Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Yue Hu has a total of 49 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Physics-guided graph Transformer for composite pipeline damage detection

    by Xuan Li, Yue Hu, Pugen Zhang, Jinsha Xu, Yi Chen, Haibin Miao, Xinguang Ren, Fucai Li

    Composite pipelines are widely used in oil and gas transportation owing to their high specific strength and corrosion resistance. However, guided-wave-based damage localization remains challenging due to multimodal wave propagation, noise interference, complex propagation-path relationships, and limited labeled damage locations. To address these challenges, this study proposes a physics-guided graph Transformer (PGGT) framework for composite pipeline damage localization. In the proposed method, guided-wave signals are first preprocessed, and actuator–receiver propagation paths are represented as graph nodes. A physics-guided radius graph is constructed by incorporating the spatial configuration of piezoelectric transducers and guided-wave propagation characteristics, enabling multi-channel guided-wave signals to be mapped into a physically meaningful non-Euclidean graph structure. A convolutional module is employed to extract local temporal features from guided-wave responses, a graph attention module is utilized to model topology relationships among signal-path nodes, and a Transformer module is introduced to capture long-range global dependencies among propagation paths. Comparative experiments are conducted under four training–testing split strategies with unseen damage locations. The proposed PGGT is compared with VGG16, ResNet18, ChebyNet, GCN, GAT, Transformer, and PGGDL in terms of mean absolute percentage error (MAPE), mean absolute error (MAE), and root mean square error (RMSE). The results demonstrate that PGGT achieves superior localization accuracy and generalization capability under limited labeled data. In the most challenging case, PGGT achieves a MAPE of 0.035 (that is, 3.5%), a MAE of 5.38 mm, and an RMSE of 6.23 mm. The proposed method provides an effective physics-informed learning framework for guided-wave-based damage localization of composite pipelines. Conflict of Interest Statement The authors declare no conflicts of interest.

  • Physical-guided graph deep learning for composite pipelines structural health monitoring

    by Xiaoqian Jiang, Yue Hu, Shuai Cao, Fangsen Cui, Fucai Li, Yang Gao, Fu-Zhen Xuan

    Composite pipelines, widely applied in the oil and gas industry due to their high strength-to-weight ratio and corrosion resistance, are nonetheless vulnerable to damage under harsh operating conditions, making structural health monitoring (SHM) crucial for enhancing safety and preventing potential failures. However, the complex service environment and anisotropic properties of composite pipelines pose challenges for traditional SHM methods, which often struggle to extract underlying features and identify structural defects. To address these challenges, a new graph deep learning method, termed physical-guided graph deep learning (PGGDL), which leverages physical rules to enhance defect detection accuracy in composite pipeline health monitoring, is proposed in this study. The PGGDL method constructs graph data by integrating signals based on physical rules, including the sensor arrangement and the guided wave propagation mechanism, enabling efficient data fusion. Meanwhile, the PGGDL constructs a temporal model and a graph attention network to capture both node-level and global spatial features. Furthermore, an adaptive learning rate strategy is designed to dynamically adjust the learning rate, improving training efficiency, achieving faster convergence, and enhancing model stability. Four experimental cases are conducted to demonstrate that the PGGDL outperforms traditional deep learning models in accurately detecting and localizing pipeline defects with limited data.


  • 3-D-Printed CSRR Array Sensor With a Resnet-FC Algorithm for Simultaneous Temperature and Humidity Detection

    by Kaihao Liang, Jinglong Bi, Yang Chen, Yang Gao, Xijin Hua, Wenfei Chen, Yue Hu, Fuzhen Xuan, Fuzhen

    Complimentary split ring resonator (CSRR) based sensor have great potential in structural health monitoring due to their characteristics of not requiring power supply during operation, no need for signal transmission through leads, and lower maintenance costs. Nevertheless, for practical applications, multi-parameter acquisition ability as well as a highly cost-effective and efficient fabrication method is desired for CSRR sensors. In this paper, a three dimensional (3D) printed CSRR array sensor is developed with a Resnet-FC algorithm for simultaneous detection of temperature and humidity. Based on the high precision and fast molding speed of digital light processing 3D printing technology, the integrated manufacturing of CSRR sensors is achieved. The sensitivities of the device to temperature and humidity within the range of 40-90 °C and 42-90 %RH are 610 kHz/°C and 1.851 MHz/%RH, respectively. A Resnet-FC algorithm trained with experimental data can effectively decouple the signal responses of temperature and humidity. The predicted temperature and humidity values closely match the experimental values, with errors of 5% approximately.


  • Bayesian hierarchical hyper-Laplacian priors for high-resolution defect imaging in pipe structures

    by Yue Hu, Xiaoqian Jiang, Yanping Zhu, Shuai Cao, Fangsen Cui, Fucai Li, Yang Gao, Fu-Zhen Xuan


  • Random Load Pattern Recognition of Test Road Based on a Laser Direct Writing Carbon-Based Strain Sensor and a Deep Neural Network

    by Kuan Ju, Shuo Weng, Chun Li, Lihui Zhao, Bo Li, Yue Hu, Yang Gao, Fuzhen Xuan

    The practical application of hydrogen-based fuel cell vehicles (HFCVs) requires the design of advanced stress monitoring system on the hydrogen storage devices, relying on accurate feedback and recognition from the vehicles load conditions. Herein, we proposed an intelligent monitoring method for continuous monitoring of dynamic load parameters and load recognition by combining carbon-based strain sensor, data acquisition and wireless transmission embedded unit, and machine learning (ML). The carbon-based strain sensor, prepared by laser direct writing on polyimide film, exhibited a sensitivity of 6.08 and a frequency response of 1 Hz for effectively measuring the strain status under varying road conditions. A miniaturized and highly integrated data acquisition and wireless transmission unit was adopted to implement edge computing and wireless communication. Wavelet packet decomposition (WPD) and Hilbert-Huang Transform (HHT) were fused to extract intrinsic features in random load spectrum, and the distance-based feature evaluation method was used to evaluate sensitive features to optimize the structure of the deep neural network (DNN). The DNN model could effectively recognize six types of test road conditions, with a test accuracy over 90%. This research effectively advances the design of the loading condition monitoring system of the HFCVs hydrogen storage device.


  • Bearings Degradation Assessment Based on Signal Significance Index and Deep Cumulative Features

    by Zhihao Chen, Wenjie Bao, Gangao Zuo, Wenqiang Zhao, Yue Hu, Fucai Li

    Rolling bearings are widely applied in rotating machinery, and its performance degradation assessment is a crucial work. Accurate degradation assessment and remaining useful life (RUL) prediction are of great significance for devising maintenance scheme and preventing sudden shutdown of machinery, with the selection of representative degradation features being the key factor. In this paper, a novel performance degradation assessment model is constructed based on signal significance index (SSI) and deep cumulative features (DCF). First, succinct and fast empirical mode decomposition (SFEMD) is employed to decompose the original signal. Second, an SSI approach based on morphological pattern (MP) coding and symbolic aggregate approximation (SAX) coding is proposed to select the representative intrinsic mode function (IMF). Thirdly, a new performance degradation feature, namely deep cumulative degradation feature (DCF), is proposed, which has good monotonicity and formal consistency in different failure types. Finally, an integrated LSTM network is constructed for performance degradation assessment. The comparative experimental results on the public data set show that the proposed model can provide an effective reference for the performance degradation assessment and RUL prediction.


  • Open Access

    A Hybrid Deep Learning Framework Based on Diffusion Model and Deep Residual Neural Network for Defect Detection in Composite Plates

    by Tianrui Huang, Yang Gao, Zhenglin Li, Yue Hu, Fuzhen Xuan

    The establishment of a structural health monitoring (SHM) system for the damage and defects of composite structures is of great theoretical and engineering value to ensure their production and operational safety. Advanced machine learning technologies, such as deep learning, have become one of the main driving forces for state monitoring and predictive analysis of these structures. However, it is difficult to obtain sufficient data to train the deep learning model, which may fail to build an accurate and efficient SHM model. To overcome this problem, a new method based on Lamb waves and the diffusion model (DM) is proposed to realize the identification and classification of different defects for carbon-fiber-reinforced polymer (CFRP) structures. In this study, DM is used as the generation model of data enhancement, and the optimized and improved DDPM model is constructed in this experiment. The deep residual neural network (DenseNet) is used to identify and classify the defect features from the Lamb wave signals. Experimental and test results show that the deep learning framework designed in this study based on DenseNet classification and DDPM data enhancement can accurately detect and classify damage signals of common defects in CFRP composite plates.


  • Guided Wave Damage Location of Pressure Vessel Based on Optimized Explainable Convolutional Neural Network for Multivariate Time Series Classification Neural Network

    by Junxuan Zhang, Chaojie Hu, Jianjun Yan, Yue Hu, Yang Gao, Fu-Zhen Xuan

    Guided wave is a key nondestructive technique for structural health monitoring due to its high sensitivity to structural changes and long propagation distance. However, to achieve high accuracy for damage location, large quantities of samples and thousands of iterations are typically needed for detection algorithms. To address this, in this paper, an eXplainable Convolutional neural network for Multivariate time series classification (XCM) is adopted, which is composed of one-dimensional (1D) and two-dimensional (2D) convolution layers to achieve high accuracy damage location on pressure vessels with limited training sets. By further optimizing the network parameters and network structure, the training time is greatly reduced and the accuracy is further improved. The optimized XCM improves the damage location precision from 95.5% to 98% with small samples (training set/validation set/testing set = 23/2/25) and low training epochs (under 100 epochs), suggesting that the XCM has great advantages in pressure vessel's damage location classification its potential for guided wave-based damage detection techniques in structural health monitoring.


  • Generalized Transient-Extracting Transform and Its Accurate Signal Reconstruction

    by Wenjie Bao, Yue Hu, Fucai Li

    Traditional time–frequency analysis (TFA) methods can effectively obtain instantaneous frequency (IF) features of nonstationary signals by constructing the signal model in the time domain. However, they fail to deal with the transient signal because the IF of the transient signal is discontinuous in the time domain and is a multivalued function with respect to time. Thus, in this article, we introduce a new TFA method, termed generalized transient-extracting transform (GTET), to obtain sharper time–frequency representation (TFR) of the transient signal by constructing frequency-domain model. First, we propose a new computational framework for group delay (GD). By deriving the general explicit formula for the N th-order GD, the programming implementation of any order GD can be achieved. Next, the concentrated TFR of GTET can be generated by extracting the energy from the generalized GD estimation. Finally, the signal reconstruction of GTET is derived from a new perspective. It overcomes the difficulty that the reconstruction method of transient-extracting transform cannot be generalized to high-order algorithms. The analysis result of the simulated signal with high-order GD shows that GTET has better performance than other TFA methods. Experimental results demonstrate that GTET can be used to analyze practical transient signals.


  • Bayesian Estimation of Instantaneous Speed for Rotating Machinery Fault Diagnosis

    by Yue Hu, Fangsen Cui, Xiaotong Tu, Fucai Li

    The instantaneous speed (IS), i.e., instantaneous frequency, is an important feature to exploit the time-frequency characteristic of nonstationary signals. The IS estimation is widely used in various fields of signal processing. The IS estimation can also be regarded as a useful tool for rotating machinery fault diagnosis under nonstationary conditions. However, most of the current IS estimation methods improve the IS estimation accuracy at the cost of increasing the computational time. In this article, a method based on Bayesian estimation is proposed for estimating the fast time-varying IS in a computationally efficient manner. A linear chirp signal is modeled within the Bayesian framework, where the likelihood function and the prior probability function of the IS are derived. The past information is used as the prior to guarantee the continuity of the IS ridge, which improves the robustness against the noise or other interferences. The maximum a posteriori rule is used to estimate the IS. The proposed method not only obtains a high IS estimation accuracy for the signal whose IS is fast time-varying and broadband but also improves the computational efficiency. A simulated signal and two experimental signals are considered for investigating the performance of the proposed method.