7 Dead Time Compensator
7 Dead Time Compensator
Smith predictor
Date
Theory:
The Smith predictor is a model-based controller that is effective for processes with long
dead time. Figure 1 illustrates the block diagram representation of Smith predictor. It has an
inner loop with a main controller that can be simply designed without the dead time. The effects
of load disturbance and modeling error are corrected through an outer loop. The Smith predictor
can also be used for processes with significant non-minimum phase dynamics and for high
order systems that exhibit apparent dead time.
Reducing the blocks using block diagram reduction technique results in,
It can be noted from the closed loop transfer function given in (1) that, using the Smith
predictor, the effect of time delay is eliminated in the characteristic equation.
Problem statement:
Certain three identical tank systems connected in non-interacting mode results in following
transfer function.
1
𝐺(𝑠) =
(𝑠 + 1)3
The plant is controlled by a PID controller whose controller gains are Kp=Ki=Kd=1. Design a
Smith predictor for the above plant which is subjected to a dead time of 1s and compare the
performance of the system with and without Smith predictor.
Inference
As visualized in the scope, the smith predictor was successfully able to compensate for the time
delay introduced by dead time