一种基于模糊分类网络的卷积码解码方法

本文提出一种基于模糊神经分类网络的卷积码译码方法，把译码工作转换为网络的分类工作。网络按照卷积码的编码方式由聚类自动生成，并且在网络中使用逻辑算子，因而网络的训练速度非常快，只需一次或几次迭代。而且对每个隐节点均定义了模糊隶属度函数，借以增加网络的联想能力，从而提高网络的纠错能力。在小约束度情况下，我们测试了该算法的性能，并与Ｖｉｔｅｒｂｉ算法的性能进行了比较。     

Optimal binary covering codes of length 2j

The minimum size of a binary covering code of length n and covering radius r is denoted by K(n,r), and codes of this length are called optimal. For j > 0 and n = 2^j, it is known that K(n,1) = 2 · K(n?1,1) = 2^{n ? j}. Say that two binary words of length n form a duo if the Hamming distance between them is 1 or 2. In this paper, it is shown that each optimal binary covering code of length n = 2^j, j > 0, and covering radius 1 is the union of duos in just one way, and that the closed neighborhoods of the duos form a tiling of the set of binary words of length n. Methods of constructing such optimal codes from optimal covering codes of length n ? 1 (that is, perfect single‐error‐correcting codes) are discussed. The paper ends with the construction of an optimal covering code of length 16 that does not contain an extension of any optimal covering code of length 15. © 2005 Wiley Periodicals, Inc. J Combin Designs     

Some constructions for block sequences of Steiner quadruple systems with error correcting consecutive unions

In this article, we investigate a block sequence of a Steiner quadruple system which contains the blocks exactly once such that the collection of all blocks together with all unions of two consecutive blocks of the sequence forms an error correcting code with minimum distance four. In particular, we give two recursive constructions and obtain infinitely many such sequences by utilizing individual sequences as starters of the recursions. © 2007 Wiley Periodicals, Inc. J Combin Designs 16: 152–163, 2008     

基于PSO-BP算法的压力传感器温度补偿研究

硅压力传感器在工业环境中使用时,尤其是用于测量油井井下压力的时候,环境温度变化范围通常比较大.硅压力传感器由于其自身结构的原因,输出压力值会呈现非线性变化,大大的降低了压力传感器的测量精度.本文是基于PSO-BP神经网络方法对压力传感器在温度变化时产生的误差进行补偿修正,以达到系统精度要求.PSO-BP算法的本意是使用PSO算法用于对BP神经网络的初始权值和阈值进行改进和筛选,然后再使用BP网络对样本进行训练,以提高系统的泛化能力和稳定性.     

Asymptotic Nonlinearity of Boolean Functions

Boolean functions on the space are not only important in the theory of error-correcting codes, but also in cryptography. In these two cases, the nonlinearity of these functions is a main concept. Carlet, Olejár and Stanek gave an asymptotic lower bound for the nonlinearity of most of them, and I gave an asymptotic upper bound which was strictly larger. In this article, I improve the bounds and get an exact limit for the nonlinearity of most of Boolean functions. This article is inspired by a paper of G. Halász about the related problem of real polynomials with random coefficients. AMS Classification (2000) Primary: 11T71 · Secondary: 06E30 · 42A05 · 94C10     

Directed t-Packings and Directed t-Steiner Systems

The aim of this paper is an investigation of directed t-packings and in particular of directed t-Steiner systems. A new upper bound on the number of points k for directed t-Steiner systems T(t,k,k) is obtained. We disprove a conjecture of Levenshtein on T(t,k,k) for t 3 by showing that a T(4,6,6) exists. Furthermore, it is proved that the symmetric group S ₆ can be partitioned into 30 disjoint T(4,6,6)s. Extensive computer search shows that the tight upper bound on K for t =4,5 is 6 and for t=6 is 7. The non-existence of further small directed t-Steiner systems is established, and large directed t-packings for t,4,5,6 are constructed.     

基于McEliece纠错码的公钥密码体制的研究

纠错码与密码的结合是代数编码理论和密码学发展的必然产物。深入研究了M公钥密码与MS公钥密码体制的几个性能指标,给出了它们的计算机模拟曲线;分析了M公钥密码与MS公钥密码体制通过有扰信道时的正确解密概率及计算复杂度;得到了一些有价值的结果。     

用共线点列标定摄像机镜头畸变

提出一种用共线点列标定摄像机镜头畸变参数的方法,先根据交叉比不变性确定畸变中心,再利用直线的中心投影仍为直线这一性质确定畸变系数。仿真表明,只要采用相应方法降低像点坐标的误差,畸变中心的误差小于2个像素,畸变系数的相对误差不超过0.1。实验表明,用该方法得到的畸变参数,能够很好地校正图像。该方法简单易行,可应用于摄影测量、计算机视觉等领域。     

椭偏仪数据处理及误差修正

提出了一套对椭偏仪的测量误差进行有效修正的方案,设计了一个Windows版的椭偏仪测厚数据处理软件,并在该软件中嵌入了对测量数据的误差进行修正的方法,使椭偏仪的最终测量结果更加精确。制作了30片厚度梯度分布的标准样片（厚度20nm～1μm）,用于从“软”、“硬”件两个方面对椭偏仪进行误差修正,使最终的误差小于1％。本文所提出的修正方案具有一定的普适性、实用性。     

位相测量轮廓术中的相位修正及系统标定的新方法

介绍了在位相测量轮廓术中的一种相位修正及系统参数标定的新方法。对参考面(标定面)上点的位置_相位关系进行全场拟合,并根据该相互关系对参考面相位进行修正,能够非常有效地去除相位误差;提出的参数标定方法能精确得到投影系统参数的组合。实验证明了此方法具有良好的实用性和测量精度。