Damage detection in structures using modified back-propagation neural networks

A nonparametric structural damage detection methodology based on neural networks method is presented for health monitoring of structure-unknown systems. In this approach appropriate neural networks are trained by use of the modal test data from a ‘healthy’ structure. The trained networks which are subsequently fed with vibration measurements from the same structure in different stages have the capability of recognizing the location and the content of structural damage and thereby can monitor the health of the structure. A modified back-propagation neural network is proposed to solve the two practical problems encountered by the traditional back-propagation method, i.e., slow learning progress and convergence to a false local minimum. Various training algorithms, types of the input layer and numbers of the nodes in the input layer are considered. Numerical example results from a 5-degree-of-freedom spring-mass structure and analyses on the experimental data of an actual 5-storey-steel-frame demonstrate that neural-networks-based method is a robust procedure and a practical tool for the detection of structural damage, and that the modified back-propagation algorithm could improve the computational efficiency as well as the accuracy of detection Project supported by the National Natural Science Foundation of China (No. 59908003) and the Natural Science Foundation of Hubei Province (No. 99J035).     

Efficient implementation of the proper outlet flow conditions in blood flow simulations through asymmetric arterial bifurcations

We present an efficient implementation of the proper (in vivo) outlet boundary conditions in detailed, three‐dimensional (3D) and time‐periodic simulations of blood flow through arteries. This is achieved through the intermediate use of an approximate ‘simulant’ model of the outlet pressure/flow relationship corresponding to the full 3D and time‐dependent numerical simulation. This model allows us to efficiently couple the 3D outlet pressure/flow conditions to the equivalent relations due to the downstream arterial network, as obtained from a one‐dimensional approximate model in the form of Fourier frequency impedance coefficients. An adjustable time‐periodic function correction term in the simulant model requires input from the full 3D model that has to run iteratively until convergence. The advantage of the proposed numerical scheme is that it decouples the upstream detailed simulation from the downstream approximate network model offering exceptional versatility. This approach is demonstrated here in a series of detailed 3D simulations of blood flow, performed using the commercial software FLUENT?, through an asymmetric arterial bifurcation. Two cases are considered: first a healthy system patterned after the left main coronary arterial bifurcation, and second a diseased case where an occlusion has developed in one of the daughter vessels, resulting in strengthening the asymmetry of the bifurcation. Rapid convergence of the iterative process was achieved in both cases. Subtle changes occur in the shear patterns of the daughter vessels, whereas the flow distribution is quite different. In the presence of a stenosis additional regions of low shear develop due to inertial effects. Copyright © 2010 John Wiley & Sons, Ltd.     

Radial basis function neural network model for mean velocity and vorticity of capillary flow

The radial basis function neural network (RBFNN) simulation has been designed to simulate and predict the mean velocity of capillary flow in transition from laminar to turbulent flow and the root‐mean‐square vorticity as a function of wall‐normal position at different values of Reynolds number. The system was trained on the available data of the two cases. Therefore, we designed the system to work in automatic way for finding the best network that has the ability to have the best test and prediction. The proposed system shows an excellent agreement with that of an experimental data in these cases. The technique has been also designed to simulate the other distributions not presented in the training set and predicted them with effective matching. Copyright © 2010 John Wiley & Sons, Ltd.     

Numerical simulation of the process of autoregulation of the arterial blood flow

For describing the autoregulation of the blood flow in an artery under constant transmural pressure a mathematical model that takes into account the multilayer structure of the arterial wall, the diffusion and kinetic processes in the wall, and the nonlinear viscoelastic properties of the wall material is proposed. The limiting case of a thin-walled artery is studied analytically. The arterial-wall viscosity range on which the equilibrium state of the system is stable is estimated. Accurate stationary distributions of the nitric oxide, calcium and myosin concentrations in the arterial wall are found. Numerical simulation of the autoregulation process demonstrated the possibility of arterial adaptation to radius perturbations, the existence of slow oscillations, and system transition to a new equilibrium state with change in blood flow level. The results are in good agreement with the experimental data.     

Modeling and prediction of chaotic systems with artificial neural networks

This study of chaotic systems and their prediction is motivated by the fact that many phenomena, both natural and man‐made, are of a chaotic nature. Such phenomena include but are not limited to earthquakes, laser systems, epileptic seizures, combustion, and weather patterns. These phenomena have previously been thought to be unpredictable. However, it is indeed possible to predict time series generated by chaotic systems. The primary objective of this study is to develop a system that would train the artificial neural network (ANN) and then predict the future data of the process. In the present application, the chosen chaotic data set was obtained by solving Lorenz's equations. To predict the future data, the concept of a multilayer feed‐forward ANN with nonlinear auto‐regressive moving averages with exogenous input is used. A Backpropagation algorithm is used to train the network for the chaotic data. The final updated weights from the trained network were then used for the prediction of the future values of the system. Lyapunov exponents, phase diagrams and statistical analyses were used to evaluate the neural network output. A correlation of 94% and a negative Lyapunov exponent indicate that the results obtained from ANN are in good agreement with the actual values. Copyright © 2009 John Wiley & Sons, Ltd.     

计算动脉血管发展流动的新差分格式

在引用守恒型N—S方程和SIMPLE方法的基础上,对血管发展流动的控制方程提出了新的差分格式。算例结果表明：本差分格式对血管发展流动的数值研究是可行的。     

A versatile incompressible Navier–Stokes solver for blood flow application

We present in this paper an integrated approach to compute quickly an incompressible Navier–Stokes (NS) flow in a section of a large blood vessel using medical imaging data. The goal is essentially to provide a first‐order approximation of some main quantities of interest in cardiovascular disease: the shear stress and the pressure on the wall. The NS solver relies on the L₂ penalty approach pioneered by Caltagirone and co‐workers and combines nicely with a level set method based on the Mumford–Shah energy model. Simulations on stenosis cases based on angiogram are run in parallel with MatlabMPI on a shared‐memory machine. While MatlabMPI communications are based on the load and save functions of Matlab and have high latency indeed, we show that our Aitken–Schwarz domain decomposition algorithm provides a good parallel efficiency and scalability of the NS code. Copyright © 2006 John Wiley & Sons, Ltd.     

Theoretical and numerical study of Couette-Taylor flow with an axial flow using lattice Boltzmann method

In this study, the numerical models for swirling flows developed by Li et al and Zhou for lattice Boltzmann method (LBM) are chosen. These models were firstly validated using the Couette-Taylor flow between two concentric cylinders simulations. Numerical results showed the efficiency of the Zhou's model. Numerical simulation results using LBM are in good agreement for the steady and unsteady regimes compared to the literature review. In a second step, the Zhou model was then adopted to our study to determine the Couette-Taylor instabilities with an axial flow. Two protocols are tested. The first one (direct protocol) starts with an azimuthal flow without any axial flow (Re = 0). Once the regime is established, an axial flow is then superposed to the Couette-Taylor flow (with a sudden or a progressive manner). The second protocol (inverse protocol) starts with an axial flow at a given Reynolds number (Poiseuille flow). Once the regime is established, an azimuthal flow is the executed (with a sudden or a progressive manner). The effect of various parameters controlling the physical situation is also discussed. The increase of the azimuthal velocity mainly led to the emergence and development of Taylor vortices. Its influence decreases when the axial Reynolds number increases. The relevant result for this study is the change of the critical axial Reynolds number Re_c (total disappearance of instabilities) with both protocols and both manners.     

GLOBAL ASYMPTOTIC STABILITY CONDITIONS OF DELAYED NEURAL NETWORKS

1　IntroductionandProblemEductionRecently,thestudiesofthestabilityforcellularneuralnetworks (CNNs)anddelayedcellularneuralnetworks (DCNNs)haveattractedattentionsofresearchersandseveralimportantresultshavebeenobtained .MostpapersdealtwithcompletelystableCNNsandDCNNsthataresuitableforimageprocessingapplications.CNNshavebeenwidelyappliedtoimageprocessing ,toprocessmovingimages,onemustintroducedelaysinthesignalstransmittedamongthecells.Buttimedelaysmayleadtoanoscillationphenomenonand ,furt…     

Boundary layer development of pulsatile blood flow in a tapered vessel

Assuming that the tapered angle is small,the problems of developing flow under unsteady oscillatory condition are studied in this paper.The formula of velocity distribution is obtained.The analyses for the results show that the blood flow in a converging tapered vessel remains a developing flow throughout the length,and the effects of tapered angle on the developing flow are increased with the increment of the tapered angle.     

地震作用下高层建筑结构的分散神经网络振动控制研究

针对地震作用下高层建筑振动神经网络控制问题,将神经网络理论与分散控制理论相结合,提出分散神经网络振动控制方案,并应用于高层结构地震反应振动控制中。利用多层前馈神经网络建立结构模型,预测结构的振动响应。基于NARMA-L2的神经自校正控制系统设计BP神经网络控制器,研究分散神经网络振动控制效果,并与神经网络集中控制进行比较。对某20层Benchmark结构模型进行数值模拟分析,结果表明,本文提出的分散神经网络振动控制方法简化了神经网络的结构,可有效控制结构振动和消除时滞;同时,相对于集中控制的单一失效,本文方法的可靠性更强且可以保证振动控制系统的实时响应。     

Electroosmotic oscillatory flow of micropolar fluid in microchannels: application to dynamics of blood flow in microfluidic devices

The electroosmotic flow of a micropolar fluid in a microchannel bounded by two parallel porous plates undergoing periodic vibration is studied. The equations for conservation of linear and angular momentums and Gauss' s law of charge distribution are solved within the framework of the Debye-H¨uckel approximation. The fluid velocity and microrotation are assumed to depend linearly on the Reynolds number. The study shows that the amplitude of microrotation is highly sensitive to the changes in the magnitude of the suction velocity and the width of the microchannel. An increase in the micropolar parameter gives rise to a decrease in the amplitude of microrotation. Numerical estimates reveal that the microrotation of the suspended microelements in blood also plays an important role in controlling the electro-osmotically actuated flow dynamics in microbio-fluidic devices.     

Dynamics of the blood flow in the curved artery with the rolling massage

Arterial wall shear stress and flow velocity are important factors in the development of some arterial diseases. Here, we aim to investigate the dynamic effect of the rolling massage on the property of the blood flow in the curved artery. The distributions of flow velocity and shear stress for the blood flow are computed by the lattice Boltzmann method, and the dynamic factors under different rolling techniques are studied numerically. The study is helpful to understand the mechanism of the massage and develop the massage techniques.     

Bifurcation of flow and mass transport in a curved blood vessel

A numerical analysis of flow and concentration fields of macromolecules in a, slightly curved blood vessel was carried out. Based on these results, the effect of the bifurcation of a flow on the mass transport in a curved blood vessel was discussed. The macromolecules turned out to be easier to deposit in the inner part of the curved blood vessel near the critical Dean number. Once the Dean number is higher than the critical number, the bifurcation of the flow appears. This bifurcation can prevent macromolecules from concentrating in the inner part of the curved blood vessel. This result is helpful for understanding the possible correlations between the blood dynamics and atherosclerosis. The project supported by National Natural Science Foundation of China (10002003), JSPS Postdoctoral Fellowship for Foreign Researcher and Foundation for University Teachers, the Ministry of Education     

Construction of hybrid 1D-0D networks for efficient and accurate blood flow simulations

The one-dimensional (1D) modeling of blood flow in complex networks of vessels and cardiovascular models can result in computationally expensive simulations. The complexity of such networks has significantly increased in the last years, in terms of both enhanced anatomical detail and modeling of physiological mechanisms and mechanical characteristics. To address such issue, the main goal of this work is to present a novel methodology to construct hybrid networks of coupled 1D and 0D vessels and to perform computationally efficient and accurate blood flow simulations in such networks. Departing from both the 1D and lumped-parameter (0D) nonlinear models for blood flow, we propose high-order numerical coupling strategies to solve the 1D, 0D, and hybrid coupling of vessels at junctions. To effectively construct hybrid networks, we explore different a-priori model selection criteria focusing in obtaining the best possible trade-off between computational cost of the simulations and accuracy of the computed solutions for the hybrid network with respect to the 1D network. The achievement of the expected order of accuracy is verified in several test cases. The novel methodology is applied to two different arterial networks, the 37-artery network and the reduced ADAN56 model, where, in order to identify the best performing a-priori model selection criteria, the quantitative assessment of CPU times and errors and the qualitative comparison between results are carried out and discussed.     

A comparative study of poiseuille flow of a polar fluid under various boundary conditions with applications to blood flow

In this paper, Poiseuille flow of a polar fluid (model of a red blood cell suspension) under various boundary conditions at the wall, viz., slip or no-slip in the axial velocity and couple stresses zero or non-zero at the boundary, is considered from the point of view of its applications to blood flow. Analytic expressions for axial and rotational velocities, flow rate, effective viscosity and stresses are obtained. The magnitudes of the length ratioL and the coupling number N are determined in accordance with concentration and tube radius (in the existing literature, values ofL andN are chosen arbitrarily). Velocity profiles (both axial and rotational) and the variation of the effective viscosity with concentration, tube radius and for various values of the boundary condition parameters are shown graphically. The analytic results obtained are compared with experimental results (for blood flow). It is found that they are in a reasonably good agreement. The effective viscosity exhibits the Inverse Fahraeus-Lindquist Effect in all the cases (including the slip or no-slip in the velocity fields). A method is given for determining the non-zero couple stress boundary condition for a given concentration. Applications of this theory to blood flow are briefly discussed.     

改进遗传人工神经网络在组合导航中的应用

鉴于常规卡尔曼滤波算法组合导航系统数据融合算法中,存在易于发散的缺陷,尝试将遗传优化人工神经网络引入组合导航系统中.针对传统遗传算法存在的易早熟、算法稳定性差、固定的交叉和变异概率影响收敛效果等缺点,采用浮点式编码方式,两两竞争的选择策略、引入突变操作、重新定义交叉算子和自适应的交叉变异算子等措施进行了遗传算法的改进.仿真结果表明,改进后的算法更为有效,并且精度与常规卡尔曼滤波算法相当.     

GLOBAL DYNAMICS OF DELAYED BIDIRECTIONAL ASSOCIATIVE MEMORY （BAM） NEURAL NETWORKS

IntroductionConsiderthebidirectionalassociativememory (BAM )neuralnetworkswithconstanttransmissiondelaysdescribedbyasystemofdelaydifferentialequationsoftheform[1,2 ]:dxi(t)dt =-aixi(t) nj=1bijfj(yj(t-σij) ) Ii,　　i=1 ,2 ,… ,m ,dyj(t)dt =-cjyj(t) mi=1djigi(xi(t-τji) ) Jj,　　j=1 ,2 ,… ,n ,fort >0 .Thesystem ( 1 )consistsoftwosetsofneurons (orunits)arrangedontwolayers,namely ,I_layerandJ_layer.Inthesystem ( 1 ) ,xi( ·)andyj( ·)denotemembranepotentialoftheithneuronsfromtheI_laye…     

磁场和Hall电流对狭窄动脉中血液流动的影响

A micropolar model for blood simulating magnetohydrodynamic flow through a horizontally nonsymmetric but vertically symmetric artery with a mild stenosis is presented. To estimate the effect of the stenosis shape, a suitable geometry has been considered such that the horizontal shape of the stenosis can easily be changed just by varying a parameter referred to as the shape parameter. Flow parameters, such as velocity, the resistance to flow （the resistance impedance）, the wall shear stress distribution in the stenotic region, and its magnitude at the maximum height of the stenosis （stenosis throat）, have been computed for different shape parameters, the Hartmann number and the Hall parameter. This shows that the resistance to flow decreases with the increasing values of the parameter determining the stenosis shape and the Hail parameter, while it increases with the increasing Hartmann number. The wall shear stress and the shearing stress on the wall at the maximum height of the stenosis possess an inverse characteristic to the resistance to flow with respect to any given value of the Hartmann number and the Hall parameter. Finally, the effect of the Hartmann number and the Hall parameter on the horizontal velocity is examined.