how to design a FIR filter.

6 次查看(过去 30 天)
paul
paul 2011-5-12
评论: AJAY KUMAR 2014-5-29
DESING OF FIR FILTER WITH BOTH HAMMING AND BLACKMAN WINDOW METHOD

回答(2 个)

Emanuel
Emanuel 2011-5-12
Hello Paul,
I think it can help you!
__________________________________________________
function [b,a] = my_fir_emanuel(Ap,As,fc,fs,Fs)
close all
%This routine will be used to obtain the coefficient of a FIR filter.
%After entering the required parameters the type of filter (low pass,
%high pass, band-pass or band-stop) and the window function is
%automatically selected.
%For details on FIR filter specifications recommended reading the book
% E.C. Ifeachor, Digital Signal Processing - A Pratical Approach,
%Pearson Education, 2ª Edition, 2002.
%Especificações do Filtro FIR (FIR Filter Specifications
)
% fc = [fc1 fc2]
% fs = [fs1 fs1]
% Ap [1x1]
% As [1x1]
% Fs [1x1]
%Obs.: fc2 e fs2 apenas para filtros passa-faixa ou rejeita-faixa
Note: fc2 and fs2 only for band-pass filters or band-stop.
%Frequencia normalizada(Normalized Frequency)
fcn = fc/Fs;
fsn = fs/Fs;
f0n = 1000/Fs;
%Faixa de transição(Transition range)
df= abs(fc(1)-fs(1));
dfn= df/Fs;
%Seleciona a função janela (select de window function)
%
if As<=21 && Ap>=0.7416
N=ceil(0.9/dfn);
w=rectwin(N);%Função ajanelamento(Window Function)
disp('Rectangular Window Chosen')
elseif As<=44 && Ap>=0.0546
N=ceil(3.1/dfn) %a função ceil aredonda os coeficientes para mais ex: 3.2 aredonda para 4
w = hanning(N);%Função de Ajanelamento
disp('Hanning window chosen')
elseif As<=53 && Ap>=0.0194
N=ceil(3.3/dfn)
w = hamming(N);%Função de Ajanelamento
disp('Hamming window chosen')
elseif As<=75 && Ap>=0.0017
N=ceil(5.5/dfn)
w = blackman(N);%Função de Ajanelamento
disp('Blackman window chosen')
elseif As<90 && Ap>=0.000275
N=ceil(4.71/dfn)
w = kaiser(N);%Função de Ajanelamento
disp('Kaiser window chosen')
end
n= 1: (N-1)/2;
% Função resposta ao impulso ideal
% Selects the filter type
%Low-Pass Filter
if fc(1)<fs(1)&& length(fc)==1
hi0= 2*fcn(1);
hip = (2*fcn(1)*sin(2*pi*fcn(1)*n)./(2*pi*fcn(1)*n));
hin = fliplr(hip);
hi = [hin hi0 hip];
stem(hi)
disp('Low-Pass Filter selected')
%High-Pass Filter
elseif fc(1)>fs(1) && length(fc)==1
if length(fc)==2
hi0= 1-2*fcn(1);
hip = (-2*fcn(1)*sin(2*pi*fcn(1)*n)./(2*pi*fcn(1)*n));
hin = fliplr(hip);
hi = [hin hi0 hip];
stem(hi)
disp('High-Pass Filter Selected')
end
%Band-Stop Filter
elseif (length(fc)==2) && (length(fs)==2)
if fc(1)<fs(1)&& fc(2)>fs(2) && (length(fc)==2)
hi0= 1-2*(fcn(2)-fcn(1));
hip = (2*fcn(1)*sin(2*pi*fcn(1)*n)./(2*pi*fcn(1)*n))-(2*fcn(2)*sin(2*pi*fcn(2)*n)./(2*pi*fcn(2)*n));
hin = fliplr(hip);
hi = [hin hi0 hip];
stem(hi)
disp('Band-Stop Filter Selected')
end
%Band-Pass Filter
elseif (length(fc)==2) && (length(fs)==2)
if fc(1)>fs(1)&& fc(2)<fs(2) %&&
hi0= 2*(fcn(2)-fcn(1));
hip = (2*fcn(2)*sin(2*pi*fcn(2)*n)./(2*pi*fcn(2)*n))-(2*fcn(1)*sin(2*pi*fcn(1)*n)./(2*pi*fcn(1)*n));
hin = fliplr(hip);
hi = [hin hi0 hip];
stem(hi)
disp('Band-Pass Filter Selected')
end
disp(' ')
disp(' ')
elseif fc(1)>fs(1)& fc(2)> fs(2)
error('Impossível selecionar o tipo de filtro com os dados de entrada fc e fs.Digite: Help my_fir_emanuel para instruções')
elseif fc(1)==0 & fc(2)==1
error('Unable to select filters with the input data')
elseif fs(1)==0 & fs(2)==1
error('Unable to select filters with the input data')
end
%if size(hi,1)
% Desired impulse response function
if length(hi) == length(w)
b = hi.*w';
a=1;
figure(1)
stem(b)
figure(2)
freqz(b,a,1000,Fs);
%error('Impossível determinar a resposta em frequência para os parâmetros escolhidos')
end

Wayne King
Wayne King 2012-1-28
You can use fdesign
You do not give any information about your filter. Here is a lowpass filter with a cutoff frequency of 0.2 radians/sample and an order of 20.
d = fdesign.lowpass('N,Fc',20,0.2);
Hd1 = window(d,'window',@hamming);
Hd2 = window(d,'window',@blackman);

类别

Help CenterFile Exchange 中查找有关 Filter Analysis 的更多信息

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by