Tutorial 4

EE3114
Systems and Control
Tutorial - 4
Semester A, 2018-19
2018-9-26
Recall: Standard Second-Order System
U(s) Y(s)
2
Y (s) n
U (s) s2 2 ns
2
n
Damping Ratio Un-Damped Natural Frequency

Exercise: (5 minutes)
U(s) Y(s)
!(#)
(1) =?
%(#)
(2) Assume that 𝑈 is the Unit Ramp Input,

and all poles are on LHP. Find:
lim 𝑒 𝑡 = ?
!→#
Answer:
U(s) Y(s)
!(#) #(#&'()!)
(1) =# "
%(#) #&'()! &*)!
(2) Assume that 𝑈 is the Unit Ramp Input,

and all poles are on LHP. Find:
&()*+! )*
lim 𝑒 𝑡 = lim 𝑠𝐸 𝑠 = lim & " = ,+ (Larger 𝐾 à
#→% &→' &→' &()*+! (,+! !Smaller error)
Return to:
U(s) Y(s)
2
Y (s) n
U ( s) s2 2 ns
2
n
Recall: Poles = − 𝜁𝜔% ± 𝜔% 𝜁 & − 1
1. If 0 ≤ 𝜁 < 1, complex poles: −𝜎 ± 𝑗𝜔!

where 𝜎 = 𝜁𝜔% and 𝜔' = 𝜔% 1 − 𝜁 &
2. If 𝜁 = 1, double poles: −𝜔"
3. If 𝜁 > 1, real poles: – ( 𝜁 ± 𝜁 & − 1 ) 𝜔%
Impulse Response
2
n
U(s) = 1 Y 2 2 2
(s n ) n (1 )
(! *,! 𝑠𝑖𝑛 𝜔 𝑡 , for 𝑡 ≥ 0
1. If 0 ≤ 𝜁 < 1, 𝑦 𝑡 = "
𝑒 '
)*+
2. For very low damping: 𝜁~0
à 𝜔' = 𝜔% 1 − 𝜁 &~𝜔% , 𝜎 = 𝜁𝜔% ~0 à oscillatory
"
(!
3. If 𝜁 = 1, 𝑌 𝑠 = " à 𝑦 𝑡 = 𝜔%& 𝑡 𝑒 *+(! !
-.+(!
𝜔+ = 𝜔, 1 − 𝜁 ' = 0
3. If 𝜁 > 1, over damped: (no oscillation ↑↓ )

1 *+( ! ( + " *) ! +" *) !
𝑦 𝑡 = 𝑒 ! 𝑒 ! − 𝑒 *(!
2𝜔% 𝜁 & − 1
Step Response
1) Under Damped: 0 ≤ 𝜁 < 1
Y(s)
Response
y(t)
Error 𝑒 𝑡 = 1 − 𝑦(𝑡)
e(t)
For 𝜁 = 0 : 𝜔! = 𝜔" 1 − 𝜁 # = 𝜔"

y(t)
In Lecture-4, we will further analyze the Unite Step Response
nt 2 2
y (t ) 1 e cos( n 1 t) sin( n 1 t)
2
1
Next, consider a related topic:

General Unity-Feedback Control Systems
• How well does the system output “follow” the reference?
• To obtain some general ideas by applying Step, Ramp, and

Parabolic reference inputs to the unity feedback system:
𝐺 𝑠
𝐶 𝑠 = 𝑅(𝑠)
1+𝐺 𝑠
First, classify systems into different types
• A stable feedback system is of Type-N, if
/(0! 123)⋯(0" 123)
𝐺 𝑠 = 1 #(0 123)⋯(0
$ % 123)
where 𝑇$ are distinct constants
• Example:
(123)(126)
𝐺 𝑠 = 1 &(127)(128) à Type-2
• As N increases, accuracy is improved but stability is reduced
Stability – To be discussed later

Steady-State Errors
8 -
C 𝑠 = ).8 𝑅 𝑠
-
1
𝐸 𝑠 = 𝑅(𝑠)
1+𝐺 𝑠
• Consider error e(t) = r(t) – c(t) for tracking performance

( e(t) = 0 à perfect tracking: response tracks reference )
1
𝐸 𝑠 =𝑅 𝑠 −𝐶 𝑠 = 𝑅(𝑠)
1+𝐺 𝑠
• Steady-state tracking error (by Final Value Theorem)
𝑠𝑅(𝑠)
𝑒-- = lim 𝑒(𝑡) = lim 𝑠𝐸(𝑠) = lim
!→# -→7 -→7 1 + 𝐺(𝑠)
• Note: Steady-state error analysis is meaningful only for

stable systems (output of unstable systems will blow up)
Exercise:
(10 minutes)
𝐺(𝑠)
1 + 𝐺(𝑠)
𝑅 is the Unit Ramp reference input

-(#)
(1) =?
.(#)
(2) T9 𝑠 : = 𝑅 𝑠 − 𝐶 𝑠 = ?
(3) lim 𝑇: 𝑡 = lim [𝑟 𝑡 − 𝑐 𝑡 ] = ?
!→# !→#
(4) For what values of 𝑘, one can have

lim 𝑇: 𝑡 = 0 ?
!→#
Answer:
𝐶 𝑠
𝐺 𝑠
= 1 + 𝑘𝑠 𝑅(𝑠)
1+𝐺 𝑠
𝐺(𝑠)
1 + 𝐺(𝑠)
𝑇#
Return to: Steady-State Errors
8 - )
C 𝑠 = 𝑅 𝑠 𝐸 𝑠 = 𝑅(𝑠)
).8 - ).8 -
• Consider error e(t) = r(t) – c(t) for tracking performance

1
𝐸 𝑠 =𝑅 𝑠 −𝐶 𝑠 = 𝑅(𝑠)
1+𝐺 𝑠
• Steady-state tracking error
𝑠𝑅(𝑠)
𝑒-- = lim 𝑒(𝑡) = lim 𝑠𝐸(𝑠) = lim
!→# -→7 -→7 1 + 𝐺(𝑠)
1. Static Position Error Constant Kp
Static Position Error Constant: 𝐾< = lim 𝐺(𝑠)

-→7
For Step Input (position input):
# 1 1
𝑒## = lim 1&2(#) # = 1&*
#→0 7
- For Type-0 systems:

𝐾(𝑇9 𝑠 + 1) ⋯ (𝑇: 𝑠 + 1)
𝐾8 = lim =𝐾
&→' (𝑇; 𝑠 + 1) ⋯ (𝑇< 𝑠 + 1)
- For Type-1 or higher systems (𝑁 ≥ 1):

=(> -.))⋯(> -.))
𝐾< = lim -% (># -.))⋯(>$ -.)) → ∞
-→7 & !
2. Static Velocity Error Constant Kv
Static Velocity Error Constant: 𝐾@ = lim 𝑠𝐺(𝑠)

-→7
For Ramp Input (velocity input):
- ) ) )
𝑒-- = lim ).8(-) -" = lim -.-8(-) =
-→7 -→7 ='
-=(># -.))⋯(>$ -.))

- For Type-0 systems: 𝐾@ = lim (> -.))⋯(> -.)) =0
-→7 & !
-=(># -.))⋯(>$ -.))
- For Type-1 system: 𝐾@ = lim =𝐾
-→7 -(>& -.))⋯(>! -.))
- For Type-2 (or higher) systems (𝑁 ≥ 2):
-=(> -.))⋯(> -.))
𝐾< = lim -% (># -.))⋯(>$-.)) → ∞
-→7 & !
3. Static Acceleration Error Constant Ka
Static Acceleration Error Constant: 𝐾3 = lim 𝑠 ' 𝐺(𝑠)

#→0
For Parabolic Input (acceleration input):

!%
𝑟 𝑡 = (𝑡 ≥ 0)
&
Since
𝑠 1 1 1
𝑒-- = lim A = lim & & =
-→7 1 + 𝐺(𝑠) 𝑠 -→7 𝑠 + 𝑠 𝐺(𝑠) 𝐾B
0 𝑓𝑜𝑟 𝑁 = 0,1
𝑠 ) 𝐾(𝑇9 𝑠 + 1) ⋯ (𝑇: 𝑠 + 1)
𝐾9 = lim = = 6 𝐾 𝑓𝑜𝑟 𝑁 = 2
&→' 𝑠 (𝑇; 𝑠 + 1) ⋯ (𝑇< 𝑠 + 1)
∞ 𝑓𝑜𝑟 𝑁 ≥ 3
Summary
Steady-State Errors for Unity Feedback Systems
Type 3 system 0 0 0
Higher N All 0
Unity versus Non-Unity Feedback Systems
Y(s)
H(s)
Example:
Consider a non-unity feedback system with plant
3
transfer function 𝐺 𝑠 = 1(0123) , 𝑇 > 0, and
feedback transfer function H(s) = h, h > 0
Determine the system response to the reference input

Unity versus Non-Unity Feedback Systems
Y(s)
H(s)
)
𝐺 𝑠 = -(>-.)) , T > 0 , H(s) = h, h > 0, 𝑅(𝑠) is Unit Step input
Define System tracking error 𝑇: (t):= r(t) – y(t)

Exercise: (5 minutes)
!(#)
(1) =?
.(#)
(2) 𝑇4 (∞) ≔ lim 𝑇4 𝑡 = ?

5→6
Answer:
System tracking error: 𝑇: (t) = r(t) – y(t)
𝐺 𝑠 1+ ℎ−1 𝐺 𝑠
𝑇& 𝑠 : = 𝑅 𝑠 − 𝑌 𝑠 = 𝑅 𝑠 − 𝑅 𝑠 = 𝑅(𝑠)
1 + ℎ𝐺 𝑠 1 + ℎ𝐺 𝑠
For Unit Step input, R(s) = 1/s , tracking error:

- >-.) .C*) C*)
𝑇: ∞ = lim =
-→7 - >-.) .C C
Unity feedback ( h = 1) à 𝑇: (∞) = 0 à Type-1 system

Non-unity feedback ( ℎ ≠ 1 ) à 𝑇: (∞) ≠ 0 < ∞ à Type-0 system
à Non-unity feedback may change the system type

END

Tutorial 4

Uploaded by

Copyright:

Available Formats

Tutorial 4

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tutorial 4

Uploaded by

Copyright:

Available Formats

EE3114

Systems and Control

Damping Ratio Un-Damped Natural Frequency

(2) Assume that 𝑈 is the Unit Ramp Input,

(2) Assume that 𝑈 is the Unit Ramp Input,

Recall: Poles = − 𝜁𝜔% ± 𝜔% 𝜁 & − 1

1. If 0 ≤ 𝜁 < 1, complex poles: −𝜎 ± 𝑗𝜔!

3. If 𝜁 > 1, over damped: (no oscillation ↑↓ )

For 𝜁 = 0 : 𝜔! = 𝜔" 1 − 𝜁 # = 𝜔"

Next, consider a related topic:

• How well does the system output “follow” the reference?

• To obtain some general ideas by applying Step, Ramp, and

• As N increases, accuracy is improved but stability is reduced

Stability – To be discussed later

• Consider error e(t) = r(t) – c(t) for tracking performance

• Note: Steady-state error analysis is meaningful only for

𝑅 is the Unit Ramp reference input

(4) For what values of 𝑘, one can have

• Consider error e(t) = r(t) – c(t) for tracking performance

Static Position Error Constant: 𝐾< = lim 𝐺(𝑠)

- For Type-0 systems:

- For Type-1 or higher systems (𝑁 ≥ 1):

Static Velocity Error Constant: 𝐾@ = lim 𝑠𝐺(𝑠)

-=(># -.))⋯(>$ -.))

Static Acceleration Error Constant: 𝐾3 = lim 𝑠 ' 𝐺(𝑠)

For Parabolic Input (acceleration input):

Determine the system response to the reference input

Define System tracking error 𝑇: (t):= r(t) – y(t)

(2) 𝑇4 (∞) ≔ lim 𝑇4 𝑡 = ?

For Unit Step input, R(s) = 1/s , tracking error:

Unity feedback ( h = 1) à 𝑇: (∞) = 0 à Type-1 system

à Non-unity feedback may change the system type

You might also like