How to do filtering using convolution and impulse response ?

Question

Guruprasad Madhale Jadav 2016-12-26

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/318239-how-to-do-filtering-using-convolution-and-impulse-response

Original.mat

I want to replicate a filters output by using convolution of the original signal with the impulse response generated by the same filter.Attached is the original data, now when I load the data and run the below filter script we should get a variable called filtered. If i replace the original data with just zero's in all the rows and columns and insert 1's for example in 12000th column, I am hoping that I am creating an impulse. Now if I give this impulse to the filter I should get impulse response. Well my question is when i use convolution as below

Output = conv2(Original,ir,'same');

The Output is not same as the filtered original signal. Can you please tell where am I doing wrong ? How to get the correct result ? Below is the script for filtering. Thank you in advance for helping me out.

clc
Fs = 128;                    % sample rate in Hz
N = 24031;                     % number of signal samples
rng default;
for channel = 1:14
    sig=Original(channel, :);
    xsig = sig';
    tv = (0:N-1)/Fs;              % time vector
    Fnorm = 5/(Fs/2);           % Normalized frequency
%    df = designfilt('bandpassiir','FilterOrder',4, ...
%         'PassbandRipple',0.500000,'CutoffFrequency1',1.000000,'CutoffFrequency2',20.000000, ...
%         'SampleRate',Fs);
    df = designfilt('bandpassiir', 'FilterOrder', 4, 'PassbandFrequency1', ...
                1, 'PassbandFrequency2', 20, 'PassbandRipple', ...
                0.500000, 'SampleRate', Fs);
    grpdelay(df,2048,Fs);   % plot group delay
    %BPF = fvtool(df);
    D = mean(grpdelay(df)); % filter delay in samples
    ysig = filter(df,[xsig; zeros(round(D),1)]); % Append D zeros to the input data
    ysig = ysig(D+1:end);
    filtered(channel, :) = ysig'    
end