XUHUI HUANG

Dr. XUHUI HUANG

orcid

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
30 September 2024 in Applied Sciences
Verified
Journal Article
Open access
PDF file
XHXUHUI HUANG
XUHUI HUANG
MHMing Han
Ming Han
YDYiming Deng
Yiming Deng
In this paper, we propose a novel approach to coordinate-based acoustic emission (AE) source localization to address the challenges of limited and imbalanced datasets from fiber-optic AE sensors used for structural health monitoring (SHM). We have developed a hybrid deep learning model combining four generative adversarial network (GAN) variants for data augmentation with an adapted inception neural network for regression-based prediction. The experimental setup features a single fiber-optic AE sensor based on a tightly coiled fiber-optic Fabry-Perot interferometer formed by two identical fiber Bragg gratings. AE signals were generated using the Hsu-Nielsen pencil lead break test on a grid-marked thin aluminum plate with 35 distinct locations, simulating real-world structural monitoring conditions in bounded isotropic plate-like structures. It is demonstrated that the single-sensor configuration can achieve precise localization, avoiding the need for a multiple sensor array. The GAN-based signal augmentation expanded the dataset from 900 to 4500 samples, with the Wasserstein distance between the original and synthetic datasets decreasing by 83% after 2000 training epochs, demonstrating the high fidelity of the synthetic data. Among the GAN variants, the standard GAN architecture proved the most effective, outperforming other variants in this specific application. The hybrid model exhibits superior performance compared to non-augmented deep learning approaches, with the median error distribution comparisons revealing a significant 50% reduction in prediction errors, accompanied by substantially improved consistency across various AE source locations. Overall, this developed hybrid approach offers a promising solution for enhancing AE-based SHM in complex infrastructures, improving damage detection accuracy and reliability for more efficient predictive maintenance strategies.

15 July 2025 in NDT & E International
Verified
Journal Article
Closed access
LPLei Peng
Lei Peng
NZNa Zhang
Na Zhang
XHXUHUI HUANG
XUHUI HUANG
ZLZi Li
Zi Li
YDYiming Deng
Yiming Deng

31 May 2024 in Engineering Research Express
Verified
Journal Article
Open access
XHXUHUI HUANG
XUHUI HUANG
OEObaid Elshafiey
Obaid Elshafiey
SMSubrata Mukherjee
Subrata Mukherjee
FKFarzia Karim
Farzia Karim
YZYupeng Zhu
Yupeng Zhu
YDYiming Deng
Yiming Deng
Structural health monitoring aims to detect damage progression in materials. This study focuses on categorizing crack stages, a critical aspect of monitoring structural integrity. By leveraging acoustic emission (AE) monitoring, cracks can be analyzed in a data-driven manner. However, applying AE analysis poses several challenges, including discrepancies between simulated AE data from models and experimental data from the field, as well as class imbalance in crack progression data, with a scarcity of late-stage data. To bridge the gap between theory and experiments, our approach employs domain adaptation to synchronize simulated and actual AE data. The model learns robust domain-invariant features through meticulous experimentation across training epochs. Quantitative analysis of the model’s performance provides key insights. F1 scores vary with feature counts, and domain adaptation outperforms by 20% on highly imbalanced datasets. This emphasizes the model’s adaptability for precise crack classification, even with underrepresented damage classes. In summary, this study advances structural health monitoring by offering a solid AE analysis approach. Core contributions include reconciling simulated and experimental data discrepancies, tackling class imbalance, optimizing feature extraction, and demonstrating robust crack stage categorization. The insights gained highlight the merits of domain adaptation and data-driven AE analysis for predicting crack progression.

20 February 2024 in Research in Nondestructive Evaluation
Verified
Journal Article
Closed access
XHXUHUI HUANG
XUHUI HUANG
LPLei Peng
Lei Peng
SMSubrata Mukherjee
Subrata Mukherjee
CHCiaron Hamilton
Ciaron Hamilton
XSXiaodong Shi
Xiaodong Shi
YDYiming Deng
Yiming Deng

20 November 2024 in NDT & E International
Verified
Journal Article
Closed access
XHXUHUI HUANG
XUHUI HUANG
ZLZi Li
Zi Li
LPLei Peng
Lei Peng
YUFEI CHUYUFEI CHU
YUFEI CHU
ZMZebadiah Miles
Zebadiah Miles
YDYiming Deng
Yiming Deng

7 June 2021 in Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems
Verified
Journal Article
Open access
PDF file
SMSubrata Mukherjee
Subrata Mukherjee
XHXUHUI HUANG
XUHUI HUANG
LULalita Udpa
Lalita Udpa
YDYiming Deng
Yiming Deng
Systems in service continue to degrade with the passage of time. Pipelines are among the most common systems that wear away with usage. For public safety, it is of utmost importance to monitor pipelines. Magnetic flux leakage (MFL) testing is a widely used nondestructive evaluation (NDE) technique for defect detections within the pipelines, particularly those composed of ferromagnetic materials. Pipeline inspection gauge (PIG) procedure based on line scans can collect accurate MFL readings for defect detection. However, in real world, applications involving large pipe sectors such as extensive scanning techniques are extremely time consuming and costly. In this article, we develop a fast and cheap methodology that does not need MFL readings at all the points used in traditional PIG procedures but conducts defect detection with similar accuracy. We consider an under-sampling based scheme that collects MFL at uniformly chosen random scan points over large lattices instead of extensive PIG scans over all lattice points. On the basis of readings from the chosen random scan points, we use kriging to reconstruct MFL readings. Thereafter, we use thresholding-based segmentation on the reconstructed data for detecting defective areas. We demonstrate the applicability of our methodology on synthetic data generated using finite element models and on MFL data collected via laboratory experiments. In these experiments, spanning a wide range of defect types, our proposed novel MFL-based NDE methodology is witnessed to have operating characteristics within the acceptable threshold of PIG-based traditional methods and thus provide an extremely cost-effective, fast procedure with competing error rates.