yichuan

功率谱估计的应用范围很广，在各学科和应用领域中受到了极大的重视。在《现代信号处理》课程中讲述了经典谱估计和现代谱估计这两大类谱估计方法；经典谱估计是基于傅立叶变换的，虽然具有运算效率高的优点，但是频谱分辨率低同时旁瓣泄漏严重，对长序列有着良好的估计。为了克服经典谱估计的缺点，人们开展了对现代谱估计方法的研究。现代谱估计是以随机过程的参数模型为基础的，有最大似然估计法、最大熵法、AR模型法、预测滤波器法。现代谱估计对短序列的估计精度高，同经典谱估计互为补充。在认真学习了现 代谱估计方法后，我选择了现代谱估计中的AR模型法的仿真作为题目。下面给出AR模型的相关理论和仿真实现。(Power Spectral Estimation of very extensive, in all disciplines and fields of application of a great deal of attention. The "modern signal processing" on the curriculum of classical and modern spectral estimation spectrum is estimated that the two types of spectrum estimation method; Classical spectrum estimation is based on Fourier transform, although high computing efficiency advantages, but also low-resolution spectrum Sidelobe serious leakage of long sequences have good estimates. In order to overcome the classic shortcomings of the spectrum estimation, there have been a pair of modern spectral estimation methods. Modern spectral estimation is the random process model parameters based on the maximum likelihood estimation, maximum entropy method, AR model, forecast filter. Spe)

语音增强算法，基于LsA-MMSE方法，并对其先验信噪比估计器进行了修改，具体过程参看Isal的论文。语音间隙间的音乐噪声也消除了。(speech enhancement algorithms, based on LsA- MMSE, and signal-to-noise ratio of its prior estimate for the changes, See specific process Isal theses. Voice gap between the music also eliminate the noise.)

This program is used for face recognition

1、 生成多周期线性调频信号，并对其进行频谱分析； 2、 对仿真生成的信号利用两种窗口函数进行STFT变换生成时频分析图(1, to generate multi-cycle chirp signal, and its spectrum analysis 2, the signal generated by the simulation window function using two STFT transform to generate time-frequency analysis chart)

对标幺化感应电机的控制，依次分块地对磁通，转矩，转矩设计控制器，并对模型简化和未简化的情况做对比，适合于对感应电机控制各个环节的控制有更深的理解。(Induction motor control, in turn block the flux, torque, torque design of the controller, and the model is simplified and not simplified contrast, is suitable for the control of all aspects of a deeper understanding of the induction motor control. The outcome of the study in the experimental class)

说明：  单纯形法 主要进行最优求解，环境为MATLAB.以一个例子进行简单说明。(Simplex method for the optimal solution of the main, the environment for MATLAB. To an example of a simple description.)

说明：  This program is of optimization in Cognitive radio in spectrum sensing

本程序实现了基于相关矩阵法的LTE及4G中MIMO多径信道的仿真，最终获得了MIMO信道的传播矩阵。（第一步，选择应用场景（确定信道参数和天线配置参数）；第二步，计算相关矩阵，并做Cholesky分解；第三步,利用Jakes模型产生多普勒随机变量；第四步,形成TDL；第五步，根据前四步准备，最终得到MIMO多径信道的信道传输矩阵）(This procedure achieved based on the correlation matrix method and 4 g LTE MIMO in multipath channel simulation, finally obtained the spread of the MIMO channel matrix.)

说明：  这个程序用于语音段的过零率求取，基于matlab，可以用来做端点检测。(The program segment for voice strike zero rate, based on matlab, can be used for endpoint detection.)

极值学习机，一种最近流行的算法。它的使用，使得计算过程大大的简化，值得交流分享。(Extreme Learning Machine, a recently popular algorithms. Its use, so that the calculation process is greatly simplified, worth sharing.)