Question

The PID control architecture might be popular, but that does not mean it can be used for everything. This question explores when the PID controller is (in)sufficient: when it is too much, when it is "goldilocks", and when it is not enough. Consider a plant given by P'(s)-SaAs n increases, so does the order of the system. (sta)" Let's consider three orders and how PID controller design works for them (or does not) (a) First Order (a-l , n-l ): It has been said that PI control is adequate for processes where the dynamics is (essentially) first order. Please explain why this is the case. (b) Simple Second Order (а 0, n-2 ): It has been said that a double integrator can not be controlled by a PI controller. (1) Please explain why this is the case. (2) Also, could a PID controller be sufficient for this case? [Possible Hint: For part (1), if it helps, think about what might happen if a 1,n 2] (c) Third Order (a-l ,n-3 ): It has been said that when a process is more than second order, that a PID might be workable, but insufficient. Please explain why this might be the case (d) Delay-Dominated First Order: Control of systems with large time delays can be difficult.Consider a variation on P with large delay (TD,TD>a): Qs) It is has been said that for such systems that adding derivative action does not help much. Please explain why this might be the case.

Please explain how to do part (b), (c) and (d) only

Answer 1

Root ow

P-I comtrol action Ti.s Loed o accn er. --) Soit n Seon

→ Basically Dterm in ensd a-j S2 alus ays changiny 'So stead-stette eAaT 너