Article
Engineering, Electrical & Electronic
Mingxin Zhao, Junbo Peng, Shuangming Yu, Liyuan Liu, Nanjian Wu
Summary: This paper focuses on penalty-based pruning method for already compact networks, proposing a novel penalty term shaped like an upside-down Laplace distribution to apply more pressure on potential weak channels. Furthermore, the paper addresses the residual block pruning problem by eliminating a scaling factor, a skill often overlooked in other research.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2022)
Article
Computer Science, Interdisciplinary Applications
Seyed Omid Sajedi, Xiao Liang
Summary: This paper explores the use of deep learning and Bayesian inference in structural health monitoring, addressing prediction uncertainty and conducting three case studies. The uncertainty metrics show correlations with misclassifications, and a surrogate model is proposed to trigger human interventions.
COMPUTER-AIDED CIVIL AND INFRASTRUCTURE ENGINEERING
(2021)
Article
Engineering, Electrical & Electronic
Hiroki Kuroda, Daichi Kitahara
Summary: This paper presents a convex recovery method for block-sparse signals whose block partitions are unknown a priori. A nonconvex penalty function is introduced to adapt the block partition for the signal of interest. By exploiting a variational representation of the l(2) norm, the proposed penalty function as a convex relaxation of the nonconvex one is derived. An iterative algorithm is developed for the block-sparse recovery model designed with the proposed penalty, which guarantees convergence to a globally optimal solution. Numerical experiments demonstrate the effectiveness of the proposed method.
IEEE TRANSACTIONS ON SIGNAL PROCESSING
(2022)
Article
Materials Science, Characterization & Testing
Roberto Miorelli, Clement Fisher, Andrii Kulakovskyi, Bastien Chapuis, Olivier Mesnil, Oscar D'Almeida
Summary: This paper presents an automatic defect localization and sizing procedure for Structural Health Monitoring using guided waves imaging, applied to an aluminum plate with active piezoelectric sensors. The strategy utilizes a convolutional neural network trained on numerical simulations of guided wave signals and processed by the delay and sum imaging algorithm, showing effectiveness in inverting both synthetic and experimental data.
NDT & E INTERNATIONAL
(2021)
Article
Engineering, Multidisciplinary
Hedong Li, Demi Ai, Hongping Zhu, Hui Luo
Summary: An innovative compressed sensing-based approach is proposed for data recovery in concrete structural health monitoring, with the orthogonal matching pursuit algorithm shown to be superior to convex optimization in terms of calculation consumption and recovered errors during electromechanical admittance data loss recovery.
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
(2021)
Article
Nanoscience & Nanotechnology
Eunsik Choi, Kunsik An, Kyung-Tae Kang
Summary: This study proposes a deep-learning-based method to identify the droplet jetting status in inkjet printing process. By using a convolutional neural network based on the MobileNetV2 model and optimized hyperparameters, the method can classify images captured with a CCD camera and evaluate jetting conditions with high accuracy.
ACS APPLIED MATERIALS & INTERFACES
(2022)
Article
Engineering, Multidisciplinary
Qiuyue Pan, Yuequan Bao, Hui Li
Summary: This paper proposes a novel approach for data anomaly detection based on transfer learning, which utilizes the similarity of anomalous patterns across different bridges and shares the knowledge incorporated in a deep neural network to achieve high-accuracy data anomaly identification for bridge groups.
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
(2023)
Article
Computer Science, Artificial Intelligence
Lanfan Jiang, Wenxing Zhu
Summary: This article introduces a novel iterative weighted group thresholding method for signal recovery from underdetermined linear systems. Extensive computational experiments show that the proposed method is competitive in terms of group selection, estimation accuracy, and computation time compared to state-of-the-art methods.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Xiaoqiang Li, Chengyu Guo, Yifan Wu, Congcong Zhu, Jide Li
Summary: This paper proposes a Group-aware Contrastive Network (GACN) that extracts discriminative features using contrastive learning for robust age estimation in lightweight models.
IET IMAGE PROCESSING
(2022)
Article
Multidisciplinary Sciences
Yuan Bao, Zhaobin Liu, Zhongxuan Luo, Sibo Yang
Summary: In this paper, a novel smooth group L-1/2 (SGL(1/2)) regularization method is proposed for pruning hidden nodes in the fully connected layer of convolutional neural networks. The SGL(1/2) method approximates the weights to 0 at the group level to enable pruning, and numerical results demonstrate its superiority in terms of sparsity.
Article
Engineering, Chemical
Wenbo Zhu, Jinhong Zhang, Jose Romagnoli
Summary: This work focuses on developing a process data analytics model with better generalization ability and transferability. By using convolutional neural networks for transitional invariant feature extraction, the trained model can be reused on different variables or different processes.
INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH
(2022)
Article
Engineering, Electrical & Electronic
Bin Zhang, Xiaobin Hong, Yuan Liu
Summary: The proposed deep convolutional neural network probability imaging algorithm provides an automatic high-level damage index extraction method for guided wave imaging, overcoming the limitations of manual feature extraction and showing good generalization and performance in detecting damage.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2021)
Article
Computer Science, Artificial Intelligence
Zhuangzhi Chen, Jingyang Xiang, Yao Lu, Qi Xuan, Zhen Wang, Guanrong Chen, Xiaoniu Yang
Summary: This article studies the graph structure of the neural network and proposes a regular graph pruning (RGP) method to achieve one-shot neural network pruning. The experiments show that the average shortest path-length (ASPL) of the graph is negatively correlated with the classification accuracy of the neural network, and RGP has a strong precision retention capability with high parameter and FLOPs reduction.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Mechanics
Thanh N. Huynh, Jaehong Lee
Summary: This article presents a two-stage optimization approach for finding optimal blended composite laminate designs by predicting the optimal thickness distribution. By predicting the optimal thicknesses, the method simplifies the blending optimization problem and improves the combinatorial optimization efficiency.
COMPOSITE STRUCTURES
(2024)
Article
Computer Science, Artificial Intelligence
Srikrishna Iyer, T. Velmurugan, A. H. Gandomi, V. Noor Mohammed, K. Saravanan, S. Nandakumar
Summary: The proposed system presents a multi-robot-based fault detection system for railway tracks, utilizing a hardware prototype that implements a master-slave robot mechanism, combining ultrasonic sensor inputs and image processing to classify surface defects, with the CNN model performing well. Fault location and status can be relayed to a central location using GSM, GPS, and cloud storage technologies. The system is extended to optimize energy utilization and increase the overall network throughput by simulating the LEACH protocol with 100 robot nodes.
NEURAL COMPUTING & APPLICATIONS
(2021)
Article
Engineering, Multidisciplinary
Yuequan Bao, Zhiyi Tang, Hui Li, Yufeng Zhang
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
(2019)
Article
Construction & Building Technology
Zhiyi Tang, Zhicheng Chen, Yuequan Bao, Hui Li
STRUCTURAL CONTROL & HEALTH MONITORING
(2019)
Article
Engineering, Multidisciplinary
Yuequan Bao, Zhicheng Chen, Shiyin Wei, Yang Xu, Zhiyi Tang, Hui Li
Article
Engineering, Multidisciplinary
Yuequan Bao, Zhiyi Tang, Hui Li
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
(2020)
Article
Engineering, Mechanical
Zhicheng Chen, Zhiyi Tang, Jiahui Chen, Hui Li
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2020)
Article
Engineering, Industrial
Zhengliang Xiang, Yuequan Bao, Zhiyi Tang, Hui Li
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2020)
Article
Construction & Building Technology
Dawei Liu, Zhiyi Tang, Yuequan Bao, Hui Li
Summary: This study proposes a machine-learning-based approach to identify the modal parameters of output-only data for structural health monitoring, utilizing the independence of modal responses and the principles of unsupervised learning. The designed neural network uses a complex loss function to extract modal responses from structural vibration data and constrain the training process. The approach is able to blindly extract modal information from system responses, as demonstrated through numerical examples and verification with actual dataset from a cable-stayed bridge.
STRUCTURAL CONTROL & HEALTH MONITORING
(2021)
Article
Chemistry, Multidisciplinary
Jiaxing Guo, Zhiyi Tang, Changxing Zhang, Wei Xu, Yonghong Wu
Summary: Structural health monitoring systems can continuously monitor the operational state of structures, generating a large amount of monitoring data. Identifying extreme events in the presence of faulty data is challenging. This study proposes a deep learning-based method with visual interpretability to identify seismic data under sensor faults interference, which effectively identifies seismic data mixed with various types of faulty data while providing good interpretability.
APPLIED SCIENCES-BASEL
(2023)