Journal of University of Chinese Academy of Sciences

Well-posedness of 3D incompressible generalized Navier-Stokes system in Fourier-Triebel-Lizorkin spaces

MIN Dezai, WU Gang, YAO Zhuoya

2023, 40 (2): 145-154. DOI: 10.7523/j.ucas.2021.0065

In this paper, we consider the Cauchy problem for the 3D incompressible generalized Navier-Stokes system and study the well-posedness in critical Fourier-Triebel-Lizorkin spaces $\widehat {\dot F}_{p,q}^{4 - \alpha - \frac{3}{p}}$($\mathbb{R}^3$). Making use of Fourier localization method and Banach fixed point theorem, we proved that if $p > \frac{3}{{5 - \alpha }}$, q ≥ 1, the system is locally well-posed for large initial data as well as global well-posed for small initial data. Also we established same result for $p = \frac{3}{{5 - \alpha }}$，q∈[$\frac{3}{{5 - \alpha }}$，$\frac{6}{{5 - \alpha }}$].

Experimental investigation on the typical flow structures of liquid metal thermal convection

CHENG Youji, CHEN Xinyuan, YANG Juancheng, NI Mingjiu

2023, 40 (2): 155-164. DOI: 10.7523/j.ucas.2021.0033

Abstract ( 346 ) PDF (0KB) ( 0 )

In the present paper, we built a rectangle cell with an aspect ratio of 1:1:2, where the height of the cell is two times the width of the cell. The thermal convection of liquid Galinstan with a low Prandtl number, 0.03, is systematically investigated with/without the influence of the magnetic field. During the experiments, two thermostatic water baths are adopted to keep the upper copper plate and lower copper plate with constant temperature difference where the temperature of the upper copper plate is always lower than that of the lower copper plate. Thermoprobes are inserted in copper plates to record the temperature signal in the copper and liquid metal, while the ultrasonic sensors are inserted through the sidewall of the cell to measure the velocity distribution of liquid metal. From the velocity distribution in the cell, we observe three typical flow structures. Without the influence of the magnetic field, time-dependent large-scale convection is observed in the liquid metal thermal convection, with large velocity fluctuations near the cold and hot copper plates. The characteristics of velocity are similar to the thermal turbulence obtained from the normal liquid which is called as flow structure I. Under the influence of the horizontal magnetic field, the thermal convection is strongly suppressed by the magnetic field, the stable single roll convection, namely the flow structure II, and the double oscillation roll convection, namely the flow structure III are observed. Furthermore, the characteristics of these two flow structures with the action of the horizontal magnetic field are investigated in detail through the velocity information obtained from three ultrasonic sensors.

Molecular simulation study of the effect of iron clusters on the viscosity of liquid lithium

XU Bo, LIU Songchang, YU Xingang

2023, 40 (2): 165-172. DOI: 10.7523/j.ucas.2021.0062

Abstract ( 470 ) PDF (0KB) ( 12 )

Liquid lithium has a wide range of applications in magnetic confinement fusion devices, such as blanket, divertor, etc. Due to the active physical and chemical properties of liquid lithium, its compatibility with structural materials has always been one of the important issues in the field of nuclear fusion. Considering that stainless steel is usually used as the structural material of the liquid lithium circuit in engineering, the compatibility between lithium and iron has attracted the attention of many scholars. In view of this, the molecular dynamics method was used to simulate the liquid lithium containing iron clusters. The effects of iron clusters on the viscosity of liquid lithium and the micromechanism were analyzed. The effects of cluster size, shape, and concentration were investigated. The results show that the presence of iron clusters can significantly change the viscosity of liquid lithium, and the viscosity gradually increases with the increase of cluster size. The cluster shape also has a certain effect on the viscosity. Under the condition of the same volume, the larger the surface area, the greater the effect. In addition, we have also developed a computational model for the viscosity of nanofluid, which can well fit the simulation results of molecular dynamics.

Virtual screening and molecular dynamics simulation studies to identify potential inhibitors of Furin

SUN Fenglei, LI Xiaoyi

2023, 40 (2): 173-178. DOI: 10.7523/j.ucas.2021.0044

Abstract ( 480 ) PDF (0KB) ( 0 )

In this study, virtual screening based on Furin receptor was carried out by means of molecular docking and molecular dynamics simulations to find potential inhibitors of Furin. The pharmacophore model based on Furin receptor was constructed by Sybyl. Drug molecules were screened from ZINC database according to the pharmacophore model. The screened drug molecules were docked with Furin protein. Finally, four qualified drug molecules were screened. The four complexes were calculated by molecular dynamics simulations for 100ns. RMSD and RMSF were calculated. The binding energies of Furin receptors with drug molecular ligands were analyzed by MMGBSA. It was found that the binding energy of Furin-ZINC7664 was the lowest one, which indicated that the drug molecule ZINC7664 is the most stable potential inhibitor to Furin and could be used to design the effective anti-Furin drugs.

Stress drop distribution of 2004 M_w 6.0 Parkfield earthquake based on the elastic dislocation theory

DOU Tiantian, CHENG Huihong, SHI Yaolin

2023, 40 (2): 179-190. DOI: 10.7523/j.ucas.2021.0041

Abstract ( 362 ) PDF (0KB) ( 0 )

The occurrence of earthquakes is accompanied by regional stress state adjustment. In seismology, stress drop is usually used to characterize the stress release level after an earthquake. As one of the important parameters, stress drop is widely used to judge the type of earthquake, analyze the stress state after earthquake, and predict the rupture propagation. At present, seismologists often give an average stress drop value based on the earthquake corner frequency. However, the rock strength and stress state of the seismic fault plane are inhomogeneities in real terms. Besides, a single average value is difficult to show the spatial variation in stress changes, which could not reflect the stress adjustment across fault plane. Meanwhile, there are great differences among different researches due to the limited observation stations or different source spectrum data or other related calculation parameters. In this paper, from the point of view of mechanics, we propose the method of adoption the Okada's dislocation theory to calculate the shear stress change of the fault plane, that is, based on the slip model to obtained the distribution of fault plane. From the results of numerical calculation, it is found that the occurrence of an earthquake releases the concentrated stress of fault plane, due to the presence of obstacles on fault or uneven slip distribution, the stress release in the local area with large dislocation will increase the stress concentration in the adjacent area, showing the phenomenon of non-uniform distribution of stress drop, and increasing the rupture tendency of local section. Moreover, the non-uniformity distribution of stress drop and the uneven fault geometry determine the non-uniform slip behavior of fault. Taking the 2004 M_w 6.0 Parkfield earthquake as an example, the stress drop distribution of the fault plane was calculated. The maximum stress drop was about 9.2MPa, which near the source. But the stress drop increased in some sections of the fault plane, reaching -3.5MPa. Compared with the average stress drop, the distribution of stress drop on fault plane calculated by the dislocation model can better reflect the source rupture process and predict the aftershock evolution.

Untangling of northern midaltitude near surface air temperature and snow cover frequency trends with multidimensional ensemble empirical mode decomposition

LIU Quan, YAO Fengmei

2023, 40 (2): 191-202. DOI: 10.7523/j.ucas.2021.0060

Abstract ( 263 ) PDF (0KB) ( 0 )

The Arctic has warmed more rapidly than the globe as a whole, while the northern midlatitudes have experienced more frequent severe winters. The anomalous climate pattern has drawn wide attentions. Here we improved the method fast multidimensional ensemble empirical mode decomposition (Fast-MEEMD) to extract the long-term trends of near surface air temperature (SAT) and snow cover frequency (SCF). The results show evident "warming arctic-cooling continent (WACC)" pattern. From the perspective of cumulative change, SAT over central Eurasia has decreased, the Siberia High has amplified and the SCF over Eastern Asian has increased since 1990s. while in term of changing rate, this accelerated cooling tendency over midlatitudes is mainly occurred between 1990s-2000s. Therefore, the cooling trends may develop as short rather than long-term trends. The study provides new materials to evaluate the roles of other factors in driving midlatitudes cooling.

Low power clock tree

ZHU Jiaqi, CHEN Lan, WANG Haiyong

2023, 40 (2): 203-207. DOI: 10.7523/j.ucas.2021.0040

Abstract ( 317 ) PDF (0KB) ( 0 )

This paper proposes a clock tree design method that meets timing as much as possible and minimizes power consumption. This method uses fanout number and driver selection strategy as the optimization variables for low-power clock tree design. For different fanout numbers, we take the driver selection strategy of selecting all inverters/buffers in the standard cell library as the reference strategy,and comparing and analyzing the three driver selection strategies with partial inverters/buffers proposed in this article. At the same time, the merit factor composed of the clock skew and power consumption of the clock tree is proposed as the criterion for evaluating various driver selection strategies. The experimental results show that with the merit factor as the evaluation criterion, the optimal fanout number in the clock tree design has little correlation with the driver selection strategy, and the three driver selection strategies proposed in this paper are better than the reference strategy. In the strategy with the best merit factor, the power consumption of the clock tree is reduced by 5.82% typically. Finally, this paper presents a low power clock tree design method based on merit factor.

Field vehicle signal classification based on FVC-CNN

LI Xiang, WANG Yan, LI Baoqing

2023, 40 (2): 208-216. DOI: 10.7523/j.ucas.2021.0038

Abstract ( 271 ) PDF (0KB) ( 0 )

Aiming at the problem that single channel vehicle acoustic signal is seriously affected by wind noise and has low classification performance, a one-dimensional convolutional neural network model FVC-CNN (convolutional neural network for field vehicle classification, FVC-CNN) based on four channel synchronous acquisition signal of acoustic array is proposed in this paper. The model uses the idea of weighted average of attention mechanism to improve the structure of Inception network. As the input layer, it extracts the features of four channel acoustic signals with different time scales to suppress noise interference. According to the distribution characteristics of different vehicle acoustic signals, three feature extraction networks, SWNet, LWNet, and TNet, are trained to extract the characteristics of the corresponding vehicle, finally, the extracted features are fused with multi branches and multi dimensions for classification. Verified on the same data set, the experimental results show that the total recognition rate of FVC-CNN model can reach 94.22%, which is 14.08% higher than the traditional method, and the classification effect is better.

Suitability analysis and evaluation of MODTRAN night radiance mode

HUANG Yanyan, ZHANG Yu, QIU Shi

2023, 40 (2): 217-226. DOI: 10.7523/j.ucas.2021.0059

Abstract ( 849 ) PDF (0KB) ( 4 )

The night radiance mode of MODTRAN (moderate resolution atmospheric transmission) can simulate the lunar irradiance transfer, but the lunar-source model of MODTRAN simplifies the calculation of the lunar phase function, lunar albedo, sun-moon distances, and earth-moon distances. This simplicity can lead to some errors in the result, which affects the calibration accuracy. So, the sensitivity of MODTRAN lunar-source model to lunar model parameters is evaluated by using perturbation analysis method in this paper. The results show that the most important parameters affecting the accuracy of lunar irradiance simulation from MODTRAN are earth-moon distances, lunar phase function, and lunar albedo. This paper used lunar irradiance model MT2009 (Miller-Turner 2009) and SeaWiFS (sea-viewing wide field-of-view sensor) satellite data as benchmarks to compare and analyze lunar-source function of MODTRAN. The results showed that:1) Lunar irradiance calculated by MODTRAN were in better agreement with the values from MT2009 and SeaWiFS when wavelength was greater than 0.675μm; 2) When the absolute value of lunar phase angle was less than 30° and earth-moon distance was 390000km, the relative differences between MODTRAN and benchmarks were less than 10%; 3) When the absolute value of lunar phase angle was greater than 30° and the moon was closed to apogee, the relative differences between lunar irradiance model of MODTRAN and benchmarks were light. While lunar phase angle was between -41° and -30° or between 30° and 41° and the moon was closed to apogee, maximum relative difference was about 10%. This paper indicates that MODTRAN lunar-source model is suitable for low-light remote sensing calibration under specific conditions. More research work on MODTRAN lunar-source model will be carried out in the future.

An unsupervised representation learning approach for modelling forest landform characteristics and fire susceptibility assessment

ZHUANG Zijun, YUAN Xiaobing, PEI Jun, WANG Guohui, LIU Jianpo

2023, 40 (2): 227-239. DOI: 10.7523/j.ucas.2021.0057

Abstract ( 300 ) PDF (0KB) ( 0 )

Destructive wildfires have caused extraordinary losses in both economic and natural property worldwide with an even increasing frequency in recent decades. One practical approach in forest fire susceptibility prediction is using statistical learning methods to learn from historic data. Conventional methods use handcraft feature to reduce data dimension. With the continuous development of remote sensing technology, the difficulty of obtaining high-precision gridded multi-dimensional forest landform information is constantly decreasing. It is difficult to make full use of such data through handcraft features, which limits the performance of conventional methods when applied in the real world. This paper introduces a novel approach to model forest geographic information through deep representation learning, which is, leveraging deep convolutional neural network and state-of-the-art representation learning methods to extract feature embedding for a given area of interest. Fire susceptibility assessment experiments are used to evaluate the proposed method and compare the unsupervised learning and its supervised counterpart to show its effectiveness.

An efficient link-building strategy for cross seam in LEO satellite network

LIU Yuting, TIAN Feng, LI Guotong

2023, 40 (2): 240-249. DOI: 10.7523/j.ucas.2021.0039

Abstract ( 775 ) PDF (0KB) ( 0 )

Recently, the low earth orbit (LEO) satellite communication system has received wide attention from academia and industry, due to its global coverage and flexible construction. In the LEO satellite communication system, satellites are connected to a dynamic network through the inter-satellite link (ISL). How to reduce the communication delay of the network is an urgent research problem to be solved. For the problem of cross seam link-building of the Walker polar orbit constellation, this paper proposes a star cross looping link-building strategy (SCLBS). SCLBS divides the orbits into multiple cross link-building areas on both sides of cross seam and the transmission delay can be optimized by dynamically established ISL, providing a stable link-building scheme for polar orbit cross seam. SCLBS is applied to actual business scenarios such as the airplane and ship, and the test results show that SCLBS can effectively reduce the transmission delay by 30% to 60% compared with the link-building strategies based on the nearest satellite criterion and the longest visible time criterion, and significantly improve system performance.

Skeleton detection based on local short-connection unidirectional fusion networks

QIAO Yang, XIAO Shixiang, LIU Yue, JIAO Jianbin

2023, 40 (2): 250-257. DOI: 10.7523/j.ucas.2021.0048

Abstract ( 285 ) PDF (0KB) ( 0 )

In recent years, skeleton detection based on side-output network has shown significant effectiveness. However, the existing methods are still unable to tackle the problem of image distortion in side-output structure caused by high-multiplier up/down-sampling, and the fixed receptive field limits the feature expression of the networks. To solve these problems, this paper proposes a local short-connection unidirectional fusion network based on side-output connection, which includes a feature extraction network and a side-output connection network. The feature extraction network is a deep convolutional neural network, which is used for multi-layer feature extraction. The side-output connection network consists of a local short-connection unit and a unidirectional fusion network. The local short-connection gradually constructs the continuous large receptive field features by integrating the adjacent features of receptive field, while the unidirectional fusion of multi-scale features from deep to shallow can achieve the characterization of the object from rough to fine. Experimental results on four commonly used skeleton detection datasets demonstrate the effectiveness of the proposed method.

3D point cloud registration via matching multi types of geometric primitives

ZHANG Long, XIAO Jun, CHENG Xiaolong, WANG Ying

2023, 40 (2): 258-267. DOI: 10.7523/j.ucas.2021.0047

Abstract ( 645 ) PDF (0KB) ( 0 )

3D point cloud registration is the fundamental of a large number of applications in computer vision, computer graphics, and remote sensing, etc. However, the existence of noises and a low overlapping ratio in the scanning data poses a great challenge to the existing registration methods. Facing the point clouds of man-made objects or urban scenes that are likely to have the aforementioned issues, we propose a registration method of 3D point clouds via matching multi types of geometric primitives. Our method first extracts common geometric primitives from raw point clouds and further builds the feature descriptors from their effective combinations. Then, under the matching of the descriptors, our method realizes the matching of primitives and acquires the transformation parameters from them. Finally, based on the global evaluation for every candidate transformation, the best transformation is identified and applied to achieve the registration. Our method fully inherits the advantages of multiple types of primitives and has stronger robustness and efficiency. Experiments on various benchmarks demonstrate that our method achieves state-of-the-art registration performance.

Recognition-oriented facial 3D information estimation

CHEN Hanqin, QIN Jin, ZHAO Tong, YAN Yao

2023, 40 (2): 268-279. DOI: 10.7523/j.ucas.2021.0058

Abstract ( 420 ) PDF (0KB) ( 0 )

3D face recognition has the advantages of recognition accuracy and anti-counterfeiting strength over the popular 2D face recognition, and represents the development direction of face recognition. Due to the high cost of 3D facial acquisition, 3D facial recognition can not establish and optimize face recognition algorithms directly rely on massive face data as 2D facial recognition methods. How to obtain augmented training data for 3D faces accurately and efficiently is the most pressing problem in driving the development of 3D face recognition applications. A large amount of related research focuses on how to get better 3D face reconstruction visualization,but does not give much consideration to the subsequent recognition task,so that the recognition accuracy of 3D face recognition algorithms trained with these reconstructed images is much lower than expected. To address this problem, a recognition-oriented method for estimating facial 3D information is proposed. Different from the general method, this method directly builds an interactive bridge between information estimation and subsequent recognition:during the training process of 3D face information estimation, it directly bases on the corresponding recognition network to supervise and improve the estimation effect of 3D face information. For this purpose, we first construct a 3D face information representation, the depth-surface normal vector map (DN map), and then train a facial CycleGAN model with a real 3D dataset to learn a mapping from 2D face to DN map with preserved identity information and represent it in the form of U-Net network. Experiments are conducted on five datasets to compare with other methods, and the improvement is particularly significant in the ND-2006 dataset, with an improvement of 31.8%. In addition, experiments on performance improvement under data augmentation are conducted, here the performance improvement of the augmentation method based on the facial CycleGAN is more obvious under the same conditions of data augmentation, with a maximum improvement of 14.9% on the CASIA 3D dataset.

How does IT-business alignment influence digital innovation: the moderating role of digital business intensity

XU Tian, WEI Shaobo, YIN Jinmei

2023, 40 (2): 280-288. DOI: 10.7523/j.ucas.2021.0064

Abstract ( 207 ) PDF (0KB) ( 0 )

With the rapid development of digital business, digital innovation has become a critical source for firms to gain competitive advantage. Whilst previous studies have examined the impact of IT-business alignment on firm performance, few studies have explored its impact on digital innovation. Drawing on resource-based view, our study investigates how different types of IT-business alignment (i.e. intellectual and social alignment) affect digital innovation independently and interactively and explores the moderating role of digital business intensity (DBI) further. Using 143 match-paired questionnaires collected from Chinese firms, our results show that intellectual and social alignments are positively associated with digital innovation, and their interactions generate synergistic effects to promote digital innovation. Furthermore, DBI positively and negatively moderates the relationship between intellectual alignment and digital innovation and between social alignment and digital innovation, respectively. Our study also discusses the theoretical and practical implications of the research.

Current Issue