Article
Automation & Control Systems
Haidong Shao, Min Xia, Guangjie Han, Yu Zhang, Jiafu Wan
Summary: A new framework for fault diagnosis of rotor-bearing system under varying working conditions is proposed using modified CNN and transfer learning. Infrared thermal images are collected to characterize the health condition, and modified CNN is developed with stochastic pooling and Leaky ReLU. The proposed method outperforms other cutting edge methods in fault diagnosis of rotor-bearing system by adapting to limited available training data in different working conditions.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Automation & Control Systems
Fei Wang, Ruonan Liu, Qinghua Hu, Xuefeng Chen
Summary: A cascade CNN (C-CNN) with progressive optimization is proposed for motor fault diagnosis in nonstationary conditions, addressing the limitations of traditional CNNs. Through physical characteristics of nonstationary vibration signals, the C-CNN achieves better performance in both constant and variable speed scenarios compared to existing methods.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Computer Science, Artificial Intelligence
Xiaoli Zhao, Jianyong Yao, Wenxiang Deng, Peng Ding, Yifei Ding, Minping Jia, Zheng Liu
Summary: This paper proposes an intelligent fault diagnosis method for gearboxes under variable working conditions based on adaptive intraclass and interclass convolutional neural network (AIICNN). By applying intraclass and interclass constraints to improve sample distribution differences, and using an adaptive activation function to enlarge the heterogeneous distance and narrow the homogeneous distance of samples, the feasibility of the proposed method is verified through experimental data.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Automation & Control Systems
Xinyu Pang, Xuanyi Xue, Wangwang Jiang, Kaibo Lu
Summary: An intelligent fault diagnosis approach based on deep CNNs and vibration BSP is proposed to improve efficiency and accuracy in diagnosing planetary gearboxes. The method achieves high accuracy in identifying various gear faults, with TL further enhancing diagnostic performance. This study contributes to the development of BSP-based CNN models and extensive evaluation of CNN-TL methods for gear monitoring and diagnosis.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2021)
Article
Automation & Control Systems
Yadong Xu, Xiaoan Yan, Beibei Sun, Jinhui Zhai, Zheng Liu
Summary: Recent progress on intelligent fault diagnosis, mainly driven by the development of convolutional neural networks (CNNs), has resulted in CNN-based fault diagnosis models that extract features from measured vibration signals but are limited in their ability to extract discriminative features under strong noise conditions. To address this challenge, a new deep CNN model called MF-DRCN is proposed, which enhances feature extraction and anti-interference capabilities with a multireceptive field denoising (MFD) block and an adaptive feature integration (AFI) module. Experimental results demonstrate the effectiveness of MF-DRCN in achieving high diagnostic accuracy in noise conditions, making it a promising approach for intelligent fault diagnosis.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2022)
Article
Engineering, Environmental
Zhijian Wang, Wenlei Zhao, Wenhua Du, Naipeng Li, Junyuan Wang
Summary: This study introduces a new fault diagnosis method based on data-driven approach using AlexNet CNN. The method achieves high prediction accuracy for fault classification in different bearing datasets and has been proven feasible in engineering practice.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2021)
Article
Automation & Control Systems
Tingli Xie, Xufeng Huang, Seung-Kyum Choi
Summary: In this article, a novel intelligent fault diagnosis method based on multisensor fusion and convolutional neural network is explored. The proposed method converts multisignal data into RGB images and uses an improved CNN for classification, resulting in higher accuracy in fault diagnosis.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Automation & Control Systems
Zhiqian Zhao, Yinghou Jiao
Summary: This article proposes a efficient and lightweight convolutional neural network diagnostic model, MIXCNN, which maintains the same size of the output throughout the network. MIXCNN uses depthwise convolution to increase discrimination ability in spatial locations and traditional convolution for cross-channel interaction. Experimental validation shows that MIXCNN has higher accuracy and generalization capability compared to existing methods.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Engineering, Electrical & Electronic
Biao Chen, Tingting Liu, Chao He, Zecheng Liu, Li Zhang
Summary: This paper introduces the application of deep learning methods based on vibration signals in fault diagnosis and proposes a method to address the challenges in fault diagnosis. The method focuses on the more important segments of the vibration signal using attention mechanism and capsule network, and also proposes a visualization method to interpret the attention distribution. Experimental results demonstrate that the method has good generalization and robustness.
IEEE SENSORS JOURNAL
(2022)
Article
Automation & Control Systems
Lv Chen, Kang An, Dali Huang, Xiaoxian Wang, Min Xia, Siliang Lu
Summary: In this article, a noise-boosted CNN (NBCNN) model is proposed to improve the training speed and recognition accuracy with limited training samples. The optimal injected noise accelerates the convergence of model training and improves the accuracy of motor fault diagnosis. The effectiveness and superiority of the proposed NBCNN model are validated by benchmark datasets, and the training speed of the NBCNN is nine times faster than the conventional CNN.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Engineering, Electrical & Electronic
Xianling Li, Hao Hu, Shichao Zhang, Gang Tang
Summary: This article proposes a fault diagnosis method for rotating machinery using a semi-supervised graph convolutional network (SSGCN) and 2-D images converted from vibration signals. The method addresses the difficulty of collecting sufficient labeled samples for practical engineering applications.
IEEE SENSORS JOURNAL
(2023)
Article
Computer Science, Artificial Intelligence
Huan Wang, Zhiliang Liu, Dandan Peng, Mei Yang, Yong Qin
Summary: This article proposes a novel multitask attention convolutional neural network (MTA-CNN) that can automatically give feature-level attention to specific tasks. The architecture allows the FLA-module to learn the features of specific tasks from globally shared features, sharing information among different tasks. The multitask learning mechanism improves the results of each task, as demonstrated by the better performance compared to state-of-the-art deep learning methods on wheelset and motor bearing data sets.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Engineering, Electrical & Electronic
Shuzhi Gao, Zhiming Pei, Yimin Zhang, Tianchi Li
Summary: This paper proposes an adaptive convolutional neural network based on Nesterov momentum for rolling bearing fault diagnosis, which improves both the accuracy and convergence of neural networks compared to traditional methods by replacing the traditional momentum and utilizing an adaptive learning rate rule to enhance the generalization ability of the network.
IEEE SENSORS JOURNAL
(2021)
Article
Automation & Control Systems
Wenliao Du, Pengjie Hu, Hongchao Wang, Xiaoyun Gong, Shuangyuan Wang
Summary: In this article, a one-dimension in tandem with 2-D joint convolutional neural network (1D-2D JCNN) is proposed for rotating machinery fault diagnosis. It uses 1-D convolution to obtain multiscale feature vectors, constructs them into 2-D maps, and feeds them into a 2-D convolutional neural network. Experimental results show the excellent classification performance of 1D-2D JCNN in bearing and gear fault diagnosis.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2023)
Article
Automation & Control Systems
Daichao Wang, Yibin Li, Lei Jia, Yan Song, Tao Wen
Summary: The fault diagnosis of bearings is crucial for improving the reliability and safety of rotating mechanical equipment. Existing feature-fusion-based methods fail to fully extract and fuse complementary fault information. This article proposes a new approach based on mutual attention and bilinear model to effectively extract and fuse complementary fault features, achieving a diagnosis accuracy of 99.86%.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2023)
Article
Automation & Control Systems
Heng Zhang, Guangxing Niu, Bin Zhang, Qiang Miao
Summary: This article proposes a cost-effective fault diagnosis and prognosis method for lithium-ion batteries based on Lebesgue Sampling LSTM (LS-LSTM) network. The method uses a Lebesgue time model to describe the fault growth process and employs the Monte Carlo method to handle uncertainties. It achieves unified and as needed implementation of FDP for lithium-ion batteries.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2022)
Article
Engineering, Electrical & Electronic
Jianguo Miao, Jianyu Wang, Dingcheng Zhang, Qiang Miao
Summary: This article proposes a data augmentation method based on an improved variational autoencoding generative adversarial network to address the issue of limited and imbalanced data in fault diagnosis of rotating components. The method improves data generation capability and training stability of the generative model by introducing self-attention mechanism and Wasserstein distance with gradient penalty. It generates high-quality samples to enrich the dataset and improve fault diagnosis accuracy.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Computer Science, Artificial Intelligence
Jianguo Miao, Congying Deng, Heng Zhang, Qiang Miao
Summary: This paper proposes an improved intelligent fault detection method for rotating component based on interactive channel attention (ICA) with two submodules to consider the channel correlation of both global and local channels. Two more replaceable submodules based on adaptive multi-scale kernel (AMK) and self-attention (SA) are provided to consider the correlation of neighbor channels and long-term channels. Experimental results demonstrate the effectiveness of the proposed method, especially in the occasions with strong noise and limited data.
APPLIED SOFT COMPUTING
(2023)
Article
Engineering, Aerospace
Jianyu Wang, Heng Zhang, Qiang Miao
Summary: This paper proposes a source free unsupervised domain adaptation framework for Electro-Mechanical Actuator (EMA) fault diagnosis. The framework combines feature network and classifier, and includes steps such as training the source model, transferring it to the target domain, filtering pseudo labels, and fine-tuning the transferred model. Experimental results demonstrate the effectiveness of the proposed method.
CHINESE JOURNAL OF AERONAUTICS
(2023)
Article
Automation & Control Systems
Zhenling Mo, Zijun Zhang, Qiang Miao, Kwok-Leung Tsui
Summary: This article introduces a new dynamic bandit tree (DBT) algorithm to help achieve more adaptive filters and reduce the burden of parameter tuning in frequency band searching. By optimizing the boundaries of Meyer wavelet filters, this method can better identify demodulated fault frequencies and outperform other optimization algorithms and fault diagnosis methods in tests.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2023)
Article
Thermodynamics
Guangzheng Lyu, Heng Zhang, Qiang Miao
Summary: This paper proposes a novel SOH estimation method for lithium-ion batteries based on multi-category and multi-stage features and an input optimization interpretable multivariable LSTM model. The method improves the accuracy of SOH estimation by constructing degradation features and optimizing model input, providing guidance for battery utilization and maintenance strategies.
Article
Engineering, Electrical & Electronic
Yinxue Zeng, Yujie Zhang, Xingyou Yan, Qiang Miao
Summary: This article proposes an improved multivariate state estimation technique (MSET) with a composite operator (CO-MSET) for health indicator (HI) extraction of electro-mechanical actuators (EMAs). Experimental results demonstrate that the proposed method has a better performance in HI extraction for EMAs under various operation conditions.
IEEE SENSORS JOURNAL
(2023)
Article
Engineering, Industrial
Guangzheng Lyu, Heng Zhang, Qiang Miao
Summary: This paper proposes a lithium-ion battery early-cycle stage RUL prediction method based on Lebesgue sampling parallel state fusion LSTM (LS-PSF-LSTM). By selecting similar samples and using parallel state fusion LSTM algorithm, the prediction accuracy and efficiency are improved.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2023)
Article
Engineering, Electrical & Electronic
Wei Zhang, Xia Fang, Chengjun Wu, Mei Wang, Jie Wang, Dingcheng Zhang, Qiang Miao
Summary: This article proposes an adaptive label assignment detection framework to address difficulties in defect detection of microarmatures, including spatial feature loss of small targets and tuning the reference of sample selection strategy. The framework uses an anchor-free object detection network as the backbone, and introduces feature compensation and adaptive sample selection strategies to improve the detection performance. Experimental results show that the framework outperforms common object detection methods and improves the accuracy and efficiency of defect detection of nonstandard microworkpiece.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Engineering, Electrical & Electronic
Jiawei Liao, Haitao Xu, Xia Fang, Qiang Miao, Gaoyi Zhu
Summary: This article proposes a framework for quantitatively assessing the non-structural bird's nest risk information of transmission towers using high-resolution UAV panoramic images. The integrity of the tower's structure information is maintained by using panoramic UAV images. The proposed method involves estimating the tower's pose and determining whether each key point of the tower has a bird's nest, transforming the non-structural bird's nest locations and risk information into structured data. A risk score is used to classify the level of potential risk caused by bird's nests. The experiments demonstrate that the proposed framework can accurately detect the key points of five types of towers, with an F1-score of 83.09% for bird's nest detection and an accuracy of 91.59% for risk assessment.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Engineering, Electrical & Electronic
Guangzheng Lyu, Heng Zhang, Qiang Miao
Summary: In this article, a RUL prediction method for lithium-ion batteries in the early-cycle stage based on similar sample fusion is proposed. The Lebesgue sampling framework is used to transform the data structure of similar samples to ensure that the fused points are at the same degradation state. A linear fitting is performed, and the fitting results are used to construct a particle filter model for capacity degradation and RUL prediction.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Engineering, Electrical & Electronic
Chenyang Jiao, Dingcheng Zhang, Xia Fang, Qiang Miao
Summary: This article proposes a fault diagnosis method for planetary gearboxes based on a simplified graph neural network. By simplifying the design and using diverse feature extraction, the proposed method achieves higher accuracy and robustness in fault diagnosis.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Engineering, Electrical & Electronic
Fei Peng, Xuanyu Zheng, Qiang Miao
Summary: This article presents a novel approach to improve both the dynamic range and anti-fading performance of phi-OTDR by utilizing a conventional phi-OTDR structure and a convolutional neural network. The trained model significantly enhances the performance of phi-OTDR and has demonstrated promising results in experiments.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Automation & Control Systems
Zhenling Mo, Zijun Zhang, Qiang Miao, Kwok-Leung Tsui
Summary: This work introduces a sparsity-constrained invariant risk minimization (SCIRM) framework for machinery fault diagnosis. By integrating sparsity constraints, SCIRM develops machine learning models with better generalization capacities and achieves higher accuracy and generalization performance on real datasets.
IEEE TRANSACTIONS ON CYBERNETICS
(2022)
Article
Engineering, Electrical & Electronic
Jianyu Wang, Zhiguo Zeng, Heng Zhang, Anne Barros, Qiang Miao
Summary: This article proposes an improved deep learning-based fault diagnosis framework for electromechanical actuators (EMAs) using triplet network with coupled cluster losses. Unlike traditional methods, the framework learns to predict the similarity between samples instead of directly predicting fault labels. Experiments demonstrate that the developed framework can enhance the performance of traditional deep learning-based approaches in situations with limited and unbalanced training datasets and different working conditions.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
