Search

This study aims to propose and evaluate alternative configurations for producing maleic anhydride (MA) through the n-butane oxidation pathway to enhance overall environmental friendliness. Seven alternative configurations, which vary in the number of non-adiabatic reactors and heat exchange methods, as well as the stripping agents used for recovering MA in the purification section, are rigorously designed and comprehensively compared. Among the proposed configurations, we have found that the one utilizing single reactor with co-current cooling and water as the stripping agent (Scheme 2) shows the most promise. It has an optimal yield of 70.0 %, an energy efficiency of 97.52 %, and exergy efficiency of 44.92 %. A cradle-to-gate life cycle impact assessment determines a Global Warming Potential (GWP) of 0.091 kg-CO2eq/kg-MA from this scheme. In contrast, a techno-economic evaluation determines a minimum required selling price (MRSP) of 0.965 USD/kg-MA from this scheme. All these indicators are superior compared to those documented in the existing literature, suggesting a higher level of environmental friendliness. Subsequently, a suitable control strategy is proposed for the optimized Scheme 2and evaluated by rejecting various kinds of feed disturbances (i.e., ± 10 % change in feed flowrate, ± 5 % change in molten salt flowrate, 1 % and 2 % of i-butane as feed impurity, and catalyst deactivation up to 40 %). Enhanced controllability is achieved by regulating the flowrate ratio between the flashed liquid and the n-butane (FL/NBU), which promptly indicates the reaction conversion during production. Based on the findings, we have concluded that overall environmental friendliness can be significantly enhanced with moderate modifications in the process configuration, which are easily feasible in the industry.

Highly substituted naphthalene derivatives were prepared via palladium‐catalyzed reaction of 5‐(2‐bromophenyl)pent‐3‐en‐1‐ynes with arylboronic acids. Typically, reaction of 3‐bromo‐4‐(3‐(4‐chlorophenyl)‐5‐phenyl‐1‐(p‐tolyl)pent‐2‐en‐4‐yn‐1‐yl)‐N‐methylaniline (1a) with PhB(OH)2in the presence of Pd(PPh3)4as the catalyst under basic conditions provided 8‐benzhydryl‐7‐(4‐chlorophenyl)‐N‐methyl‐5‐(p‐tolyl)naphthalen‐2‐amine (2a) quantitatively. Overall, 19 new naphthalene compounds were obtained by this methodology and their structures were characterized by spectroscopic methods. In addition, crystal structure of 2jwas determined to confirm the structural details. Reaction pathway leading to the desired product is also discussed. Using palladium‐catalyzed reaction of 5‐(2‐bromophenyl)pent‐3‐en‐1‐ynes with arylboronic acids was studied. Reaction proceeds with C‐Br bond activation, migratory insertion, metathesis with arylboronate, and reductive elimination to yield the desired naphthalene derivatives.

ABSTRACTAs advancements push for larger and more complex Artificial Intelligence (AI) models to improve performance, preventing the occurrence of overfitting when training overparameterized Deep Neural Networks (DNNs) remains a challenge. Despite the presence of various regularization techniques aimed at mitigating this issue, poor generalization remains a concern, especially when handling diverse and limited data. This paper explores one of the latest and most promising strategies to address this challenge, Sharpness Aware Minimization (SAM), which concurrently minimizes loss value and sharpness-related loss. While this method exhibits substantial effectiveness, it comes with a notable trade-off in increased training time and is founded on several approximations. Consequently, several variants of SAM have emerged to alleviate these limitations and bolster model performance. This survey paper examines the significant advancements achieved by SAM, delves into its constraints, and categorizes recent progressive variants that further enhance current State-of-the-Art results.

Herein, we developed a ligand-promoted Rh(III)-catalyzed C(sp3)–H thiolation of 8-methylquinolines. The effect of ligands on improving the activity of the catalytic centers has been studied in detail and proven to be significant. Various substituents are well tolerated under this reaction condition to provide potential precursors for organic synthesis. The mechanistic study suggests that the reaction may proceed through a five-membered rhodacycle intermediate via thiolation twice.

A palladium-promoted cascade cyclization of 5-(2-bromophenyl)pent-3-en-1-ynes is developed for the synthesis of benzo[a]fluorene derivatives. The reaction proceeds with oxidative addition of C–Br, insertion, C–H activation, and reductive elimination in sequential steps.

AbstractRecent progress in data-driven turbulence modeling has shown its potential to enhance or replace traditional equation-based Reynolds-averaged Navier-Stokes (RANS) turbulence models. This work utilizes invariant neural network (NN) architectures to model Reynolds stresses and turbulent heat fluxes in forced convection flows (when the models can be decoupled). As the considered flow is statistically one dimensional, the invariant NN architecture for the Reynolds stress model reduces to the linear eddy viscosity model. To develop the data-driven models, direct numerical and RANS simulations in vertical planar channel geometry mimicking a part of the reactor downcomer are performed. Different conditions and fluids relevant to advanced reactors (sodium, lead, unitary-Prandtl-number fluid, and molten salt) constitute the training database. The models enabled accurate predictions of velocity and temperature, and compared to the baseline $k - \tau $k−τturbulence model with the simple gradient diffusion hypothesis, do not require tuning of the turbulent Prandtl number. The data-driven framework is implemented in the open-source graphics processing unit–accelerated spectral element solver nekRS and has shown the potential for future developments and consideration of more complex mixed convection flows.

AbstractMixed convection of low and unitary Prandtl fluids in a vertical passage is fundamental to passive heat removal in liquid metal and gas-cooled advanced reactor designs. Capturing the influence of buoyancy in flow and heat transfer in engineering analysis is hence a cornerstone to the safety of the next-generation reactor. However, accurate prediction of the mixed convection phenomenon has eluded current turbulence and heat transfer modeling approaches, yet further development and validation of modeling methods is limited by a scarcity of high-fidelity data pertaining to reactor heat transfer. In this work, a series of direct numerical simulations was conducted to investigate the influence of buoyancy on descending flow of liquid sodium, lead, and unitary Prandtl fluid in a differentially heated channel that represents the reactor downcomer region. From time-averaged statistics, flow-opposing/aiding buoyant plumes near the heated/cooled wall distort the mean velocity distribution, which gives rise to promotion/suppression of turbulence intensity and modification of turbulent shear stress and heat flux distribution. Frequency analysis of time series also suggests the existence of large-scale convective and thermal structures rising from the heated wall. As a general trend, fluids of lower Prandtl number were found to be more susceptible to the buoyancy effect due to stronger differential buoyancy across the channel. On the other hand, the effectiveness of convective heat transfer of the three studied fluids showed a distinct trend against the influence of buoyancy. Physical reasoning on observation of the Nusselt number trend is also discussed.

AbstractDeveloping heat transfer correlations for buoyancy-driven flows and mixed convection is challenging, especially if the fluid’s Prandtl (Pr) number is not close to 1. For advanced nuclear reactor (Generation IV) designs, the downcomer plays a crucial role in normal operation and loss-of-power scenarios. The fluid-flow behavior in the downcomer can involve forced, mixed, or natural convection. Characterizing the heat transfer for these changing regimes is a serious challenge, especially in the heat transfer deterioration region. In this paper, the downcomer is simplified to heated parallel plates. The high–Pr number fluid FLiBe (a mixture of lithium fluoride and beryllium fluoride) is considered for all simulations. Direct numerical simulations using the graphics processing unit–based spectral element code NekRS are performed for a wide range of the Richardson number, from 0 to 400, at two different FLiBe Pr numbers (12 and 24). This results in an unprecedented 74 cases in total. Each case’s Nusselt number is calculated to evaluate existing heat transfer correlations.Moreover, we propose several new modifications for cases without satisfactory choice. As a result, several novel mixed-convection heat transfer correlations have been built for high–Pr number fluids. The correlations are expressed as a function of the buoyancy number, covering several mixed-convection regimes. The Pr number effect on the Nusselt number behavior is also analyzed in detail. We also propose a novel method to evaluate the heat transfer deterioration region. Modified Reynolds-Gnielinski forced-convection correlations are defined for the laminarization region, and a free-convection correlation is used for the natural-convection-dominated region. These correlations can describe well the trend in the heat transfer–deficient region.

AbstractStable density stratification in a large enclosure could significantly hamper the effectiveness of natural convection cooling in pool-type liquid metal or gas-cooled advanced reactors. In addition, accurate prediction of stratified front behavior remains to be a challenging task for turbulence modeling. With the rapid growth of high-performance-computing capabilities in recent years, conducting high-fidelity simulations for a large-timescale transient has become more affordable and hence a valuable data source to support turbulence modeling as well as to gain further physical insights. In this work, direct numerical simulation is performed at experiment-consistent conditions to simulate the density stratification transient High-Resolution Jet (HiRJET) facility. Specifically, we focus on the case where an injected aqueous sugar solution has 1.5% density higher than that in the enclosure. In the early stage of the transient, the impingement of the denser jet to the bottom surface of the enclosure promoted turbulent mixing locally. This rendered the establishment of the mixture layer, formation and swift upward propagation of the stratified front, and elevation with (locally) the highest vertical concentration gradient. As the front rose, the diminishing turbulent mass flux slowed down the propagation, and a larger vertical concentration gradient was established. In this stage, both the velocity and the concentration scalar showed large-timescale fluctuation behavior around the stratified front. For the concentration time signal, the characteristic frequency in the power spectral density was found to agree well with the Brunt-Väisällä frequency. The preliminary validation endeavor showed that the stratified front location and the corresponding concentration gradient magnitude in the simulation agreed well with the experiment data. Further validation will mainly revolve around benchmarking against high-resolution density measurement and high-order flow statistics.

The present work describes a preclinical trial (in silico, in vivoand in vitro) protocol to assess the biomechanical performance and osteogenic capability of 3D-printed polymeric scaffolds implants used to repair partial defects in a sheep mandible. The protocol spans multiple steps of the medical device development pipeline, including initial concept design of the scaffold implant, digital twin in silicofinite element modeling, manufacturing of the device prototype, in vivodevice implantation, and in vitrolaboratory mechanical testing. First, a patient-specific one-body scaffold implant used for reconstructing a critical-sized defect along the lower border of the sheep mandible ramus was designed using on computed-tomographic (CT) imagery and computer-aided design software. Next, the biomechanical performance of the implant was predicted numerically by simulating physiological load conditions in a digital twin in silicofinite element model of the sheep mandible. This allowed for possible redesigning of the implant prior to commencing in vivoexperimentation. Then, two types of polymeric biomaterials were used to manufacture the mandibular scaffold implants: poly ether ether ketone (PEEK) and poly ether ketone (PEK) printed with fused deposition modeling (FDM) and selective laser sintering (SLS), respectively. Then, after being implanted for 13 weeks invivo, the implant and surrounding bone tissue was harvested and microCT scanned to visualize and quantify neo-tissue formation in the porous space of the scaffold. Finally, the implant and local bone tissue was assessed by in vitrolaboratory mechanical testing to quantify the osteointegration. The protocol consists of six component procedures: (i) scaffold design and finite element analysis to predict its biomechanical response, (ii) scaffold fabrication with FDM and SLS 3D printing, (iii) surface treatment of the scaffold with plasma immersion ion implantation (PIII) techniques, (iv) ovine mandibular implantation, (v) postoperative sheep recovery, euthanasia, and harvesting of the scaffold and surrounding host bone, microCT scanning, and (vi) in vitrolaboratory mechanical tests of the harvested scaffolds. The results of microCT imagery and 3-point mechanical bend testing demonstrate that PIII-SLS-PEK is a promising biomaterial for the manufacturing of scaffold implants to enhance the bone-scaffold contact and bone ingrowth in porous scaffold implants. MicroCT images of the harvested implant and surrounding bone tissue showed encouraging new bone growth at the scaffold-bone interface and inside the porous network of the lattice structure of the SLS-PEK scaffolds.

Catalytic asymmetric 1,2-allylation of aurone-derived azadienes is very difficult to achieve due to the driving force for aromatization and the greater steric hindrance of 1,2-addition compared with 1,4-addition. By taking advantage of the ability of nitrogen ligated metal complexes, we successfully demonstrated the first example of copper-catalyzed 1,2-allylation of azadienes with allylboronates for the highly enantioselective synthesis of homoallylic amines. Meanwhile, the enantioenriched 1,4-addition products could also be obtained through a subsequent 3,3-sigmatropic rearrangement of the 1,2-addition products. Extensive DFT calculations were carried out to elucidate the origins of high regioselectivity (1,2- vs 1,4-) and enantioselectivity.

ABSTRACTIn-process monitoring and quality control are the most critical aspects of the manufacturing industry, especially in ultra-precision machining (UPM) at an industrial scale. However, in-process ensuring product quality has been difficult, as any subtle change in the process influences the UPM process dynamics and the process outcome. In order to meet the increasingly soaring demand for precision components, intelligent monitoring of the machining process is essentially important and much needed. Capturing complex signal patterns through conventional signal processing for the UPM process is often challenging due to the comparably high noise levels in the industrial environment. Signals obtained during UPM are inherent transients and non-stationary, necessitating extensive and accurate features for classification. Accurate detection of anomalies may allow for quick corrective actions, reducing the degree of damage. Earlier research revealed multi-sensor analysis, which yields richer signal feature information, but the unavoidable sensor failure in conjunction with heterogeneous sensing made it challenging. In order to address the challenges, this paper investigates the feasibility of convolution neural network (CNN) for classifying abnormal and normal machining in the UPM process. The vibrational signals obtained from B&J 4533-B accelerometer during diamond turning are transformed into time-frequency-based log-spectrogram images. These images are classified using CNN, and the results show that a proposed convolutional neural network algorithm has demonstrated an accuracy of 85.92% in classifying images and thus the corresponding in-process machining status.

This letter demonstrates a 1200-V E-mode GaN-on-sapphire power transistor based on active passivation technique. The active passivation concept utilizes a thin p-GaN layer extending from the p-GaN gate towards near the drain to screen the surface traps. The fabricated active-passivation HEMT (AP-HEMT) with ${L}_{\text {GD}}$ of $27 ~\mu \text{m}$ exhibits a low ${R}_{\text {ON}}$ of $16.9 \Omega \cdot $ mm, corresponding to a specific ${R}_{\text {ON}}$ $({R}_{\text {ON, SP}})$ of 6.42 $\text{m}\Omega \cdot $ cm2. A breakdown voltage (BV) over 2000 V is obtained for the AP-HEMT. Besides, the AP-HEMT showcased excellent dynamic performance due to the surface shielding effect provided by the active passivation. Dynamic ON-resistance was characterized $120 ~\mu \text{s}$ after a 10-ms VDS-OFF stress. At VDS-OFF = 1200 V, the ratio of dynamic ${R}_{\text {ON}}$ to static ${R}_{\text {ON}}$ is 1.09. The results highlight the superior capabilities of active passivation technique for 1200-V GaN power transistors.

Aims:Targeted temperature management is recommended for at least 24 h in comatose survivors of in-hospital cardiac arrest (IHCA) after the return of spontaneous circulation; however, whether an extension for 72 h leads to better neurological outcomes is uncertain. Methods:We included data from the Qilu Hospital of Shandong University between July 20, 2019, and June 30, 2022. Unconscious patients who had return of spontaneous circulation lasting >20 consecutive min and received endovascular cooling (72 h) or normothermia treatment were compared in terms of survival-to-discharge and favorable neurological survival. Propensity score matching was used to formulate balanced 1:3 matched patients. Results:In total, 2,084 patients were included. Sixteen patients received extended endovascular cooling and 48 matched controls received normothermia therapy. Compared with the normothermia group, patients who received prolonged endovascular cooling had a higher survival-to-discharge rate. However, good neurological outcomes did not differ significantly. Before matching, Cox regression analysis, using mortality as the event, showed that extended endovascular cooling independently affected the survival of IHCA patients. Conclusions:Among comatose patients who had been resuscitated from IHCA, the use of endovascular cooling for 72 h might confer a benefit on survival-to-discharge.

Tumor metastasis and cancer recurrence are often a result of cell heterogeneity, where specific subpopulations of tumor cells may be resistant to radio- or chemotherapy. To investigate this physiological and phenotypic diversity, single-cell metabolomics provides a powerful approach at the chemical level, where distinct lipid profiles can be found in different tumor cells. Here, we established a highly sensitive platform using nanoflow liquid chromatography (nLC) combined with multinozzle emitter electrospray ionization mass spectrometry for more in-depth metabolomics profiling. Our platform identified 15 and 17 lipids from individual osteosarcoma (U2OS) and glioblastoma (GBM) cells when analyzing single-cell samples. Additionally, we used the functional single-cell selection (fSCS) pipeline to analyze the subpopulations of cells with a DNA damage response (DDR) in U2OS cells and fast migration in GBM cells. Specifically, we observed a down-regulation of polyunsaturated fatty acids (PUFAs) in U2OS cells undergoing DDR, such as fatty acids FA 20:3; O2 and FA 17:4; O3. Furthermore, ceramides (Cer 38:0; O3) and triglycerides (TG 36:0) were found to be down-regulated in fast-migrating GBM cells compared to the slow-migrating subpopulation. These findings suggest the potential roles of these metabolites and/or lipids in the cellular behavior of the subpopulations.

The inversion of transient electromagnetic (TEM) data entails a complex nonlinear problem with high dimensionality and ill-posedness. All convolutional neural networks (CNNs) for TEM inversion only utilize single-scale convolution kernels to extract data features, thereby restricting the model’s ability to perceive global information. Here, we propose a TEM inversion algorithm based on multiscale dilated CNNs (MDCNNs). The algorithm leverages dilated convolutions to expand the receptive field without introducing additional computational complexity or memory consumption. Simultaneously, it integrates multiscale features fusion modules to extract features from TEM responses, aiming to fully leverage the multiscale information within the data and enhance the reliability of inversion. Furthermore, we customize the design of the MDCNN and explore various data preprocessing methods to make it more suitable for TEM inversion. Synthetic data tests demonstrate that, compared to CNNs, the proposed method can provide more reliable inversion results across the entire depth range of subsurface models. Testing on noisy data also proves its superior generalization capability and noise robustness. In addition, we successfully applied the MDCNN to the measured data, and the forward response generated from its inversion results aligns well with the measured data, demonstrating its effectiveness in inverting real-world data. In summary, the proposed MDCNN can rapidly provide reliable inversion results of TEM data, holding significant potential for real-time imaging of subsurface structures.

The navigation and positioning system of mobile robots using multisensor fusion has become a research hotspot in the aspects of high accuracy, low computational complexity, and strong stability. In order to improve the accuracy of sensor asynchronous information fusion and meet the application requirements of geometric structure feature degradation in warehousing logistics, an adaptive weighted factor graph (AWFG) positioning method using inertial measurement unit (IMU), LiDAR, and stereo camera is proposed. Combining the dominant features of three sensors and factor graph theory, a new multisensor fusion factor graph model is established. By dynamically adjusting the reliability of sensor measurement information, an adaptive factor weight function is designed to improve positioning accuracy and system stability under abnormal sensor or environmental interference conditions. Besides, sliding window optimization is added to limit the factor scale, and the variable elimination algorithm is combined to optimize the factors in the window to further reduce the computational complexity. Compared with extended Kalman filter (EKF) and particle swarm optimization (PSO) algorithms, simulation and experimental results show that the proposed method not only reduces the mean location error by about 30%, but also effectively enhances the computational efficiency and system stability.

Inexpensive, lightweight, sensitive, compact, easy-to-install, and mass-producible vibration sensors are the future wave. This article validates the use of a 3-D-printed doublet bracket that turns out to be a scissor-like X-structure with scissor-like kinematics when vibrating. The main concept is simple: an intact standard single-mode fiber (SMF) is folded into a droplet-bending shape and embedded in an arch bridge-type assembly connected to an X-shaped structure. When regular vibrations are transmitted to the X-structure, the continuous scissor motion of the X-structure drives the arch bridge-type assembly to regularly extend and shrink, thereby changing the size of the fiber bending, creating variations in the optical loss, and allowing for the measurement of the vibration. In addition, the sensitivity of the vibration measurement can be adjusted by altering the torque by varying the intersection angle of the doublet bracket in the X-structure or the fiber loss by varying the number of turns of fiber bending. In this way, the sensitivity can be adjusted according to the magnitude of the vibration amplitude of the object being measured in various measurement scenarios. For instance, if the vibration amplitude is weak, a high-sensitivity frame will be utilized for the measurement, and vice versa. Furthermore, this 3-D-printed-based X-structure and fiber-bending-based vibration sensor can be integrated with free-space optics (FSOs) for more versatile measurements in various applications. In summary, this scheme showcases an affordable, lightweight, variable-sensitivity, simple, and easily mass-producible fiber-optic vibration sensing technology.

Digital elevation models (DEMs) are crucial geographical data sources, whereas the resolution of commonly used DEM products is low and cannot meet the requirement of detailed geo-related applications. Deep learning-based methods have demonstrated to be effective in super-resolution (SR) techniques, which reconstruct high-resolution (HR) images from low-resolution (LR) images. However, existing deep learning methods have not fully considered the multiscale spatial heterogeneity and topographic knowledge of DEM data that differentiate them from traditional images. These defects inevitably lead to the localized smoothing of the reconstructed DEM and influence the reliability of downstream geographical analysis. This study proposes a transformer-based topography neural network (TTSR) for DEM SR incorporating a local-global deformable block (LGDB) for capturing the multiscale spatial heterogeneity and topographic knowledge, a spatio-channel coupled channel attention (SimAM) mechanism for reallocating channel weights and providing a supplement of the global spatial features, and an improved terrain loss function (iLoss) for mitigating noise across datasets. TTSR was validated using two publicly available real-world DEM datasets for recovering DEM from 30 to 10 m. The root mean square error (RMSE) of the proposed method was reduced by approximately 6%–30%, 4%–16%, and 1%–9% in elevation accuracy, slope accuracy, and aspect accuracy, respectively, compared to the best one of those state-of-the-art methods. This research provides new insights for improving the accuracy of the DEM SR, which will help generate global HR terrain products for geographical studies in the future.

Micro-milling technology is an efficient method for machining titanium alloy thin-walled micro parts. However, the machinability of thin-walled microstructures can be greatly affected by chatters due to their weak stiffness, especially in difficult-to-machine materials like titanium alloys. In order to obtain high-quality thin-walled micro parts, a chatter mitigation strategy is presented and its mechanism is analyzed in this paper. The titanium alloy workpiece is submerged into a kind of high-viscosity fluid, which can improve machining stability by increasing the damping of the micro-milling process. The stability lobe diagrams (SLD) are established, and the machinability is analyzed and compared under different machining conditions. The cutting force, surface roughness, micro-tool wear, and dimension accuracy of thin-walled micro parts are selected as the evaluation criteria for the machinability assessment. Well-designed micro-milling experiments are carried out under dry, lubrication, and chatter mitigation conditions using the same cutting process parameters. The results show that the use of cutting fluid can effectively improve machinability under stable cutting processes, but the effect of cutting fluid is weakened under chatter machining. The proposed chatter mitigation strategy can suppress regenerative chatter throughout the micro-milling process for thin-walled micro parts and improve their machinability comprehensively.

This letter reports on a high-linearity and low-leakage current AlN/GaN/InGaN coupling-channel HEMT (CC-HEMT) utilizing the $\text{N}_{{2}}\text{O}$ oxidation treatment process. Attributed to the channel coupling effect and the gate oxidation process, there is a significant improvement in both the leakage current and linearity figure of merit. The fabricated HEMT exhibited a low off-state leakage of 1.6 nA/mm, which achieves a high on/off current ratio of 109 for AlN-based transistors. The extrinsic ${f}_{\text {T}}$ and ${f}_{\text {max}}$ of the proposed CC-HEMT are 53 and 118 GHz, respectively. The two-tone power level linearity of the proposed CC-HEMT was characterized. The Ka-band output third-order intermodulation point (OIP3) and linearity figure of merit (LFoM) of OIP3/ ${P}_{\text {DC}}$ are 36.4 dBm and 12.9 dB, respectively. These results highlight the application prospects of the AlN/GaN/InGaN CC-HEMT system for high-linearity and low-power consumption power amplifiers.

When infrared small-dim target images under strong background clutters are employed to train a deep learning (DL)-based detection network, the model becomes biased toward the clutters, negatively impacting detection performance. While background estimation is able to address this issue, most convolutional neural network (CNN)-based ways require manual foreground mask extraction during the learning phases. In unsupervised tactics, it is often assumed that the background in frames is captured by a stationary camera. Howbeit, lots of small target sequences have dynamic backgrounds due to motion in the imaging platform, challenging this hypothesis. There is limited focus on unsupervised background estimation for small target images with sensor motion. To address this gap, a learning-based model, named background estimation for multiframe small-dim target detection (BEmST), is raised. BEmST combines a variational autoencoder with stable principal component pursuit (SPCP) optimization for unsupervised deep background modeling. Target detection is then performed using U-net++ on differences between the modeled background and input images. This innovative tactic integrates unsupervised probabilistic background estimation with supervised dense classification for bettered small target detection. Extensive qualitative/quantitative experiments on public datasets validate that BEmST not only outperforms state-of-the-art tactics in availably and robustly detecting small-dim target images across various challenging scenarios, but also achieves superior detection performance, such as higher probabilities of detection, lower false alarm rates, and larger areas under receiver operating characteristic (ROC) curves. The results pave a way for the future utilization of small-dim target images in a more efficient manner.

The inversion of transient electromagnetic (TEM) data presents a complex nonlinear problem, and traditional inversion methods encounter certain limitations. Data-driven machine learning techniques, such as convolutional neural network (CNN), provide an alternative approach to TEM inversion. However, conventional CNN faces challenges, such as limited generalization capability. Here, we propose a novel TEM inversion method called CNN-LightGBM that combines the strengths of CNN and the light gradient boosting machine (LightGBM). Specifically, CNN is employed to extract crucial features from TEM responses, which are then regressed using LightGBM to improve the reliability of the inversion results. The proposed method is tested on synthetic and measured data, and its performance is compared against the existing methods. Synthetic tests demonstrate that, compared with CNN, the proposed model provides more reliable inversion results in the entire depth range, exhibiting superior generalization capability and noise robustness. Meanwhile, the inversion efficiency of this method is remarkably high, allowing for the inversion of 7200 points of TEM data in just 1 s on a common PC. In addition, the successful application of the trained CNN-LightGBM in the inversion of measured data also proves the effectiveness of the method. The model can rapidly provide reliable inversion results and has tremendous potential for near real-time imaging of subsurface structures.

This paper verifies the feasibility and effectiveness of utilizing an integration technology of free-space optics (FSO) and satellite communications for an open radio access network (O-RAN) system while simultaneously transmitting fiber-optic sensing signals over the same FSO link. The achievement of O-RAN is to improve the flexibility, openness, and cost-effectiveness of current and next-generation wireless networks. Compared with radio wave-based communication, optical wireless communication (OWC), such as FSO, has many advantages, such as high-speed transmission, data confidentiality, and freedom from electromagnetic interference. Therefore, the use of FSO as a transmission link between ground radar transceivers for satellite communication and O-RAN systems will enhance the communication quality of satellite communication-based O-RAN systems. In particular, FSOs are small and easy to place at ground radar transceiver stations, satellite trucks, and communication target points. This makes it easier to quickly switch the transmission route of the radar transceiver station to different communication transmission and sensing monitoring target scenarios than physical optical fiber cable. In addition, in the fiber optic sensing part, use a machine learning model that combines advanced conditional generative adversarial network (CGAN) and convolutional neural network (CNN) models. This model aims to generate a large amount of sensor data and improve the sensing performance of fiber sensing systems. The experimental findings unambiguously show that the proposed integrated system achieves normal functioning in the communication part and outstanding temperature sensing measurement accuracy in the sensing part.

Edge termination is the key building block in power devices to enable near-ideal, avalanche breakdown voltage (BV). This work presents the design, fabrication, and physics of a GaN guard ring (GR) edge termination formed by selective-area nitrogen implantation through an epitaxial p-GaN layer. The fabrication of this termination only includes a single implantation step that does not require precise control of implant depth, rendering a large process latitude. The selective-area implantation produces p-GaN rings that are separated by the implanted, semi-insulating regions. The number and spacing of the p-type rings are found to determine the BV of the vertical GaN p-n diode. The 16-ring structure enables a BV of 1800 V, being 88% of the theoretical 1-D parallel-plane limit. Avalanche characteristics are observed in devices with a large variety of GR designs. Finally, we present a comprehensive survey on the efficiency, fabrication complexity, real estate, and avalanche capability of various edge termination techniques that have been reported in vertical GaN devices. The high efficiency (among the highest reported in avalanche-capable GaN terminations), simple and robust fabrication process, and uniform avalanche capability make this implanted GR a promising edge termination for high-voltage GaN devices.

An enhancement-mode (E-mode) p-type gallium nitride (p-GaN) gate high electron mobility transistor with a gate termination extension (GTE-HEMT) has been developed on a sapphire substrate, intended for power switching applications of kilovoltage (kV) level. The GTE-HEMT device exhibits a low ON-resistance ( ${R} _{ \mathrm{\scriptscriptstyle ON}}$ ) of $19.3~\Omega \cdot $ mm, corresponding to a specific ${R} _{ \mathrm{\scriptscriptstyle ON}}$ ( ${R} _{ \mathrm{\scriptscriptstyle ON},\text {SP}}$ ) of 7.33 $\text{m}\Omega \cdot $ cm2. Notably, the GTE-HEMT has achieved a breakthrough by increasing the breakdown voltage (BV) to 2573 V, surpassing the conventional p-GaN gate HEMT (Conv-HEMT) with its BV of 1679 V. The enhanced BV is ascribed to the gate termination extension (GTE) that reduces the gate-edge electric field peak, similar to a junction termination extension (JTE) in traditional Si power devices. In addition, the dynamic ${R} _{ \mathrm{\scriptscriptstyle ON}}$ of the GTE-HEMT has been improved, with a normalized dynamic ${R} _{ \mathrm{\scriptscriptstyle ON}}$ (dynamic ${R} _{ \mathrm{\scriptscriptstyle ON}}$ /static ${R} _{ \mathrm{\scriptscriptstyle ON}}$ ) of 1.4 and 2.6 for 650- and 1200-V OFF-state stress, respectively. The GTE-HEMT has been benchmarked to the state-of-the-art GaN power transistors, demonstrating its potential for high-voltage applications.

This study aimed to determine the effect of drilling parameters on hole cylindricity and surface roughness in drilling applications on aircraft wings. It was determined that increasing feed rate resulted in poor hole cylindricity and surface roughness. The ideal feed rate for better hole cylindricity is 0.230 mm/rev, while the ideal feed rate for better surface roughness is 0.200 mm/rev. Both hole cylindricity and surface roughness have a minimum correlation with each other in producing a good quality of drilled holes.

Aircraft is still dominated by aluminum alloys despite the growth of composites and other lightweight materials due to their fatigue crack resistance, superior damage tolerance, and mature production methods. Assembly process of the aircraft's structures, must be involved for rivets and bolts, required a precise dimension compliance into the aviation guideline. Poor hole quality is one of the major challenges might cause cracks in the aircraft structure and reduce their reliability thus, manufacturing costs are directly affected by this rejection of parts at the assembly stage. The aim of this paper is to study on feed rate, towards surface roughness and Built Up Edges (BUE), and drilling of aluminium alloys used in aircraft component. An experimental approach was used to determine the most efficient feed rate for drilling operations. Observations were made on the BUE formation and roughness of the surface after each drill had completed up to 30 holes in order to meet the requirements of machined parts. From the experiment's findings, it was observed that the BUE formation and surface roughness both marginally changed as feed rates were varied. This experiment's most efficient feed rate was 0.15 m/min, and met the experiment's goal of determining the most efficient drilling feed rate. In conclusion, as feed rate increased, Ra (surface roughness) increased as well, while BUE decreased. While, Ra was set to a high feed rate, its value increased marginally. Drilling on aluminum alloy T6-6061 was accomplished at an efficient feed rate, resulting in improved product quality and reduced tool and workpiece damage.

In this contribution, we rationally designed and controllably fabricated a NiMo/Al2O3-montmorillonite (3D-NiMo/Al2O3-MMT) monolithic catalyst via a 3D printing strategy with economical montmorillonite (MMT) as a binder. The catalytic performance of the resulting NiMo/Al2O3-MMT for 4,6-dimethyldibenzothiophene (4,6-DMDBT) hydrodesulfurization (HDS) was evaluated. The experimental results unveil that the 3D-NiMo/Al2O3-MMT monolithic catalyst exhibits robust stability and exceptional HDS activity with 99.2% 4,6-DMDBT conversion (residual 4 ppm of S), which is remarkably superior to that of conventional NiMo/Al2O3(61.5%), NiMo/MMT (63.2%), and even NiMo/Al2O3-MMT (76.5%) prepared by the mechanical-mixing method. This should be ascribed to the synthetic effect between the MMT binder and the Al2O3substrate, which effectively weakens the interaction between the Mo species and the Lewis acids on the original Al2O3surface, thereby significantly increasing the content of reducible Mo species and considerably facilitating the formation of more highly active NiMoS phase (Type II) with optimal average stacking layers and thereafter remarkably enhancing the ultradeep HDS activity of the 3D-NiMo/Al2O3-MMT monolithic catalyst.

The rising on-demand food-delivery (ODFD) industry has raised road safety concerns because of the often aggressive and risky driving behavior among ODFD drivers, especially those on motorcycles. This study aims to characterize their aggressive and risky driving behavior with respect to microscopic kinematic characteristics and vehicular interactions. We employed traffic videos collected by unmanned aerial vehicles at two intersections in Taipei, Taiwan, with a 0.1 s time resolution. We extracted vehicular trajectory data using artificial intelligence-based video recognition algorithms to obtain microscopic kinematic variables. We compared microscopic traffic flow characteristics (i.e., speed, lateral velocity, and acceleration) and microscopic interactions (i.e., weaving maneuver frequency, safety gap, and time-to-collision) presented by ODFD and non-ODFD drivers. In addition, we compared the difference in their driving behaviors between non-meal-peak hours and meal-peak hours, hypothetically caused by the platform-employed incentive program. We found that, compared to non-ODFD drivers, ODFD drivers could be more likely to perform aggressive and risky driving, indicated by their faster longitudinal and lateral speed, harsher acceleration/deceleration, more frequent weaving maneuvers, shorter safety gaps, and shorter time-to-collision. Also, their aggressive and risky driving behavior could more likely occur during meal-peak hours. To our knowledge, this research is the first study using naturalistic traffic data to investigate the revealed driving behavior of ODFD drivers, contributing to kinematic/quantitative understandings of ODFD drivers’ aggressive and risky driving. Based on these findings, policymakers and platform companies can prescribe countermeasures and devise training programs to improve public road safety and the occupational safety of ODFD drivers, a vulnerable occupational group emerging worldwide.

Micro-flotation experiments, Fourier transform infrared spectroscopy (FT-IR) tests, Zeta potential tests, X-ray photoelectron spectroscopy (XPS) analysis and molecular dynamics simulation were used to study the flotation behavior, separation and depressing mechanisms of graphite and sphalerite in the presence of H-acid monosodium salt (HAMS). Under the optimum flotation conditions, the recovery of sphalerite and graphite in the concentrate of mixed mineral flotation was 93.37% and 4.98%, respectively. Zeta potential tests and XPS analysis indicated that HAMS was considerably adsorbed on graphite surfaces but virtually absent on sphalerite surfaces. The FT-IR tests revealed that HAMS had no significant chemisorption on graphite surfaces. Molecular dynamics simulation inferred that the hydrophobic interaction between the naphthalene ring in HAMS and graphite was responsible for the adsorption of HAMS on the graphite surface.

In this article, we propose a novel meta-learning approach for improving the sensing quality of free space optics (FSO)-based multichannel fiber Bragg grating (FBG) sensor systems. Since it is difficult and time-consuming to collect large amounts of training data using real FBG sensor experiments, the proposed meta-learning can provide opportunities for reducing the amount of training data needed by learning from previous experience or knowledge. This makes our proposed meta-learning method suitable for scenarios, where a large amount of FBG sensor training data collection is challenging, limited, or expensive. Thus, we use a meta-learning approach to solve crosstalk problems and accurately predict the peak wavelength of each FBG sensor ( ${S}_{n}$ ) from the overlap spectra. The demonstrated results proved that using only a small amount of target data, the proposed meta-learning method can adapt more quickly and improve the central wavelength detection performance in the FBG sensor system than the traditional methods. Besides, the proposed FSO method addresses costs and topological limitations for the installation of fiber cables. Therefore, our proposed system is flexible, faster, cost-effective, adaptable, reduces the amount of training data needed, and can improve the accuracy of wavelength detection for FBG sensor systems.

Aerostatic spindles are commonly employed in ultra-precision machines. The aerostatic spindle, as a critical key component for the development of industrial ultra-precision grinding machines, requires an industrial feasible precision engineering approach to robust design and optimization in order to render the high-precision spindle system. This paper presents such an approach to the design and development of aerostatic spindles and the associated digital virtual spindle systems based on multiscale modelling and analysis. Multiscale modelling and analysis combined with the virtual spindle simulation are used as the kernel of the virtual spindle system, including innovative design on the spindle system structure, fluid dynamics modelling, drive and control system, and the integration of the spindle electromechanical system, which can be used to systematically model and simulate both the static and dynamic performances of the aerostatic spindle system. Experiments are carried out to evaluate and validate the above approach and the virtual spindle simulations, which can be further utilized for the development of next-generation high-precision aerostatic spindle systems.