Question about an LTI System and its Frequency Response

Click For Summary
SUMMARY

This discussion centers on the challenges of analyzing a nonlinear system mistakenly treated as a Linear Time-Invariant (LTI) system. The user attempts to find the impulse response h(t) using MATLAB's Fast Fourier Transform (FFT) on sinusoidal inputs and outputs. The key issue identified is that the output frequencies do not match the input frequencies, violating the fundamental properties of LTI systems. The advice given emphasizes the importance of understanding LTI systems and generating appropriate impulse responses for accurate analysis.

PREREQUISITES
  • Understanding of Linear Time-Invariant (LTI) systems
  • Familiarity with MATLAB, specifically its Fast Fourier Transform (FFT) and convolution functions
  • Knowledge of sinusoidal functions and their properties in signal processing
  • Basic concepts of frequency response and impulse response
NEXT STEPS
  • Study the properties of Linear Time-Invariant (LTI) systems in detail
  • Learn how to correctly apply MATLAB's convolution function for signal analysis
  • Explore the derivation and significance of impulse responses in system analysis
  • Review textbooks or online resources focused on signal processing fundamentals
USEFUL FOR

Students and researchers in signal processing, engineers working with LTI systems, and anyone involved in analyzing system responses using MATLAB.

ntran26
Messages
2
Reaction score
0
Hello All,
I am trying to learn about lti systems and what not for my research work but before I can apply my research I have to understand the fundamentals, which is what this question is kind of about!

I want to find h(t), the impulse response of an LTI system where my input and output are sinusoids. So I have made two functions.
Input(t)=4*SIN(3*t+PI/8)+9*SIN(4*t-PI()/6)
Output(t)=2*SIN(8*t+PI/3)+5*SIN(2*t-PI()/7)
I did use matlab's fast Fourier transform to bring results of n^10 data points to the frequency domain.
I then do output(w)/input(w) ,output and input in frequency domain, to get H(w), the frequency response. I then use MATLAB to apply the inverse fast Fourier transform to H(w) to obtain h(t).

I use matlabs convolution function which should result in I(t)*h(t)=o(t)

The problem is that this is not matching up.

Am i doing something wrong with my thought process? Thanks!
 
Engineering news on Phys.org
Your system is nonlinear, so it is not LTI. In particular, the output frequencies of a linear system must match those of the input (another way of saying that sines and cosines are the eigenfunctions of an LTI system).

It is invalid to arbitrarily make up an output function. I suggest you read and understand LTI systems, then you can generate an appropriate impulse response and watch what it does to your input signal. (Alternately, work the exercises in your textbook...)
 
marcusl said:
Your system is nonlinear, so it is not LTI. In particular, the output frequencies of a linear system must match those of the input (another way of saying that sines and cosines are the eigenfunctions of an LTI system).

It is invalid to arbitrarily make up an output function. I suggest you read and understand LTI systems, then you can generate an appropriate impulse response and watch what it does to your input signal. (Alternately, work the exercises in your textbook...)
I see, thank you! Unfortunately, I do not have any textbooks on this subject and have just been using online resources. I am using a reservoir simulator, where I apply a pressure pulse that is sinusoidal and observing pressures at another location. The reservoir is supposed to be an LTI system, but I cannot seem to get the convolution correctly. Thank you for your help!
 

Similar threads

  • · Replies 5 ·
Replies
5
Views
4K
  • · Replies 7 ·
Replies
7
Views
3K
  • · Replies 1 ·
Replies
1
Views
4K
  • · Replies 2 ·
Replies
2
Views
5K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 4 ·
Replies
4
Views
4K
  • · Replies 4 ·
Replies
4
Views
2K
Replies
10
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K