Article
Automation & Control Systems
Hui Hou, Hongquan Ji
Summary: A novel feature selection strategy is proposed to improve the multiclass support vector data description (SVDD) algorithm for planetary gearbox fault diagnosis. By selecting features sensitive to faults and developing an improved multiclass SVDD algorithm, the fault diagnosis task is effectively completed.
CONTROL ENGINEERING PRACTICE
(2021)
Article
Engineering, Electrical & Electronic
Bing Sun, Xiaofeng Liu
Summary: Wheelset bearing is a critical component in high-speed trains for safe and efficient operation. However, the Support Vector Machine (SVM) method for bearing health monitoring can lead to overfitting when outliers are present in the training dataset. In order to address this issue, an improved Significance SVM (SSVM) is proposed that assigns significant coefficients to samples in the model training process, giving less attention to outlier samples. The experiments on HST bearing vibration dataset demonstrate the effectiveness and stability of the proposed method under different noise levels.
IEEE SENSORS JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Anurag Choudhary, Deepam Goyal, Shimi Sudha Letha
Summary: The study proposed a method based on 2D-DWT infrared thermography for diagnosing bearing faults in induction motors. By using PCA and MD methods to extract and rank the most relevant features, SVM was employed for fault classification. The results showed that this method performed well in identifying bearing faults.
IEEE SENSORS JOURNAL
(2021)
Article
Engineering, Electrical & Electronic
Vikash Kumar, Subrata Mukherjee, Alok Kumar Verma, Somnath Sarangi
Summary: This article presents an AI-based nonparametric filter technique for fault diagnosis of gearboxes. The proposed technique improves the existing technique by adding proper features and can provide higher classification accuracy on signals acquired at an affordable sampling rate.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Engineering, Electrical & Electronic
Mingliang Cui, Youqing Wang, Xinshuang Lin, Maiying Zhong
Summary: The study proposes a feature distance stack autoencoder (FD-SAE) for rolling bearing fault diagnosis, which has stronger feature extraction ability and faster network convergence speed compared to existing methods.
IEEE SENSORS JOURNAL
(2021)
Article
Chemistry, Analytical
Yuanjing Guo, Shaofei Jiang, Youdong Yang, Xiaohang Jin, Yanding Wei
Summary: This paper proposes a novel approach for gearbox fault feature extraction and fault diagnosis based on improved VME. By using STFT and a new index SDE, which solves the problem of extracting weak impulse features, the feasibility and effectiveness of the proposed approach are verified through simulation and experimental results.
Article
Acoustics
Jiachi Yao, Chao Liu, Keyu Song, Chenlong Feng, Dongxiang Jiang
Summary: A fault diagnosis method based on acoustic signals for planetary gearbox is proposed, using FDM decomposition and RF classification algorithm to improve fault diagnosis accuracy.
Article
Engineering, Mechanical
Adel Afia, Fawzi Gougam, Chemseddine Rahmoune, Walid Touzout, Hand Ouelmokhtar, Djamel Benazzouz
Summary: In this paper, an intelligent algorithm based on robust empirical mode decomposition, time domain features, and equilibrium optimizer is proposed for gear and bearing fault diagnosis. The experimental results demonstrate the effectiveness of the proposed approach in detecting, identifying, and classifying all gear and bearing defects even under different operating modes.
JOURNAL OF VIBRATION ENGINEERING & TECHNOLOGIES
(2023)
Article
Computer Science, Information Systems
Yuhan Wu, Xianbo Sun, Yi Zhang, Xianjing Zhong, Lei Cheng
Summary: This paper proposes a power transformer fault diagnosis method based on KPCA and TISOA-SVM, which optimizes the parameters of SVM to build the optimal diagnosis model, achieving higher diagnostic accuracy and efficiency.
Article
Chemistry, Analytical
Cong Dai Nguyen, Cheol Hong Kim, Jong-Myon Kim
Summary: This paper proposes a new fault diagnostic model for multi-degree tooth-cut failures in a gearbox using ANCT and DRPCA. Experimental results show that the proposed model outperforms other methods in terms of identification accuracy.
Article
Engineering, Electrical & Electronic
Yi Liao, Weiguo Huang, Changqing Shen, Zhongkui Zhu, Jianping Xuan, Lingfeng Mao
Summary: This paper introduces a multivariate non-convex logarithm penalty based on generalized infimal convolution smoothing for vibration signal denoising and compound fault diagnosis in gearboxes. By ensuring the global minimum of the overall cost function and deriving a convexity condition for the non-convex penalty, an optimal sparse solution can be calculated using a convex algorithm. The proposed method effectively induces sparsity and enhances estimation accuracy of signal components compared to classical convex L1 norm and newly developed non-convex generalized minimax-concave penalties.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2021)
Article
Engineering, Mechanical
Junchuan Shi, Dikang Peng, Zhongxiao Peng, Ziyang Zhang, Kai Goebel, Dazhong Wu
Summary: Traditional model-based fault diagnosis techniques struggle to automatically extract spatial and temporal features, while deep learning methods are capable of automatic feature extraction but struggle to extract spatial and temporal features simultaneously.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Chemistry, Analytical
Yang Feng, Xiangfeng Zhang, Hong Jiang, Jun Li
Summary: This study proposes a compound fault diagnosis method for wind turbine gearboxes based on MOMEDA and RSSD, which effectively extracts fault characteristics and identifies fault types through signal preprocessing, decomposition, and envelope demodulation.
Article
Automation & Control Systems
Yu Wei, Yuantao Yang, Minqiang Xu, Wenhu Huang
Summary: A novel signal processing scheme combining refined composite hierarchical fuzzy entropy and random forest is proposed for fault diagnosis of planetary gearboxes. The proposed method outperforms existing methods in identifying fault types of planetary gearboxes through effective fault pattern identification.
Article
Engineering, Electrical & Electronic
Zhiqiang Zhang, Qingyu Yang, Yanyang Zi, Zongze Wu
Summary: Single-layer representation learning (SLRL) shows great potential for gearbox fault diagnosis. However, existing methods may not be able to learn sufficiently discriminative features from complex vibration signals. In this study, we propose a new SLRL method called discriminative sparse AE (DSAE) that incorporates label information to enhance feature learning. Experimental results on a gearbox dataset demonstrate that DSAE achieves superior performance in terms of diagnosis accuracy, stability, and anti-noise capability compared to state-of-the-art methods. Moreover, the discriminator in DSAE also exhibits better feature selection ability than traditional feature selectors.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Engineering, Multidisciplinary
Zhongxin Chen, Feng Zhao, Jun Zhou, Panling Huang, Wenping Song
Summary: This study introduces a new strategy of applying SMOTE and new selecting strategy to SVM algorithm, which improves labeling efficiency by preprocessing unlabeled samples and determining the number of labeling samples with experts based on the current hyperplane to achieve efficient and accurate classification of micro-defects on different batches of pistons.
Article
Automation & Control Systems
Zhongxin Chen, Yongwei Tang, Zenglin Gao, Jun Zhou, Panling Huang
Summary: This paper proposes an active learning framework called CALF to improve the efficiency of moisture content prediction in fluidized bed granulation with fewer query samples. The framework utilizes conditional variational auto-encoder and selective ensemble algorithm to effectively incorporate the information in query samples. The effectiveness of the proposed method is verified through large batch experiments.
JOURNAL OF PROCESS CONTROL
(2022)
Article
Automation & Control Systems
Zhongxin Chen, Yiran Shen, Binbin Chen, Jun Zhou, Panling Huang, Hengchang Zang, Yongxia Guan
Summary: This paper proposes a new online TCM composition analysis platform for high-quality online spectral data acquisition. The generative framework VasLine, based on deep learning, is designed to estimate multiple critical quality attributes (CQAs). The results show that VasLine outperforms state-of-the-art regression approaches for all 7 different CQAs, with R2 metrics higher than 0.95.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
