral modeling integrated with weakly supervised few-shot learning for planetary gearbox fault diagnosis
📝 Admin July 02, 2025 Comments
Our Lab has presented research introducing a novel physics-aware framework for industrial fault diagnosis in Measurement Science and Technology. The paper, titled “Physics-aware Dynamic Spectral Modeling with Weakly Supervised Few-shot Learning (PADSM-WSFL) for Planetary Gearbox Fault Diagnosis,” (DOI: 10.1088/1361-6501/adfc3b) addresses a critical challenge in industrial maintenance – the need for large amounts of labeled data in AI-based fault detection systems.
The PADSM-WSFL framework represents a significant advancement in machinery health monitoring by combining physics-based modeling with cutting-edge deep learning techniques. This approach uniquely integrates weakly supervised and few-shot learning methodologies, allowing effective fault diagnosis with extremely limited labeled data samples. The framework’s graph-based feature extraction module has demonstrated superior performance in capturing complex fault patterns, while its physics-aware components enhance the model’s interpretability and reliability. Experimental validation across multiple datasets has shown PADSM-WSFL outperforming existing state-of-the-art methods, marking a crucial step forward in making AI-driven fault diagnosis more practical and accessible for real-world industrial applications.
Zhigang Ma Celebrates Master of Science Graduation
📝 Admin June 24, 2025 Comments
Zhigang Ma has successfully earned his Master of Science degree from S. T. University, demonstrating exceptional academic performance throughout his program. His research contributions and scholastic achievements were particularly noted during the graduation ceremony, where he delivered a heartfelt address expressing appreciation to his supporters. Ma’s celebration was attended by proud friends, and S3DP-Lab members, marking the culmination of his graduate studies.
We wish Zhigang Ma continued success in his future pursuits!
Journal Paper: Metric-guided graph contrastive learning: An unsupervised approach for few-shot gearbox fault diagnosis
📝 Admin June 20, 2025 Comments
Our Lab have published a study in Measurement Science and Technology introducing a novel AI framework for planetary gearbox fault diagnosis. The paper, titled “Metric-guided graph contrastive learning: An unsupervised approach for few-shot gearbox fault diagnosis” (DOI: 10.1088/1361-6501/ade7a7), presents the Metric-guided Graph Contrastive Learning (MGCL) framework.
This innovative approach tackles key challenges in traditional fault diagnosis methods, particularly in critical applications like wind turbines, helicopters, and hybrid vehicles. The MGCL framework combines feature-decoupled pre-training, advanced distance metrics, and a two-stage training paradigm to deliver more accurate and reliable fault detection with limited data. This advancement promises to enhance safety and efficiency in industrial operations where gearbox reliability is crucial.
Journal Paper: Progressive contrastive representation learning for defect diagnosis in aluminum disk substrates with a bio-inspired vision sensor
📝 Admin May 22, 2025 Comments
Our latest research work, “Progressive contrastive representation learning for defect diagnosis in aluminum disk substrates with a bio-inspired vision sensor, (DOI: 10.1016/j.eswa.2025.128305) ” has been published in the prestigious Expert Systems With Applications journal (IF (2025) = 7.5, Top Journal).
The innovative system achieves superior detection accuracy for microscopic surface defects, real-time processing capabilities for production environments, reduced false positive rates compared to traditional methods, and enhanced robustness against varying lighting conditions.
This breakthrough leverages biomimetic principles by incorporating a bio-inspired vision sensor that mimics natural visual processing systems. The developed framework demonstrates particular effectiveness in identifying surface scratches, material inconsistencies, structural anomalies, and pattern deviations.
Congratulations to Mr. Zhigang Ma for Successfully Defending His Master’s Thesis Admin May 15, 2025 Comments
Mr. Zhigang Ma successfully defended his master’s thesis titled “Research on Adaptive Anti-interference Mechanism and Remote Collaborative Diagnosis of Mechanical Key Components (机械关键部件的自适应抗干扰机理与远程协同诊断方法研究)” on May 15, 2024, marking an achievement in the field of mechanical engineering and smart diagnostics. His research work has garnered notable recognition through publications in prestigious journals, including the “IEEE Internet of Things Journal”, “IEEE Transactions on Instrumentation and Measurement”, and “Nondestructive Testing and Evaluation”. Ma’s contributions advance the understanding of mechanical component diagnostics and anti-interference mechanisms, demonstrating innovative approaches to remote collaborative diagnosis systems that have important implications for industrial applications.
S3DP-Lab Advances Team Synergy at Strategic N. A. Island Retreat
📝 Admin April 26, 2025 Comments
S3DP-Lab members strengthened team dynamics during an intensive two-day retreat on a N. A. island. The program featured collaborative challenges that leveraged the isolated coastal setting to enhance team communication and partnership. The carefully orchestrated experience delivered measurable improvements in team cohesion and collaboration, setting a new benchmark for the lab’s collective performance.
Journal Paper: Semi-supervised Consistency Models for Automated Defect Detection in Carbon Fiber Composite Structures with Limited Data
📝 Admin Mar. 13, 2025 Comments
Our group has published a paper “Semi-supervised Consistency Models for Automated Defect Detection in Carbon Fiber Composite Structures with Limited Data” in Measurement Science and Technology introducing a novel semi-supervised approach for detecting defects in Carbon Fiber Composite Structures (CFCS), particularly focusing on ACCC wires. The paper (DOI: 10.1088/1361-6501/adc031) addresses a critical challenge in the non-destructive testing industry: the scarcity of annotated failure data for training deep learning models.
The proposed methodology combines synthetic data generation using consistency strategies, transformer-based feature fusion, and a DenseNet-based detection module. Our experiments demonstrate significant improvements in defect detection accuracy through this hybrid approach of synthetic and real-world data, making deep learning-based inspection systems more practical for industrial applications with limited training data availability.
亮点工作·工学院青年教师风采 | 陈鹏——工业装备物联网智能监测与诊断
📝 Admin Mar. 13, 2025 Comments
转引“工学院青年教师风采报道”：🖇️ 原文[PDF]
亮点工作 | 聚焦工业装备智能监测与诊断领域，攻克工业物联网场景下多源干扰、传输失真及复杂工况带来的技术挑战。创新性地提出抗失真在线监测体系突破无线感知网络中的传输衰落与设备泛化瓶颈，开发多模态适应性诊断框架，实现风电变工况下的干扰特征解耦与噪声主动抑制，构建尺度感知迁移学习模型，攻克关键部件复杂表面损伤检测的跨域适配难题，相关成果形成覆盖信号采集-特征解析-决策优化的全链路监测方案，为高端装备可靠性提升提供全新途径。
Journal Paper: Adaptive signal regime for identifying transient shifts: A novel approach toward fault diagnosis in wind turbine systems
📝 Admin Feb. 24, 2025 Comments
We have developed an adaptive signal processing technique that enhances fault detection in offshore wind turbines, detailed in our new paper “Adaptive signal regime for identifying transient shifts: A novel approach toward fault diagnosis in wind turbine systems, (DOI: 10.1016/j.oceaneng.2025.120798) ” has been published in the prestigious Ocean Engineering journal (IF (2025) = 4.6, Top Journal).
This research addresses wind turbine bearing fault detection challenges caused by variable operating conditions and environmental noise through an adaptive signal processing framework combining real-time transient shift tracking, a Dynamic Markov Transition Frequency model, and a Multi-Period Weighted Average Framework. Validated with operational turbine data, the method achieves superior detection accuracy in non-Gaussian/transient noise environments compared to conventional approaches.
Journal Paper: Interference Suppression of Non-stationary Signals for Bearing Diagnosis under Transient Noise Measurements
📝 Admin Jan. 07, 2025 Comments
We are delighted to announce our latest research paper: “Interference Suppression of Non-stationary Signals for Bearing Diagnosis under Transient Noise Measurements, (DOI: 10.1109/TR.2025.3527739) ” has been published in the prestigious IEEE Transactions on Reliability journal (IF (2024) = 5.0, Top Journal).
This publication represents a significant contribution to the field of signal processing and bearing diagnostics, offering new approaches for noise suppression in challenging measurement conditions. Congratulations to all contributors on this outstanding achievement!
Journal Paper: Step-wise contrastive representation learning for diagnosing unknown defective categories in planetary gearboxes
📝 Admin Dec. 10, 2024 Comments
Exciting Research Breakthrough! 🎉 🎉 🎉 We have achieved a breakthrough in planetary gearbox fault detection. Our latest research, “Step-wise contrastive representation learning for diagnosing unknown defective categories in planetary gearboxes, (DOI: 10.1016/j.knosys.2024.112863) ” has been published in the prestigious Knowledge-Based Systems journal (IF (2024) = 7.2, Top Journal).
The innovative research introduces a novel approach to identifying unknown defects in planetary gearboxes, potentially revolutionizing predictive maintenance in industrial applications. This publication marks a notable contribution to the field of intelligent diagnostic systems.
Journal Paper: Scale-aware Domain Adaptation for Surface Defects Detection on Machine Tool Components in Contaminant Measurements
📝 Admin Oct. 28, 2024 Comments
Exciting news from our research lab! We’ve just published our research work “Scale-aware Domain Adaptation for Surface Defects Detection on Machine Tool Components in Contaminant Measurements“ (DOI: 10.1109/TIM.2024.3502888) in IEEE Transactions on Instrumentation and Measurement (IF (2024) = 5.6, Top Journal), one of the top journals in our field. Our research introduces a smart new technology called Scale-aware Domain Adaptation (SADA) that’s like giving machines super-vision to spot defects on industrial equipment, even when parts are covered in oil or other contaminants. This innovation is particularly useful for detecting damage on ball screws in CNC machines, making industrial maintenance more reliable and efficient.
S3DP-Lab Welcomes New Researchers for Autumn 2024
📝 Admin Oct. 19, 2024 Comments
The S3DP-Lab is delighted to welcome three new researchers for 2024: Mr. Yazheng Wang, Mr. Qingsheng Wei, and Mr. Enyu Yang. These talented individuals join our team to contribute to ongoing projects in smart sensing, signal processing, and deep learning. Their join strengthens our research capabilities and we look forward to their valuable contributions to our lab’s mission.
2024 Global Reliability & Prognostics and Health Management (GlobalRel&PHM) Conference
📝 Admin Oct. 14, 2024 Comments
Beijing, China – October 11-13, 2024: Dr. Peng Chen participated in the 2024 Global Reliability & Prognostics and Health Management (GlobalRel&PHM) Conference, held in Beijing from October 11-13. The event, co-sponsored by the IEEE Reliability Society, attracted experts from around the world to discuss advancements in reliability engineering and health management technologies.
During the conference, Dr. Chen delivered a presentation on “Markov Latent Frequency Transition Analysis for Robust Bearing Diagnosis in Transient Noise Scenarios.” His insightful talk focused on innovative methods to enhance diagnostic capabilities and reliability of bearings in challenging environments, a critical component in various industrial applications.
The GlobalRel&PHM Conference continues to serve as a premier platform for sharing cutting-edge research and fostering collaboration among professionals in the field.
Chaojun Xu Celebrates Master of Science Graduation
📝 Admin June 20, 2024 Comments
Congratulations to Chaojun Xu, who recently graduated with a Master of Science (M.S.) degree from S. T. Univ. Known for his dedication and academic excellence, Chaojun completed his studies with distinction, earning high praise for his research.
The graduation ceremony was a joyous event, with family, friends, and faculty celebrating Chaojun’s hard work and achievements. In his speech, Chaojun expressed gratitude to his supporters and shared his excitement for future endeavors.
We wish Chaojun Xu continued success in his future pursuits!
Visiting the University of Oxford’s Computer Science Department
📝 Admin June 14, 2024 Comments
S3DP-Lab Members Enhance Team Spirit in N. A. Island Retreat
📝 Admin May 24, 2024 Comments
S3DP-Lab members successfully concluded a two-day team-building event on an island in N. A. The retreat aimed to foster collaboration and strengthen the bonds among lab members through a series of structured activities and workshops. Participants engaged in team challenges, professional development sessions, and recreational activities, all designed to enhance teamwork and communication skills. The event’s scenic location and well-planned itinerary ensured that lab members returned with renewed motivation and a stronger sense of camaraderie.
Journal Paper: Self-supervised transfer learning for remote wear evaluation in machine tool elements with imaging transmission attenuation
📝 Admin Mar. 26, 2024 Comments
We are pleased to announce the publication of our paper “Self-supervised transfer learning for remote wear evaluation in machine tool elements with imaging transmission attenuation“(DOI: 10.1109/JIOT.2024.3382878) in IEEE Internet of Things Journal (IF (2023) = 10.6, Top Journal). This research proposes a self-supervised transfer learning model integrated with image capture technology, revolutionizing remote wear evaluation for CNC machines by eliminating the reliance on extensive historical data and ensuring reliability despite compromised data transmission environments.
Congratulations to Mr. Chaojun Xu for Successfully Defending His Master’s Thesis
📝 Admin May 22, 2024 Comments
We are pleased to announce that Mr. Chaojun Xu has successfully defended his master’s thesis on May 22, 2024. His thesis, titled “Toward Fault Diagnosis of Wind Turbine Gearbox via Diffusion Representation and Scale Sensing Stack Modeling,” presents groundbreaking research in the field of renewable energy technology.
Mr. Xu’s work focuses on enhancing the reliability and efficiency of wind turbines through advanced fault diagnosis techniques. His innovative approach combines diffusion representation and scale sensing stack modeling to accurately detect and diagnose faults in wind turbine gearboxes, potentially leading to significant improvements in the maintenance and operation of wind turbines.
We extend our heartfelt congratulations to Mr. Xu on this significant achievement and wish him continued success in his future endeavors.
TEPEN2024-IWFDP: International Workshop on Fault Diagnostics and Prognostics
📝 Admin May 11, 2024 Comments
Qingdao, China, May 8th-11th, 2024: Dr. Peng Chen chaired a session on “Artificial Intelligence and Machine Learning” (Certificate of Session Chair) at the international conference in Qingdao. He presented research on “Semi-Supervised Gearbox Diagnostics via Few-Shot Learning with DGAT Algorithm, ” as well as on remote intelligent diagnosis.
Journal Paper: Markov Modeling of Signal Condition Transitions for Bearing Diagnostics under External Interference Conditions
📝 Admin Feb. 15, 2024 Comments
We are delighted to announce the publication of our paper “Markov Modeling of Signal Condition Transitions for Bearing Diagnostics under External Interference Conditions“(DOI: 10.1109/TIM.2024.3383497) in IEEE Transactions on Instrumentation and Measurement (IF (2023) = 5.6, Top Journal). This innovative method enhances rolling bearing fault diagnosis by mitigating transient noise interference, improving demodulation band selection accuracy. The study’s results validate its effectiveness in identifying fault-characteristic bands, showcasing the quality of research from our lab.
New Researchers Join Our Team for Autumn 2023
Admin Oct. 27, 2023 Comments
The Autumn 2023 semester brings a wave of fresh talent to our research group, enhancing our capabilities and expanding our horizons. Our new researchers, experts in a variety of fields, are set to invigorate our projects and foster interdisciplinary collaboration. As we embark on this exciting journey with our new team members, we anticipate a season of exploration, discovery, and academic achievement. We invite you to stay tuned for updates on our research group’s activities as we continue to push the boundaries of knowledge.
Congratulations to Outstanding Students on Masters Scholarships for the 2022-2023 academic year!
Admin Oct. 19, 2023 Comments
In a remarkable achievement, three outstanding students, Mr. Chaojun Xu, Mr. Jia Gao, and Mr. Ruijin Zhang, have secured the prestigious First-Class Master’s Scholarships for the 2022-2023 academic year. Additionally, Mr. Zhigang Ma, Mr. Junxiao Ma, and Mr. Yuhao Wu have been awarded Second-Class Master’s Scholarships, demonstrating their exceptional dedication to academic excellence. Congratulations to these scholars on their well-deserved honors!
Dr. Peng Chen Fosters Collaborative Discussions at Cambridge University’s Department of Engineering Admin June 17, 2023 Comments
Seminar on Innovation and Application of Heavy Duty Mobile Robots
📝 Admin May 30, 2023 Comments
Dr. Peng Chen, accompanied by Mr. Chaojun Xu and Mr. Zhigang Ma, attended the Seminar on Innovation and Application of Heavy Duty Mobile Robots at the Foshan Institute of Intelligent Equipment Technology. This seminar, which took place on May 30, 2023, served as a platform for industry leaders, researchers, and enthusiasts to delve into the latest advancements in heavy-duty robotics and their practical applications. Participants had the opportunity to engage in idea exchange, networking, and gain valuable insights into the potential of heavy-duty robots.
Journal Paper: A Mixed Samples-driven Methodology based on Denoising Diffusion Probabilistic Model for Identifying Damage in Carbon Fiber Composite Structures
📝 Admin Apr. 16, 2023 Comments
We are thrilled to announce that our research lab has recently published a journal paper titled “A Mixed Samples-driven Methodology based on Denoising Diffusion Probabilistic Model for Identifying Damage in Carbon Fiber Composite Structures” (doi: 10.1109/TIM.2023.3267522) in the prestigious IEEE Transactions on Instrumentation and Measurement (IF (2022) = 5.33).
We would like to extend our heartfelt congratulations to the entire team of researchers involved in this project, whose dedication and hard work have made this achievement possible. This publication is a testament to the outstanding quality of research being conducted in our lab and will undoubtedly contribute to the advancement of the specific field. This achievement is a testament to the exceptional research being conducted in our lab, and we look forward to further contributions to the field.
New Researchers Join Our Lab to Drive Innovation
📝Admin Apr. 06, 2023 Comments
We are pleased to announce the newest members of our research lab. Mr. Junxiao MA, Mr. Ruijin ZHANG, Mr. Jia GAO, and Mr. Yuhao WU have recently joined our team, bringing with them fresh perspectives and valuable skills.
We believe that the addition of these four individuals will further enhance our capabilities and enable us to achieve even greater success in our ongoing research projects. Please join us in welcoming Mr. Junxiao MA, Mr. Ruijin ZHANG, Mr. Jia GAO, and Mr. Yuhao WU to our research lab. We are excited to have them on board and look forward to working with them to advance our research goals.