LFLuo_1

Matlab implementation of the ES (extensive search algorithm ) used in motion estimation of macro blocks.

generating gold sequence

S4VM 对传统的S3VM 进行了改进。传统的S3VM 基于低密度假设，它试图找到一个低密度的分界线， 也就是更倾向于决策边界穿过特征空间的低密度区域。S4VM 和S3VM 的不同点在于，S3VM 试图把注意力 集中在一个最优的低密度分界线上，而S4VM 则同时关注多个可能的低密度分界线。之所以这样做，是因为 给定一些有标记的点和大量未标记的点，可能存在着不止一个“间隔”较大的低密度分界线（如图2 ），基于 有限的标记样本，很难决定哪个是最好的。虽然这些低密度分界线都与有限的标记样本吻合，但它们之间差 异很大，因此如果选择错了，会有一个很大的损失，导致性能下降。所以S4VM 试图考虑所有可能的低密度 分界线。 (Over the past decade, many improved versions are proposed to solve every different specific problem of semi-supervised support vector machines, for example: to the inefficiency issue, meanS3VM is introduced to performance degenerate problem, S4VM is proposed)

一种实现复矩阵的SVD分解的算法，使用Matlab进行仿真，亲测可用。(SVD decomposition algorithm to achieve complex matrix, using Matlab to simulate pro-test is available.)

说明：  利用遗传算法解决列车优化运行问题，为了简化问题考虑只有一个牵引和惰性的过程。(Genetic algorithm is used to solve the problem of train operation optimization. In order to simplify the problem, there is only one process of traction and inertia.)

五点差分法代码，可以自动运行的，可以试试(Five-point difference method code, can automatically run, you can try)

吴振杨数字信号处理 实验指导书 是学习matlab不错的入门资料(DSP WUZHENYANG )

二维和三维的离散双树小波正变换和逆变换，具有多方向分辨特性(Doubletree two-dimensional and three-dimensional discrete wavelet transform and inverse transform is, to distinguish features with multi-directional)

小波变换进行变速箱故障诊断 首先将s1和s2信号时域谱（s1代表维修前的振动信号，s2代表维修后的振动信号），画出来进行对比，会发现s1信号的振动幅值会比较高；然后对s1和s2分别进行小波分解并画出各层的时域谱，只对比s1和s2分后的最高时域谱，s2更加的具有周期性，且幅值较小，s2周期性不够明显，且振动过于密集。分析完时域谱后开始分析频谱，首先对原信号s1和s2fft转换，得到频谱图，如图，然后再对小波分析后高频部分进行fft转换，做出频谱图。对比后我们发现，在f 2375HZ出（也就是齿轮的啮合频率），s1信号频谱幅值会有急剧的升高，然后，其谐波的幅值也会增大。所以可以推断出该维修前的增速箱的齿轮可能出现问题。(Gearbox fault diagnosis based on wavelet transform s1 and s2 first time domain signal spectrum (s1 behalf of the vibration signal before maintenance, vibration signal s2 representative of repaired), drawn comparison, you will find a signal s1 vibration amplitude will be higher then s1 and s2 respectively wavelet decomposition and draw the layers of time-domain spectrum to only a maximum time domain spectroscopy s1 and s2 points after, s2 is more cyclical and smaller magnitude, s2 cyclical enough obviously, and the vibration is too dense. End domain spectral analysis after the beginning of the spectrum analysis, the first of the original signal s1 and s2fft conversion to give the spectrum, as shown, and then after the high frequency part of wavelet analysis fft transformation to make spectrum. We found that after the comparison, at f 2375HZ out (i.e. the gear mesh frequency), there will be a sharp increase in the amplitude spectrum of the signal s1, and then, the amplitude of the ha)

一个仿真的论文，关于svpwm的异步电动机调速(A simulation of the papers, on the SVPWM induction motor speed control)