检索结果-南通市图书馆

Multilayer perceptron neural network activated by adaptive Gaussian radial basis function and its application to predict lid-driven cavity flow

在线全文

学校读者我要写书评

暂无评论

Acta Mechanica Sinica 2021年第12期37卷 1757-1772页

作者： Qinghua Jiang Lailai Zhu chang shu Vinothkumar Sekar Department of Mechanical Engineering National University of Singapore10 Kent Ridge CrescentSingapore 117576Singapore

To improve the performance of multilayer perceptron(MLP)neural networks activated by conventional activation functions,this paper presents a new MLP activated by univariate Gaussian radial basis functions(RBFs)with adaptive centers and widths,which is composed of more than one hidden *** the hidden layer of the RBF-activated MLP network(MLPRBF),the outputs of the preceding layer are first linearly transformed and then fed into the univariate Gaussian RBF,which exploits the highly nonlinear property of *** RBFs might address the issues of saturated outputs,low sensitivity,and vanishing gradients in MLPs activated by other prevailing nonlinear ***,we apply four MLP networks with the rectified linear unit(ReLU),sigmoid function(sigmoid),hyperbolic tangent function(tanh),and Gaussian RBF as the activation functions to approximate the one-dimensional(1D)sinusoidal function,the analytical solution of viscous Burgers’equation,and the two-dimensional(2D)steady lid-driven cavity *** the same network structure,MLP-RBF generally predicts more accurately and converges faster than the other ***-RBF using less hidden layers and/or neurons per layer can yield comparable or even higher approximation accuracy than other MLPs equipped with more layers or neurons.

关键词： Multilayer perceptron neural network Activation function Radial basis function Numerical approximation

Machine learning of synaptic structure with neurons to promote tumor growth

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Applied Mathematics and Mechanics(English Edition) 2020年第11期41卷 1697-1706页

作者： Erhui WANG Xuelan ZHANG Liancun ZHENG chang shu School of Energy and Environmental Engineering University of Science and Technology BeijingBeijing 100083China School of Mathematics and Physics University of Science and Technology BeijingBeijing 100083China State Key Laboratory of Cardiovascular Disease Center of Vascular SurgeryFuwai HospitalNational Center for Cardiovascular DiseasesChinese Academy of Medical Sciences and Peking Union Medical CollegeBeijing 100037China

In this paper,we use machine learning techniques to form a cancer cell model that displays the growth and promotion of synaptic and electrical ***,such a technique can be applied directly to the spiking neural network of cancer cell *** results show that machine learning techniques for the spiked network of cancer cell synapses have the powerful function of neuron models and potential supervisors for different *** changes in the neural activity of tumor microenvironment caused by synaptic and electrical signals are *** can be used to cancer cells and tumor training processes of neural networks to reproduce complex spatiotemporal dynamics and to mechanize the association of excitatory synaptic structures which are between tumors and neurons in the brain with complex human health behaviors.

关键词： machine learning technique computational hemodynamics electrodiffusive activity complex synaptic dynamics

Design and optimization of reactive distillation: a review

在线全文

学校读者我要写书评

暂无评论

Frontiers of Chemical Science and Engineering 2022年第6期16卷 799-818页

作者： chang shu Xingang Li Hong Li Xin Gao School of Chemical Engineering and Technology National Engineering Research Center of Distillation TechnologyCollaborative Innovation Center of Chemical Science and Engineering(Tianjin)Tianjin UniversityTianjin 300072China

Reactive distillation process,a representative process intensification technology,has been widely applied in the chemical ***,due to the strong interaction between reaction and separation,the extension of reactive distillation technology is restricted by the difficulties in process analysis and *** overcome this problem,the design and optimization of reactive distillation have been widely studied and illustrated for plenty of reactive mixtures over the past three *** design and optimization methods of the reactive distillation process are classified into three categories:graphical,optimization-based,and evolutionary/heuristic *** primary objective of this article is to provide an up-to-date review of the existing design and optimization *** and output information,advantages and limitations of each method are stated,the modification and development for original methodologies are also *** on future research on the design and optimization of reactive distillation method are proposed for further research.

关键词： reactive distillation process intensification design method reactive phase diagram optimization algorithm

Fluid Dynamics of Flapping Insect Wing in Ground Effect

在线全文

学校读者我要写书评

暂无评论

Journal of Bionic Engineering 2014年第1期11卷 52-60页

作者： Jie Wu chang shu Ning Zhao Weiwei Yan Department of Aerodynamics Nanjing University of Aeronautics and Astronautics Nanjing 210016 P. R. China Department of Mechanical Engineering National University of Singapore 10 Kent Ridge Crescent 119260 Singapore College of Metrology and Measurement Engineering China Jiliang University Hangzhou 310018 P. R. China

The fluid dynamics of flapping insect wing in ground effect is investigated numerically in this study. To model the insect wing cross-section in forward-flight mode, the laminar flow over a NACA0012 airfoil animated by a combination of harmonic plunge and pitch rotation is considered. To implement the simulation, the proposed immersed boundary-lattice Boltzmann method is employed. By fixing the Reynolds number and the amplitude of motion, we systematically examine the influences of the distance between the foil and the ground and the flapping frequency on the flow behaviors. As compared to the situation out of ground effect, the forces for foil placed in close proximity to the ground show some differences. The mean drag coefficient is increased at low frequency and decreased at high frequency. Meanwhile, the mean lift coefficient is increased at both low and high frequencies and decreased at middle frequency. Moreover, an interesting phenomenon with oblate vortices due to vortex interaction with the ground is observed.

关键词： flapping insect wing ground effect drag reduction lift enhancement

Magnetic nanoparticle drug targeting to patient-specific atherosclerosis: effects of magnetic field intensity and configuration

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Applied Mathematics and Mechanics(English Edition) 2020年第2期41卷 349-360页

作者： Xuelan ZHANG Mingyao LUO Peilai TAN Liancun ZHENG chang shu School of Energy and Environmental Engineering University of Science and Technology BeijingBeijing 100083China School of Mathematics and Physics University of Science and Technology BeijingBeijing 100083China Department of Vascular Surgery Fuwai HospitalNational Center for Cardiovascular DiseasesChinese Academy of Medical Sciences and Peking Union Medical CollegeBeijing 100037China

Nanoparticle-mediated drug delivery is recognized as a promising option for targeted treatment of atherosclerosis. In this paper, the Eulerian-Lagrangian technique is adopted to simulate the delivery of drug-loaded nanoparticles to patient-specific atherosclerotic plaque with the aid of an external magnetic field. Plaques and vascular walls are introduced as porous media formulated by the Darcy-Forchheimer model in this targeted transport process. The results demonstrate that the delivery efficiency of particles to atherosclerosis depends on the external magnetic field, such as configuration and intensity, in which the configuration angle of the current wire is a key factor and the double current wires have advantages over the single current wire. Meanwhile, the delivery efficiency gradually decreases as the distance between the plaque cap and the current wire increases. Further, although augmenting the current or magnetic susceptibility can generally improve the delivery efficiency of nanoparticles, this increase is not apparent when small-sized nanoparticles are employed as drug transport particles. The results obtained can potentially serve as the guideline to optimize regimens for the targeted therapy of atherosclerosis.

关键词： atherosclerosis nanoparticle-mediated drug delivery magnetic field targeted delivery efficacy computational fluid dynamics

On the CR Poincaré–Lelong Equation, Yamabe Steady Solitons and Structures of Complete Noncompact Sasakian Manifolds

在线全文

学校读者我要写书评

暂无评论

Acta Mathematica Sinica,English Series 2018年第8期34卷 1313-1344页

作者： Der Chen chang shu Cheng chang Ying Bo HAN Chien LIN Department of Mathematics and Statistics Georgetown University Washington D. C. 20057 USA Graduate Institute of Applied Science and Engineering Fu Jen Catholic University Taipei 242 Taiwan China Yau Mathematical Sciences Center Tsinghua University Beijing 100084 P. R. China School of Mathematics and Statistics Xinyang Normal University Xinyang 464000 P. R. China Department of Mathematics National Tsing Hua University Hsinchu 30013 Taiwan China

In this paper, we solve the so-called CR Poincare-Lelong equation by solving the CR Poisson equation on a complete noncompact CR （2n ＋ 1）-manifold with nonegative pseudohermitian bisectional curvature tensors and vanishing torsion which is an odd dimensional counterpart of Kahler geometry. With applications of this solution plus the CR Liouvelle property, we study the structures of complete noncompact Sasakian manifolds and CR Yamabe steady solitons.

关键词： CR Poisson equation CR Poincare-Lelong equation CR Yamabe solitons

Electroaeroelastic Modeling and Analysis for Flow Energy Piezoelectric Harvester

在线全文

学校读者我要写书评

暂无评论

Transactions of Nanjing University of Aeronautics and Astronautics 2017年第1期34卷 9-14页

作者： Zhang Jiantao Wu Song shu chang Li Chaodong School of Mechatronic Engineering and Automation Shanghai UniversityShanghai 200072P.R.China

An electroaeroelastic model for wind energy harvesting using piezoelectric generators is *** flow field is mapped in *** force which the fluid flow exerts on the generator is *** output voltage levels generated from the mechanical strain within the piezoelectric elements are *** analytical model is developed with consideration of the interactions between the fluid,solid and *** analytical results are obtained,such as flow velocity contour and pressure contour for the flow,moving trajectories,stress contour and output voltage of the harvester.A prototype is fabricated and *** simulation result is close to the experimental *** model developed in this paper can predict the performance and behavior of different energy *** it also can be used as a design tool for optimizing the performance of the harvester.

关键词： wind energy harvester piezoelectric aeroelasticity electromechanics

SENSITIVE NUMERICAL SIMULATION AND ANALYSIS OF RAINSTORM USING NESTED WRF MODEL

在线全文

学校读者我要写书评

暂无评论

Journal of Hydrodynamics 2006年第5期18卷 578-586页

作者： WANG shu-chang HUANG Si-xun LI Yi Institute of Meteorology PLA University of Science and Technology Nanjing 211101 China

Nowadays, the nested grid model is widely used to increase the resolution of numerical model. An attempt to investigate the influence of the newly developed nesting technique of WRF model on the mesoscale numerical simulation of severe storm is described in this article. The results of sensitivity tests infer that the WRF model give more accurate and vivid prediction of severe storm with nesting scheme. As is shown in the tests, simulating value of the precipitation centers using nesting scheme are larger than those without using nesting scheme, which indicates that the simulating precipitation of nested storm is closer to the real rainfall. And there is obvious improvement of distribution of rain area and position of precipitation centers. According to the high resolution output data of the WRF model, deep and detailed analysis of characters of the mesoscale system that triggers the rainstorm is reported in this article, and the thermal and dynamical characters of this Meiyu front severe storm are also discussed.

关键词： WRF nesting numerical simulation mesoscale analysis

Optical biometry intraocular lens power calculation using different formulas in patients with different axial lengths

在线全文

学校读者我要写书评

暂无评论

International Journal of Ophthalmology(English edition) 2013年第2期6卷 150-154页

作者： Jia-Kang Wang shu-Wen chang Department of Ophthalmology Far Eastern Memorial Hospital Oriental Institute of Technology Department of Medicine National Yang Ming University

AIM: To investigate the predictability of intraocular lens (IOL) power calculation using the IOLMaster and different IOL power calculation formulas in eyes with various axial length (AL). METHODS: Patients were included who underwent uneventful phacoemulsification with IOL implantation in the Department of Ophthalmology, Far Eastern Memorial Hospital, Taipei, Taiwan, China from February 2007 to January 2009. Preoperative AL and keratometric values (Ks) were measured by IOLMaster optical biometry. Patients were divided into 3 groups based on AL less than 22mm (Group 1), 22-26mm (Group 2), and more than 26mm (Group 3). The power of the implanted IOL was used to calculate the predicted postoperative spherical equivalence (SE) by various formulas: the Haigis, Hoffer Q, Holladay 1, and SRK/T. The predictive accuracy of each formula was analyzed by comparing the difference between the actual and predicted postoperative SE (MedAE, median absolute error). All the patients had follow-up periods exceeding 3 months. RESULTS: Totally, there were 200 eyes (33 eyes in Group 1, 92 eyes in Group 2, 75 eyes in Group 3). In all patients, the Haigis had the significantly lower MedAE generated by the other formulas (P<0.05). In Group 1 to 3, the MedAE calculated by the Haigis was either significantly lower or comparable to those calculated by the other ***: Compared with other formulas using IOLMaster biometric data, the Haigis formula yields superior refractive results in eyes with various AL.

关键词： intraocular lens optical biometry calculation formula