基于多相滤波器组的语音基频检测方法

基于多相滤波器组的语音基频检测方法，运用多相滤波器组分解语音信号频谱，然后利用声带震动的能量准周期性在各子带进行峰值搜索，并综合这些子带的搜索结果计算基音周期，最后根据先验知识以及一种新的清浊音判定方法对结果进行校正。基于标准TIMIT语料库子集的实验结果证明了此方法的有效性。     

蚁群算法求解连续空间优化问题

借鉴蚁群算法的进化思想，提出一种求解连续空问优化问题的蚁群算法。该算法主要包括全局搜索、局部搜索和信息素强度更新规则。在全舄搜索过程中，利用信息素强度和启发式函数确定蚂蚁移动方向。在局部搜索过程中，嵌入了确定性搜索，以改善寻优性能，加快收敛速率。通过一个实例问题的求解表明了该算法的有效性。     

Probabilistic proximity search: Fighting the curse of dimensionality in metric spaces

Proximity searches become very difficult on “high dimensional” metric spaces, that is, those whose histogram of distances has a large mean and/or a small variance. This so-called “curse of dimensionality”, well known in vector spaces, is also observed in metric spaces. The search complexity grows sharply with the dimension and with the search radius. We present a general probabilistic framework applicable to any search algorithm and whose net effect is to reduce the search radius. The higher the dimension, the more effective the technique. We illustrate empirically its practical performance on a particular class of algorithms, where large improvements in the search time are obtained at the cost of a very small error probability.     

Orthogonal queries in segments

We consider theorthgonal clipping problem in a set of segments: Given a set ofn segments ind-dimensional space, we preprocess them into a data structure such that given an orthogonal query window, the segments intersecting it can be counted/reported efficiently. We show that the efficiency of the data structure significantly depends on a geometric discrete parameterK named theProjected-image complexity, which becomes Θ(n ²) in the worst case but practically much smaller. If we useO(m) space, whereK log^4d−7 n≥m≥n log^4d−7 n, the query time isO((K/m)^1/2 log^{max{4, 4d−5}} n). This is near to an Ω((K/m)^1/2) lower bound.     

Application of design parameters space search for belt conveyor design

Belt conveyor design is examined as an application of a proposed Design Parameters Space Search technique. First, the main characteristics of the belt-conveyor design process are presented as they appear in the current literature. Furthermore, a proposed general knowledge-representation platform is described, and its ability to house the relevant conveyor design knowledge is also shown.

Next, the extended search technique of the design space is discussed, and an integrated example of a belt-conveyor design is presented, based on the proposed representation platform and the extended search technique. Finally, it is shown that the design of belt conveyors according to the proposed approach presents the following significant advantages:

• • Due to the knowledge-representation scheme adopted, both qualitative and quantitative knowledge can be used within the same platform.

• • Multiple solutions can be easily produced through user-defined design criteria. These solutions can be further modified and/or evaluated to produce more-specific designs.

• • The required for user-input data is kept to a minimum. Due to the applied extended search method, semiautomatic design can be achieved. As a consequence, the design process is completed in less time than that required by the conventional methods.

     

一种求解W(s)=0的有效算法

复变函数的幅角原理[1]给出了方程W（s）=0的根存在的充分和必要条件，但如何有效地确定这些根，还需要进一步探索。本文给出一种在复域解决这一问题的数值方法─—正交搜索法。该方法利用解析函数保角映射的特性，通过沿复变量s的实部和虚部交替搜索，逐步逼近和达到所求方程的根．该方法不需计算任何导数，直观、简便，且能在计算机上自动完成。应用该方法解决某工程问题，获得了满意的结果．     

无网格线探索布线算法

提出的一种高效快速的无网格线探索面线算法适用于焊盘外形尺寸不一，线宽及线间距离可变的印制电路板及ＭＣＭ电路布线，精心设计的数据结构及灵活的绕障探索方式可大大地提高搜索效率；多级多扁优化策略为改善布线结果提供了可靠的保证，该算法已成功地应用于通用印制电路板设计系统ＭＰＣＢ，取得了令人满意的效果。     

基于进化策略的抽题算法设计

基于进化算法的工作原理，设计了智能试题库系统中的抽题算法，提出了基于进化策略的抽题系统的数学模型，分析了算法的设计方案，该算法在全局寻优过程中收敛速度快，效率较高，具有良好的用户适用性，测试结果表明该系统能根据用户期望进行智能抽题。     

具有子串检索功能的可扩展分布式数据结构

EH*S是可扩展分布式数据结构EH*的一个改进，增加了子串检索功能。通过对子串和关键字计算描述符向量，为EH*文件中的每个桶添加一个桶描述符向量，然后把子串描述符向量分别与关键字和桶的描述符向量进行比较，得到包含子串的关键字集。     

用边缘金字塔结构实现Hausdorff距离匹配

为了快速准确地在海量图像中搜索模板的所有实例，提出运用边缘金字塔结构来实现Hausdorff距离匹配．根据模板的边缘自适应地确定边缘金字塔分解次数和抽取模式，然后通过金字塔结构的引导搜索策略实现快速匹配．对以水平和垂直边缘为主的集成电路显微图像，文中算法体现出优越的匹配性能．在集成电路显微图像和一般图像上分别测试的实验结果表明：文中算法在获得良好的匹配效果的同时，大幅减少了匹配时间．