blind-sources-separation-

利用小波的对图像的分解和重构，然后对重构图像消噪，并比较消噪前后的图像。(the use of wavelet image decomposition and reconstruction, and then reconstructed image denoising, and compare the before and after denoising images.)

说明：  本代码描述的是有关灰狼算法的研究，这个代码有详细的内容说明与简介，网络上的代码普遍有错误，本人将所有的代码全部实现并改正了错误。供大家学习参考(This code describes the research on the gray wolf algorithm. This code has detailed description and introduction. The code on the network is generally wrong. I have implemented all the code and corrected the error. For everyone to learn reference)

以中等纯度的精馏塔为研究对象，考虑到不等分子溢流的影响和非理想的汽液平衡，可以得到塔顶产品轻组分含量Y与回流量L之间的传递函数为： 其中由于现场环境干扰，输出带有测量噪声是（0，1）的正态分布序列，它的方差为ɑ=0.5。由于输出中带有很大的噪声信号，故将数字滤波技术中常见的低通滤波器由于偏差控制，滤掉其中的噪声信号，然后在对其进行PID计算，得到实际的控制量。在该方法在噪声较强的环境下，可以得到较好的控制效果。 控制要求： 1、 采用带低通滤波器的增量式PID将塔顶轻组分含量控制在0.99 2、 用ISTE法整定PID参数 3、 在控制过程中，到150周期加入幅度为-20 的阶跃干扰，在第300周期干扰消失。分析PID参数的抗干扰性。 (Medium-purity distillation column, taking into account ranging from molecular overflow and non-ideal vapor-liquid equilibrium, the transfer function between the light component content of the top product Y and the back flow L can be obtained as: Because the field of environmental interference, output with the measurement noise is a normal distribution (0,1) sequence, its variance ɑ = 0.5. As with a lot of noise in the output signal, it will be common in low-pass filter in the digital filtering techniques due to the bias control, filter out the noise signal, then its PID, the actual amount of control. In the strong noise environment, you can get better control effect. Control requirements: 1, using the incremental PID with a low-pass filter to the top of the tower content of light components in the 0.99 2, the method of ISTE tuning PID parameters 3, in the control process, to 150 cycles join the range of-20 step disturbance, disappearance of interference in the first 300 c)

绘制一个位于坐标轴中的目标按钮，通过改变滑动条来移动目标按钮的位置，并同时显示它的坐标。(Draw a destination button located in the axis, and by changing the slider to move the position of the target button and its coordinates are displayed simultaneously.)

可视化的进度条，能够在程序运行期间实时查看代码运行的进度(Visualization of the progress bar)

wlan物理层仿真文件的说明，便于读者理解上手(Description wlan physical layer simulation files, easier to understand to get started)

说明：  Multirate Filtering for Digital Signal Processing

this is the active noise developed by running an active sysetem

灰色verhust模型的matlab程序在给定一组序列的情况下求得其预测值，残差，相对误差(Gray verhust model matlab program to obtain their predictive value, residuals, the relative error in the sequence of a given set of circumstances)

Solving TSP using Neural Networks