神经网络余／十转换的数学原理

本文提出了孙子定理的微分方程求解法，讨论了系数为求余算子的非线性微分方程的稳定平衡点和最佳解的误差。本文构造的非线性微分方程可以唯一地收敛于孙子定理的解，其误差可任意小。该微分方程可用人工神经网络实时计算，实现余／十转换。计算机模拟结果证实了本文理论的正确性。     

采用基本门单元完全测试矢量的测试生成算法

文章介绍了一种采用基本逻辑门单元的安全测试矢量集生成测试矢量的方法，该方法可以将搜索空间限制在2(n 1）种组合内。它采用故障支配和故障等效的故障传播、回退等技术，建立了一套从局部到全局的测试生成新方法。同时，利用基本门单元安全测试矢量的规律性，可以实现最小的内存容量要求。在一些基准电路的应用实例中，得到了满意的结果。     

运动估计块匹配两级搜索算法

根据视频图像相邻块在空间上具有很大的相关性,吸收了矢量预测块匹配的核心思想,提出一种两级块匹配运动估计算法,在第一级中,首先根据矢量预测确定一个小区域,然后在此区域内进行快速搜索,如果在第一级中没有搜索到匹配块则进行第二级搜索,该算法可大大提高搜索速度.     

模糊串匹配算法及其应用

本文分析了模糊字符串几种类型和它的表示方法，并给出模糊匹配的算法及Ｃ语言的源程序，同时也指出了在文枉编辑器的应用例子。     

一种新的环网故障诊断与重构算法

为提高环网的容错性能，文献（1）提出了一种基于事件驱动的故障重构算法，其缺点是该算法只能容忍连线故障。本文给出一种新的采用时钟起时驱动的重构算法，该算法既能容忍连线故障也能容忍站（或节点）故障，且令牌控制较文献（1）的简单。     

An Optimal Model and Algorithm of Distributed Information System

Distributed information system makes itself be placed in changing file storage position according to the users＇ request pattern. In this paper, we rebuild the model for a management system to turn the process of file managing into a 0-1 programming problem, and present a new individual form to improve the operating efficiency. Aiming at the model, we define a neighborhood span to make segmentation for searching space by using the fitness, based on the region contraction algorithm, present a new evolution algorithm which has the capability of self-adaptively generating new individuals, and ultimately solve the management problem of the distributed file system.     

水轮机调速器参数自整定遗传算法的研究

水轮机调速器的参数整定直接关系到水轮机调节系统的稳定运行及调节品质，进而会影响到电能的质量.应用可变交叉概率和变异概率的改进遗传算法对水轮机PID调速器参数进行优化，并将改进遗传算法的计算结果进行仿真分析，结果表明，该方法有效地消除了对参数初值的依赖性，使得寻优效率大大提高，改善了收敛性，是一种较好的PID参数寻优方法。     

Database Encoding and A New Algorithm for Association Rules Mining

This paper proposes a method to encode database. By this way, a record is denoted by only one binary number, so the size of the database is reduced sharply. If some known modified algorithms are used on the database encoded, the efficiency will be improved significantly. At the meantime, a new algorithm based on the proposed encoding method is introduced too. By using some properties of numbers, the itemsets of the database can be converted into numerical fields. Different from the Apriori, the new algorithm discovers the association rules from the largest frequent itemset at first, and then all subitemsets, which are also frequent, will be gotten without any calculation, and all the other small none-frequent itemsets that must be generated in the Apriori will be omitted, and the scan times of the database are also reduced. Test results show the new algorithm based on the encoding database has a lower complexity of time and space.     

抑制微光波前传感器随机噪声的方法研究

在自适应光学系统微光波前传感器中采用像增强器后，高帧频CCD输出的图像中会产生严重的随机噪声，影响光斑质心计算精度，从而使自适应光学系统波前探测精度降低。本文根据微光波前传感器中CCD输出图像的特点对中值滤波，均值滤波，自相关，延时相关等多种方法进行了仿真和实验比较，并提出了一系列评价方法。研究表明延时相关算法是最佳选择。