What effects does have a different input such as impulse, step, sine sweep etc. on performing a system identification?
I mean what is up/down side, and which one is best, and why chose one rather than the other.
For now I see the sine sweep as the most reliable for determining a system since the other ones doesn't seem to hit nearly all modes of a system.
But what do you say? I am not an expert at all and I am a bit confused here.
Both Impulse and Step Response can help identify a system, however it might be hard to actually perform the identification. You'd have to fourier transform the response. Basically, this is only possible using digital signal processing. Using a sine sweep, the result is easily read, but the sweep will take a relatively long time.
What effects does have a different input such as impulse, step, sine sweep etc. on performing a system identification?
These are not the only signals in use but it really depends on what you are trying to do.
Theoretically, exciting a Linear Time Invariant system with the unit impulse will result in its impulse response. In practice however, what would be applied as input to a system would be the unit pulse.
Exciting a system with the unit impulse will result in a time series called its "Impulse Response".
Exciting a system with a chirp and integrating its output will result, directly, to its frequency amplitude response.
Other typical signals in use are, the step function, the ramp and the parabolic. Exciting a system with these inputs and obtaining the outputs can lead to identification of various parameters. Very briefly, the step function is used to characterise the transient response of a system as well as its "position" error, the ramp is used to determine the "velocity" error and the parabolic is used to determine the "accelleration" error. For more information please see this link. For much more information it would be worth having a look at the books "Signals and Systems" (by Oppenheim, Willsky and Hamid-Nawab) and "Feedback Control of Dynamic Systems" (by Franklin, Powell and Emami-Naeini).
Hope this helps.
