低速模拟中叶型最大厚度模化研究

介绍了低速模拟中低速叶型的设计准则,分析说明了叶型喉道宽度与最大厚度、最大厚度位置存在线性关系.建立了最大厚度模型,用以预估低速叶型最大厚度.为了验证最大厚度模型的适用性,进行了一组典型CDA叶型的低速模拟设计.数值计算结果显示,低速模拟叶型的特性和叶表无量纲速度分布较为一致,用以判断最大厚度模型适用性的吸力面峰值速度...     

基于高次余弦散射分布的空间卫星可见光特性

基于几何光学和辐射理论,研究了空间卫星的可见光散射特性.空间卫星的背景辐射主要包括太阳的直接辐射和地球及大气的散射辐射,根据目标的结构特性与背景特性建立了空间卫星的几何模型和光照模型.分析目标表面状况,入射到目标表面的光线近似服从高次余弦散射分布,根据能量守恒定理及表面材料的高次余弦散射分布特性建立了目标散射特性的数学模型.通过矢量坐标变换确定太阳、地球、观测卫星在目标本体坐标系下的相对位置关系.根据给定的几何尺寸和表面物性参量仿真获得了目标在探测器入瞳处的能量分布及星等特征,目标本体与太阳帆板在探测器入瞳处的辐照度最大量级均为10-12 W/m2.仿真结果表明太阳帆板在目标特性分析时不可忽略,为空间目标的可见光探测提供参考数据.     

基于柯西分布的视频图像序列背景建模和运动目标检测

提出了一种用于视觉监视系统的基于柯西分布的发光模型的光照不变变化检测方法.假定视频图像序列中每个背景图像像素点灰度观测值的时序变化由白噪声引起,利用建立的初始化背景高斯统计模型对每帧图像进行归一化,得到了背景像灰度比值的分布符合标准柯西分布的结论,解决了柯西分布的模型参量估计问题.在变化检测的基础上,YCbCr颜色空间的亮度、色调和饱和度被用来识别和消除由阴影和反光等引起的变化区域.结果表明,提出的背景建模方法对场景中各种光线变化、小的背景扰动等噪声具有稳健性,可以较为可靠地检测前景目标,识别和去除阴影和反光.     

On Physically Unacceptable Numerical Solutions Yielding Strong Chaotic Signals

Physically unacceptable chaotic numerical solutions of nonlinear circuits and systems are discussed in this paper. First, as an introduction, a simple example of a wrong choice of a numerical solver to deal with a second-order linear ordinary differential equation is presented. Then, the main result follows with the analysis of an ill-designed numerical approach to solve and analyze a particular nonlinear memristive circuit. The obtained trajectory of the numerical solution is unphysical (not acceptable), as it violates the presence of an invariant plane in the continuous systems. Such a poor outcome is then turned around, as we look at the unphysical numerical solution as a source of strong chaotic sequences. The 0–1 test for chaos and bifurcation diagrams are applied to prove that the unacceptable (from the continuous system point of view) numerical solutions are, in fact, useful chaotic sequences with possible applications in cryptography and the secure transmission of data.     

Microscale Thermophoresis and Molecular Modelling to Explore the Chelating Drug Transportation in the Milk to Infant

The microscale thermophoresis (MST) technique was utilized to investigate lactoferrin–drug interaction with the iron chelator, deferiprone, using label-free system. MST depends on the intrinsic fluorescence of one interacting partner. The results indicated a significant interaction between lactoferrin and deferiprone. The estimated binding constant for the lactoferrin–deferiprone interaction was 8.9 × 10⁻⁶ ± 1.6, SD, which is to be reported for the first time. Such significant binding between lactoferrin and deferiprone may indicate the potentiation of the drug secretion into a lactating mother’s milk. The technique showed a fast and simple approach to study protein–drug interaction while avoiding complicated labeling procedures. Moreover, the binding behavior of deferiprone within the binding sites of lactoferrin was investigated through molecular docking which reflected that deferiprone mediates strong hydrogen bonding with ARG121 and ASP297 in pocket 1 and forms H-bond and ionic interaction with ASN640 and ASP395, respectively, in pocket 2 of lactoferrin. Meanwhile, iron ions provide ionic interaction with deferiprone in both of the pockets. The molecular dynamic simulation further confirmed that the binding of deferiprone with lactoferrin brings conformational changes in lactoferrin that is more energetically stable. It also confirmed that deferiprone causes positive correlation motion in the interacting residues of both pockets, with strong negative correlation motion in the loop regions, and thus changes the dynamics of lactoferrin. The MM-GBSA based binding free energy calculation revealed that deferiprone exhibits ∆G TOTAL of −63,163 kcal/mol in pocket 1 and −63,073 kcal/mol in pocket 2 with complex receptor–ligand difference in pocket 1 and pocket 2 of −117.38 kcal/mol and −111.54 kcal/mol, respectively, which in turn suggests that deferiprone binds more strongly in the pocket 1. The free energy landscape of the lactoferrin–deferiprone complex also showed that this complex remains in a high energy state that confirms the strong binding of deferiprone with the lactoferrin. The current research concluded that iron-chelating drugs (deferiprone) can be transported from the mother to the infant in the milk because of the strong attachment with the lactoferrin active pockets.     

Novel Design of RNA Aptamers as Cancer Inhibitors and Diagnosis Targeting the Tyrosine Kinase Domain of the NT-3 Growth Factor Receptor Using a Computational Sequence-Based Approach

Aptamers, the nucleic acid analogs of antibodies, bind to their target molecules with remarkable specificity and sensitivity, making them promising diagnostic and therapeutic tools. The systematic evolution of ligands by exponential enrichment (SELEX) is time-consuming and expensive. However, regardless of those issues, it is the most used in vitro method for selecting aptamers. Therefore, recent studies have used computational approaches to reduce the time and cost associated with the synthesis and selection of aptamers. In an effort to present the potential of computational techniques in aptamer selection, a simple sequence-based method was used to design a 69-nucleotide long aptamer (mod_09) with a relatively stable structure (with a minimum free energy of −32.2 kcal/mol) and investigate its binding properties to the tyrosine kinase domain of the NT-3 growth factor receptor, for the first time, by employing computational modeling and docking tools.     

高阶常微分方程的动态演化建模

提出了采用高阶常微分方程模型代替传统时序分析中所用的ARMA模型来实现一维时间序列的建模和预报．设计的将遗传程序设计与遗传算法相嵌套的动态演化建模算法,用遗传程序设计优化模型结构,以遗传算法优化模型参数,边收集数据边建模边预报,首次成功地实现了时间序列实时预报的程序自动化．两个时间序列的应用实例表明采用此算法可获得较好的实时预报效果．     

耦合FE/WB法在声分析中的应用

简要描叙FE法(finite element method)和WB法(wave based method)的理论背景以及耦合FE/WB法的数学基础.耦合FE/WB法利用两者的优势——FE法的广泛应用和WB法的高收敛特性,将FE模型中较大且几何简单的部分采用WB法代替.耦合模型具有相对较少的自由度.对于较高的频率还可以进行细分得到更高的计算精度,并利用模态缩减法进一步减少自由度数.数值算例结果表明,该耦合方法有潜力覆盖中频段的声分析.     

汽车整车非线性动力计算

通过轿车副车架橡胶支承的动态特性试验和理论分析,实现橡胶支承迟滞特性的数学建模和参数识别。利用所建模型重构恢复力-位移迟滞回线,并与试验结果比较表明所建模型的正确性。介绍了具有线性和非线性连接子结构的自由界面模态综合法的基本原理,并采用该方法建立了考虑副车架橡胶支承迟滞特性的整车非线性动力学模型,用所建模型对汽车动力特性进行了数值仿真,并通过试验对仿真结果进行验证。     

基于SARIMA-BP神经网络组合方法的MODIS叶面积指数时间序列建模与预测

植被叶面积指数(LAI)时间序列的建模及预测是陆面过程模型和遥感数据同化方法的重要组成部分。MODIS数据产品MOD15A2是目前应用最为广泛的LAI数据源之一,然而MODIS LAI时间序列产品包含了一些低质量的数据,例如由于云层、气溶胶等的影响,该产品在时间和空间上缺乏连续性。MODIS LAI时间序列包含线性部分和外在干扰产生的非线性部分,单一的线性方法或非线性方法都不能对其精确建模和预测。首先利用Savitzky-Golay(SG)滤波和线性插值平滑受到干扰的LAI时间序列,然后采用季节自回归积分滑动平均(SARIMA)方法、BP神经网络方法及二者的组合方法(SARIMA-BP)对MODIS LAI时间序列进行建模及预测。在SARIMA-BP神经网络组合方法中,各自在线性与非线性建模的优势得以充分发挥,其中SARIMA方法用于建模及预测LAI时间序列中的线性部分,BP神经网络方法用于对非线性残差部分进行建模及预测。实验结果显示：SG滤波和线性插值后的LAI时间序列比原LAI时间序列更平滑;SARIMA-BP神经网络组合方法的决定系数为0.981,比SARIMA和BP神经网络的0.941和0.884更接近于1;SARIMA-BP神经网络组合方法的预测值同观测值之间的相关系数为0.991,高于SARIMA(0.971)和BP神经网络(0.942)的相关系数。由此得出结论：SARIMA-BP神经网络组合方法对MODIS LAI时间序列具有更好的适应性,其建模和预测准确性高于SARIMA方法或BP神经网络方法。