Search

A deep-seated landslide could release numerous microseismic signals from creep-slip movement, which includes a rock-soil slip from the slope surface and a rock-soil shear rupture in the subsurface. Machine learning can effectively enhance the classification of microseismic signals in landslide seismic monitoring and interpret the mechanical processes of landslide motion. In this paper, eight sets of triaxial seismic sensors were deployed inside the deep-seated landslide, Jiuxianping, China, and a large number of microseismic signals related to the slope movement were obtained through 1-year-long continuous monitoring. All the data were passed through the seismic event identification mode, the ratio of the long-time average and short-time average. We selected 11 days of data, manually classified 4131 data into eight categories, and created a microseismic event database. Classical machine learning algorithms and ensemble learning algorithms were tested in this paper. In order to evaluate the seismic event classification performance of each algorithmic model, we evaluated the proposed algorithms through the dimensions of the accuracy, precision, and recall of each model. The validation results demonstrated that the best performing decision tree algorithm among the classical machine learning algorithms had an accuracy of 88.75%, while the ensemble algorithms, including random forest, Gradient Boosting Trees, Extreme Gradient Boosting, and Light Gradient Boosting Machine, had an accuracy range from 93.5% to 94.2% and also achieved better results in the combined evaluation of the precision, recall, and F1 score. The specific classification tests for each microseismic event category showed the same results. The results suggested that the ensemble learning algorithms show better results compared to the classical machine learning algorithms. [ABSTRACT FROM AUTHOR]

Understanding how Aspergillus oryzae responds to light is critical for developing efficient light regulation strategies in the brewing and waste treatment industries. Although continuous light is known to restrict A. oryzae, little is known about A. oryzae's sensitivity to light with photoperiod. In this study, we used pulse wave modulation (PWM) to generate nine pulsed blue light (PBL) treatments with varying peak light intensities and frequencies. The effect of PBL on A. oryzae was then compared to that of continuous blue light (CBL). Our findings showed that A. oryzae GDMCC 3.31 mycelium developed faster and produced more conidia under PBL with specific peak intensities and frequencies than under CBL treatment when the light dose and average light intensity were held constant. The colony diameter and conidia count under the two PBL treatments (PL-20_40%_1 Hz and PL-400_20%_10 kHz) were 1.13 and 1.22 times greater than under the CBL treatments, respectively. This different response may be mainly attributed to A. oryzae's adaptation to the light–dark cycles in nature. Furthermore, an interactive effect was found between peak light intensity and frequency. This work includes pulsed wave modulation as a new factor that influences the A. oryzae photoresponse and recommends it in the development of light regulation methods for fermentation. [ABSTRACT FROM AUTHOR]

Radio astronomy uses radio telescopes to detect very faint emissions from celestial objects. However, human-made radio frequency interference (RFI) is currently a common problem faced by most terrestrial radio telescopes, and it is getting worse with the development of the economy and technology. Therefore, it is essential to monitor and evaluate interference during the planning, construction, and operation stages of the radio telescope and protect the quiet radio environment around the radio astronomical site. In this paper, we present a software for an RFI analysis of the radio environment around the telescope. In this software, information has been collected, including the location of the site; the technical specifications, such as aperture and the frequency range of the radio telescopes; and the terrain around the site. The software and its modules are composed of telescope, geographic, and meteorological databases, and analysis modules of terrestrial and space-based RFI. Combined with the propagation characteristics of radio waves, we can analyze and evaluate RFI on the ground and in space around the radio telescope. The feasibility of the software has been proved by the experimental implementation of the propagation properties and RFI source estimation. With this software, efficient technical support can be expected for protecting the radio environment around the telescope, as well as improving site selection for planned radio astronomical facilities. [ABSTRACT FROM AUTHOR]

Antimicrobial blue light (aBL) therapy is a novel non-antibiotic antimicrobial approach which works by generating reactive oxygen species. It has shown excellent antimicrobial ability to various microbial pathogens in many studies. However, due to the variability of aBL parameters (e.g., wavelength, dose), there are differences in the antimicrobial effect across different studies, which makes it difficult to form treatment plans for clinical and industrial application. In this review, we summarize research on aBL from the last six years to provide suggestions for clinical and industrial settings. Furthermore, we discuss the damage mechanism and protection mechanism of aBL therapy, and provide a prospect about valuable research fields related to aBL therapy. [ABSTRACT FROM AUTHOR]

The continued scaling-down of nanoscale semiconductor devices has made it very challenging to obtain analytic surface potential solutions from complex equations in physics, which is the fundamental purpose of the MOSFET compact model. In this work, we proposed a general framework to automatically derive analytical solutions for surface potential in MOSFET, by leveraging the universal approximation power of deep neural networks. Our framework incorporated a physical-relation-neural-network (PRNN) to learn side-by-side from a general-purpose numerical simulator in handling complex equations of mathematical physics, and then instilled the "knowledge" from the simulation data into the neural network, so as to generate an accurate closed-form mapping between device parameters and surface potential. Inherently, the surface potential was able to reflect the numerical solution of a two-dimensional (2D) Poisson equation, surpassing the limits of traditional 1D Poisson equation solutions, thus better illustrating the physical characteristics of scaling devices. We obtained promising results in inferring the analytic surface potential of MOSFET, and in applying the derived potential function to the building of 130 nm MOSFET compact models and circuit simulation. Such an efficient framework with accurate prediction of device performances demonstrates its potential in device optimization and circuit design. [ABSTRACT FROM AUTHOR]

Mesocarbon microbead-silicon carbide (MCMB-SiC) composites were prepared by hot-press sintering (2100 °C/40 MPa/1 h) with two different graphitized MCMBs as the second phase, which exhibited good self-lubricating properties. The effects of the graphitization degree of the MCMBs on the microstructure and properties of the composites were investigated contrastively. The results showed that the composites that added raw MCMBs with a low degree of graphitization had excellent self-sintering properties, higher densities, and better mechanical properties; by comparison, the composites that added mature MCMBs with a high degree of graphitization, which has regular and orderly lamellar structures, not only had good mechanical properties but also exhibited a lower and more stable dry friction coefficient (0.35), despite the higher wear rate (2.66 × 10−6 mm3·N−1·m−1). Large amounts of mature MCMBs were peeled off during the friction process to form a uniform and flat graphite lubricating film, which was the main reason for reducing the dry friction coefficient of the self-lubricating composites and making the friction coefficient more stable. [ABSTRACT FROM AUTHOR]

Needle beams have received widespread attention due to their unique characteristics of high intensity, small focal size, and extended depth of focus (DOF). Here, a single–layer all–dielectric metalens based on Pancharatnam–Berry (PB) was used to efficiently generate and focus an azimuthally polarized vortex beam at the same time. Then, additional phase or amplitude modulation was respectively adopted to work with the metalens to produce optical needles. By decorating the PB metalens with the binary optical element (BOE), an optical needle with full–width–at–half–maximum (FWHM) of 0.47 λ and DOF of 3.42 λ could be obtained. By decorating the PB metalens with an annular aperture, an optical needle with long DOF (16.4 λ) and subdiffraction size (0.46 λ) could be obtained. It is expected that our work has potential applications in super–resolution imaging, photolithography, and particle trapping. [ABSTRACT FROM AUTHOR]

In recent studies, self-supervised learning methods have been explored for monocular depth estimation. They minimize the reconstruction loss of images instead of depth information as a supervised signal. However, existing methods usually assume that the corresponding points in different views should have the same color, which leads to unreliable unsupervised signals and ultimately damages the reconstruction loss during the training. Meanwhile, in the low texture region, it is unable to predict the disparity value of pixels correctly because of the small number of extracted features. To solve the above issues, we propose a network—PDANet—that integrates perceptual consistency and data augmentation consistency, which are more reliable unsupervised signals, into a regular unsupervised depth estimation model. Specifically, we apply a reliable data augmentation mechanism to minimize the loss of the disparity map generated by the original image and the augmented image, respectively, which will enhance the robustness of the image in the prediction of color fluctuation. At the same time, we aggregate the features of different layers extracted by a pre-trained VGG16 network to explore the higher-level perceptual differences between the input image and the generated one. Ablation studies demonstrate the effectiveness of each components, and PDANet shows high-quality depth estimation results on the KITTI benchmark, which optimizes the state-of-the-art method from 0.114 to 0.084, measured by absolute relative error for depth estimation. [ABSTRACT FROM AUTHOR]

Periodontal disease is the most common oral chronic inflammatory disease in humans. Recent studies have indicated that red light Photobiomodulation (PBM) could inhibit cell inflammation effectively, but the effect of different doses of PBM on the treatment of inflammation has to be improved. Thus, this study was aimed to investigate the effects of various doses of PBM (630 ± 30 nm, (1) 5 mW/cm2, 1 J/cm2, 200 s; (2) 5 mW/cm2, 3 J/cm2, 600 s; (3) 5 mW/cm2, 9 J/cm2, 1800 s; (4) 5 mW/cm2, 18 J/cm2, 3600 s; (5) 5 mW/cm2,36 J/cm2, 7200 s) on the anti-inflammatory response of human gingival fibroblasts. Our results suggested that PBM (630 ± 30 nm) with doses of 18 J/cm2 and 36 J/cm2 could significantly inhibit the production of inflammatory cytokines such as Prostaglandin E2 (PGE2) and IL-8, presumably due to the fact that a high dose of PBM treatment could reduce intracellular Reactive oxygen species (ROS) in human gingival fibroblasts, thus reducing the expression of COX-2 enzyme. In addition, it was found that treatment with different doses of PBM (630 ± 30 nm) did not result in reduced mitochondrial membrane potential and mitochondrial dysfunction in human gingival fibroblasts. Our study provides a theoretical reference for the selection of PBM parameters and the application of PBM in the clinical treatment of periodontitis. [ABSTRACT FROM AUTHOR]

In traditional Reinforcement Learning (RL), agents learn to optimize actions in a dynamic context based on recursive estimation of expected values. We show that this form of machine learning fails when rewards (returns) are affected by tail risk, i.e., leptokurtosis. Here, we adapt a recent extension of RL, called distributional RL (disRL), and introduce estimation efficiency, while properly adjusting for differential impact of outliers on the two terms of the RL prediction error in the updating equations. We show that the resulting "efficient distributional RL" (e-disRL) learns much faster, and is robust once it settles on a policy. Our paper also provides a brief, nontechnical overview of machine learning, focusing on RL. [ABSTRACT FROM AUTHOR]

With the development of low-power technology in electronic devices, the wireless sensor network shows great potential in applications in health tracing and ocean monitoring. These scenarios usually contain abundant low-frequency vibration energy, which can be collected through appropriate energy conversion architecture; thus, the common issue of limited battery life in wireless sensor devices could be solved. Traditional energy-converting structures such as the cantilever-beam type or spring-mass type have the problem of high working frequency. In this work, an eccentric pendulum-based electromagnetic vibration energy harvester is designed, analyzed, and verified with the finite element analysis method. The pendulum that contains alternative distributed magnets in the outer side works as a rotor and has the advantages of a simple structure and low center frequency. The structure size is well scalable, and the optimal output performance can be obtained by optimizing the coil thickness and width for a given diameter of the energy harvester. The simulation results show that the energy harvester could work in ultra-low frequencies of 0.2–3.0 Hz. A full-scale prototype of the energy harvester is manufactured and tested. The center working frequency is 2.0 Hz with an average output power of 8.37 mW, which has potential for application in driving low-power wireless sensor nodes. [ABSTRACT FROM AUTHOR]

The equation of state and stability of topaz at high-pressure/high-temperature conditions have been investigated by in situ synchrotron X-ray diffraction (XRD) and Raman spectroscopy in this study. No phase transition occurs on topaz over the experimental pressure–temperature (P-T) range. The pressure–volume data were fitted by the third-order Birch–Murnaghan equation of state (EoS) with the zero-pressure unit–cell volume V0 = 343.86 (9) Å3, the zero-pressure bulk modulus K0 = 172 (3) GPa, and its pressure derivative K'0 = 1.3 (4), while the obtained K0 = 155 (2) GPa when fixed K'0 = 4. In the pressure range of 0–24.4 GPa, the vibration modes of in-plane bending OH-groups for topaz show non-linear changes with the increase in pressure, while the other vibration modes show linear changes. Moreover, the temperature–volume data were fitted by Fei's thermal equation with the thermal expansion coefficient α300 = 1.9 (1) × 10−5 K−1 at 300 K. Finally, the P-T stability of topaz was studied by a synchrotron-based single-crystal XRD at simultaneously high P-T conditions up to ~10.9 GPa and 700 K, which shows that topaz may maintain a metastable state at depths above 370 km in the upper mantle along the coldest subducting slab geotherm. Thus, topaz may be a potential volatile-carrier in the cold subduction zone. It can carry hydrogen and fluorine elements into the deep upper mantle and further affect the geochemical behavior of the upper mantle. [ABSTRACT FROM AUTHOR]

Crystal-chemical properties of synthetic Almandine-Pyrope (Alm-Pyr) solid solutions were investigated by X-ray single-crystal diffraction and Raman spectroscopy. Garnet solid solution with different compositions were synthesized from powder at 4.0 GPa and annealed at 1200 °C for 48 h by a multi-anvil pressure apparatus. Garnet crystals with different sizes (about 60–1000 μm) were obtained from synthesis. The results of X-ray single-crystal diffraction show that the unit cell constants decrease with increasing Pyr contents in the synthetic Alm-Pyr crystals due to the smaller ionic radius of Mg2+ in eightfold coordination than that of Fe2+. The data exhibit obviously positive deviations from ideal mixing volumes across the Alm-Pyr join which may be caused by the distortion of the SiO4 tetrahedron. Moreover, the significant decrease in the average M-O bond length and volume of the [MgO8]/[FeO8] dodecahedron with increasing Pyr contents are the most important factors to the decrease in the Alm-Pyr crystal unit cell constant and volume. On the other hand, selected bond distances (average , , and distances) have a linear correlation with the unit-cell parameter, but the distance has nonlinear correlation. With increasing the unit-cell parameter, the average distance increases significantly, followed by the average and distances. While the distance changes negligibly further confirming the conclusion that the significant decrease of the average M-O bond length of the [MgO8]/[FeO8] dodecahedron with increasing Pyr contents are the most important factors to the decrease in the Alm-Pyr crystal unit cell volume. In the Raman spectra collected for the Alm-Pyr solid solutions, Raman vibration mode assignments indicate that the Raman vibrational spectra change along the Alm-Pyr binary solution. The mode frequencies of Si-O stretching, Si-O bending, and the rotation of the SiO4-tetrahedron (R(SiO4)) decrease linearly, while the translational modes of the SiO4-tetrahedron (T(SiO4)) increase with increasing Alm contents. [ABSTRACT FROM AUTHOR]