Showing papers by "National University of Defense Technology published in 2023"
TL;DR: In this paper , a multi-channel gliding arc (MCGA) plasma was employed to enhance combustion near the flame blowout limit (BL) in a C2H4-fueled and cavity-based model scramjet combustor.
4 citations
TL;DR: Wang et al. as discussed by the authors proposed a new method combining the advantages of remote sensing cloud computing, long-term groundwater modeling simulation and GIS technology to solve it efficiently, and achieved a multivariate, spatial and long time series groundwater pollution risk assessment system, which is a new groundwater environmental impact assessment (GEIA).
4 citations
TL;DR: In this article , a new algorithm is designed that uses an improved range normalization and global matrix approach to address range dependence in ocean environments, and the main steps of the algorithm are parallelized, so OpenMP multithreading technology is also applied for further acceleration.
Abstract: The stepwise coupled-mode model is a classic approach for solving range-dependent sound propagation problems. Existing coupled-mode programs have disadvantages such as high computational cost, weak adaptability to complex ocean environments and numerical instability. In this paper, a new algorithm is designed that uses an improved range normalization and global matrix approach to address range dependence in ocean environments. Due to its high accuracy in solving differential equations, the spectral method has recently been applied to range-independent normal modes and has achieved remarkable results. This algorithm uses the Chebyshev--Tau spectral method to solve for the eigenmodes in the range-independent segments. The main steps of the algorithm are parallelized, so OpenMP multithreading technology is also applied for further acceleration. Based on this algorithm, an efficient program is developed, and numerical simulations verify that this algorithm is reliable, accurate and capable. Compared with the existing coupled-mode programs, the newly developed program is more stable and efficient at comparable accuracies and can solve waveguides in more complex and realistic ocean environments.
3 citations
TL;DR: In this paper , a nonlinear acoustic metamaterial (NAM) sandwich plate was proposed to suppress low-frequency and broadband vibration of light-weight and high-stiffness honeycomb sandwich plates with small mass cost.
2 citations
01 Jan 2023-Journal of Materials Chemistry C. Materials for Optical, Magnetic and Electronic Devices
TL;DR: In this paper , a CsCu 2 I 3 SA-based passively mode-locking fiber laser was demonstrated, which could operate stably for at least five months and could potentially unlock the pathways for ultrafast photonics based on lead-free halide perovskites.
Abstract: A CsCu 2 I 3 SA-based passively mode-locking fiber laser was demonstrated. The fiber laser could operate stably for at least five months, which will potentially unlock the pathways for ultrafast photonics based on lead-free halide perovskites.
2 citations
TL;DR: In this article , vanadium doped Ca0.5Ti2 (PO4)3/C nanofibers have been used to boost the ionic/electronic conductivities.
2 citations
TL;DR: In this article , the Hartley structure image-reject mixing (IRM) was introduced to compensate the amplitude and phase imbalance of the IQ channel, resulting in a high image rejection ratio up to around ∼50 dB for all channels.
2 citations
TL;DR: Zhang et al. as discussed by the authors introduced weight prediction into the AdamW optimizer to boost its convergence when training the deep neural network (DNN) models, and they performed extensive experimental evaluations on image classification and language modeling tasks to verify the effectiveness of their proposal.
Abstract: In this paper, we introduce weight prediction into the AdamW optimizer to boost its convergence when training the deep neural network (DNN) models. In particular, ahead of each mini-batch training, we predict the future weights according to the update rule of AdamW and then apply the predicted future weights to do both forward pass and backward propagation. In this way, the AdamW optimizer always utilizes the gradients w.r.t. the future weights instead of current weights to update the DNN parameters, making the AdamW optimizer achieve better convergence. Our proposal is simple and straightforward to implement but effective in boosting the convergence of DNN training. We performed extensive experimental evaluations on image classification and language modeling tasks to verify the effectiveness of our proposal. The experimental results validate that our proposal can boost the convergence of AdamW and achieve better accuracy than AdamW when training the DNN models.
1 citations
TL;DR: In this article , a model of coherent beam combining (CBC) architecture was proposed to customize the exotic light beams array with controllable structures, and a proof-of-concept experiment was carried out for further verification.
Abstract: The customization of optical vortices with controllable structures is of fundamental importance for promoting various applications such as materials processing, optical communication, and optical tweezers. Here, a model of coherent beam combining (CBC) architecture was proposed to customize the exotic light beams array with controllable structures. The simulated results revealed that the optical vortex lattice (OVL) with multi-modes can be generated in the far field, and a proof-of-concept experiment was carried out for further verification. The experiment results were in a high degree of fidelity with the simulated results, indicating that the controllable structures could be generated efficiently by using the CBC architecture. The phenomenon of generated hollow beam having no orbital angular momentum (OAM) and the OVL copying can be observed. Furthermore, the phase detection experimental results verified the OAM states, the topological charge (TC) could be modulated from −2 to + 2. This work may present a new understanding of generating and manipulating the structured light fields.
1 citations
TL;DR: In this article , a hollow chamber with different fuel injection positions was used to achieve continuous rotating detonation (CRD) in the hollow chamber, where the axial heat release rate of the former two modes were rapid, indicating that CRD waves propagated near the orifices.
1 citations
TL;DR: In this article , the effects of the O2 content (XO2) in the oxidizer on the MILD combustion behaviors of large-proportion semicoke mixtures, such as those involving temperature, MILD regime recognition, char surface consumption, homo/heterogeneous Damkhöler number, and NOx emission, under O2/N2 and O 2/CO2 atmospheres, were investigated.
01 Jan 2023
TL;DR: In this article , a large-mode-area multi-resonant all-solid photonic bandgap fiber (AS-PBGF) is proposed, with a hexagon core that is measured to be 46 μm in the corner-to-corner direction.
Abstract: In this paper, we have experimentally demonstrated a novel large-mode-area multi-resonant all-solid photonic bandgap fiber (AS-PBGF) which is specifically optimized for conventional applications of fiber lasers around 1 μm, with a hexagon core that is measured to be 46 μm in the corner-to-corner direction and a broad bandwidth from 970 nm to 1180 nm in the 3rd photonic bandgap (PBG). Bending loss lower than 0.1 dB/m of a 5-m-length fiber is obtained in the case of bending radius R greater than 25 cm. Robust single-mode (SM) operation of the fiber is verified within the whole PBG. In addition, we also report a phenomenon about the abnormal variation of M2 within the low loss region of the 3rd PBG. The evolution of beam quality within the 3rd PBG appears as a U-shaped curve rather than a constant. Finally, a preliminary conclusion about the relationship between the beam quality and SM characteristics in AS-PBGF which is quite different from that in conventional SIFs is verified.
TL;DR: In this paper , the iridium doped Co3O4 (Ir-Co 3O4) with low Ir content of 2.88 wt% was developed for efficient acidic OER.
TL;DR: In this paper , a two-layer path planning method for a cooperative ground vehicle (GV) and drone system is investigated, where the GV and the drone are used to conduct multiple area covering tasks collaboratively.
Abstract: A novel two-layer path planning method for a cooperative ground vehicle (GV) and drone system is investigated, where the GV and the drone are used to conduct multiple area covering tasks collaboratively. The GV acts as the mobile platform of the drone and takes the drone to visit a set of discrete areas. The drone takes off from the GV at potential nodes around target areas and scans each area for collecting information. The drone can be recharged in the GV during the time when it travels between different areas. The objective is to optimize the drone’s scanning path for area coverage and the GV’s travel path for visiting all areas. A 0–1 integer programming model is developed to formulate the problem. A two-stage heuristic based on cost saving strategy is designed to quickly construct a feasible solution, then the Adaptive Large Neighborhood Search (ALNS) algorithm is employed to improve the quality of the solution. A simulation experiment based on the parks in Changsha, China, is presented to illustrate the application of the method. Random instances are designed to further test the performance of the proposed algorithm.
TL;DR: In this article , the authors established the two-degree-of-freedom radial dynamic models of ball screw pairs under normal and faulty conditions to analyze the dynamic characteristics that may occur in the vibration responses of normal and fault-prone ball screw pair.
TL;DR: This article showed that overediting of GLI1 induced an arginine-to-glycine substitution, augmenting tumor-initiating potential and exhibiting a more aggressive phenotype.
08 May 2023-Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science
TL;DR: OmniLeg as mentioned in this paper is a concentrated-driven omnidirectional legged locomotion bionic leg, where all the motor actuators are centrally installed in the hip plate of the proposed leg to reduce the rotational inertia.
Abstract: The Flexible, fast and lightweight leg mechanism is the key to improve legged robot’s mobility, efficiency and adaptability on complex terrain. Inspired by the motion mechanism of legged animals, we report a concentrated-driven omnidirectional legged locomotion bionic leg – OmniLeg. All the motor actuators are centrally installed in the hip plate of the proposed leg to reduce the rotational inertia. Meanwhile, the special 3-Dimensional dual-parallelogram-linkages transmission mechanism enables it to have the ability of omnidirectional legged locomotion in 3-Dimensional space and makes the posture of leg’s endpoint does not change with the motion of the leg mechanism which simplifying the motion control. In this paper, firstly the theoretical model, including mobility, kinematics model, inverse kinematics model, dimension parameters and workspace model, have been derived, analyzed and verified with the design and control of the real prototype. Next, to demonstrate the ability of omnidirectional legged locomotion intuitively, we manufacture a single-legged omnidirectional mobile robot and design a quadruped robot, and the simulation and experiment result illustrate that the design and analytical mode of OmniLeg is feasible and correct, which could be widely applied in the field of legged robot.
TL;DR: Zhang et al. as mentioned in this paper used a measure, DPM, to quantitatively estimate the temporal and spatial specificity of MDDrelated genes expressed in a region during a period, based on the gene expression dataset of developing human brains from the Allen Brain Atlas.
Abstract: Major depression disorder (MDD) usually comes with structural and functional alterations of the brain, which are determined by altered gene expression patterns. We extracted threefunctional metrics, including Amplitude of Low Frequency Fluctuation, fractional Amplitude of Low Frequency Fluctuation, Regional Homogeneity from fMRI data, and one structural metrics, grey matter density from MRI data, to explore inter-group differences between MDD patients and healthy controls. Based on the association analysis of gene expression and brain imaging, we found 796 gene signatures whose expression was significantly correlated with functional or structural alteration of brain images in MDD. To understand how these genes affect the development of human brains, we used a measure, DPM, to quantitatively estimate the temporal and spatial specificity of MDD-related genes expressed in a region during a period, based on the gene expression dataset of developing human brains from the Allen Brain Atlas. We found MDD-related genes displayed more spatial specificity and less temporal specificity than control genes, suggesting MDD may occur at all periods, but only cause functional and structural changes in specific brain regions. Furthermore, MDD-related genes were found significantly overexpressed in ventrolateral prefrontal cortex, posterior superior temporal cortex, inferior parietal cortex, orbital frontal cortex, primary somatosensory cortex, primary motor cortex, primary auditory cortex, mediodorsal nucleus of the thalamus and cerebellar cortex mainly during adolescence and adulthood, associated with a decrease of grey matter density and a compensatory increase of functional activities affected by MDD. These findings can provide new insights into temporal-spatial expression pattern of MDD-related genes in human brains.
TL;DR: In this article , a deep neural network (DNN) improvement part and a reconstruction part are used to describe heterogeneous nurse rostering solutions and normalize these solutions to the same dimension, then the DNN model determines the best heuristic for approximating the local optimal solution.
TL;DR: Wu et al. as mentioned in this paper proposed a semi-supervised deep transfer learning framework for the microstructure recognition of nickel-based superalloys with different compositions and heat treatment procedures.
01 Feb 2023
TL;DR: In this paper , ground-based sampling, electron microscopy analyses, and theoretical calculations were used to investigate the interactions among the liquid-liquid phase separation (LLPS), soot redistribution in secondary particles, and fractal dimension (Df) of soot particles on the eastern rim of the Tibetan Plateau (TP).
Abstract: Abstract. Black carbon (BC, i.e., soot) absorbs radiation and contributes to glacier retreat over the Tibetan Plateau (TP). A lack of comprehensive understanding of the actual mixing state leads to large controversies in the climatic simulation of BC over the TP. In this study, ground-based sampling, electron microscopy analyses, and theoretical calculations were used to investigate the interactions among the liquid‒liquid phase separation (LLPS), soot redistribution in secondary particles, and fractal dimension (Df) of soot particles on the eastern rim of the TP. We found that more than half of the total analysed particles were soot-containing particles. One-third of soot-containing particles showed the LLPS phenomenon between organic matter and inorganic aerosols in individual particles, which further induced soot redistribution. The results show that a larger LLPS particle size, thicker organic coating, and smaller soot particles tended to drag soot from the sulfate core into the organic coating. The Df sequence is ranked as externally mixed soot (1.79 ± 0.09) < sulfate-coated soot (1.84 ± 0.07) < organic-coated soot (1.95 ± 0.06). We concluded that the soot redistribution process and high RH both promoted the morphological compaction of soot particles. This study indicates that soot-containing particles experienced consistent ageing processes that induced a more compact morphology and soot redistribution in the LLPS particles on the remote eastern rim of the TP. Understanding the microscopic changes in aged soot particles could further improve the current climate models and evaluations of BC’s radiative impacts on the eastern TP and similar remote air.
TL;DR: Zhang et al. as discussed by the authors proposed a two-stage concept prerequisite learning model (TCPL) to integrate domain-related knowledge to enhance pre-trained language models, and optimized the encoder of the resource-concept graph with hinge loss as an auxiliary training objective.
Abstract: Prerequisite chains are crucial to acquiring new knowledge efficiently. Many studies have been devoted to automatically identifying the prerequisite relationships between concepts from educational data. Though effective to some extent, these methods have neglected two key factors: most works have failed to utilize domain-related knowledge to enhance pre-trained language models, thus making the textual representation of concepts less effective; they also ignore the fusion of semantic information and structural information formed by existing prerequisites. We propose a two-stage concept prerequisite learning model (TCPL), to integrate the above factors. In the first stage, we designed two continual pre-training tasks for domain-adaptive and task-specific enhancement, to obtain better textual representation. In the second stage, to leverage the complementary effects of the semantic and structural information, we optimized the encoder of the resource–concept graph and the pre-trained language model simultaneously, with hinge loss as an auxiliary training objective. Extensive experiments conducted on three public datasets demonstrated the effectiveness of the proposed approach. Our proposed model improved by 7.9%, 6.7%, 5.6%, and 8.4% on ACC, F1, AP, and AUC on average, compared to the state-of-the-art methods.
01 Jan 2023
TL;DR: In this paper , a feature extraction of walkie-talkie signals is carried out to improve the accuracy of individual identification, where the original signal spectrum is moved to zero-frequency, filtered by a low-pass filter and resampled.
Abstract: In order to improve the accuracy of walkie-talkie individual identification, feature extraction of walkie-talkie signals is carried out. In this paper, five walkie-talkies of the same type and production batch are used as signal sources for experiments. By performing FFT on the signals sent by the walkie-talkies when there is no human voice, it is found that there is a stable frequency deviation in the transmission source of the walkie-talkies and the causes of such deviation are briefly analyzed. In order to make this deviation separable, the original signal spectrum is moved to zero-frequency, filtered by a low-pass filter and resampled. Then the normalized FFT peak point is extracted as the fingerprint feature of the walkie-talkie, and the discriminant analysis classification method is used to classify this feature. Experimental results show that the identification accuracy of this method reaches 100% when there is no interference of other factors, and it has high identification accuracy in multiple interference environments. And through the experiment under the condition of very low SNR, it is further verified that the system has strong anti-noise interference ability, which fully proves the effectiveness of this method.
TL;DR: Yang et al. as mentioned in this paper considered a free-boundary fluid-structure interaction problem for plaque growth in a cylindrical domain with ninety-degree contact angles and obtained the existence and uniqueness of strong solutions to the model problems for the linear systems.
TL;DR: In this article , a waveform design problem based on minimizing the asymptotic mean square error of MUSIC was formulated and an alternating direction method of multipliers was developed to tackle the encountered nonconvex optimization problem.
TL;DR: In this article , a radar forward-looking imaging method with improved imaging quality and reduced system complexity is proposed, where the matrix completion method is utilized to recover the missing signals of sparse array and a fast 2-dimensional atomic norm minimization (ANM) is adopted to solve the positions and amplitudes of scatterers.
Abstract: The sparse array imaging radar could largely reduce the system complexity, but it also suffers from poor azimuthal resolution due to the limited aperture length and incomplete observation caused by non-uniform array configuration. In this letter, a radar forward-looking imaging method with improved imaging quality and reduced system complexity is proposed. Firstly, the matrix completion method is utilized to recover the missing signals of sparse array. Then, based on the plane-wave assumption, a fast 2-dimensional atomic norm minimization (ANM) is adopted to solve the positions and amplitudes of scatterers. As a continues compressed sensing method, the result of ANM-based method earns sharper main-lobes and much lower-level side-lobes compared to the fast Fourier transform -based method. Although the resolution of proposed method is still limited by Rayleigh resolution, the imaging quality can be greatly improved by emphasizing main-lobes and avoiding the perturbation of side-lobes. Simulations and experiments are carried out to demonstrate the performance of the proposed method.
TL;DR: In this paper , a 3D oblate spheroidal reconstruction algorithm was applied to 2D video disdrometer (2DVD) observations during the passages of 14 tropical cyclones.
Abstract: Raindrops are usually parameterized by oblate spheroids with axis ratios (ARs) determined from projected images. However, the adequacy of representing the 3-Dimensional (3D) nature of raindrops with their 2-Dimensional (2D) shadows remains an open question. This study reexamines raindrop AR parameterizations by applying a 3D oblate spheroidal reconstruction algorithm to 2D video disdrometer (2DVD) observations during the passages of 14 tropical cyclones. The results reveal an appreciable AR overestimation for large raindrops when retrieved from a single camera and especially for 2DVD raw products. This overestimation propagates into the simulated differential reflectivity (ZDR), resulting in an underestimation of ZDR by up to 0.5 dB at X-band when large raindrops dominate the raindrop size distribution (RSD), which would introduce an underestimation of more than 20% for heavy rain rate (>60 mm hr−1) estimation by polarimetric radar. In precipitating systems with drastically changing RSDs, the importance of adopting more realistic ARs is warranted.
TL;DR: In this paper , a microsecond pulse test platform was established to investigate the lifetime characteristics of mica paper capacitors for Marx generators, which can output voltage up to 60 kV and operate in a repetitive rate of 20 Hz for 50 s or 10 Hz for 100 s. Voltage jitter is lower than 1%.
Abstract: In recent years, the development of mica capacitor technology has greatly improved the withstand voltage and energy storage density of capacitors, which is suitable for Marx generators. Before using mica paper capacitors to assemble Marx generators, it is important to study the electrical performance and life characteristics of the capacitors. In this paper, a repetitive-rate microsecond pulse test platform was established to research the lifetime characteristic of mica paper capacitors. The test platform is mainly divided into two parts. The first part includes an air-core pulse transformer, a primary electrolytic capacitor, a thyristor and a diode. Another part contains a mica paper capacitor, a SF6/N2 gas switch and a water dummy load. This platform can output voltage up to 60 kV and operate in a repetitive rate of 20 Hz for 50 s or 10 Hz for 100 s. Voltage jitter is lower than 1%. When working continuously, it can charge mica capacitor 2000 pulses for a time. With an interval of about 30 s each time, the test platform can work continuously for one hour.
TL;DR: In this article , a model is built to evaluate the heat induced by fusion splices quantitatively and predict the temperature rise in high-power fiber laser systems, and a real-time in-situ temperature measurement experiment is carried out, and the actual STF value is obtained by fitting the model parameters with the experimental results.