matlabshipingchexiangtouduqu

求解各种时延微分方程的Matlab M文件，在一定程度上比SIMULINK更方便(Delay differential equations to solve a variety of Matlab M file, to a certain extent, more convenient than the SIMULINK)

主要是看如何产生矩形波，定义了矩形波的持续时间，采样点数(width points)

说明：  一个用于计算六方点阵的单光束干涉方法，MATLAB编写的(A six-dot matrix used to calculate the single-beam interference method, MATLAB prepared)

用matlab仿真滤波器，fs=1.ws=0.2,wp=0.3(Filter simulation using matlab, fs = 1.ws = 0.2, wp = 0.3)

给出了不确定广义系统的鲁棒预测控制仿真程序(Given the uncertainty of the generalized system robust predictive control simulation program)

说明：  进退法，也成成功--失败法，是最优化方法的一种，用于确定区间(Advance and retreat method, have become a success- failure of law is a kind of optimization method used to determine the range)

最速下降法和牛顿法解决最优化问题 dampnm.m - 牛顿法 grad.m - 最速下降法(The steepest descent method and Newton' s method to solve optimization problems dampnm.m- Newton grad.m- the steepest descent method)

论文使用一种经典的特征提取方法—主成分分析法（PCA）进行特征提取，其基本思想是降维。降维后的数据除了计算工作量减少之外不会减少原始数据所包含的有效信息量。(This paper use a classical method for feature extraction—Principal Component Analysis(PCA)with the basic idea of dimensionality reduction(it still contains all valid information).)

这是一本全面介绍MATLAB建模与仿真的书籍，很难找的，在学校图书馆找到的，赶紧下(This is a comprehensive introduction to modeling and simulation of MATLAB books, hard to find in the school library to find and quickly under)

gm2对上下变动的数据，预测效果比gm1好，但对趋势较好的数据，预测精度没有gm1高。(gm2 on the up and down movements of the data to predict better than gm1 good, but better data on trends, forecast accuracy not gm1 high)