exa130402.m

此程序是关于小波变换用于故障信号分析和处理的程序。

%------------------------------------------------------------------- 
%    exa130402, De-noising by wavelet
%  本程序先生成一个“Blocks”信号，然后加上高斯白噪声，再用'db3'小波进行分解，
%  以说明三种取阈值方法和三种噪声标志对去噪性能的影响。
%  注：在该程序中，用到了子程序 GWNoisy.m，
%      该程序请读者在如下的网站上下载:
%                      http://www-stat.stanford.edu/~wavelab/
%      因为该网站上的小波分析软件并没有列入MATLAB的工具箱，涉及到知识产权问题，
%      因此不能将其列入国内公开出版物上，但读者可以自由下载。
%-------------------------------------------------------------------
clear all;

% 生成“Blocks”信号
M = 1/1024;
N = 1/M;
t =0:M:1-1/1024;
TPoint = [0.1,0.13,0.15,0.23,0.25,0.40,0.44,0.65,0.76,0.78,0.81];
h = [4,-5,3,-4,5,-4.2,2.1,4.3,-3.1,2.1,-4.2];
[aaa,TSize] = size(TPoint);
FBlock = zeros(N,1);
for j = 1:N
    for i = 1:TSize
        if (t(j) - TPoint(i))> 0
            sgnt = 1;
        elseif (t(j) - TPoint(i))< 0
            sgnt  = -1;
        else
            sgnt =0;
        end
        FBlock(j) = FBlock(j) +h(i)*(1+sgnt)/2;
    end
end
%sig=FBlock';
%subplot(321)
%plot(FBlock)
%title(' 原信号');  
%grid on;

% 加上白噪声
sigma=0.2;
sig= GWNoisy(FBlock',sigma);
%subplot(323)
%plot(sig)
%title(' 带噪信号');  
%grid on;

figure(1);
subplot(331), plot(sig);
%axis tight;
%title('the initial signal');

% 小波分解,分别使用不同的阈值选择方法和噪声标志
[xds,cxds,lxds]=wden(sig,'sqtwolog','s','one',4,'db3');
subplot(332);plot(xds)
%title('sqtwolong');
[xds,cxds,lxds]=wden(sig,'sqtwolog','s','mln',4,'db3');
subplot(333);plot(xds)
%title('sqtwolong');

[xds,cxds,lxds]=wden(sig,'rigrsure','s','one',4,'db3');
subplot(334);plot(xds)
%title('rigrsure');
[xds,cxds,lxds]=wden(sig,'rigrsure','s','sln',4,'db3');
subplot(335);plot(xds)
%title('rigrsure');
[xds,cxds,lxds]=wden(sig,'rigrsure','s','mln',4,'db3');
subplot(336);plot(xds)
%title('rigrsure');

[xds,cxds,lxds]=wden(sig,'heursure','s','one',4,'db3');
subplot(337);plot(xds)
%title('heursure');
[xds,cxds,lxds]=wden(sig,'heursure','s','sln',4,'db3');
subplot(338);plot(xds)
%title('heursure');
[xds,cxds,lxds]=wden(sig,'heursure','s','mln',4,'db3');
subplot(339);plot(xds)
%title('heursure');