查看更多>>摘要：Convolutional neural network (CNN) has been extensively used in pattern recognition of partial discharge (PD) in gas-insulated switchgear (GIS) because of its powerful feature extraction ability. However, at this stage, manual trial and error is needed to construct the CNN. Moreover, the model is designed for specific datasets, which will cause domain bias when applied to a new dataset. Therefore, a novel differentiable neural architecture search method is proposed to automatically construct a GIS PD pattern recognition model. First, a factorized hierarchical search space is used to design the CNN architecture. Then, a discrete search space is relaxed into a continuous search space through a search strategy based on Gumbel-softmax. Experiments show that the recognition accuracy of the proposed method can reach 97.625%. Furthermore, the proposed method has strong robustness and high precision against noise and strong tolerance for unbalanced datasets.
查看更多>>摘要：Positioning seabed objects in ultra-deep waters is critical in the mapping, exploration, and industrial exploitation of oceans. This paper presents research on a subsea wellhead positioning process for the marking of seabed positions and its subsequent as-built verification. The results indicate that the ultra-short baseline (USBL) acoustic technique can provide static positioning accuracy and precision comparable to those of a long-baseline method (LBL) and significantly reduces the time and cost of the offshore project. The new approach was tested at different stages of the process. First, an innovative offset transformation method, which increases the precision of alignments, was proposed, followed by a step-by-step procedure for determining the positions of marker buoys on the seabed based on spin tests and cardinal heading calibration. Finally, the M-estimation method, which has not yet been applied in subsea engineering, was implemented and proved to be a suitable alternative to the ordinary least squares method.
Seeneevassen, SeydirenLeong, AinanKashan, Mohammad Ali MohammadzadehSwamy, Varghese...
8页
查看更多>>摘要：Nitrogen and synthetic air have been used as standard effluent gases in relative humidity (RH) sensing experiments, especially to characterise the dynamic sensing capabilities of RH sensors utilising graphene oxide (GO) as the sensing material. In the present work, we have used GO coated quartz crystal microbalance (QCM) as a tool to investigate the mass adsorption characteristics of samples subjected to multiple humidification-dehumidification cycles using N-2 or synthetic air as the effluent gas. We identified that the residual mass adsorbed and retained by the GO film is on average 71.6% higher when using N-2 in the initial humidification-dehumidification cycle and this subsequently impacted the hysteresis characteristics until the GO film saturated. To negate the impact of the effluent gas, the GO-based sample should be treated with multiple humidification and dehumidification cycles using either effluent gas.
Ismail, AfiqahSafuan, A. Rashid AhmadSa'ari, RadzuanRasib, Abd Wahid...
20页
查看更多>>摘要：Recent advancements in remote sensing techniques have made it possible to overcome the risky issues using conventional methods and have opened up new opportunities for collecting data on discontinuity characteristics. This research compares the application of a Terrestrial Laser Scanner (TLS) and Unmanned Aerial Vehicle (UAV) in rock slope stability analysis. Two case studies, Ulu Choh Quarry and Jelapang Rock Slope are analysed using Kinematic Analysis and Slope Mass Rating (SMR) to depict rock through the 3D point cloud. The techniques are compared with the hand mapping method concerning the accuracy of the data acquisition in assessing the rock slope. The standard deviation for the dip and dip direction between digital capture and hand mapping is 2.97 degrees and 2.86 degrees, which falls under the acceptable tolerance limit. The integration of UAV and TLS also generates excellent clear point cloud data from the top to the toe of the slope.
查看更多>>摘要：As one of the essential parts of the mechanical transmission system, rolling bearing is vital to ensure the safe operation of mechanical equipment. The rolling bearing goes through four stages from its installation to the end of its life: normal operation, early weak failure, serious failure, and failure. If faults can be found in the early failure stage of the whole life cycle and maintenance strategies can be adopted in time, the safe and trouble-free operation of the equipment can be guaranteed. However, the fault features are not apparent in the early failure stage of the bearing's full life cycle. Moreover, being completely submerged in strong background noise can easily occur, making early fault diagnosis challenging. This study presents a new sparse enhancement model based on kurtosis-wavelet total variation denoising (Kurt-WATV) for early fault feature extraction. Firstly, a sparse optimization model is constructed to extract the early fault feature, and the original signal is decomposed by the over-complete rational discrete wavelet transform (ORDWT). Then a fast iterative algorithm is deduced to solve the established sparse optimization model, and the optimal wavelet subband is selected by Kurt-WATV, which is reconstructed to the fault signal. Finally, the bearing test data from bearing's full-life cycle are adopted to illustrate the effectiveness and robustness of the proposed method. Results confirm that the established method can achieve excellent performance in early fault feature extraction.
查看更多>>摘要：Neural component analysis (NCA) is one of the latest nonlinear multivariate statistical process control (MSPC) methods, which consists in combining artificial neural networks (ANNs) with principal component analysis (PCA). However, NCA cannot handle the non-Gaussian feature and the extracted principal components (PCs) in NCA may not be the key information in the process data. Herein, we propose an improved NCA (INCA) which introduces a new cost function based on kurtosis to restrict the Gaussianity of PCs. We also propose a novel PC selection mechanism based on the information of PCs in the original data space rather than in the PC data space. INCA achieves almost 100% detection rates in three different types of faults in a simulation model test, and it can detect the fault in the thermal power plant process more than 1 min ahead of orthogonal nonlinear PCA (ONLPCA) and NCA.
查看更多>>摘要：In this paper, a modal decomposition imaging (MDI) algorithm for detecting delamination defects in carbon fiber composite plates using air-coupled Lamb waves was proposed. Compared with the traditional time-domain imaging method using amplitude difference, the proposed MDI algorithm contained modal decomposition process and rotating scanning defect probability imaging method, which was more suitable for analyzing the nonlinear and non-stationary leakage Lamb waves signal. The cross-correlation coefficient of the instantaneous energy was constructed to be the damage index, which was obtained by the pretreatment of the relatively pure A0 mode Lamb waves. The effectiveness of MDI algorithm proposed in this paper for realizing the delamination defects of carbon fiber composite plates was verified by qualitative imaging and quantitative characterization comparing with the defect probability imaging (DPI) algorithm. It would be suitable for achieving accurate characterization of defects in situ testing and large-area rapid scanning of aerospace composite plates after quickly scanning.
查看更多>>摘要：Multi-wavelength absorption in uncomplicated system is pursued continuously due to its advantage on measuring thin-film material dispersion. Here, we experimentally demonstrate a single-periodic metamaterial with polarization-insensitive multi-wavelength absorption for the infrared thin-film material measurement. The meta-atom of metamaterial absorber is constructed by holed cylindrical gold disk and gold ground separated by a ZnS substrate. Simulation shows that the multi-wavelength absorption is related to the multiple resonant ei-genmodes. The metamaterial absorber is proposed through photolithography process and multi-wavelength absorptions occur at 3.95 mu m, 4.27 mu m and 10.90 mu m, respectively, in the experiment. We show that the multi-wavelength absorption is very sensitive to the dielectric property of the coating thin-film material, which indicates the proposed absorber can be used to retrieve the complex refractive index of coating thin film at various wavelengths. Our work provides a convenient way to measure refractive indices at both mid-infrared and far-infrared regimes.
查看更多>>摘要：As an essential component of a gearbox, gears can damage a structure or even an entire gear transmission system in case of failures. As a result, advanced fault diagnosis methods are crucial to system's operation. Currently, single-signal-driven gear fault diagnosis techniques have been applied in many fields, but multipath noise and single-sensor sampling errors inevitably affected the accuracy of diagnosis. This paper proposes a gear fault diagnosis method based on a novel accommodative random weighting theory and a balanced binary one dimension ternary pattern (BB-1D-TP) model. It can accurately diagnose the types of gear failures under the circumstances of multiple channels and strong background noise. The novel accommodative random weighting algorithm reduces the total mean-square error (MSE) by adaptively adjusting the proportional connection between a measured value at a present state and a historical state. Then the balanced binary algorithm extracts texture features of fault signals for signal enhancement. In the end, the classification is done by using Support Vector Machine (SVM) method. The result of experiments demonstrated that the method in this article effectively improves accuracy and efficiency of gear fault identification.
NSTL
Elsevier