Calculate Wavelength from velocity FFT

Hello all,

I have multiple text files, first column positon that makes a line, 2nd is corresponding velocity at each point, third is time (repeated time for all rows so i take just the first row of each file)

I am using this code:

for i = 1:40

% check if file exists

filename = sprintf('data_%d.txt', i);

if exist(filename, 'file')

% read data from file

data = dlmread(filename, '\t');

velocity = data(:,2); % velocity data is in the second column

time = data(1,3); % time data is in the 3rd column

% perform FFT on velocity

fft_velocity = fft(velocity);

% find maximum peak frequency

[max_peak, max_peak_index] = max(abs(fft_velocity));

max_freq = max_peak_index;

% calculate wavelength

wavelength = 1/max_freq;

% save wavelength and time in a matrix

results = [results; wavelength, time];

last_wavelength = wavelength; % update last non-missing point

else

% file doesn't exist, use last non-missing point

results = [results; last_wavelength, time];

disp(sprintf('File %s does not exist, using last non-missing point...', filename))

end

end

The plot is not correct becuase of the values of amplitude

I have tried also this:

% perform FFT on velocity

[fft_position, fft_velocity] = fft(velocity);

% find maximum peak frequency

[max_peak, max_peak_index] = max(abs(fft_velocity));

max_freq = fft_position(max_peak_index);

But I am getting this error:

Too many output arguments.

Error in multiplelines (line 27)

[fft_position, fft_velocity] = fft(velocity);

How do I solve it, or better is the method I am using to find the wavelength correct?

Thanks

Answer 1

[Sulaymon Eshkabilov]

the Matlab fcn fft() computes the complex valued discrete Fourier Transform values. Thus, there is an error in your code.

To obtain frequency values you'd follow these steps. Note the lowest frequency value shows the longest wavelength and the highest frequency shows the shortest wavelegnth. Here is the simple example code showing these.

fs = 2000;    % Sampling frequency  
t =0:1/fs:5;  % Time 
F1 = 5;       % Lowest frequency
F2 = 17;      % Mid frequency
F3 = 55;      % Highest frequecy    
X = 3*sin(2*pi*t*F1)+ cos(2*pi*t*F2)+ 0.5*sin(2*pi*t*F3)+randn(size(t))*.5;
nexttile
plot(t, X)
xlim([0, 1])
xlabel('Time, [s]')
ylabel('Signal, x(t)')
N = 2^nextpow2(length(X));
Y = fft(X, N);               % FFT of the signal x(t)
freq = fs*(0:(N/2))/N;       % Freq values
Yp = abs(Y/N);               % Absolute value or magnitude of the signal spectrum
nexttile
plot(freq,Yp(1:N/2+1)) 
title('FFT analysis of a signal')
xlabel('Frequency, [Hz]')
ylabel('Magnitude of signal = |Y(f)|')
grid on
xlim([0, 65])

[Hussein Kokash]

Thank you for the reply, Can that be applied to the data detailed in the question?

[Sulaymon Eshkabilov]

Yes.