小波包重构EEG的α信号和β信号

Question

我对采样率为128HZ的EEG信号进行5层小波包分解，重构得到α信号（8-12HZ)，但是频域显示是10-20HZ，好久了，一直解决不了.

%% 8-12HZ重构信号  
nodes=[31;32;33;34;35;36;37;38;39;40;41;42;43;44;45;46;47;48;49;50;51;52;53;54;55;56;57;58;59;60;61;62];   %第5层的节点号
ord=wpfrqord(nodes);  %小波包系数重排，ord是重排后小波包系数索引构成的矩阵　
nodes_ord=nodes(ord); %重排后的小波系数
firstChanneldata = EEG.data(1,:);
x = firstChanneldata;
t = wpdec(x,5,'dmey');
% cfs31=wpcoef(t,nodes_ord(1));
% cfs32=wpcoef(t,nodes_ord(2));
% cfs33=wpcoef(t,nodes_ord(3));
% cfs34=wpcoef(t,nodes_ord(4));
cfs35=wpcoef(t,nodes_ord(5));
cfs36=wpcoef(t,nodes_ord(6));
% cfs37=wpcoef(t,nodes_ord(7));
% cfs38=wpcoef(t,nodes_ord(8));
% cfs39=wpcoef(t,nodes_ord(9));
% cfs40=wpcoef(t,nodes_ord(10));
% cfs41=wpcoef(t,nodes_ord(11));
% cfs42=wpcoef(t,nodes_ord(12));
% cfs43=wpcoef(t,nodes_ord(13));
% cfs44=wpcoef(t,nodes_ord(14));
% cfs45=wpcoef(t,nodes_ord(15));
% cfs46=wpcoef(t,nodes_ord(16));
% cfs47=wpcoef(t,nodes_ord(17));
% cfs48=wpcoef(t,nodes_ord(18));
% cfs49=wpcoef(t,nodes_ord(19));
% cfs50=wpcoef(t,nodes_ord(20));
% cfs51=wpcoef(t,nodes_ord(21));
% cfs52=wpcoef(t,nodes_ord(22));
% cfs53=wpcoef(t,nodes_ord(23));
% cfs54=wpcoef(t,nodes_ord(24));
% cfs55=wpcoef(t,nodes_ord(25));
% cfs56=wpcoef(t,nodes_ord(26));
% cfs57=wpcoef(t,nodes_ord(27));
% cfs58=wpcoef(t,nodes_ord(28));
% cfs59=wpcoef(t,nodes_ord(29));
% cfs60=wpcoef(t,nodes_ord(30));
% cfs61=wpcoef(t,nodes_ord(31));
% cfs62=wpcoef(t,nodes_ord(32));



sNod = read(t,'sizes',[31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62]);
cfs31=zeros(sNod(1,:));
cfs32=zeros(sNod(2,:));
cfs33=zeros(sNod(3,:));
cfs34=zeros(sNod(4,:));
% cfs35=zeros(sNod(5,:));
% cfs36=zeros(sNod(6,:));
cfs37=zeros(sNod(7,:));
cfs38=zeros(sNod(8,:));
cfs39=zeros(sNod(9,:));
cfs40=zeros(sNod(10,:));
cfs41=zeros(sNod(11,:));
cfs42=zeros(sNod(12,:));
cfs43=zeros(sNod(13,:));
cfs44=zeros(sNod(14,:));
cfs45=zeros(sNod(15,:));
cfs46=zeros(sNod(16,:));
cfs47=zeros(sNod(17,:));
cfs48=zeros(sNod(18,:));
cfs49=zeros(sNod(19,:));
cfs50=zeros(sNod(20,:));
cfs51=zeros(sNod(21,:));
cfs52=zeros(sNod(22,:));
cfs53=zeros(sNod(23,:));
cfs54=zeros(sNod(24,:));
cfs55=zeros(sNod(25,:));
cfs56=zeros(sNod(26,:));
cfs57=zeros(sNod(27,:));
cfs58=zeros(sNod(28,:));
cfs59=zeros(sNod(29,:));
cfs60=zeros(sNod(30,:));
cfs61=zeros(sNod(31,:));
cfs62=zeros(sNod(32,:));

t3=write(t,'cfs',31,cfs31,'cfs',32,cfs32,'cfs',33,cfs33,'cfs',34,cfs34,'cfs',35,cfs35,'cfs',36,cfs36,'cfs',37,cfs37,'cfs',38,cfs38,'cfs',39,cfs39,'cfs',40,cfs40,'cfs',41,cfs41,...
         'cfs',42,cfs42,'cfs',43,cfs43,'cfs',44,cfs44,'cfs',45,cfs45,'cfs',46,cfs46,'cfs',47,cfs47,'cfs',48,cfs48,'cfs',49,cfs49,'cfs',50,cfs50,'cfs',51,cfs51,'cfs',52,cfs52,'cfs',53,cfs53, ...
         'cfs',54,cfs54,'cfs',55,cfs55,'cfs',56,cfs56,'cfs',57,cfs57,'cfs',58,cfs58,'cfs',59,cfs59,'cfs',60,cfs60,'cfs',61,cfs61,'cfs',62,cfs62);
t4 = wprec(t3);
%  
% % plot(x,'b');
% % title('原始信号');
% % hold on
% % % subplot(2,1,2);
% % plot(t4,'r');
% % title('α信号波形');
% % 
% % % 查看频谱范围
signal=t4;       
fs=128;
N=length(signal); %采样点个数
signalFFT=abs(fft(signal,N));%真实的幅值
Y=2*signalFFT/N;
f=(0:N/2)*(fs/N);
figure;
plot(f,Y(1:N/2+1));
ylabel('amp'); 
xlabel('frequency');
title('输入信号的频谱');

Answer 1

第一，你的采样频率是128Hz，所以你的信号最高频率为64Hz。
第二，你进行5层小波包分解，节点为32个，每个节点是2Hz，你找对应那个节点进行重构，应该就可以了