自適應(yīng)濾波的MATLAB實現(xiàn)(共6頁)

1、精選優(yōu)質(zhì)文檔-傾情為你奉上自適應(yīng)濾波的MATLAB實現(xiàn) 發(fā)表于: 2006-12-06 08:56 來源: 中國振動聯(lián)盟1.LMS算法的仿真程序：%lms 算法clear allclose allhold off%系統(tǒng)信道權(quán)數(shù)sysorder = 5 ;%抽頭數(shù)N=1000;%總采樣次數(shù)inp = randn(N,1);%產(chǎn)生高斯隨機系列n = randn(N,1);b,a = butter(2,0.25);Gz = tf(b,a,-1);%逆變換函數(shù)h= 0.0976;0.2873;0.3360;0.2210;0.0964;%信道特性向量y = lsim(Gz,inp);%加入噪聲n = n

2、 * std(y)/(10*std(n);%噪聲信號d = y + n;%期望輸出信號totallength=size(d,1);%步長N=60 ; %60節(jié)點作為訓(xùn)練序列%算法的開始w = zeros ( sysorder , 1 ) ;%初始化for n = sysorder : Nu = inp(n:-1:n-sysorder+1) ;% u的矩陣y(n)= w' * u;%系統(tǒng)輸出e(n) = d(n) - y(n) ;%誤差if n < 20mu=0.32;elsemu=0.15;endw = w + mu * u * e(n) ;%迭代方程end%檢驗結(jié)果for n

3、= N+1 : totallengthu = inp(n:-1:n-sysorder+1) ;y(n) = w' * u ;e(n) = d(n) - y(n) ;%誤差endhold onplot(d)plot(y,'r');title('系統(tǒng)輸出') ;xlabel('樣本')ylabel('實際輸出')figuresemilogy(abs(e) ;% e的絕對值坐標title('誤差曲線') ;xlabel('樣本')ylabel('誤差矢量')figure%作圖plo

4、t(h, 'k+')hold onplot(w, 'r*')legend('實際權(quán)矢量','估計權(quán)矢量')title('比較實際和估計權(quán)矢量') ;axis(0 6 0.05 0.35)2.NLMS算法的仿真程序：%lms 算法clear allclose allhold off%系統(tǒng)信道權(quán)數(shù)sysorder = 5 ;%抽頭數(shù)N=1000;%總采樣次數(shù)inp = randn(N,1);%產(chǎn)生高斯隨機系列n = randn(N,1);b,a = butter(2,0.25);Gz = tf(b,a,-1);%逆變換

5、函數(shù)h= 0.0976;0.2873;0.3360;0.2210;0.0964;%信道特性向量y = lsim(Gz,inp);%加入噪聲n = n * std(y)/(10*std(n);%噪聲信號d = y + n;%期望輸出信號totallength=size(d,1);%步長N=60 ; %60節(jié)點作為訓(xùn)練序列%算法的開始w = zeros ( sysorder , 1 ) ;%初始化for n = sysorder : Nu = inp(n:-1:n-sysorder+1) ;% u的矩陣y(n)= w' * u;%系統(tǒng)輸出r(n)=u'*u;%自相關(guān)矩陣e(n) =

6、 d(n) - y(n) ;%誤差fai=.0001;%修正參數(shù)，為防止u'*u過小導(dǎo)致步長值太大而設(shè)置的if n < 20mu=0.32;elsemu=0.15;endw = w + mu * u * e(n)/(r(n)+fai) ;%迭代方程end%檢驗結(jié)果for n = N+1 : totallengthu = inp(n:-1:n-sysorder+1) ;y(n) = w' * u ;e(n) = d(n) - y(n) ;%誤差endhold onplot(d)plot(y,'r');title('系統(tǒng)輸出') ;xlabel

7、('樣本')ylabel('實際輸出')figuresemilogy(abs(e) ;% e的絕對值坐標title('誤差曲線') ;xlabel('樣本')ylabel('誤差矢量')figure%作圖plot(h, 'k+')hold onplot(w, 'r*')legend('實際權(quán)矢量','估計權(quán)矢量')title('比較實際和估計權(quán)矢量') ;axis(0 6 0.05 0.35)3.RLS算法的仿真程序：% RLS算法ra

8、ndn('seed', 0) ;rand('seed', 0) ;NoOfData = 8000 ; % Set no of data points used for trainingOrder = 32 ; % 自適應(yīng)濾波權(quán)數(shù)Lambda = 0.98 ; % 遺忘因子Delta = 0.001 ; % 相關(guān)矩陣R的初始化x = randn(NoOfData, 1) ;%高斯隨機系列h = rand(Order, 1) ; % 系統(tǒng)隨機抽樣d = filter(h, 1, x) ; % 期望輸出% RLS算法的初始化P = Delta * eye ( Orde

9、r, Order ) ;%相關(guān)矩陣w = zeros ( Order, 1 ) ;%濾波系數(shù)矢量的初始化% RLS Adaptationfor n = Order : NoOfData ;u = x(n:-1:n-Order+1) ;%延時函數(shù)pi_ = u' * P ;%互相關(guān)函數(shù)k = Lambda + pi_ * u ;K = pi_'/k;%增益矢量e(n) = d(n) - w' * u ;%誤差函數(shù)w = w + K * e(n) ;%遞歸公式PPrime = K * pi_ ;P = ( P - PPrime ) / Lambda ;%誤差相關(guān)矩陣w_er

10、r(n) = norm(h - w) ;%真實估計誤差end ;% 作圖表示結(jié)果figure ;plot(20*log10(abs(e) ;%| e |的誤差曲線title('學習曲線') ;xlabel('迭代次數(shù)') ;ylabel('輸出誤差估計') ;figure ;semilogy(w_err) ;%作實際估計誤差圖title('矢量估計誤差') ;xlabel('迭代次數(shù)') ;ylabel('誤差權(quán)矢量') ;4.自適應(yīng)均衡器的仿真程序：% Illustration of the co

11、nventional RLS algorithmclose all;W=2.9;Nexp=10;N=2000;Nmc=1; % Number of ensemble realizationsM=11;%抽頭系數(shù)lambda=0.99;%遺忘因子varv=0.001;%噪聲方差h=zeros(3,1);%h的初始化er=zeros(N,Nmc);%er的初始化h(1)=0.5*(1+cos(2*pi*(1-2)/W);h(2)=0.5*(1+cos(2*pi*(2-2)/W);h(3)=0.5*(1+cos(2*pi*(3-2)/W);% 學習曲線hc=0 h(1) h(2) h(3)'

12、n0=7;t=(1:N)'for i=1:Nmcy=sign(rand(N,1)-0.5);%輸入信號v=sqrt(varv)*randn(N,1);%噪聲信號x=filter(hc,1,y)+v;%信號混合x=zeros(M-1,1);x;%x矩陣yd=zeros(N+M-1,1); %延遲信號初始化e=yd;yd(n0+M-1:N+M-1)=y(1:N-n0+1);% CRLS 算法% Initializationlambda=0.98;P=(10-3)*eye(M,M);c=zeros(M,1);g=c;glambda=g;% 迭代范圍for n=M:M+N-1xn=flipud

13、(x(n-M+1:n);glambda=P*xn;alphal=lambda+conj(glambda')*xn;g=glambda/lambda;a(n)=1-conj(g')*xn;P=(P-g*conj(glambda')/lambda;P=(P+P')/2;e(n)=yd(n)-conj(c')*xn;c=c+g*conj(e(n);endendeplot=e(M:M+N-1).2;subplot(2,1,1), plot(t,abs(eplot)ylabel('|e(n)|2');xlabel('n');subp

14、lot(2,1,2), plot(t,a(M:M+N-1);ylabel('alpha(n)');xlabel('n');5.自適應(yīng)陷波器的仿真程序：N=400; %總采樣長度t=0:N-1; %時間的變化范圍s=sin(2*pi*t/20); %輸入信號A=0.5; %干擾信號的幅值fai=pi/3;%干擾信號的相移n=A*cos(2*pi*t/10+fai);%干擾信號x=s+n;%信號混合subplot(2,2,1);%作第一子圖plot(t,s);subplot(2,2,2); %作第二子圖plot(t,x);x1=cos(2*pi*t/10);x2=sin(2*pi*t/1