I want FFT plot of a motor data mat-file. It is of 10kHz frequency but I'm not getting the peaks of the plot at 50 Hz frequency. What corrections are needed? And how to write comments for the FFT plots drawn between frequency vs amplitude?

Question

Deepika 2013-10-2

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/88894-i-want-fft-plot-of-a-motor-data-mat-file-it-is-of-10khz-frequency-but-i-m-not-getting-the-peaks-of

评论： Ganavi Mg 2018-2-6

load t10k50sno.mat;
>> n = 1000;
>> ts =0.0001;
>> ws = 2*pi/ts;
>> f = fft(t10k50sno.Y(1).Data(1:1000));
>> fc=fftshift(f)*ts;
>> w = ws*(-n/2:(n/2)-1)/n;
>> plot(w,abs(fc))
>> xlabel('frequency');
>> ylabel('amplitude');
>> title('fast fourier transform of healthy motor at no-load condition');

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Ganavi Mg 2018-2-6

Hi Even I want fft plot of ppg data mat file. In your code specially in this line-f = fft(t10k50sno.Y(1).Data(1:1000));how you had choosen Y(1).

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Wayne King 2013-10-2

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/88894-i-want-fft-plot-of-a-motor-data-mat-file-it-is-of-10khz-frequency-but-i-m-not-getting-the-peaks-of#answer_98438

编辑：Wayne King 2013-10-2

在 MATLAB Online 中打开

Without your data it is difficult to say exactly, but your frequency vector is in radians/second, not cycles/second, so could you just not be looking for a peak at the correct frequency.

In radians/second, you should be getting peaks at (2*pi*50) radians/second (approx 314)

If you want you can just modify your example a little bit to see if that is the issue. Since I don't have your data, I have to make up a signal here.

n = 1000;
ts =0.0001;
ws = 2*pi/ts;
t = 0:ts:0.1-ts;
x = cos(2*pi*1000*t)+randn(size(t));
f = fft(x);
fc=fftshift(f)*ts;
w = ws*(-n/2:(n/2)-1)/n;
plot(w,abs(fc))
xlabel('frequency');
ylabel('amplitude');

The above plot is in radians/second. Now to convert to Hz.

n = 1000;
ts =0.0001;
ws = 2*pi/ts;
t = 0:ts:0.1-ts;
x = cos(2*pi*1000*t)+randn(size(t));
f = fft(x);
fc=fftshift(f)*ts;
w = ws*(-n/2:(n/2)-1)/n;
plot(w./(2*pi),abs(fc))
xlabel('frequency');
ylabel('amplitude');