How to use FFT on my acceleration data in Matlab?

Hello, in a bit of pinch here. I have a single-axis acceleration dataset collected at 100 Hz and wish to perform an FFT on it. The dataset contains one column of acceleration values and one column of time values. I've attached an example of this data in an excel workbook.

I am quite new to this. While I have a fundamental understanding of what FFT does, I can't figure out how to implement it using this data or if this is even possible/useful.

I would like to perform an FFT, filter out some frequencies, and replot the filtered data in the time domain. Any help would be greatly appreciated, thank you!

Answer 1

[Star Strider]

Try something like this —

T1 = readtable('https://www.mathworks.com/matlabcentral/answers/uploaded_files/1203773/example.xlsx', 'VariableNamingRule','preserve')

T1 = 244×2 table
    Acceleration Forward (cm/s²)    UTC Time
    _____________________________    ________

                 -39                      0  
                 -89                   0.01  
                -147                   0.02  
                -187                   0.03  
                -134                   0.04  
                -126                   0.05  
                -126                   0.06  
                 -57                   0.07  
                 -23                   0.08  
                   0                   0.09  
                  47                    0.1  
                   3                   0.11  
                  83                   0.12  
                  60                   0.13  
                  12                   0.14  
                  14                   0.15  

Acc = T1{:,1};
t = T1{:,2};
Fs = 1/(t(2)-t(1));
Fn = Fs/2;
L = size(t,1);
NFFT = 2^nextpow2(L);
FTAcc = fft(Acc-mean(Acc),NFFT)/L;
Fv = linspace(0, 1, NFFT/2+1)*Fn;
Iv = 1:numel(Fv);

figure
plot(Fv, abs(FTAcc(Iv))*2)
grid
xlabel('Frequency')
ylabel('Maagnitude')
title('Fourier Transform of Acceleration Data')

Your intent to ‘filter out some frequencies’ can be done in a relatively straightforward way.

There are a number of filtering functions such as bandstop (and related functions) that you can use for this, although the signal appears to be relatively ‘clean’. It would be nice to have more details.

.

[Sam Osborne]

Thank you very much. This is extremely helpful. A couple of questions. Why are the Fv and Iv calculations necessary?

[Star Strider]

My pleasure.

They are there for both necessity and convenience. The ‘Fv’ vector is the vector of frequencies (for the plot and other purposes if necesary, since this is a one-sided Fourier transform), and ‘Iv’ is the index vector into ‘FTAcc’ corresponding to them..

[Sam Osborne]

I sincerely appreciate your kind and explicit help. I was able to apply your solution (now that I understand it) to a separate dataset and successfully dampen the signal to my liking. Your support jump started my learning moreso than any video.

[Star Strider]

As always, my pleasure!

I very much appreciate your compliment!