matlab-ECG信号自适应滤波处理去噪，对比LMS,NLMS以及RLS三种自适应滤波算法-源码

clc; clear; close all; warning off; addpath(genpath(pwd)); load('ECG_database.mat'); fs=1000; d1=Org_Data1(2,:)/200; d2=Org_Data2(2,:)/200; d3=Org_Data3(2,:)/200; noise_wn = wn/10; noise_bwn = Org_bw_noise(1,1:5000)/200; noise_emn = Org_em_noise(1,1:5000)/200; noise_man = Org_ma_noise(1,1:5000)/200; noise_pli = 0.1*sin(2*pi*60*(0:length(d1)-1)/fs); x1=d1+noise_wn; x2=d1+noise_bwn; x3=d2+noise_emn; x4=d1+noise_man; x5=d1+noise_pli; lamda = 1; %RLS mu = 0.02; %LMS a=0.1; %NLMS mu2 = 0.02; %for NLMS M1 =2;%LMS M2 =2;%NLMS M3 =2;%RLS %----for white noise---% [e1_wn,y1_wn,w1_wn] = myLMS(x1,noise_wn, mu, M1); [e2_wn,y2_wn,w2_wn] = myNLMS(x1, noise_wn, mu2, M2, a); [e3_wn,y3_wn,w3_wn] = myRLS(x1,noise_wn,lamda, M3); figure; subplot(411); plot(x1); title('ECG corrupted by White Gaussian Noise'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(412); plot(e1_wn); title('LMS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(413); plot(e2_wn); title('NLMS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(414); plot(e3_wn); title('RLS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid figure; plot(y1_wn'-noise_wn,'r'); hold on; plot(y2_wn'-noise_wn,'b'); hold on; plot(y3_wn'-noise_wn,'g'); hold on; title('Error convergence curve of three algorithms for WN'); xlabel('Samples (n)'); ylabel('Error Magnitude'); legend('LMS','NLMS','RLS'); %-----for bwn---% [e1_bwn,y1_bwn,w1_bwn] = myLMS(x2,noise_bwn, mu, M1); [e2_bwn,y2_bwn,w2_bwn] = myNLMS(x2, noise_bwn, mu2, M2, a); [e3_bwn,y3_bwn,w3_bwn] = myRLS(x2,noise_bwn,lamda, M3); figure; subplot(411); plot(x2); title('ECG corrupted by Baseline Wander Noise'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(412); plot(e1_bwn); title('LMS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(413); plot(e2_bwn); title('NLMS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(414); plot(e3_bwn); title('RLS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid %-----for emn---% [e1_emn,y1_emn,w1_emn] = myLMS(x3,noise_emn, mu, M1); [e2_emn,y2_emn,w2_emn] = myNLMS(x3, noise_emn, mu2, M2, a); [e3_emn,y3_emn,w3_emn] = myRLS(x3,noise_emn,lamda, M3); figure; subplot(411); plot(x3); title('ECG corrupted by Electrode Motion Artifact'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(412); plot(e1_emn); title('LMS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(413); plot(e2_emn); title('NLMS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(414); plot(e3_emn); title('RLS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid % figure; % plot(y1_emn'-noise_emn,'r'); hold on; % plot(y2_emn'-noise_emn,'b'); hold on; % plot(y3_emn'-noise_emn,'g'); hold on; % title('Error convergence curve of three algorithms for EMN'); % xlabel('Samples (n)'); % ylabel('Error Magnitude'); % legend('LMS','NLMS','RLS'); %----for man----% [e1_man,y1_man,w1_man] = myLMS(x4,noise_man, mu, M1); [e2_man,y2_man,w2_man] = myNLMS(x4, noise_man, mu2, M2, a); [e3_man,y3_man,w3_man] = myRLS(x4,noise_man,lamda, M3); figure; subplot(411); plot(x4); title('ECG corrupted by Muscle Artifacts Noise'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(412); plot(e1_man); title('LMS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(413); plot(e2_man); title('NLMS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(414); plot(e3_man); title('RLS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid %----for PLI----% [e1_pli,y1_pli,w1_pli] = myLMS(x5,noise_pli, mu, M1); [e2_pli,y2_pli,w2_pli] = myNLMS(x5, noise_pli, mu2, M2, a); [e3_pli,y3_pli,w3_pli] = myRLS(x5,noise_pli,lamda, M3); figure; subplot(411); plot(x5); title('ECG corrupted by 60Hz PLI'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(412); plot(e1_pli); title('LMS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(413); plot(e2_pli); title('NLMS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid subplot(414); plot(e3_pli); title('RLS Filter Response'); xlabel('Samples (n)'); ylabel('Amplititude (mV)') ;grid %--error--% figure; plot(y1_pli'-noise_pli,'r'); hold on; plot(y2_pli'-noise_pli,'b'); hold on; plot(y3_pli'-noise_pli,'g'); hold on; title('Error convergence curve of three algorithms for 60Hz PLI'); xlabel('Samples (n)'); ylabel('Error Magnitude'); legend('LMS','NLMS','RLS'); figure; plot(d1); title('Clean ECG signal (1)'); xlabel('Samples (n)'); ylabel('Amplititude (mV)');