Timeline for Calculating numerator and denominator polynomials of a transfer function
Current License: CC BY-SA 4.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 7, 2021 at 14:15 | vote | accept | whoknowsmerida | ||
| Dec 7, 2021 at 14:09 | comment | added | ZR Han | I’ve already linked two hyperlinks for system identification using LMS algorithm and H1 estimator. If you need some further reading for IIR modeling of an FIR, you can have a look at the documentation of prony and stmcb in MATLAB, that will help and contains references. | |
| Dec 7, 2021 at 14:02 | comment | added | whoknowsmerida | thank you. You've been very helpful :) . Just to to finish, can you refer me to some reading material so I can learn more about the steps you mentioned in your first answer. I'd like to learn more about it. | |
| Dec 7, 2021 at 14:00 | comment | added | ZR Han | Yes I think the IFFT result is exactly the impulse response of your system. Now you can model it as an IIR filter using prony or stmcb to obtain the numerator and the denominator polynomials. | |
| Dec 7, 2021 at 13:41 | comment | added | whoknowsmerida | you are right. Sorry, what I meant was the impulse peak was at the first sample in the time domain after iFFT ( so I likened it to the DC component, which when I think about makes no sense. Sorry about that) . just to help me understand. probably that peak itself is the impulse, but what conclusion, if any, can I derive when the first sample is the peak of the impulse response. | |
| Dec 7, 2021 at 13:30 | comment | added | ZR Han | It seems that you are confused with impulse response $h(n)$ and frequency response $H(k)$, the former is in the time domain and therefore has no “DC component”, the latter is in the frequency domain. | |
| Dec 7, 2021 at 13:23 | comment | added | whoknowsmerida | I did plot it. There is a peak at the DC component. But the remainder was more or less just flat with no 'impulse' like phenomenon. | |
| Dec 7, 2021 at 10:33 | comment | added | ZR Han | Plot the impulse response. You should see something like an impulse. | |
| Dec 7, 2021 at 10:27 | comment | added | whoknowsmerida |
@ZR_Han took your advice and went with the H1 estimator approach, H_k = zeros(1,length(out_freq_data)); for ii=1:length(out_freq_data) num = out_freq_data(ii)*in_freq_data(ii)'; den = in_freq_data(ii)*in_freq_data(ii)'; H_k(ii) = num/den; end When i take the IFFT to check the impulse response (with linear gain or NO gain), the values of the impulse response are close to 0 , in the range of 10^-17 . is this expected have I done anything wrong with my implementation ?
|
|
| Dec 7, 2021 at 7:51 | history | answered | ZR Han | CC BY-SA 4.0 |