基于scoke通信

一个很好用的bmp图像转protelPCB封装库，如果有需要在pcb中插入图像，汉字的朋友可以下载(bmp to protel99，it s a good tool，welcom to download)

dfasdffsdfdsfsdfdsfsdaafdsadfadfdfadfdfadfasdfasdfafdsfdsfasfdsfffffffffffffffffffffffffffffffffffff

基于小波包特征向量与神经网络的滚动轴承故障诊断。：基于故障轴承的特征提取，提出了将小波包分析与神经网络结合的滚动轴承故障诊断方法。对滚动轴承信号进行3层小波包分解，构造小波包特征向量作为故障样本，用训练好的BP神经网络进行故障诊断，试验结果表明，该方法能够有效地诊断出滚动轴承的故障类型。(Fault Diagnosis of Rolling Bearings Based on W avelet Packet Energy Eigenvector and Neural Network)

基于核密度估计的层次聚类算法基于核密度估计的层次聚类算法(hierarchcial clustering with kernel density estimation)

设用7个短线段构成1,2,3,4,5,6,7,8,9,10共10个数码图形，令这7个线段分别用一个矢量 来代表，又设对数码图形中用到的线段，相应分量取值为1，未用到的线段相应的分量取值为0，因此每个数码图形分别可由一个矢量表示，其顺序编号为： ，试设计一神经网络，能够区分奇数码和偶数码。(Set with seven short line segments 1, 2, 3, 4, 5, 10 digital graphics, 7 segment represented by a vector, and digital graphics used in the segment, the corresponding component value of 1, and the unused segments corresponding component value of 0 can be a vector, respectively, so that each digital graphics, its sequence number is: trial design of a neural network, it is possible to distinguish the odd and even digital.)

别人的一个代码，看着还行。给大家看看。基本就是 看看设计的手法-C#

％ 奇异值分解 (sigular value decomposition,SVD) 是另一种正交矩阵分解法；SVD是最可靠的分解法， ％ 但是它比QR 分解法要花上近十倍的计算时间。[U,S,V]=svd(A)，其中U和V代表二个相互正交矩阵， ％ 而S代表一对角矩阵。 和QR分解法相同者， 原矩阵A不必为正方矩阵。 ％ 使用SVD分解法的用途是解最小平方误差法和数据压缩。用svd分解法解线性方程组，在Quke2中就用这个来计算图形信息，性能相当的好。在计算线性方程组时，一些不能分解的矩阵或者严重病态矩阵的线性方程都能很好的得到解( Singular value decomposition (sigular value decomposition, SVD) is another orthogonal matrix decomposition method SVD decomposition is the most reliable method, but it takes more than QR decomposition near ten times the computing time. [U, S, V] = svd (A), in which U and V on behalf of two mutually orthogonal matrix, and the S on behalf of a diagonal matrix. And QR decomposition are the same, the original matrix A is no need for the square matrix. The use of SVD decomposition method are used as a solution of least squares error method and data compression. Using SVD decomposition solution of linear equations, in Quke2 on to use this information to calculate the graphics performance quite good. In the calculation of linear equations, some indecomposable matrix or serious pathological matrix of linear equations can be a very good solution)

说明：  说明： 压缩BMP图片，图片质量有些损失，但压缩率很大(Explanation: Compressing BMP pictures may cause some loss of image quality, but the compression rate is very high.)

本书全面系统地讲述了软件工程的概念、原理和典型的方法学,并介绍了软件项目的管理技术。本书正文共13章,第1章是概述,第2章至第8章顺序讲述软件生命周期各阶段的任务、过程、结构化方法和工具,第9章至第12章分别讲述面向对象方法学引论、面向对象分析、面向对象设计和面向对象实现,第13章介绍软件项目管理。正文后面有两个附录,分别讲述了用面向对象方法和结构化方法开发两个软件的过程,对读者深入理解软件工程学很有帮助,也是上机实习的好材料。 本书可作为高等院校“软件工程”课程的教材或教学参考书,也可供有一定实际经验的软件工作人员和需要开发应用软件的广大计算机用户阅读参考。 (This book comprehensively and systematically describes the concept of software engineering principles and typical methodologies, and introduces software project management techniques. Text book of 13 chapters, Chapter 1 is an overview, Chapter 2 to Chapter 8 describes the sequence of the various stages of the software lifecycle tasks, processes, methods and tools for structured, Chapter 9 to Chapter 12 were about object-oriented approach Introduction to science, object-oriented analysis, object-oriented design and object-oriented implementation, Chapter 13 describes software project management. There are two appendices behind the text, were about object-oriented methods and structured approach to develop two software products, for readers in-depth understanding of software engineering is helpful, but also on the local practice good material. This book can serve as institutions of higher learning " software engineering" course materials or teaching reference books are also avai)

三电平逆变器SVPWM控制，不带解耦功能，谐波含有百分之5(Three level inverter SVPWM control without decoupling.)