Scribd
Upload
78 views14 pages

Dead Time Compensator

Here are the key steps to solve this problem: (1) The open loop transfer function is: gOL(s) = kc e^-s / (0.5s + 1) The characteristic equation is: 1 + gOL(s) = 0 1 + kc e^-s / (0.5s + 1) = 0 Using the approximation e^-s ≈ 1 - s for low frequencies, the crossover frequency (ωc) and ultimate gain (Ku) can be found. (2) With a perfect Smith predictor, the effective process is just G(s) = kc / (0.5s + 1)

Uploaded by

Meena
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
78 views14 pages

Dead Time Compensator

Here are the key steps to solve this problem: (1) The open loop transfer function is: gOL(s) = kc e^-s / (0.5s + 1) The characteristic equation is: 1 + gOL(s) = 0 1 + kc e^-s / (0.5s + 1) = 0 Using the approximation e^-s ≈ 1 - s for low frequencies, the crossover frequency (ωc) and ultimate gain (Ku) can be found. (2) With a perfect Smith predictor, the effective process is just G(s) = kc / (0.5s + 1)

Uploaded by

Meena
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 14

Dead Time Compensator

Prof. Sneh. K. Soni


Assistant Professor (Inst. & Control)
Room No. D306 (Third Floor)
[email protected]
Complex Control Schemes

o Cascade Control
o Control for large dead time process
o Feedforward-Feedback Control
o Selective Control
• Override type
• Auctioneering type
o Ratio Control
o Split Range Control
o Anti Reset Windup
o Inferential Control
o Adaptive Control
o Inverse Response System Control
Dead Time Process

What is dead time process?


There are several reasons behind this delay to be added in the
control loop.
• Sensor is required to be installed away from the process
• Analyzers take some finite time to yield their measurement
results. It means that the output of the analyzer is actually
delayed information.
If the delay is added with the first order it will become First
Order Plus Dead Time (FOPDT)
 D s
ke
G (s) 
Tp s  1
Dead Time Process

Step Input Response of FOPDT process


Example of Dead Time Process
Example of Dead Time Process
Control of Dead Time Process
Control of Dead Time Process

Why it is difficult to control the dead time process?


Transfer function of the time delay part is

L[ y (t  t d )
G ( s)   e t d s
L[ y (t )]

If you consider the magnitude and phase equation of this


transfer function,

G ( jw)  1
  t d w
Control of Dead Time Process

k c =1.5
AR
kc e  Ds
gOL (s )  k c =8.5

0.5s  1 1
k c =1.5

Bode plot shows that


 ..as the process dead-time


increases, the crossover 
frequency decreases.
-180 o
 ..as the process dead-time D=1 D=0.1 D=0.01
increases, the ultimate gain
decreases. 2.3 17 160 
Dead Time Compensator (Smith Predictor)

Feedback control for process with dead time


Dead Time Compensator (Smith Predictor)

(s)=(1- )G(s)

y(s)= G(s)
Dead Time Compensator (Smith Predictor)

∗
𝑦 ( 𝑠 )=𝐺𝑐 ( 𝑠 ) 𝐺 ( 𝑠 ) 𝐸 (𝑠 )

Effective Control Loop


Performance of Smith Predictor

Performance of the smith predictor is largely depends upon the


accuracy of the model (Process + Dead Time)
Example

An open loop transfer function of a control system with dead


time as shown below:
-s
k e
gOL (s) = c
0.5s +1

(1) Find out the cross over frequency and ultimate gain of the
control loop.
(2) Prove that if the perfect smith predictor is implemented
than the ultimate gain will become infinite and such
system can never be unstable for any value of gain,

You might also like