Identification of Coating Layer Pipeline Defects Based on the GA-SENet-ResNet18 Model DOI

S.. Wang,

Wei Liang,

Fang Shi

et al.

International Journal of Pressure Vessels and Piping, Journal Year: 2024, Volume and Issue: unknown, P. 105327 - 105327

Published: Sept. 1, 2024

Language: Английский

Interpretable contour encoding network customized for acoustic emission adaptive cepstrum in laser shock peening monitoring DOI
Rui Qin, Zhifen Zhang, Jing Huang

et al.

Journal of Manufacturing Processes, Journal Year: 2024, Volume and Issue: 132, P. 224 - 237

Published: Nov. 4, 2024

Language: Английский

Citations

3
Multi-point estimation weldment recognition and estimation of pose with data-driven robotics design DOI Creative Commons

Xiang-Yi Meng

Nonlinear Engineering, Journal Year: 2025, Volume and Issue: 14(1)

Published: Jan. 1, 2025

Abstract In robotic welding systems, weldment recognition and pose estimation play crucial roles in achieving precision efficiency. Weldment involves identifying classifying different types of weld joints components with high accuracy, often employing computer vision techniques machine learning algorithms trained on diverse datasets. Concurrently, determines the precise position orientation torch relative to weldment, which is for ensuring proper alignment execution tasks. Hence, this study proposed a multi-point entropy (MPEE) model estimation. The MPEE computes design data-driven points. estimates features. With estimated points Weldmart, are tracked fault detection. Through approach, employed robotics. specifically addresses challenge focuses estimating multiple within design. By leveraging integrates models, enhances accuracy reliability point results stated that dataset comprising joint variations, system achieves over 95% real-time applications. geometric hashing iterative closest enables an average error margin less than 1 mm.

Language: Английский

Citations

0
A novel fault early warning method for centrifugal blowers based on stacked denoising autoencoder and transfer learning DOI
Zhang You, Congbo Li, Ying Tang

et al.

Journal of Manufacturing Systems, Journal Year: 2024, Volume and Issue: 76, P. 443 - 456

Published: Aug. 22, 2024

Language: Английский

Citations

2
An incremental learning framework for pipeline weld crack damage identification and leakage rate prediction DOI Creative Commons
Jing Huang, Zhifen Zhang, Yanlong Yu

et al.

Engineering Applications of Computational Fluid Mechanics, Journal Year: 2024, Volume and Issue: 18(1)

Published: Sept. 26, 2024

Language: Английский

Citations

2
Deciphering laser shock peening quality monitoring: Wavelet-driven network with interpretability DOI
Rui Qin,

Zhifen Zhang,

Jing Huang

et al.

Advanced Engineering Informatics, Journal Year: 2024, Volume and Issue: 62, P. 102917 - 102917

Published: Oct. 1, 2024

Language: Английский

Citations

2
Accelerable adaptive cepstrum and L2-Dual Net for acoustic emission-based quality monitoring in laser shock peening DOI
Rui Qin, Zhifen Zhang, Jing Huang

et al.

Journal of Manufacturing Systems, Journal Year: 2024, Volume and Issue: 77, P. 301 - 319

Published: Sept. 27, 2024

Language: Английский

Citations

1
Large language models enabled intelligent microstructure optimization and defects classification of welded titanium alloys DOI Open Access

Suyang Zhang,

William Yi Wang, Xinzhao Wang

et al.

Journal of Materials Informatics, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 31, 2024

The quick developments of artificial intelligence have brought tremendous attractive opportunities and changes to smart welding technology. In the present work, a novel model, ConvNeXt, which incorporates advantages convolutional neural networks (CNNs) vision transformers (ViTs), has been designed identify defects. classification accuracy pre-trained ConvNeXt based on transfer learning method reaches as high 99.52% after 500 iterations training, while traditional CNNs MobileNetV2 ResNet34 achieve 85.94% 93.41%, respectively. Moreover, performance can be further improved through dataset optimization t-distributed stochastic neighbor embedding (t-SNE). addition, arc geometrical features are added input parameters for building back propagation network predict formation weld seam, led reduction in maximum prediction error seam thickness from 0.8 0.6 mm. Furthermore, out 28 sets experimental parameters, only four result errors exceeding 0.2 It is worth noting that large language models (LLMs) utilized facilitate automated programming defect recognition, including ChatGPT 3.5, Bing Copilot, Claude3, ERNIE Bot. LLM-aided technology applied develop image stitching programs, achieving unsupervised automatic multiple tissue images obtaining clear wide-field ones. These case studies deep technologies LLMs set up solidified block recognition during non-equilibrium solidification.

Language: Английский

Citations

1
Multi-Generator Tropical Cyclone Forecasting Based on Cross-Modal Fusion DOI
Qian Liu,

Sun Hu,

Yaocheng Gui

et al.

Published: Jan. 1, 2024

Language: Английский

Citations

0
Identification of Coating Layer Pipeline Defects Based on the GA-SENet-ResNet18 Model DOI

S.. Wang,

Wei Liang,

Fang Shi

et al.

International Journal of Pressure Vessels and Piping, Journal Year: 2024, Volume and Issue: unknown, P. 105327 - 105327

Published: Sept. 1, 2024

Language: Английский

Citations

0