Consider an ideal series PID controller with parameters Kc, Ta and Tb (satisfying some closed loop requirements) written as

PID1(s)=Kc(1+Tas)(1+Tbs)/(Tas)=

=Kc(1+1/(Tas))(1+Tbs)

It means that Ta may be interpreted as the integral time constant Ti and Tb as the derivative time constant TD.

In the linear range of control it is obviously equivalent to the controller

PID2(s)=Kc'(1+Tas)(1+Tbs)/(Tbs)=

=Kc'(1+1/(Tbs))(1+Tas)

with Kc'=KcTb/Ta, Ti=Tb, TD=Ta.

Importance of this ambiguity will be shown in constrained control. Do you know about some papers dealing with this topic?

More Mikulas Huba's questions See All
Similar questions and discussions