WIDETCTDOA

求两个信号的时延估计，通过时延估计对声源实现定位(Ask the two signal delay estimation, time delay estimation of sound source by positioning to achieve)

本人根据卡尔曼滤波原理，编写的Kalman小程序，简单可用！(I based on the principle of Kalman filtering, Kalman write small programs, simply available!)

深度学习工具箱的matlab程序，包括NN,CNN,CAE,SAE,DBN等多种深度学习算法。 (It s a deep learning toolbox of matlab, including NN, CNN, CAE, SAE, DBN and other deep learning algorithms.)

光伏并网系统单相逆变器电压前馈电流控制，基于matlab 7.3 。(photovotaic grid-connected system single-phase inverter voltage feed forward current control,based on matlab 7.3)

pwm整流器的建模仿真，通过matlab代码，信号维数的估计，构成不同频率的调制信号，从先验概率中采样，计算权重，用MATLAB实现的压缩传感。( Modeling and simulation pwm rectifier By matlab code, Signal dimension estimates, Constituting the modulated signals of different frequencies, Sampling a priori probability, calculate the weight, Using MATLAB compressed sensing.)

基于非线性模型的多公共点二维坐标转换，在测绘工程上适用性很强(Based on the nonlinear model of the public more two-dimensional coordinate transformation, on the surveying and mapping engineering applicability is very strong)

MATLAB 牛顿下山法的一个小程序，原理就不详细说了。因牛顿迭代法受初值选取的限制，为防止迭代发散，对迭代过程再附加一项要求:|f(x(k+1))|<|f(x(k))|,将牛顿法迭代的结果:x(k+1) =x(k)-f(x(k))/f (x(k))和前一近似值x(k)适当加权平均做为新的改进值:x(k+1)=u*x(k+1) +(1-u)*x(k),其中u(0<> 迭代时u取1开始，逐次减半计算，直至附加要求符合为止。实例计算中x(k)=x(0)不变，只更新u和x(k+1)，直至:|f(x(k+1))|<|f(x(k))|（即|f(x(1))|<|f(x(0))|)，然后更新下山因子为u=1,继续以牛顿法迭代。(MATLAB Newton-down a small program, the principle is not explained in detail. Newton iteration by the initial value of the selected limit, in order to prevent additional iterative divergence, the iterative process a request: | f (x (k+1)) | < | f (x (k)) | Newton iteration Results: x (k+1) ' = X (k)-f (x (k))/f' (x (k)) and a previous approximation x (k), appropriately weighted average as a new improved value: x (k+1 ) = u* x (k+1) ' + (1-u)* x (k), where u (0 < u> iteration take successive calculated at 50 until the additional requirements found so far. examples of calculation of X ( k) = x (0) unchanged profile u and x (k+1, hungry) until: | f (x (k+1)) | < | f (x (k)) | (i.e. | f (x (1)) | < | f (x (0)) |), then profile downhill factor for u = 1, to continue to Newton iteration.)

灰色预测是就灰色系统所做的预测。所谓灰色系统是介于白色系统和黑箱系统之间的过渡系统。其具体的含义是:如果某一系统的全部信息已知为白色系统，全部信息未知为黑箱系统，部分信息已知，部分信息未知，那么这一系统就是灰色系统。(Grey prediction is the prediction of grey system. The so-called grey system is a transitional system between the white system and the black box system. Its concrete meaning is: if all information of a system is known as a white system, all information is unknown as a black box system, some information is known, and some information is unknown, then this system is a grey system.)

说明：  非常有用的源代码，已经通过调试，可以放心使用asasa(A very useful source code has been through the debugger, you can rest assured that the use of asasa)

用dtw方法做的语音识别程序，可以识别简单的单个数字(dtw methods using the voice recognition procedure, a simple recognition of individual figures)