How to perform FFT analysis of a signal in MATLAB

YouTube Video

The content is basically the same, but we introduce a few details such as how to create data.

Please check this out as well!

The MATLAB manual is still confusing, so I will summarize it. You can paste the code together into the command window.

intended audience

  • Those who want to do FFT analysis in MATLAB
  • Prepare data in Excel

    As an example here, we have three sin curves in Excel, each with a different frequency.
    In this case, 1 Hz, 3 Hz, and 10 Hz were used.
    The signal in column E (yellow), which is the sum of the two, is the target of the analysis.

    Screenshot 2018 07 30 21 00 05

    Save the file as a CSV file, leaving only columns A and E.

    The formulas for each column are as follows.

    • Column B "=SIN(2*PI()*A2)"
    • Column C "=SIN(2*PI()*3*A2)"
    • Column D "=SIN(2*PI()*10*A2)"
    • Column E "=SUM(B2:D2)"

    You can copy and paste it directly into Excel. I will also upload the Excel file to my Github so you can use that as well.
    https://github.com/buono/MATLAB_FFT

    Data Import in MATLAB

    Import t and sum as arrays in MATLAB as follows, respectively.

    Screenshot 2018 07 30 21 05 22

    Screenshot 2018 07 30 21 06 23

    Set the signal and plot

    In this case, the following code was used to create signals up to 4 seconds at 0.002 second intervals.

    [crayon] T = 0.002; % sample period
    L = 1001; % Number of signal points (note that it is not time!)
    t = (0:L-1)*T; % time vector

    Y = fft(sum);
    P2 = abs(Y/L);.
    P1 = P2(1:L/2+1);.
    P1(2:end-1) = 2*P1(2:end-1);

    Fs = 1/T;.
    F = Fs*(0:(L/2))/L;.
    plot(F,P1);.
    [/crayon]

    As shown below, we see peaks at 1Hz, 3Hz, and 10Hz.

    Screenshot 2018 07 30 22 06 06