基于快速多极边界元法的沥青混凝土弹性模量预测

由于对沥青混凝土材料的研究无论是试验法还是经验法均建立在宏观层面上,无法与其细观结构建立本质的联系,因此利用快速多极边界元法(Fast Multipole Boundary Element Method,FMBEM),结合数字图像处理技术,实现沥青混凝土二维几何建模及弹性模量预测.通过数字图像处理技术识别拍摄得到的原始...     

基于改进的参考矢量选取的码字快速搜索算法

矢量量化的编码过程计算复杂性极高,为了减少编码时大量的矢量间距离计算,许多文献提出利用不等式关系以较少的计算量来估算距离的方法。在Chang等人提出的利用双限制三角不等式的快速搜索算法基础上,通过改进参考矢量的选取方法,有效提高了码字搜索的效率。实验结果表明,改进算法的码字排除率可以提高3.735%9.976%,编码时间可以减少6.03%35.25%。     

复杂背景条件下的红外快速运动小目标检测方法

针对复杂背景下红外快速运动小目标检测问题,提出一种将时域特性分析、小波包变换和管道滤波多帧检测相结合的方法。该方法首先对图像进行时域特性分析确定目标所在帧图像序列,再利用小波包变换对目标所在帧图像进行多尺度分解,提取候选目标,然后通过管道滤波从候选目标中确定小目标的位置,完成对小目标的检测。对实测数据进行仿真实验,结果表明该方法能有效检测红外快速运动小目标。     

基于SAD和SATD的H.264快速帧内预测算法

H.264在空域上进行多模式帧内预测,用RDO准则选择最优模式,提高了压缩率,但是增加了计算复杂度。该文提出一种快速帧内预测算法,利用SAD值对Intra_4×4方式进行预判,根据SATD特征排除可能性小的Intra_4×4预测模式。实验结果表明,该算法在保持峰值信噪比基本不变和码率略增的情况下,可节省69.3%的帧内编码时间。     

Accurate evaluation of Green’s functions in a layered medium by SDP-FLAM

Based on local Taylor expansions on the complex plane, a method for fast locating all modes (FLAM) of spectral-domain Green’s Functions in a planar layered medium is developed in this paper. SDP-FLAM, a combination of FLAM with the steepest descent path algorithm (SDP), is employed to accurately evaluate the spatial-domain Green’s functions in a layered medium. According to the theory of complex analysis, the relationship among the poles, branch points and Riemann sheets is also analyzed rigorously. To inverse the Green’s functions from spectral to spatial domain, SDP-FLAM method and discrete complex image method (DCIM) are applied to the non-near field region and the near filed region, respectively. The significant advantage of SDP-FLAM lies in its capability of calculating Green’s functions in a layered medium of moderate thickness with loss or without loss. Some numerical examples are presented to validate SDP-FLAM method. Supported by the National Natural Science Foundation of China (Grant No. 60621002), and the State Key Development Program for Basic Research of China (Grant No. 2009CB320200)     

基于T-S模型的自适应模糊广义预测控制

对一类非线性系统,利用一种基于模糊规则的快速模糊辨识方法建立起系统的T—S模型,并基于该模型应用局部递推最小二乘方法根据采样值对模型参数进行在线修正,根据系统动态线性化模型采取广义预测控制策略,从而实现了基于T—S模糊模型的非线性系统自适应模糊预潮控制。与以往的模糊广义预测控制算法相比,此方法简单,而且较大地减少计算量,适合于在线控制。通过仿真研究验证所提方法的有效性。     

双积分系统阶跃响应快速无超调控制: 一种增益切换非线性PD控制

采用标准PD或PID控制器镇定双积分对象时,系统的阶跃响应必然存在超调:为此,设计了一种可以切换增益的非线性PD控制器,给出了系统阶跃响应无超调的充要条件,以及减小调节时间的方法.采用的切换规律是:当误差的幅值大于设定门限时,设计好的快速PD控制器投入工作,以加快误差的收敛:当误差的幅值小于设定门限时,设计好的强阻尼PD控制器投入工作,以抑制可能出现的超调.通过这两个控制器的配合作用,既能实现阶跃响应无超调,又能有效减小调节时间.仿真验证了该结论.     

一种改进的AEI算法中初始匹配码字的快速查找方法

矢量量化（VQ）是一种高效的有损压缩技术。快速码字搜索算法是矢量量化的核心问题之一,其性能决定了编码时间。快速码字搜索算法中,绝对误差不等式删除算法(AEI)是一种典型的3步算法,其第1步查找输入矢量的初始匹配码字的方法采用了Minimax法,是整个AEI算法中计算量最大的步骤,严重影响了算法的效率。针对这个问题,提出了一种新的查找初始匹配码字的方法——Partial Minimax法。该方法在保证所找到的初始匹配码字与原始AEI算法相同并且重建图像的PSNR（峰值信噪比）值不变的前提下,可显著减小这一步骤的计算量和查找时间,从而有效地提高了算法的总体编码速度。     

Analysis of electro‐optical properties of polymer‐stabilized OCB and the application to TFT‐LCDs

Abstract— A 9‐in. full‐color polymer‐stabilized OCB TFT‐LCD with stable bend alignment in the absence of an electric field was developed. The condition of the polymer stabilization, the characteristics of UV‐curable monomers, and their influence on the configurations of the polymer network in the cell were studied. Possible models of the configuration were proposed and their relationship to the electro‐optical properties was analyzed using a novel simulation method considering the distribution of anchoring effects from both alignment surfaces and the polymer network. It was suggested that a good performance such as high contrast ratio and fast response could be expected in the polymer network originating from newly developed monomers composed of multifunctional LC acrylates due to a relatively weak‐anchoring effect and presumably its localization near the alignment surfaces. By using the newly developed monomers under the optimized polymer‐stabilizing process, a high contrast ratio of 250:1 and fast response nearly equal to that of a conventional OCB cell were achieved.     

计算机时代的设计效果图渲染新说

手工绘图时期非常注重各种绘图工具的学习,计算机的出现使得手工绘图工具对效果图的渲染变的不再重要。要想画好效果图,提高人脑的认识才是关键。本文从实践设计总结的经验出发,提出效果图渲染的学习,从认识上可以分为"全因素表现"和"快速表现";进一步又阐述了每种渲染在设计师的头脑认识上需要把握哪好些处理内容。