Related references
Note: Only part of the references are listed.
Article
Engineering, Multidisciplinary
Guolong Zhang et al.
Summary: A high-resolution and non-contact method for measuring the thickness of insulating ceramic coatings is proposed. The method utilizes a chromatic confocal sensor and eddy-current sensor to obtain distance measurements from the outer and inner surfaces of the coating respectively. The system's performance is tested and shows high precision, non-contact capability, quick speed, stability, and visual display.
Article
Computer Science, Hardware & Architecture
Jiusi Zhang et al.
IEEE Transactions on Neural Networks and Learning Systems
(2023)
Article
Engineering, Electrical & Electronic
Silei Chen et al.
Summary: This article proposes an improved method for arc fault detection in building integrated photovoltaic systems using time series analysis and the sample entropy feature and gated recurrent unit classifier. The method achieves higher detection accuracy and faster runtime compared to existing techniques.
IEEE SENSORS JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Hairui Fang et al.
Summary: Although Transformer has achieved excellent results in various tasks in industrial scenes, the fault diagnosis approaches based on Transformer face the challenges of robustness and lightweight. To address these challenges, a lightweight framework called X-self-attention convolution neural network (XACNN) was designed, which improves the robustness to noise through FFT adjustment and achieves lightweight optimization through self-attention. The effectiveness of XACNN was demonstrated through testing on a self-made dataset in various noise environments, and its feasibility as a portable fault diagnosis device was verified on a smartphone.
IEEE SENSORS JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Yong Xu et al.
Summary: In this article, a high accuracy LFDD method combining GRU-AE and RF was proposed for CRDM in PWRs. The movement sequences of the CRDM coils' current were taken as time series and analyzed using GRU-AE for data reconstruction. Abnormal data were generated and then diagnosed by RF. The results show that this combined approach achieves high-accuracy detection on real-world imbalanced datasets.
IEEE SENSORS JOURNAL
(2023)
Article
Engineering, Multidisciplinary
Hongyu Zhong et al.
Summary: This study proposes a novel fault diagnosis method, SA-SN-DCGAN-TL, which combines data augmentation and transfer learning to address the issue of insufficient training data in the big-data context of Industry 4.0, thereby improving the accuracy of deep networks.
Article
Engineering, Mechanical
Yupeng Wei et al.
Summary: Bearings are commonly used to reduce friction between moving parts. To predict the remaining useful life (RUL) of bearings, it is important to consider the correlation of features at different time points. Current data-driven methods often have limitations in processing long sequences and require longer training time.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Engineering, Electrical & Electronic
Xiaotian Zhang et al.
Summary: This article proposes an inferable deep distilled attention network (IDDAN) method, which efficiently and accurately diagnoses and classifies multiple bearing faults in various motor drive systems. The SA-based network can better extract global feature information and benefit from large amounts of pretraining data. Experimental results show that the proposed method achieves higher classification accuracy and better performance compared to other methods.
IEEE TRANSACTIONS ON TRANSPORTATION ELECTRIFICATION
(2023)
Article
Automation & Control Systems
Junjun Zhu et al.
Summary: This paper proposes a new RUL prediction method for rotating machinery using health indicators constructed by the residual hybrid network with self-attention mechanism (Res-HSA). The method addresses the problem of distribution consistency and interference from abnormal fluctuations in the health curve. Experimental results demonstrate its good performance in RUL prediction.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Automation & Control Systems
Xin Zhang et al.
Summary: This paper proposes a multi-sensor multi-head GAT model for fault diagnosis of large rotating machinery, which can dynamically fuse and mine high-level fault characteristics to improve the effectiveness of diagnosis.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Automation & Control Systems
Wenhe Chen et al.
Summary: This paper proposes a novel condition monitoring method for wind turbines using a deep learning model and dynamic kernel principal components Mahalanobis distance. The method accurately evaluates the performance of wind turbines for fault detection. It first identifies and removes outliers from raw data using a feature selection method, and then reconstructs the objective variables using a deep learning model with temporal pattern attention. The method considers the dynamic correlation between variables by constructing a condition index based on the reconstructed errors and determines the threshold using a delay perception-based method. Experimental results demonstrate the effectiveness of the proposed approach in detecting early abnormal conditions, outperforming other state-of-the-art methods.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Engineering, Electrical & Electronic
Lexuan Shao et al.
Summary: An improved GAN with filtering mechanism is proposed for fault data augmentation in this article. The self-attention mechanism and IN are introduced in the sample generation process to improve the quality of generated samples. The effectiveness of the method is verified using two public datasets for fault diagnosis and shows better performance than state-of-the-art GAN-based methods.
IEEE SENSORS JOURNAL
(2023)
Article
Automation & Control Systems
Jianghong Zhou et al.
Summary: Facing the gap in the unsupervised construction of health indicator (HI) with a uniform failure threshold, a new approach is developed by estimating the distribution of the raw vibration signal using the Gaussian mixture model and designing a distribution contact ratio metric (DCRM). A distribution contact ratio metric health indicator (DCRHI) is constructed to represent the degradation process and obtain a uniform failure threshold. Furthermore, a novel consolidated memory gated recurrent unit (CMGRU) is proposed to slow down the forgetting speed of important trend information and improve the prediction ability. The proposed methodology shows great application value in the RUL prediction.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Automation & Control Systems
Hongbing Shang et al.
Summary: This article proposes an intelligent borescope inspection method to detect surface damage on aeroengine blades. The method efficiently models pixel-to-pixel relations and handles weak damage information caused by background noise and unsatisfactory illumination. It also incorporates label relations as prior knowledge and fuses image features and label features for mode recognition and damage localization.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Engineering, Marine
Fanjie Li et al.
Summary: This study proposes a dynamic model of the offshore wind turbine rotor system, taking into account the insulating coating, inclined shaft current damage, and contact characteristics. Through numerical simulation, the mapping relationship between vibration response and damage length is investigated. The influence of the interaction between inclination angle and radial clearance on vibration response is analyzed, and the effects of the mass of the insulating coating and speed on vibration response are discussed. The practicality of the model is demonstrated through physical prototype experiments. The results show that increasing the length of shaft current damage and the coupling of inclination angle and radial clearance enhance the vibration response, with radial clearance having a more significant effect. Increasing speed worsens the operational state, transitioning from stable quasi-periodic to chaotic behavior.
Article
Engineering, Environmental
Jiaxin Zhang et al.
Summary: This study proposes a novel GRU-EDCNN model for improved fault detection and diagnosis of chemical processes. The model utilizes a maximum smooth function to improve calculation accuracy and efficiency, optimizes the convolutional layers using a decentralized structure, and tackles the issue of gradient disappearance through GRU embedment.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2023)
Article
Acoustics
Tongguang Yang et al.
Summary: In this study, a new compound faults diagnosis framework called Lightweight Spatial-Temporal Model Fusion Self-Attention Mechanism (LST-SATM-Net) is proposed for aero-engine hydraulic pipeline systems. The model extracts fine-grained spatial features and coarse-grained temporal features, and utilizes a self-attention mechanism to optimize feature learning. Experimental results demonstrate that the proposed LST-SATM-Net model achieves more accurate diagnosis of compound faults in the hydraulic pipeline system compared to other advanced methods. This model can be applied to actual condition monitoring of aero-engine hydraulic pipelines to reduce maintenance costs and downtime.
Article
Engineering, Industrial
Yong Shi et al.
Summary: Predictive Maintenance (PdM) is crucial for safety management, planning necessary maintenance in advance to prevent future breakdown. The challenge lies in predicting the Remaining Useful Life (RUL) based on historical data with long and complex patterns. To address this, a lightweight RUL prediction model called TCNASA is proposed, integrating TCN, ASA, and an autoregressive component.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2023)
Article
Green & Sustainable Science & Technology
Anqi Wang et al.
Summary: In this paper, a novel method for anomaly detection and identification of underlying causes in wind turbines is proposed. The method utilizes a self-attention mechanism and a multivariable query pattern to extract cross-variable correlations and evaluate the influences of different features on the target. Experimental results confirm the effectiveness of the proposed method for early detection of turbine faults and identification of anomaly causes.
Article
Engineering, Electrical & Electronic
Yong Xu et al.
Summary: In this study, a high accuracy LFDD method combining GRU-AE and RF was proposed for CRDM in PWRs. The GRU-AE was used to build a health operating data reconstruction model, and RF was used as a classifier to diagnose abnormal types. The results demonstrated the high accuracy detection of potential faults in real-world imbalanced datasets.
IEEE SENSORS JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Lina Tong et al.
Summary: This paper proposes a deep learning model based on inertial sensors for human activity recognition, which shows better performance and robustness compared to other methods. The impact of sensor configuration optimization is also analyzed.
IEEE SENSORS JOURNAL
(2022)
Article
Computer Science, Artificial Intelligence
Yanning Sun et al.
Summary: The study introduces a nonparametric-copula-entropy and network deconvolution method for causal discovery in complex manufacturing systems, improving association measurement accuracy between parameters and extracting direct information for optimal control. Experimental applications demonstrate the method's efficacy in revealing causal relationships and guiding optimization strategies in complex manufacturing systems.
JOURNAL OF INTELLIGENT MANUFACTURING
(2022)
Article
Computer Science, Information Systems
Dong Chen et al.
Summary: Human activity recognition in smart indoor environments is a challenging task, especially when it comes to recognizing activities of multiple people. In this study, we propose TRANS-BiGRU, a deep learning method that efficiently learns and recognizes different activities performed by multiple residents, outperforming existing models in complex activity recognition.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Computer Science, Artificial Intelligence
Yanning Sun et al.
Summary: This study proposes an adaptive fault detection and root-cause analysis scheme for complex industrial processes using moving window kernel principal component analysis (KPCA) and information geometric causal inference (IGCI). It combines a research scheme that handles the nonlinear and dynamic characteristics of complex industrial processes with an adaptive threshold, optimal hyperparameters selection using multiobjective evolutionary algorithm, and fault root-cause analysis method based on IGCI. The proposed scheme shows good performance in reducing faulty false alarms, missed detection rates, and locating fault root-cause, as tested on the Tennessee Eastman platform.
JOURNAL OF INTELLIGENT MANUFACTURING
(2021)
Article
Engineering, Electrical & Electronic
Tianyi Chen et al.
Summary: Stochastic compositional optimization generalizes classic stochastic optimization for minimizing compositions of functions, with applications in reinforcement learning and meta learning. The new Stochastically Corrected Stochastic Compositional gradient method (SCSC) ensures convergence at the same rate as traditional methods and can be accelerated with SGD techniques. Applying Adam to SCSC achieves state-of-the-art performance in stochastic compositional optimization, tested in model-agnostic meta-learning tasks.
IEEE TRANSACTIONS ON SIGNAL PROCESSING
(2021)
Article
Computer Science, Information Systems
Yipu Xu et al.
Summary: This paper investigates the impact of SiC motor drives on motor bearing currents, analyzing the mechanism and modeling of bearing currents, and assessing the effects of various factors on bearing currents through experiments.
Article
Materials Science, Ceramics
Tao Zhang et al.
CERAMICS INTERNATIONAL
(2019)
