Time delay creates a better feedback system??

Hey all,

I'm not sure if this is the right place to ask a question like this, but I'm studying control engineering and ran into a weird phenomenon while working with Simulink. If I enter a unit delay block in my feedback loop (controller is an observer) just before this enters my system/plant, performance improves (a lot less overshoot but some more oscillations).

Now my question is: Am I doing something wrong in Simulink or is it possible that entering a time delay can actually improve performance? It seems rather counterintuitive. The system that needs to be control is pretty unstable (1/(s^4+s^3+s^2) in CT).

Does anyone know if this is possible?

Thanks in advance!

Leon

Answer 1

[Azzi Abdelmalek]

This is not true. Adding an element hazardously can be beneficial or harmful. Adding a unit delay without any reason is not logical, but if your feedback is already bad, adding another error to your system, sometimes can be beneficial.

[Leon]

The thing is that I'm working on an exercise in which I need to evaluate several different possible controllers and now add a time delay to each of them. But could it be the case that, because my observer does not know the initial conditions of the system, inserting a time delay causes the "searching" of the observer state to be somewhat more balanced? Sorry if I'm not making too much sense, I'm still rather new to this field..

[Azzi Abdelmalek]

Your system is at the limit of instability (one unstable pole =0). Just make your controller, a PID for example, and try to find the appropriate parameters. If you a control system toobox, you can use a tunable PID controller