模态分解用于旋转机械的故障诊断 EMD

说明：  有两个程序，用于研究Mallet算法及滤波器设计和消噪(are two procedures for research and Mallet filter algorithm design and denoising)

说明：  主要用于实现模糊滑模控制，不仅效果明显，而且代码简单有效(fuzzy sliding control)

免疫粒子群算法的matlab源代码，自己设置优化函数(Immune particle swarm algorithm matlab source code, set up their own optimization function)

matlab 程序代码 线性系统 龙格库塔法 可以解微分方程(matlab code linear system can solve differential equations Runge-Kutta method)

The code is available for direct input image and output image histogram code for reference. Environment: matlab

读入一幅256*256图像，将它分成8*8pixel的若干块，对每块做DCT变换，保留Z型顺序的若干变换系数，进行IDCT重构图像，计算重构图像的信噪比(Reading a 256* 256 image, it will be divided into several blocks of 8*8 pixel, keep each optimal Z transformation of sequence number, IDCT transform coefficient calculation reconstruction image reconstruction images,and the signal-to-noise ratio)

利用行像素点、列像素点识别分割字符成单个字符图像----letercog.m(The row of pixels, pixel columns into a single character recognition segmentation, image---- letercog. M)

ANISODIFF - Anisotropic diffusion. Scale-space and edge detection using anisotropic diffusion. (Matlab SourceCode) (ANISODIFF- Anisotropic diffusion. Scale-space and edge detection using anisotropic diffusion. (Matlab SourceCode) )

实现了三层介质的地震记录，雷克子波，卷积，采用均方根速度计算。(Implementation of the earthquake record in three layer medium, Ricker wavelet, convolution, using the root mean square velocity calculation.)

在工程应用中 经常会遇到一些复杂的非线性系统, 这些系统状态方程复杂, 难以 用数学方 法准确建模。 在这种情况下, 可以 建立 BP 神 经网 络表达这些 非 线 性系 统。 该方法把未 知 系 统看成是一个黑箱, 首先用 系 统 输入输出 数据 训 练 BP 神 经网 络, 使网 络能 够表达该未 知 函 数, 然后就可以 用训练好的 BP 神经网络预测系统输出 。 本章拟合的非线性函 数为 y = x1^2+x2^2. 。 (In engineering applications often encounter some complex nonlinear systems, which complicated the equation of state, can not be accurately modeled mathematically. In this case, BP neural network can be established to express these nonlinear systems. The method of the unknown system as a black box, the first input and output data with systematic training BP neural network, the network can express the unknown function, then you can use the trained BP neural network forecasting system output. Chapter for fitting non-linear function y = x1 ^ 2+x2 ^ 2..)