High performance“non-local”generic face reconstruction model using the lightweight Speckle-Transformer(SpT)UNet

Significant progress has been made in computational imaging(CI),in which deep convolutional neural networks(CNNs)have demonstrated that sparse speckle patterns can be reconstructed.However,due to the limited“local”kernel size of the convolutional operator,for the spatially dense patterns,such as the generic face images,the performance of CNNs is limited.Here,we propose a“non-local”model,termed the Speckle-Transformer(SpT)UNet,for speckle feature extraction of generic face images.It is worth noting that the lightweight SpT UNet reveals a high efficiency and strong comparative performance with Pearson Correlation Coefficient(PCC),and structural similarity measure(SSIM)exceeding 0.989,and 0.950,respectively.

Non-local damage and fracture model of rock and concrete-like materials

In view of the non-local phenomena appearing in the rock and concrete-like materials, the non-local damage and fracture model of rock and concrete-like materials was established through non-local method of Gaussian weighting function. The result indicates that, the stress of one point in the material is correlated not only to its strain history, but also to the interaction of the points in its certain adjacent region of the material. Based on the established non-local model, the numerical simulation of notch containing three-point bending beam was carried out. The results show that the grid sensitivities have been avoided and the fracture direction of the material has not been influenced by the grid shape, and the model proposed can be used to better simulate the damage developing process of the rock and concrete-like materials.

NON-LOCAL CONTINUUM MODEL AND ITS NUMERICAL SIMULATION FOR LOCALIZATION PROBLEM

A non-local continuum model for strain-softening simply takingplastic strain or damage vari- able as a non-local variable isderived by using the additive decomposition principle of finitedeformation gra- dient. At the same time, variational equations,their finite element formulations and numerical convolutedintegration algorithm of the model in current configuration usuallycalled co-moving coordinate system are given. stability andconvergence of the model are proven by means of the weak convergencetheorem of gen- eral function and the convoluted integration theory.

A 2-D Non-local Closure Model for Atmospheric Boundary Layer Simulations

In this paper a new approach for PBL simulation, the non-local closure scheme based on the transient turbulence theory has been used. It was set up as an alternative to local closure schemes which physical concept is reasonable and distinct. A 2-D non-local closure model was developed in order to study the PBL structure and simulatesome interesting atmospheric processes over non-ulliform underlying surface, especially under the convective and unique weather conditions, such as sea-land circulation and the TIBL structure. The modelled results show good agreement with field measurement.

Energy Conditions and Constraints on the Generalized Non-Local Gravity Model

We study and derive the energy conditions in generalized non-local gravity, which is the modified theory of general relativity obtained by adding a term m2n-2R□-nRto the Einstein-Hilbert action. Moreover, to obtain some insight on the meaning of the energy conditions, we illustrate the evolutions of four energy conditions with the model parameter ε for different n. By analysis we give the constraints on the model parameters ε.

GURTIN-TYPE VARIATIONAL PRINCIPLES FOR DYNAMICS OF A NON-LOCAL THERMAL EQUILIBRIUM SATURATED POROUS MEDIUM

Based on the porous media theory and by taking into account the efects of the pore fuid viscidity, energy exchanges due to the additional thermal conduction and convection between solid and fuid phases, a mathematical model for the dynamic-thermo-hydro-mechanical coupling of a non-local thermal equilibrium fuid-saturated porous medium, in which the two constituents are assumed to be incompressible and immiscible, is established under the assumption of small de- formation of the solid phase, small velocity of the fuid phase and small temperature changes of the two constituents. The mathematical model of a local thermal equilibrium fuid-saturated porous medium can be obtained directly from the above one. Several Gurtin-type variational principles, especially Hu-Washizu type variational principles, for the initial boundary value problems of dy- namic and quasi-static responses are presented. It should be pointed out that these variational principles can be degenerated easily into the case of isothermal incompressible fuid-saturated elastic porous media, which have been discussed previously.

Surface effect and non-local elasticity in wave propagation of functionally graded piezoelectric nano-rod excited to applied voltage

A non-local solution for a functionally graded piezoelectric nano-rod is pre- sented by accounting the surface effect. This solution is used to evaluate the charac- teristics of the wave propagation in the rod structure. The model is loaded under a two-dimensional （2D） electric potential and an initially applied voltage at the top of the rod. The mechanical and electrical properties are assumed to be variable along the thick- ness direction of the rod according to the power law. The Hamilton principle is used to derive the governing differential equations of the electromechanical system. The effects of some important parameters such as the applied voltage and gradation of the material properties on the wave characteristics of the rod are studied.

Random noise suppression for seismic data using a non-local Bayes algorithm

For random noise suppression of seismic data, we present a non-local Bayes （NL- Bayes） filtering algorithm. The NL-Bayes algorithm uses the Gaussian model instead of the weighted average of all similar patches in the NL-means algorithm to reduce the fuzzy of structural details, thereby improving the denoising performance. In the denoising process of seismic data, the size and the number of patches in the Gaussian model are adaptively calculated according to the standard deviation of noise. The NL-Bayes algorithm requires two iterations to complete seismic data denoising, but the second iteration makes use of denoised seismic data from the first iteration to calculate the better mean and covariance of the patch Gaussian model for improving the similarity of patches and achieving the purpose of denoising. Tests with synthetic and real data sets demonstrate that the NL-Bayes algorithm can effectively improve the SNR and preserve the fidelity of seismic data.

Strain localization of Mohr-Coulomb soils with non-associated plasticity based on micropolar continuum theory

To address the problems of strain localization, the exact Mohr-Coulomb (MC) model is used based on second-order cone programming (mpcFEM-SOCP) in the framework of micropolar continuum finite element method. Using the uniaxial compression test, we focused on the earth pressure problem of rigid wall segment involving non-associated plasticity. The numerical results reveal that when mpcFEM-SOCP is applied, the problems of mesh dependency can be effectively addressed. For geotechnical strain localization analysis involving non-associated MC plasticity, mpcFEM-SOCP in conjunction with the pseudo-time discrete scheme can improve the numerical stability and avoid the unreasonable softening issue in the pressure-displacement curves, which may be encountered in the conventional FEM. It also shows that the pressure-displacement responses calculated by mpcFEM-SOCP with the pseudo-time discrete scheme are higher than those calculated by mpcFEM-SOCP with the Davis scheme. The inclination angle of shear band predicted by mpcFEM-SOCP with the pseudo-time discrete scheme agrees well with the theoretical solution of non-associated MC plasticity.

基于非局部Mohr-Coulomb模型的土体渐进破坏分析

软化塑性模型的常规有限元数值分析结果严重依赖网格尺寸,而非局部塑性模型是解决这一问题的有效方法。但现有非局部模型仅能用于von Mises准则,无法用来进行一般软化土体渐进破坏分析。提出了一种改进的针对非局部模型的全隐式应力回代迭代计算方法,该方法具有在迭代计算过程中逐步确定弹塑性点的特点,克服了现有算法误差较大及不稳定的缺点。将非局部理论推广到Mohr-Coulomb塑性模型中,使其能用来分析土体稳定性问题。采用局部和非局部模型对两个土体稳定问题,包括条形基础承载力问题和三角形荷载下边坡稳定问题进行渐进破坏分析,数值计算结果表明该方法可以消除软化塑性有限元计算的网格敏感性,起到了正则化的效果。

Nonlinear shear behavior of rock joints using a linearized implementation of the Barton-Bandis model

Experiments on rock joint behaviors have shown that joint surface roughness is mobilized under shearing,inducing dilation and resulting in nonlinear joint shear strength and shear stress vs.shear displacement behaviors.The Barton-Bandis(B-B) joint model provides the most realistic prediction for the nonlinear shear behavior of rock joints.The B-B model accounts for asperity roughness and strength through the joint roughness coefficient(JRC) and joint wall compressive strength(JCS) parameters.Nevertheless,many computer codes for rock engineering analysis still use the constant shear strength parameters from the linear Mohr-Coulomb(M-C) model,which is only appropriate for smooth and non-dilatant joints.This limitation prevents fractured rock models from capturing the nonlinearity of joint shear behavior.To bridge the B-B and the M C models,this paper aims to provide a linearized implementation of the B-B model using a tangential technique to obtain the equivalent M-C parameters that can satisfy the nonlinear shear behavior of rock joints.These equivalent parameters,namely the equivalent peak cohesion,friction angle,and dilation angle,are then converted into their mobilized forms to account for the mobilization and degradation of JRC under shearing.The conversion is done by expressing JRC in the equivalent peak parameters as functions of joint shear displacement using proposed hyperbolic and logarithmic functions at the pre-and post-peak regions of shear displacement,respectively.Likewise,the pre-and post-peak joint shear stiffnesses are derived so that a complete shear stress-shear displacement relationship can be established.Verifications of the linearized implementation of the B-B model show that the shear stress-shear displacement curves,the dilation behavior,and the shear strength envelopes of rock joints are consistent with available experimental and numerical results.

Constitutive models in stability analysis of rock slope

Equivalent Mohr-Coulomb yield criterion was established,and the relationship between different constitutive models was studied.The application of equivalent Mohr-Coulomb yield criterion in Ansys was achieved by means of transforming material parameters.The stability research aiming at the most common rock slope without conspicuous slide surface was accomplished,the methods of measurably assessing the stability of rock slope without conspicuous slide surface were explored,and the disadvantages of method of minimum slide-resisted reserve as dangerous slide path were pointed out.The results show that through the calculation and analysis of cases,the conception that measurable assessment of the stability of rock slope without conspicuous slide surface can be achieved under condition that equivalent Mohr-Coulomb yield criterion is validated.Its safety parameter formula is explicit in theory and credible in results.The results obtained are approximate to those obtained by using finite element intensity reducing method.

Establishment of tensile failure induced sanding onset prediction model for cased-perforated gas wells

Sand production is a challenging issue in upstream oil and gas industry,causing operational and safety problems.Therefore,before drilling the wells,it is essential to predict and evaluate sanding onset of the wells.In this paper,new poroelastoplastic stress solutions around the perforation tunnel and tip based on the Mohr-Coulomb criterion are presented firstly.Based on the stress models,a tensile failure induced sanding onset prediction model for cased-perforated gas wells is derived.Then the analytical model is applied to field data to verify its applicability.The results from the perforation tip tensile failure induced sanding model are very close to field data.Therefore,this model is recommended for forecasting the critical conditions of sand production analysis.Such predictions are necessary for providing technical support for sand control decision-making and predicting the production condition at which sanding onset occurs.

Existence of Travelling Waves with Strong Generic Kernel in a Vector-Disease Model

In this paper, traveling wave solutions for a vector-disease model incorporating time delay and diffusion have been studied. The existence of traveling wave solutions for the sufficiently small delays has been proved. In order to solve these problems, we are able to deal with travelling wave solutions using dynamical systems techniques, invariant manifold theory, together with linear chain techniques and the geometric singular perturbation theory. For the strong generic delay kernel, traveling wave solutions exist while the delay is sufficiently small, using the methods above.

Threshold Dynamics of an SIR Epidemic Model with Nonlinear Incidence Rate and Non-Local Delay Effect

In this paper, we are concerned with a reaction-diffusion SIR epidemic model with nonlinear incidence rate and non-local delay effect in a continuous bounded spatial domain. We introduce the basic reproduction number R_0 of the model by the idea of next generation operator. By means of the theory of dynamical systems and uniform persistence, we investigate the global dynamics of the model in terms of R_0. Finally, we implement numerical simulations to show the feasibility of our results and explore some epidemiological insights.

The Global Attractor of a Non-Local PDE Model with Delay for Population Dynamics in Rn

In this paper, we consider a non-local PDE model with delay for population dynamics in Rn. First, we prove the existence and uniqueness of weak solutions under some suitable decayed assumptions on non-local term at infinity. Then, we obtain the global attractor by proving w-limit compactness property of the solution operator semigroup.

On the non-local priors for sparsity selection in high-dimensional Gaussian DAG models

We consider sparsity selection for the Cholesky factor L of the inverse covariance matrix in high-dimensional Gaussian DAG models.The sparsity is induced over the space of L via non-local priors,namely the product moment(pMOM)prior[Johnson,V.,&Rossell,D.(2012).Bayesian model selection in high-dimensional settings.Journal of the American Statistical Asso-ciation,107(498),649-660.https://doi.org/10.1080/01621459.2012.682536]and the hierarchi-cal hyper-pMOM prior[Cao,X.,Khare,K.,&Ghosh,M.(2020).High-dimensional posterior consistency for hierarchical non-local priors in regression.Bayesian Analysis,15(1),241-262.https://doi.org/10.1214/19-BA1154].We establish model selection consistency for Cholesky fac-tor under more relaxed conditions compared to those in the literature and implement an efficient MCMC algorithm for parallel selecting the sparsity pattern for each column of L.We demonstrate the validity of our theoretical results via numerical simulations,and also use further simulations to demonstrate that our sparsity selection approach is competitive with existing methods.

Application of Disaggregated Logit Modeling with Stated Preference Method in Park and Ride Behavior for Non-local Private Car Travelers in Big Events

During big events, non-local private car travelers can be divided into two types which were returning in one day and in several days. It was demonstrated that those two kinds of travelers have distinct behavior on park and ride （P＆R）, due to their different travel demand and behavior attributes. In this paper focusing on the travelers returning in several days, the travel attributes and requirements for P＆R were analyzed with stated preference survey. A P＆R choice behavior disaggregated logit model was established and calibrated based on random utility theory. The model concludes three variables, which were travel time, parking fee and comprehensive attractiveness index for suburban satellite towns comparing to urban district. The results revealed that for travelers returning in several days the primary key point is increasing the attractiveness of suburban satellite towns.

红外交通场景下遮挡行人目标检测算法研究

针对交通十字路口等视野盲区往来行人间存在遮挡情况,如何高效准确地检测复杂道路中目标行人具有实际意义。为了实现夜间交汇路口场景行人检测,提出一种基于改进YOLOv5的行人目标检测算法,采用Non-local和PSA模块对YOLOv5原网络的Bottleneck CSP进行改进,能够有效弥补遮挡中行人特征的帧间信息交互过程,增强长程范围通道特征依赖关系。设计更深的160×160检测层和自适应anthor,提升夜间行人检测的边界回归精确度。实验结果表明,针对夜间下交通路口场景,压缩改进后模型对行人检测鲁棒性高,相较于原始算法mAP_0.5和mAP_0.5:0.95值分别提升了14.2%和12.7%,说明所提算法对夜间行人检测的有效性。

LARGE EDDY SIMULATION OF COMPLEX TURBULENT FLOWS:PHYSICAL ASPECTS AND RESEARCH TRENDS

In the current paper, we have primarily addressed one powerful simulation tool developed during the last decades-Large Eddy Simulation (LES), which is most suitable for unsteady three-dimensional complex turbulent flows in industry and natural environment. The main point in LES is that the large-scale motion is resolved while the small-scale motion is modeled or, in geophysical terminology, parameterized. With a view to devising a subgrid-scale(SGS) model of high quality, we have highlighted analyzing physical aspects in scale interaction and-energy transfer such as dissipation, backscatter, local and non-local interaction, anisotropy and resolution requirement. They are the factors responsible for where the advantages and disadvantages in existing SGS models come from. A case study on LES of turbulence in vegetative canopy is presented to illustrate that LES model is more based on physical arguments. Then, varieties of challenging complex turbulent flows in both industry and geophysical fields in the near future-are presented. In conclusion; we may say with confidence that new century shall see the flourish in the research of turbulence with the aid of LES combined with other approaches.