Professional Documents
Culture Documents
𝑎𝑦 !! 𝑡 + 𝑏𝑦 ! 𝑡 + 𝑐𝑦 𝑡 = 𝑢(𝑡)
𝑎, 𝑏, 𝑐 ∶ coefficients
1
ØDifferential equations are classified according to several categories:
𝑑 % 𝑦(𝑡) 𝑑𝑦(𝑡)
2 %
+3 + 6𝑦 𝑡 = 0
𝑑𝑡 𝑑𝑡
𝑦(𝑡): a function of a single variable; therefore, Ordinary Differential Equation
2𝑦 !! 𝑡 + 3𝑦 ! 𝑡 + 6𝑦 𝑡 = 0
2
ØIf the function in the differential equation to be found is a multi-variable one,
then the differential equation is said to be a partial differential equation.
3
ii) Linear – Nonlinear Differential Equations:
ØIf all the terms in the de appear linearly, the differential equation is called linear differential
equation
ØIf at least one term violates the linearity, the differential equation is called Non-linear
Differential Equation
#
𝑑 ! 𝑦(𝑡) 𝑑 " 𝑦(𝑡) 𝑑𝑦(𝑡)
!
+ +3 + 6𝑦 𝑡 = 𝑠𝑖𝑛(𝑡)
𝑑𝑡 𝑑𝑡 " 𝑑𝑥
#
$ ! %(')
: Violates the linearity; therefore, Non-linear Differential Equation
$' !
4
iii) Time Invariant – Time Varying Differential Equations:
ØIf all the coefficients of the differential equation are time invariant
(i.e. constant), then the differential equation is called to be an invariant
differential equation
𝑑 % 𝑦(𝑡) 𝑑𝑦(𝑡)
1 %
+5 + 3𝑦 𝑡 = 𝑡𝑎𝑛 𝑡
𝑑𝑡 𝑑𝑥
5
ØIf at least one coefficient is time varying (i.e. a function of time), then the
differential equation is called to be a time variant differential equation
𝑑 ! 𝑦(𝑡) 𝑑𝑦(𝑡)
3𝑡 ! +2 + 3𝑦 𝑡 = 𝑠𝑖𝑛 𝑡
𝑑𝑡 ! 𝑑𝑡
𝑑 ! 𝑦(𝑡) "
𝑑𝑦(𝑡)
6𝑡 !
+ 2𝑡 + 3𝑦 𝑡 = 𝑠𝑖𝑛 𝑡
𝑑𝑡 𝑑𝑡
! "
3𝑡 , 6𝑡, 2𝑡 : Time varying coefficients exist; therefore, Time Varying Differential
Equation
vIn this class, differential equations belong to class of ordinary, linear, and time
invariant will be of interest.
Definition: Power of the term with the highest order derivative defines the degree of the
differential equations
vSince we are interested in linear differential equations, the degree of the differential
equations in this course will be always one.
!! "($) !"($)
!$ !
+ 8 !& + 4𝑦 𝑡 = 𝑎𝑟𝑐𝑠𝑖𝑛 𝑡 ⟹ Order of the d.e.: 2, Degree of the d.e.: 1
'
!" "($) !! "($) !"($)
!$ "
+ !$ ! + 3 !& + 9𝑦 𝑡 = cos(𝑡) ⟹ Order of the d.e.: 3, Degree of the d.e.: 2
(
!! "($) !# "($)
!$ !
+ !$ " + 𝑦 𝑡 = 𝑡𝑎𝑛 𝑡 ⟹ Order of the d.e.: 2, Degree of the de.: 1
7
Solving Differential Equations
0 + 𝑠𝑖𝑛 𝑡 = 𝑠𝑖𝑛(𝑡)
9
ØTherefore, if 𝑦+ 𝑡 is a solution of Eq. (2), then 𝑦) 𝑡 = 𝑦* 𝑡 + 𝑦+ 𝑡 is also a solution
of Eq. (2). Since our aim is to find the complete function as the solution, then solving
Eq. (2) requires two solution process:
10
First Order Differential Equations:
ØFirst order differential equations belong to the easiest type of differential equations to deal
with. Therefore, there are techniques to solve not only the linear time invariant ones but
also some special kinds of time varying time and/or non-linear ones.
ØIn general, while solving the 1st order ordinary LTI differential equations,
ØThis class of differential equation belong to the linear time varying ones, However, if one
choses 𝑝 𝑡 as a constant (let us say a constant 𝑝), the equation becomes time invariant
12
Technique: 𝑦 ! 𝑡 + 𝑝 𝑡 . 𝑦 𝑡 = 𝑢 𝑡
Ø Multiply each side by 𝑒 -($) where µ(𝑡 ) will be chosen later.
𝑒 -($) 𝑦 ! 𝑡 + 𝑒 -($) 𝑝 𝑡 . 𝑦 𝑡 = 𝑒 -($) 𝑢 𝑡
Ø If 𝑒 -($) 𝑝 𝑡 in the second term can be made derivative of 𝑒 -($) (by appropriate selection of 𝑒 -($) ) then the
equation becomes:
𝑑(𝑒 -($) 𝑦 𝑡 )
= 𝑒 -($) 𝑢 𝑡
𝑑𝑡
⟹ V 𝑑 𝑒0 $ 𝑦 𝑡 = V 𝑒 0 $ 𝑢 𝑡 𝑑𝑡
⟹ 𝑒0 $ 𝑦 𝑡 = 𝐴 𝑡 + 𝐶
𝑒 -($) 𝑝 𝑡 = (𝑒 -($) )! ⟹ 𝑒 0 $ 𝑝 𝑡 = µ! 𝑡 𝑒0 $
𝑝 𝑡 = µ! (𝑡)
⟹ µ(𝑡) = V 𝑝 𝑡 𝑑𝑡
vThus, µ(𝑡) is known as integration factor
13
Example: 𝑦 ! 𝑡 + 3𝑦 𝑡 = 𝑠𝑖𝑛 𝑡
Ø Find 𝑒 -($)
µ 𝑡 = V 3𝑑𝑡 ⟹ µ 𝑡 = 3𝑡
"(1 !" # $ )
⟹ 𝑒 ($ 𝑦 ! 𝑡 + 𝑒 ($ 3𝑦 𝑡 = 𝑠𝑖𝑛 𝑡 𝑒 ($ ⟹ = 𝑠𝑖𝑛 𝑡 𝑒 ($
"$
⟹ ∫ 𝑑 𝑒 ($ 𝑦 𝑡 = ∫ 𝑠𝑖𝑛 𝑡 𝑒 ($ 𝑑𝑡 ⟹ 𝑒 ($ 𝑦 𝑡 = ∫ 𝑠𝑖𝑛 𝑡 𝑒 ($ 𝑑𝑡
( +
⟹𝑦 𝑡 = 𝑠𝑖𝑛 𝑡 - cos 𝑡 + 𝐶𝑒 '($
+, +,
𝑦2 𝑡 𝑦3 𝑡
forced solution homogeneous solution
𝑎𝑦 !! 𝑡 + 𝑏𝑦 ! 𝑡 + 𝑐𝑦 𝑡 = 𝑢(𝑡)
the solution process includes two steps (finding the solution of homogeneous part and finding the solution
of forced part). Additionally, depending on the function , the solution process to find the forced solution
may include tedious computations.
Ø For ordinary LTI differential equations, there is a powerful technique named Laplace Transformation.
Using this method, one can solve any order of differential equations. Furthermore, both homogeneous
and forced responses can be obtained in one step.
15
Laplace Transform
Laplace transform is a technique which transforms the differential equation (written in 𝑡 -domain) into an
algebraic equation (written in 𝑠 -domain).
4
ℒ 𝑓 𝑡 = 𝐹 𝑠 = ∫, 𝑒 '5$ 𝑓 𝑡 𝑑𝑡
Ø But this definition is rarely used to find 𝐹(𝑠) for a function 𝑓 𝑡 . In general, the Laplace table, in which
there are the transforms of frequently used functions is employed. Using Laplace Transform tables, the diff.
equation is transformed from 𝑡 domain to the 𝑠 – domain. (The function to be found 𝑦(𝑡) is transformed to
𝑌(𝑠)). However, finding 𝑌(𝑠) is much easier than finding 𝑦(𝑡).
Ø Additionally, 𝑌(𝑠) includes both 𝑌2 𝑠 𝑎𝑛𝑑 𝑌3 𝑠 components.
Ø After finding 𝑌 𝑠 , all is needed to be done is to transform 𝑌(𝑆) to 𝑦(𝑡) again, using the Inverse Laplace
Transform.
16
Some Properties of Laplace Transformation
1) Additivity:
ℒ
𝑓+ 𝑡 → 𝐹+ 𝑠
ℒ
⇒ 𝑓+ 𝑡 + 𝑓7 𝑡 → 𝐹+ 𝑠 +𝐹7 (𝑠)
ℒ
𝑓7 𝑡 → 𝐹7 (𝑠)
Example:
8$
ℒ 1
𝑓+ 𝑡 = 𝑒 → 𝐹+ 𝑠 =
𝑠 −5
ℒ 1
𝑓7 𝑡 = 𝑠𝑖𝑛𝑡 → 𝐹7 𝑠 = 7
𝑠 +1
ℒ + + 5# )5 '9
⇒ 𝑓+ 𝑡 + 𝑓7 𝑡 = 𝑒 8$ + 𝑠𝑖𝑛𝑡 → + =
5 '8 5# )+ 5! '85# )5 '8
17
2) Homogeneity:
ℒ
𝑓+ 𝑡 → 𝐹+ 𝑠
ℒ
𝑘𝑓+ 𝑡 → 𝑘𝐹+ 𝑠 (𝑤ℎ𝑒𝑟𝑒 𝑘 𝑖𝑠 𝑎 𝑠𝑐𝑎𝑙𝑎𝑟)
ℒ + ;
Example: 𝑒 ':$ → ⇒ 6. 𝑒 ':$ →
5): 5):
3) Scaling:
ℒ
𝑓+ 𝑡 → 𝐹+ 𝑠
ℒ
𝑒 '<$ 𝑓+ 𝑡 → 𝐹+ 𝑠 + 𝑎
ℒ 𝑠 $%&
ℒ (𝑠 + 5) (𝑠 + 5)
Example: cos(7𝑡) → ⇒ 𝑒 cos(7𝑡) → =
𝑠 # + 49 𝑠 + 5 # + 49 𝑠 # + 10𝑠 + 74
18
4) Derivative:
ℒ "2$ $ ℒ
𝑓+ 𝑡 → 𝐹+ 𝑠 ⇒ → 𝑠𝐹 𝑠 −𝑓 0 (𝑓𝑖𝑟𝑠𝑡 𝑜𝑟𝑑𝑒𝑟 𝑑𝑒𝑟𝑖𝑣𝑎𝑡𝑖𝑣𝑒)
"$
ℒ "(&) 2$ $ ℒ =
𝑓+ 𝑡 → 𝐹+ (𝑠) ⇒ →𝑠 𝐹 𝑠 − 𝑠 ='+ 𝑓 0 − 𝑠 ='7 𝑓 + 0 − ⋯ − 𝑠+ 0 − 𝑓 ='+ 0 (𝑛$3 𝑜𝑟𝑑𝑒𝑟 𝑑𝑒𝑟. )
"$ &
ℒ
Example: 𝑦 !! 𝑡 − 3𝑦 ! 𝑡 + 2𝑦 𝑡 = 𝑠𝑖𝑛𝑡 𝑦 0 = 1, 𝑦! 0 =4 𝑦 𝑡 →𝑌 𝑠 ?
ℒ
𝑦 𝑡 → 𝑌 𝑠 ⇒ ℒ{𝑦 !! 𝑡 − 3𝑦 ! 𝑡 + 2𝑦 𝑡 } = ℒ{𝑠𝑖𝑛𝑡}
1
⟹ [𝑠 7 𝑌(𝑠) − 𝑠𝑦(0) − 𝑦 ! (0)] − 3[𝑠𝑌 𝑠 − 𝑦 0 ] + 2 𝑌 𝑠 =
𝑠7 + 1
1
⟹ 𝑠 7𝑌 𝑠 − 𝑠 − 4 − 3𝑠𝑌 𝑠 + 3 + 2𝑌(𝑠) = 7
𝑠 +1
+
⟹ 𝑌 𝑠 𝑠 7 − 3𝑠 + 2 = + s +1
5# )+
Ø Using these properties and the Laplace tables, solving a differential equation is much easier. The
procedure can be defined as follows:
Procedure:
1. Take the Laplace transforms of the differential equation using the Table of Laplace Transforms
2. Obtain 𝑌 𝑠 , which is the transformed version of 𝑦(𝑡).
3. Transform 𝑌 𝑠 to 𝑦 𝑡 using Inverse Laplace Transform (Using Laplace table)
20
Example: Solve the given equation using Laplace Transform, and specify the homogenous and the forced solutions.
𝑦 ! 𝑡 + 3𝑦 𝑡 = 𝑠𝑖𝑛𝑡 , 𝑦 0 =5
ℒ
𝑦 𝑡 → 𝑌 𝑠 ⇒ ℒ{𝑦 ! 𝑡 + 3𝑦 𝑡 = 𝑠𝑖𝑛} = ℒ 𝑠𝑖𝑛𝑡
1 1
⇒ [𝑠𝑌 𝑠 − 𝑦 0 ] + 3 𝑌 𝑠 = 7 ⟹ 𝑠𝑌 𝑠 − 5 + 3𝑌(𝑠) = 7
𝑠 +1 𝑠 +1
1 5𝑠 7 + 6
⟹ 𝑠+3 𝑌 𝑠 = 7 + 5 ⟹ 𝑌(𝑠) =
𝑠 +1 𝑠 + 3 [𝑠 7 +1]
v To obtain 𝑦(𝑡) all is needed to be done is decomposing 𝑌 𝑠 into proper fractions, then using Laplace Table.
@ A5)B 85# );
𝑌(𝑠) = + # = ⟹ 𝐴𝑠 7 + 𝐴 + 𝐵𝑠 7 + 3𝐵 + 𝐶 𝑠 + 3𝐶 = 5𝑠 7 + 6
5)( 5 )+ (5)()(5# )+)
⟹ 𝐴 + 𝐵 𝑠 7 + 3𝐵 + 𝐶 𝑠 + 3𝐶 + 𝐴 = 5𝑠 7 + 6
𝐴+𝐵 =5 𝐴+𝐵 =5 51 1 3
=3𝐵 + 𝐶 = 0 ⟹ 𝐴 − 9𝐵 = 6 𝐴= , 𝐵=− , 𝐶=
10 10 10
3𝐶 + 𝐴 = 6 10𝐵 = −1
1 3
51/10 − 𝑠 ℒ !" 51 $'& 1 3
𝑌 𝑠 = + # 10 + #10 𝑦 𝑡 = .𝑒 − 𝑐𝑜𝑠𝑡 + 𝑠𝑖𝑛𝑡
𝑠+3 𝑠 +1 𝑠 +1 10 10 10
𝑦( 𝑡 𝑦) 𝑡 21
Example: Solve the given equation using Laplace Transform, and specify the homogenous and the forced solutions.
𝑦 !! 𝑡 + 4𝑦 ! 𝑡 + 3𝑦 𝑡 = 0 , 𝑦 0 = 2, 𝑦! 0 = 6
ℒ
𝑦 𝑡 →𝑌 𝑠 ⟹ ℒ 𝑦 !! 𝑡 + 4𝑦 ! 𝑡 + 3𝑦 𝑡 = ℒ 0 ⟹ 𝑠 7 𝑌 𝑠 − 2𝑠 + 6 + 4 𝑌 𝑠 − 2 + 3𝑌(𝑠) = 0
2𝑠 + 2 2 𝑠 + 1 2 ℒ ($
𝑌 𝑠 𝑠 7 + 4𝑠 + 3 = 2𝑠 + 2 ⟹ 𝑌 𝑠 = 7 = = 2𝑒 '($ = 𝑦 𝑡
𝑠 + 4𝑠 + 3 𝑠+3 𝑠+1 𝑠+3
𝑦 𝑡 : ℎ𝑜𝑚𝑜𝑔𝑒𝑛𝑜𝑢𝑠 𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛, 𝑡ℎ𝑒𝑟𝑒 𝑖𝑠 𝑛𝑜 𝑓𝑜𝑟𝑐𝑒𝑑 𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛
Example: 𝑦 !! 𝑡 + 4𝑦 ! 𝑡 + 3𝑦 𝑡 = 𝑒 '7$ , 𝑦 0 = 2, 𝑦 ! 0 = 6
ℒ +
𝑦 𝑡 →𝑌 𝑠 ⟹ ℒ 𝑦 !! 𝑡 + 4𝑦 ! 𝑡 + 3𝑦 𝑡 = ℒ 𝑒 '7$ ⟹ 𝑠 7 𝑌 𝑠 − 2𝑠 + 6 + 4 𝑌 𝑠 − 2 + 3𝑌(𝑠) =
5)7
+ 75# );5)8 75# );5)8 75# );5)8 @ A B
𝑌 𝑠 𝑠7 + 4𝑠 + 3 = 2𝑠 + 2 + = ⟹𝑌 𝑠 = = = + +
5)7 5)7 5)7 5# )95)( 5)( 5)+ 5)7 5)( 5)+ 5)7
⟹ 𝐴 𝑠 + 1 𝑠 + 2 + 𝐵 𝑠 + 2 𝑠 + 3 + 𝐶 𝑠 + 1 𝑠 + 3 = 2𝑠 7 + 6𝑠 + 5
5 5/2 1/2 −1
𝑠 = −3 ⇒ 2𝐴 = 18 − 18 + 5 ⇒ 𝐴 = 𝑌(𝑠) = + +
2 𝑠+3 𝑠+1 𝑠+2
1 5 '($ 1 '$
𝑠 = −1 ⇒ 2𝐵 = 1 − 6 + 5 ⇒ 𝐵 =
2
𝑦 𝑡 = 𝑒 + 𝑒 − 𝑒 '7$
2 2
𝑠 = −2 ⇒ −𝐶 = 8 − 12 + 5 ⇒ 𝐶 = −1 22
𝑦( 𝑡 𝑦) 𝑡
Example: Solve the given equation using Laplace Transform, and specify the homogenous and the forced solutions.
𝑦 !! 𝑡 + 4𝑦 ! 𝑡 + 4𝑦 𝑡 = 4 , 𝑦 0 = 1, 𝑦! 0 = 2
ℒ +
𝑦 𝑡 →𝑌 𝑠 ⟹ ℒ 𝑦 ** 𝑡 + 4𝑦 * 𝑡 + 4𝑦 𝑡 = ℒ 4 ⟹ 𝑠 # 𝑌 𝑠 − 𝑠 − 2 + 4 𝑌 𝑠 − 1 + 4𝑌(𝑠) =
,
#
4
𝑌 𝑠 𝑠 + 4𝑠 + 4 = + 𝑠 + 6
𝑠
𝑠 # + 6𝑠 + 4 𝐴 𝐵 𝐶
⟹𝑌 𝑠 = # = + +
𝑠 + 4𝑠 + 4 . 𝑠 (𝑠 + 2) (𝑠 + 2)# 𝑠
𝐴 𝑠 + 2 𝑠 + 𝐵𝑠 + 𝐶(𝑠 + 2)7 = 𝑠 7 + 6𝑠 + 4
𝐴𝑠 7 + 2𝐴𝑠 + 𝐵𝑠 + 𝐶𝑠 7 + 4𝑠𝐶 + 4𝐶 = 𝑠 7 + 6𝑠 + 4
𝐴 + 𝐶 𝑠 7 + 2𝐴 + 4𝐶 + 𝐵 𝑠 + 4𝐶 = 𝑠 7 + 6𝑠 + 4
, 7 +
C=1 𝑌 𝑠 = + + ⟹
5)7 (5)7)# 5
A=0 𝑦 𝑡 = 2𝑡𝑒 '7$ + 1
B=2
𝑦( 𝑡 𝑦) 𝑡 23
Example: Solve the given equation using Laplace Transform, and specify the homogenous and the forced solutions.
𝑦 !! 𝑡 + 2𝑦 ! 𝑡 + 10𝑦 𝑡 = 10 , 𝑦 0 = 1, 𝑦! 0 = 1
10 𝑠 7 + 3𝑠 + 10 𝐴 𝐵𝑠 + 𝐶
𝑌 𝑠 7
𝑠 + 2𝑠 + 10 = +𝑠+3 ⇒𝑌 𝑠 = 7 = + 7
𝑠 𝑠 + 2𝑠 + 10 . 𝑠 𝑠 𝑠 + 2𝑠 + 10
𝐴 𝑠 7 + 2𝑠 + 10 + 𝐵𝑠 + 𝐶 𝑠 = 𝑠 7 + 3𝑠 + 10 C=1
A=1
𝐴 + 𝐵 𝑠 7 + 2𝐴 + 𝐶 𝑠 + 10𝐴 = 𝑠 7 + 3𝑠 + 10 B=0
+ + ℒ ($ + + + '+ (
𝑌 𝑠 = + # 𝑦 𝑡 = 1 + ℒ '+ = 1 + ℒ '+ =1+ ℒ
5 5 )75)+, (5)+)# )C (5)+)# )(# ( (5)+)# )(#
1 '$
𝑦 𝑡 = 1 + 𝑒 sin(3𝑡)
3
𝑦) 𝑡 𝑦( 𝑡
24
Some Special Functions Used For Systems
i. Dirac Delta Function (Impulse Function): 𝜹(𝒕)
∞, 𝑡=0 ℒ
𝛿(𝑡) = R ⇒𝛿 𝑡 →𝛿 s = 1
0, 𝑡≠0
.
1, 𝑎<0<𝑏
U 𝛿 𝑡 𝑑𝑡 = W
- 0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
25
ii. Unit Step Function: 𝒖𝒔 𝒕
0, 𝑡<0 ℒ 1
𝑢, 𝑡 = R ⇒ 𝑢, 𝑡 → U s =
1, 𝑡>0 s
0
𝑢, 𝑡 = U 𝛿 𝑡 𝑑𝑡
$0
Ø In general, 𝑢5 𝑡 is used to decay the value of any function for 𝑡 < 0 . That is, suppose that 𝑓 𝑡 is a constant
function.
𝑓 𝑡 = 5 ⇒ It gives a value of 5 for ∀𝑡 ∈ −∞, ∞
However, 𝑓 𝑡 = 5. 𝑢5 (𝑡) ⇒ It gives a value of 5 for only +∈ [0, ∞]. For +∈ −∞, 0 , 𝑓 𝑡 = 0.
Ø While dealing with systems, we will denote the constant inputs as 5. 𝑢5 (𝑡) rather than denoting as ‘’ 𝑎 ’’ only.
26
iii. Unit Ramp Function: 𝒓 𝒕 = 𝒕. 𝒖𝒔 𝒕
0, 𝑡 < 0 ℒ 1
r 𝑡 = R ⇒ 𝑟 𝑡 → R s = ,#
𝑡, 𝑡 > 0
0
r 𝑡 = ∫$0 𝑢, 𝑡 𝑑𝑡
𝟏
iv. Parabolic Function: 𝒂 𝒕 = 𝒕𝟐 . 𝒖𝒔 𝒕
𝟐
0, 𝑡 < 0 ℒ 1
a 𝑡 = W1 # ⇒ 𝑎 𝑡 →A s = $
𝑡 , 𝑡>0 ,
#
0
a 𝑡 = ∫$0 𝑟 𝑡 𝑑𝑡
27
ØUsing Laplace Transformation, solving (ordinary, LTI) differential
equations easier; however, in most of the situations the complete
solution of the differential equation is not needed.
28
Transfer Function Approach State Space Approach (Modern Control Theory)
• The transfer function is defined as: Ø Equivalent to write an 𝑛!" order differential equations as 𝑛 first order
diff. equations
𝑌(𝑠) 𝑜𝑢𝑡𝑝𝑢𝑡
𝐺 𝑠 = = (𝑎𝑙𝑙 𝑖𝑛𝑖𝑡𝑖𝑎𝑙 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛𝑠 𝑎𝑟𝑒 𝑧𝑒𝑟𝑜) 𝑑𝑦 # (𝑡)
𝑈(𝑠) 𝑖𝑛𝑝𝑢𝑡
𝑎# #
+ 𝑎#$% 𝑦 #$% 𝑡 + … + 𝑎% 𝑦 & (𝑡) + 𝑎' 𝑦(𝑡) = 𝑢(𝑡)
𝑑𝑡
𝑌 𝑠 = 𝐺 𝑠 .𝑈 𝑠 States: 𝑥% , 𝑥( , … . , 𝑥# inputs: 𝑢 output ⇒ 𝑦 = 𝑥%
𝑥%̇
𝑥% 0
𝑥(̇
𝑈 𝑠 𝑥( 0
𝑌 𝑠 (Block diagram representation) .
𝐺(𝑠) 𝑥̇ = 𝐴. 𝑥 + 𝐵. 𝑢 ⇒ . =
. + . 𝑢
. .
.
• In other words, transfer function is another way of expressing the diff. 𝑥# 1
equation. (Assuming all the initial conditions are zero) 𝑥#̇
! !"# (𝑡) + ……… +𝑎# 𝑦 $ (𝑡) + 𝑎% 𝑦(𝑡)} = ℒ{𝑢(𝑡)}
ℒ {𝑎! 𝑦 (𝑡) + 𝑎! 𝑦 𝑥%
+ (all conditions are zero) 𝑥(
y = C 𝑥 + Du ⇒ 𝑦 = 1 0 … . . 0 . + 0.u
.
𝑎! . 𝑠 ! 𝑌 (𝑠) + 𝑎!"# 𝑠 !"#
𝑌 𝑠 + … + 𝑎# 𝑠𝑌(𝑠) + 𝑎% 𝑌(𝑠) = 𝑈(𝑠) 𝑥#
𝑌(𝑠) 𝑜𝑢𝑡𝑝𝑢𝑡
𝐺 𝑠 = = (𝑎𝑙𝑙 𝑖𝑛𝑖𝑡𝑖𝑎𝑙 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛𝑠 𝑎𝑟𝑒 𝑧𝑒𝑟𝑜)
𝑈(𝑠) 𝑖𝑛𝑝𝑢𝑡
𝑌 𝑠 = 𝐺 𝑠 .𝑈 𝑠
𝑈 𝑠 𝑌 𝑠
𝐺(𝑠) (Block diagram representation)
Ø In other words, transfer function is another way of expressing the diff. equation. (Assuming all the initial conditions are zero)
ℒ {𝑎: 𝑦 : (𝑡) + 𝑎:;<𝑦 :;< (𝑡) + ……… +𝑎<𝑦 = (𝑡) + 𝑎>𝑦(𝑡)} = ℒ{𝑢(𝑡)}
+ (all initial conditions are zero)
𝑌(𝑠) 1
= = 𝐺 𝑠 ⇒ 𝑌 𝑠 = 𝐺 𝑠 𝑈(𝑠)
𝑈(𝑠) (𝑎: . 𝑠 : + 𝑎:;<𝑠 :;< + … + 𝑎<𝑠 + 𝑎>)
30
Ø Since transfer function is based on Laplace
𝑈 𝑠 1 𝑌 𝑠 transformation, it is applicable to LTI systems, only.
(𝑎I . 𝑠 I + … + 𝑎1 𝑠 + 𝑎J ) Ø Transfer function is the way of expressing the diff.
equation regardless of the input (𝑈 𝑠 ) in a more
compact way.
K(,)
Example: 𝑦 ** 𝑡 + 3𝑦 * 𝑡 + 4𝑦 𝑡 = 𝑢 𝑡 Find the transfer function 𝐺 𝑠 = of the system.
L(,)
𝑈 𝑠 1 𝑌 𝑠
𝑌 𝑠 = 𝐺 𝑠 𝑈(𝑠)
𝑠 # + 3𝑠 + 4 v One can solve 𝑌 𝑠 for any 𝑈 𝑠 if 𝑌 𝑠 is needed.
31
Ø Transfer function approach can also be used for the systems whose dynamics are expressed more than one
diff. equation.
Example: Suppose that the dynamics of a physical system is given by the following set of equations:
D$ (5)
Assuming 𝑢(𝑡) is the input and 𝑦+ 𝑡 is output, find the transfer function of the system 𝐺 𝑠 = E(5)
As can be seen from the equations, both 𝑦+ 𝑡 and 𝑦7 𝑡 (and their derivatives) appear in both of the
D$ (5)
equations. Since𝐺 𝑠 = is desired to be found, 𝑌7 𝑠 must be eliminated.
E(5)
3𝑠 + 1 𝑈(𝑠)
𝑌1 𝑠 =
𝑠 + + 6𝑠 ' + 8𝑠 # + 18𝑠 + 17
𝑌1 (𝑠) 3𝑠 + 1
𝐺 𝑠 = =
𝑈(𝑠) 𝑠 + + 6𝑠 ' + 8𝑠 # + 18𝑠 + 17
𝑈(𝑠) 1 𝑌1 (𝑠)
3𝑠 +1
( + 3𝑠 + 4
𝑠
𝑠 @ + 6𝑠 A + 8𝑠 ? + 18𝑠 + 17
33
vUsing only the transfer function, most of the details of 𝑦(𝑡) can be obtained without solving the
differential equation. To this end, two concepts, named «zero» and «pole» of the system should be
defined.
Definition: Suppose that the transfer function of a system is given as below:
(5)9)(5)+) 𝑗𝑤
𝐺 𝑠 = ⇒ 𝑃G 𝑠 = 𝑠 + 4 𝑠 + 1 = 0 ⇒ 𝑠+ = −4, 𝑠7 = −1
5! )85# );5)9 𝑠 − 𝑃𝑙𝑎𝑛𝑒
𝜎
-4 -1
34
Definition: Suppose that the transfer function of a system is given as below:
The roots of 𝑃S (𝑠) is defined as the poles of the system and marked as «x» in the 𝑠 – plane.
𝑠+5 𝑠 − 𝑃𝑙𝑎𝑛𝑒
𝐺 𝑠 = 7
𝑠 + 4𝑠 + 3
⇒ zeros: 𝑃G 𝑠 = 0 ⇒ 𝑠 = −5
X X 𝜎
-5 -3 -1
⇒ 𝑝𝑜𝑙𝑒𝑠: 𝑃H 𝑠 = 0 ⇒ 𝑠+ = −3, 𝑠7 = −1
35
Ø Using poles and zeros, many detailed information about 𝑦(𝑡) can be obtained, without solving the diff.
equation, directly from 𝐺 𝑠 .
D(5)
Ø Using the transfer function i.e. 𝐺 𝑠 = , the diff. equation can be reconstructed and 𝑦(𝑡) can be solved
E(5)
by employing 𝑌 𝑠 = 𝐺 𝑠 𝑈(𝑆) for a specified input.
Example: For a physical system, the diff. equation between the input-output is given as below:
𝑑 7 𝑦(𝑡) 𝑑𝑦(𝑡)
7 + 6 + 8 𝑦 𝑡 = 3𝑢 𝑡
𝑑𝑡 𝑑𝑡
D(5)
i. Find the transfer function of the system i.e. 𝐺 𝑠 =
E(5)
ii. Find the poles and the zeros of the system and mark them on the s – plane.
iii. For 𝑢 𝑡 = 𝑢5 𝑡 , find 𝑦(𝑡) using the transfer function. Specify the homogenous and the forced
components of 𝑦 𝑡 .
0 0 0
ℒ
𝑖. 𝑦 𝑡 →𝑌 𝑠 ⇒ 𝑠 7 𝑌 𝑠 − 𝑠𝑦 0 − 𝑦 ! 0 + 6 𝑠𝑌 𝑠 − 𝑦 0 + 8𝑌 𝑠 = 3𝑈(𝑠)
𝑗𝑤
ℒ D(5) (
𝑢(𝑡) → 𝑈(s) 𝑌 𝑠 𝑠 7 + 6𝑠 + 8 = 3𝑈 𝑠 ⇒ = =𝐺 𝑠
E(5) 5# );5): 𝑠 − 𝑃𝑙𝑎𝑛𝑒
𝜎
ℒ ( 𝑃G 𝑠 ⇒ 𝑛𝑜 𝑧𝑒𝑟𝑜𝑠! -4 -2
𝑖𝑖. 𝑦 𝑡 →𝑌 𝑠 𝐺 𝑠 = ⇒
5# );5): 𝑃H 𝑠 ⇒ 𝑠+ = −4, . 𝑠7 = −2 36
𝑖𝑖𝑖. Since the transfer function will be used to solve 𝑌 𝑠 all the initial conditions will be
taken as zero.
𝑈, (𝑠)
𝑌(𝑠) 3 1 3
=𝐺 𝑠 ⇒𝑌 𝑠 =𝐺 𝑠 𝑈 𝑠 ⇒𝑌 𝑠 = . ⇒𝑌 𝑠 =
𝑈(𝑠) (𝑠 + 2)(𝑠 + 4) 𝑠 𝑠+2 𝑠+4 𝑠
3 𝐴 𝐵 𝐶 6 3 3
𝑌(𝑠) = = + + ⇒ 𝐴=− , 𝐵= , 𝐶=
𝑠+2 𝑠+4 𝑠 𝑠+2 𝑠+4 𝑠 8 8 8
6 '7$ 3 '9$ 3
𝑦(𝑡) = − 𝑒 + 𝑒 + 𝑢5 (𝑡)
8 8 8
𝑦( 𝑡 𝑦) 𝑡 v Note that the exponential terms directly comes from the location
of the poles.
37
Modeling of Electrical Circuits
vBy the term of «electrical circuits» it is meant that passive electrical circuits. In a passive electrical circuit
there may be the following elements.
i. Sources: +
𝑖,
Ø Independent current source. Provides a current of 𝑖, to the circuit. The voltage 𝐼,
across it is determined by the circuit 𝑉,
𝐼 = 𝐼, _
𝑑𝑉X (𝑡) ℒ
𝑽𝑪 (𝒕) 𝐼W 𝑡 = 𝐶. → 𝐼W 𝑠 = 𝑠𝐶𝑉X (𝑠)
+ - 𝑑𝑡
(All initial conditions are
1 & ℒ 1 assumed to be zero)
𝑉W 𝑡 = U 𝑖X 𝑡 𝑑𝑡 → 𝑉W 𝑠 = 𝐼 (𝑠)
𝒊𝑪 (𝒕) 𝐶 J 𝐶𝑠 X
iv. Inductor:
Ø The voltage – current relation of the inductor is given as the following equations:
𝑽𝑳 (𝒕)
+ -
𝑑𝑖Y (𝑡) ℒ
𝑉Y 𝑡 = 𝐿 → 𝑉Y 𝑠 = 𝑠𝐿𝐼Y (𝑠)
𝑑𝑡 (All initial conditions are
assumed to be zero)
1 & ℒ 1
𝑖Y 𝑡 = U 𝑉Y 𝑡 𝑑𝐿 → 𝐼Y 𝑠 = 𝑉 (𝑠)
𝒊𝑳 (𝒕) 𝐿 J 𝐿𝑠 Y
39
vAs might be observed, the 𝐿 and 𝐶 elements can be considered as if they behave like a resistor in
𝑠 − 𝑑𝑜𝑚𝑎𝑖𝑛 and they obey the Ohm’s rule. If the Ohm’s rule is generalized for 𝐿 and 𝐶 elements
in 𝑠 − 𝑑𝑜𝑚𝑎𝑖𝑛:
1
𝐼T 𝑠 = 𝑉 𝑠
𝑉T 𝑠 = 𝑅𝐼T 𝑠 𝑅 T
1
𝑉W 𝑠 = 𝐼 𝑠 𝑉𝑜𝑙𝑡𝑎𝑔𝑒 = 𝐼𝑚𝑝𝑒𝑑𝑎𝑛𝑐𝑒 𝑥 𝐶𝑢𝑟𝑟𝑒𝑛𝑡 𝐼W 𝑠 = 𝑠𝐶 𝑉W 𝑠 𝐶𝑢𝑟𝑟𝑒𝑛𝑡 = 𝐴𝑑𝑚𝑖𝑡𝑡𝑎𝑛𝑐𝑒 𝑥 𝑉𝑜𝑙𝑡
,W W
1
𝑉Y 𝑠 = 𝑠𝐿𝐼Y 𝑠 𝐼Y 𝑠 = 𝑉 𝑠
𝑠𝐿 Y
𝑍 𝑠 : 𝐼𝑚𝑝𝑒𝑑𝑎𝑛𝑐𝑒 𝑌 𝑠 : 𝐴𝑑𝑚𝑖𝑡𝑡𝑎𝑛𝑐𝑒
v Therefore, while we are deriving the transfer function of a circuit (between the defined input and the
defined output) we can consider every element as if it is a resistor while writing the circuit equations
in 𝑠 − 𝑑𝑜𝑚𝑎𝑖𝑛.
40
Ø Then we will need three facts to derive the transfer function of a circuit:
𝑖1 𝑖#
i. Kirchoff’s current law (KCL): 𝑖1 = 𝑖# + 𝑖'
𝑖'
_ _
+ +
ii. Kirchoff’s voltage law (KVL): 𝑉T + 𝑉W + 𝑉Y = 𝑉[ + 𝑉Y 𝑉T +
𝑉, 𝑉W
_ _
iii. Generalized Ohm’s Law: Writing the voltages and / or currents in terms of currents and voltages,
depends on the solution strategy. (The generalized Ohm’s law is explained above)
Ø In order to find a transfer function of a circuit, first the circuit must be analyzed. To accomplish this
analysis, two methods can be used:
i. Node voltages method (expressing all the currents in the circuit in terms of node voltages)
ii. Mesh currents method (expressing all the voltages in the circuit in terms of mesh currents)
Both methodologies facilitate the circuit analysis. After finding the node voltages (or mesh currents), any
required voltage / current can be computed.
41
• Example: By selecting 𝑉 𝑡 as the input and 𝑉I (𝑡) as the output, find the transfer function of the circuit,
JB (5)
i.e. 𝐺 𝑠 = given below:
J(5)
𝑅% = 1Ω 𝑅( = 5Ω
To solve the problem, the circuit must be analyzed as the first
+ step. The analysis can be performed by two ways.
1
𝐶= 𝐹
𝑉(𝑡) 3 𝐿 = 1𝐻 𝑉Y (𝑡)
_ i. Mesh currents method
ii. Node voltages method
1
1. 𝑉T1 + 𝑉W − 𝑉 𝑠 = 0 ⇒ 𝑅1 𝐼1 𝑠 +
,W
𝐼1 𝑠 − 𝐼# 𝑠 =𝑉 𝑠
1
2. −𝑉W + 𝑉T# + 𝑉Y 𝑡 = 0 ⇒ −
,W
𝐼1 𝑠 − 𝐼# 𝑠 + 𝐼# 𝑠 𝑅# + 𝑠𝐿𝐼# 𝑠 = 0 42
C
!
)+ '+
5BK$ )+ +
− 𝐼+ (𝑠) 𝑉(𝑠)
C C
𝐼+ (𝑠) 𝑉(𝑠)
5B 5B ! !
5# IB)5BK# )+
= = $ # D =
−
+ 𝐼7 (𝑠) 0 '+ !
5 ) !5)+ 𝐼7 (𝑠) 0
5B 5B C $
! !
5
5)( (
5
−5 𝐼+ (𝑠) 𝑉(𝑠)
5# )85)(
= (𝑉I 𝑠 = 𝑠𝐿𝐼7 𝑠 )
−
( 𝐼7 (𝑠) 0
5 5
𝑠 + 3 𝑉(𝑠)
3 3
− 0 𝑉(𝑠)
𝑠 ⟹ 𝐼# 𝑠 = 𝑠
Then solving 𝐼# (𝑠) is logical 𝐼# 𝑠 = (𝑠 + 3)(𝑠 # + 5𝑠 + 3) 9
𝑠+3 3
− − #
𝑠 𝑠 𝑠# 𝑠
3 𝑠 # + 5𝑠 + 3
−𝑠 𝑠
3
𝐼# (𝑠) 𝑠 𝐼# (𝑠) 3
⟹ = ' ⇒ = #
𝑉(𝑠) 𝑠 + 9𝑠 # + 18𝑠 𝑉(𝑠) 𝑠 + 8𝑠 + 18
𝑠 #
𝑉Y (𝑠)
𝑠𝐿 = 3 𝑉Y (𝑠) 3𝑠
⇒ = #
𝑉(𝑠) 𝑠 # + 8𝑠 + 18 𝑉(𝑠) 𝑠 + 8𝑠 + 18 43
ii) Node Voltages Method:
1 2 3
1Ω 𝑖1 (𝑡) 𝑖# (𝑡) 5Ω i. Define the nodes in the circuit and select one of them as
the reference node. (In general try to select the ground
𝑖# (𝑡)
𝑖' (t) + node as the reference)
1
𝑉(𝑡) 𝐹 1𝐻 𝑉Y (𝑡) ii. Attach the currents to the circuit
3
_
iii. Apply KCL for each node and write the equations using
generalized Ohm’s rule. While writing, if a node has only a
voltage source between the node and the reference node,
00 00 00 than the value of the node is equal to the source.
Reference node
1. 𝑉+ = 𝑉 𝑡 ⇒ 𝑉+ (𝑠) = 𝑉 𝑠
J$ 5 ' J# (5) J# 5 ' J! (5) J (5)
2. 𝑖+ = 𝑖7 + 𝑖( ⇒
K$
=
K#
+ #
LE (5)
J# 5 ' J! (5) J! (5)
3. 𝐼7 𝑠 = 𝑖7 𝑠 ⇒
K#
=
LB (5)
44
1. 𝑉1 𝑠 = 𝑉 𝑠
%
6+ 𝑠 −1 𝑉# (𝑠) 5𝑉(𝑠) 5
' = 6 + 𝑠 5𝑉(𝑠)
3
−𝑠 𝑠 + 5 𝑉' (𝑠) 0
−𝑠 0 5𝑠𝑉(𝑠)
⟹ 𝑉Y 𝑠 = ⟹ 𝑉Y 𝑠 =
5 5
6 + 3 𝑠 −1 6+ 𝑠 𝑠+5 +𝑠
𝑉' 𝑠 = 𝑉Y 𝑠 3
−𝑠 𝑠+5
𝑉Y (𝑠) 5𝑠 3𝑠
⟹ = = #
𝑉(𝑠) 5 𝑠 # + 40 𝑠 + 30 𝑠 + 8𝑠 + 18
3 3
ℒ
Input: 𝑓 𝑡 𝐹𝑜𝑟𝑐𝑒 → 𝐹 𝑠
ℒ
Output: 𝑥 𝑡 𝐷𝑖𝑠𝑝𝑙𝑎𝑐𝑒𝑚𝑒𝑛𝑡 → 𝑋 𝑠
ℒ
𝑣 𝑡 𝑉𝑒𝑙𝑜𝑐𝑖𝑡𝑦 → 𝑉 𝑠
𝑥(𝑡) Mass provides a reaction force directly proportional to the acceleration of it.
iii) Mass: ℒ
𝐹 𝑡 =𝑚𝑎 𝑡 → 𝐹 𝑠 =𝑚𝐴 𝑠
or
𝐹 𝑡 𝑑𝑣 𝑡 ℒ
𝑚 (𝑘𝑔) 𝐹 𝑡 =𝑚 → 𝐹 𝑠 = 𝑚𝑠𝑉(𝑠)
𝑑𝑡
or
]#^ & ℒ
𝑥(𝑡) 𝐹 𝑡 =𝑚 → 𝐹 𝑠 = 𝑠 # 𝑋(𝑠) 47
]& #
Ø In order to write the dynamical equations of a mechanical system, two physical principle will be used:
1) Newton’s Second Law: 𝐹 𝑡 = 𝑚 𝑎 𝑡 (We have used it for deriving the force – acceleration
relation of the mass)
2) Newton’s Third Law: ∑ 𝑎𝑐𝑡𝑖𝑜𝑛 𝑓𝑜𝑟𝑐𝑒 = ∑ 𝑟𝑒𝑎𝑐𝑡𝑖𝑜𝑛 𝑓𝑜𝑟𝑐𝑒
Example:
_"(,)
Supposing 𝐹(𝑠) is the input and 𝑋1 𝑡 is the output, obtain the transfer function of the system, i.e. 𝐺 𝑠 =
`(,)
48
v According to superposition principle, only one mass will be allowed to move while deriving the equations.
Force acting on mass 𝑚1 Force acting on mass 𝑚#
2𝐹(𝑠)
⇒ 𝑋< 𝑠 = 𝐹(𝑠) 1
1 𝑋1 (𝑠)
10𝑠 @ + 4𝑠 A + 22𝑠 ? + 4𝑠 + 12 − 4
𝑠 ( + 3𝑠 + 4
5𝑠 @ + 2𝑠 A + 11𝑠 ? + 2𝑠 + 4
𝑋< (𝑠) 1
⇒ = =𝐺 𝑠 49
𝐹(𝑠) 5𝑠 @ + 2𝑠 A + 11𝑠 ? + 2𝑠 + 4
II) Rotational Mechanical Systems
Ø As in the translational system, in general, the following quantities are taken as inputs and outputs:
ℒ
Input: 𝜏 𝑡 𝑇𝑜𝑟𝑞𝑢𝑒 → 𝑇 𝑠
ℒ
Output: 𝜃 𝑡 𝐴𝑛𝑔𝑢𝑙𝑎𝑟 𝑑𝑖𝑠𝑝𝑙𝑎𝑐𝑒𝑚𝑒𝑛𝑡 → Θ 𝑠
ℒ
𝜔 𝑡 𝐴𝑛𝑔𝑢𝑙𝑎𝑟 𝑣𝑒𝑙𝑜𝑐𝑖𝑡𝑦 → Ω 𝑠
Ø Logically, the elements define a relationship between Θ 𝑠 , (or Ω 𝑠 ) and 𝑇 𝑠 .
Ø In general, three basic elements (torsional spring, torsional viscous damper, inertia) are used in rotational
mechanical systems.
i) Torsional Spring: Torsional spring provides a reaction torque directly proportional to the angular displacement of it.
ℒ
𝜃(𝑡) 𝜏(𝑡) 𝜏 𝑡 = 𝑘 𝜃 𝑡 → 𝑇 𝑠 = 𝑘Θ 𝑠
𝑘 (𝑁. 𝑚/𝑟𝑎𝑑)
ℒ Ω 𝑠
𝜏 𝑡 = 𝑘 U 𝜔 𝑡 𝑑𝑡 → 𝑇 𝑠 = 𝑘
𝑠
50
ii) Rotational viscous damper: Torsional viscous damper provides a reaction torque directly
proportional to the angular velocity of it.
𝜃(𝑡) 𝜏(𝑡) ℒ
𝜏 𝑡 = 𝑏𝜔 𝑡 → 𝑇 𝑠 = 𝑏Ω 𝑠
𝑑𝜃(𝑡) ℒ
𝑏 (𝑁. 𝑚. 𝑠/𝑟𝑎𝑑) 𝜏 𝑡 =𝑏 → 𝑇 𝑠 = 𝑏 𝑠Θ 𝑠
𝑑𝑡
iii) Inertia: Inertia provides a reaction torque directly proportional to the angular acceleration of it.
ℒ
𝜏 𝑡 = 𝐽𝛼 𝑡 → 𝑇 𝑠 = 𝐽 𝐴 𝑠
𝐽 (𝑁. 𝑚. 𝑠 # /𝑟𝑎𝑑)
𝑑𝜔(𝑡) ℒ
𝜏 𝑡 =𝐽 → 𝑇 𝑠 = 𝐽 𝑠Ω 𝑠
𝑑𝑡
𝑑# 𝜃(𝑡) ℒ
𝜏 𝑡 =𝐽 #
→ 𝑇 𝑠 = 𝐽 𝑠#Θ 𝑠
𝑑𝑡
Ø In order to write the dynamical equations of a rotational system, two physical principle will be used:
1) Newton’s Second Law (for rotation): 𝑇 𝑡 = 𝐽 𝑎 𝑡 (We have used it for deriving the
torque – angular acceleration of the inertia)
2) Newton’s Third Law (for rotation): ∑ 𝑎𝑐𝑡𝑖𝑜𝑛 𝑡𝑜𝑟𝑞𝑢𝑒 = ∑ 𝑟𝑒𝑎𝑐𝑡𝑖𝑜𝑛 𝑡𝑜𝑟𝑞𝑢𝑒
Ø Like translational mechanical systems, these systems are considered to be linear. Therefore, superposition principle
can be used here as well. 51
Example:
Supposing 𝜏(𝑡) is the input and 𝜃1 (𝑡) is
the output, obtain the transfer function
of the system. i.e.
Θ<(𝑠)
𝐺 𝑠 =
𝑇(𝑠)
v According to superposition principle, only one inertia will be allowed to move while deriving the equations:
𝐽1 𝐽#
1) 2𝑠 # + 𝑠 + 3 Θ1 𝑠 − 𝑠 + 1 Θ# 𝑠 = 0 2) − 𝑠 + 1 Θ1 𝑠 + 2𝑠 # + 𝑠 + 1 Θ# 𝑠 = 𝑇(𝑠)
52
2𝑠 7 + 𝑠 + 3 −(𝑠 + 1) Θ+ 𝑠 0
=
−(𝑠 + 1) 2𝑠 7 + 𝑠 + 1 Θ7 𝑠 𝑇(𝑠)
0 −(𝑠 + 1)
𝑇(𝑠) 2𝑠 7 + 𝑠 + 1 𝑠 + 1 𝑇(𝑠)
Θ+ 𝑠 = ⇒ Θ+ 𝑠 =
2𝑠 7 + 𝑠 + 3 −(𝑠 + 1) 4𝑠 9 + 4𝑠 ( + 8𝑠 7 + 4𝑠 + 3 − (𝑠 7 + 2𝑠 + 1)
−(𝑠 + 1) 2𝑠 7 + 𝑠 + 1
53
Gears in Rotational Mechanical Systems
Ø In rotational mechanical systems, gears are used frequently in order to amplify / transmit the torques.
Ø In this course, all the gear mechanisms will be considered to be ideal.
For an ideal gear pair: Ø The total displacements (translational) are equal: 𝑟+ 𝜃+ = 𝑟7 𝜃7
𝜃+ 𝑟7
𝜃1 =
𝜃7 𝑟+
𝜃#
𝑟1 𝑟# Ø The energy consumed in both gearwheels are equal: 𝜏+ 𝜃+ = 𝜏7 𝜃7
𝜃+ 𝜏7
=
𝜃7 𝜏+
𝑁# 𝑟1 𝑁1 𝜏1 𝜃#
𝑁1 Ø If those equations are combined: = = =
(number of teeth: 𝑁) 𝑟# 𝑁# 𝜏# 𝜃1
Θ# (𝑠) 1
𝐺 𝑠 = = #
𝑇# (𝑠) 𝐽𝑠 + 𝑏𝑠 + 𝑘
𝑁?
a"(,) $R" $R# a"(,) R#⁄R" #
Ø However, if 𝐺 ** 𝑠 = is required: 𝐽𝑠 # + 𝑏𝑠 + 𝑘 Θ1 𝑠 = 𝑇1 𝑠 ⇒ 𝐺 ** 𝑠 = =
b"(,) R# R" b"(,) d, #M.,Me
# # #
𝑁1 𝑁1 𝑁1
𝐽* = 𝐽 , 𝑏* = 𝑏 , 𝑘* = 𝑘
𝑁# 𝑁# 𝑁#
55
Example: Find the transfer functions of the system given below
𝜃1 (𝑠)
𝑖. 𝐺1 𝑠 =
𝑇1 (𝑠)
𝜃# (𝑠)
𝑖𝑖. 𝐺# 𝑠 =
𝑇# (𝑠)
𝜃' (𝑠)
𝑖𝑖𝑖. 𝐺' 𝑠 =
𝑇1 (𝑠)
1
2
3
≡
56
Θ𝟏 𝒔
𝒊) 𝑮𝟏 𝒔 =
𝑻𝟏 𝒔
4𝑠 # + 4𝑠 −4𝑠 Θ1 (𝑠) 𝑇1 (𝑠)
1. 4𝑠 # + 4𝑠 Θ1 𝑠 − 4𝑠Θ# 𝑠 = 𝑇1 𝑠 ⇒ =
−4𝑠 4𝑠 # + 4𝑠 + 8 Θ# (𝑠) 0
2. −4𝑠Θ1 𝑠 + 4𝑠 # + 4𝑠 + 8 Θ# 𝑠 = 0
𝑇1 (𝑠) −4𝑠
0 4𝑠 #
+ 4𝑠 + 8 𝑇1 𝑠 (4𝑠 # + 4𝑠 + 8)
Θ1 𝑠 = ⇒ Θ1 𝑠 =
4𝑠 # + 4𝑠 −4𝑠 4𝑠 # + 4𝑠 4𝑠 # + 4𝑠 + 8 − 16𝑠 #
−4𝑠 4𝑠 # + 4𝑠 + 8
𝑁1 𝑁
Θ# 𝑠 = − Θ1 𝑠 , Θ<(𝑠) − ? Θ?(𝑠) 𝑁? ?
Θ?(𝑠) (𝑠 ? + 𝑠 + 2)
𝑁# 𝑁<
⇒ = = =
𝑁# 𝑇<(𝑠) 𝑁< 𝑁< 𝑇?(𝑠) 4𝑠 @ + 8𝑠 A + 8𝑠 ? + 8𝑠
𝑇# 𝑠 = − 𝑇1 (𝑠) − 𝑇?(𝑠)
𝑁?
𝑁1
+ M# (5) (5# )5)7) M# (5) 9(5# )5)7) (5# )5)7)
⇒ = ⇒ = =
9 N# (5) 95F ):5! ):5# ):5 N# (5) 95F ):5! ):5# ):5 5F )75! )75# )75 57
Θ𝟑 𝒔
𝒊𝒊𝒊) 𝑮𝟑 𝒔 =
𝑻𝟏 𝒔
≡
≡
1 1
1. 16𝑠 # + 16𝑠 Θ# 𝑠 − 16𝑠Θ' 𝑠 = − 𝑇1 𝑠 1. 𝑠 # + 𝑠 Θ# 𝑠 − 𝑠Θ' 𝑠 = − 𝑇 𝑠
# '# 1
2. −16𝑠Θ# 𝑠 + 16𝑠 # + 16𝑠 + 32 Θ' 𝑠 = 0 2. −𝑠Θ# 𝑠 + 𝑠 # + 𝑠 + 2 Θ' 𝑠 = 0
1
𝑠# + 𝑠 −𝑠 Θ# (𝑠) − 𝑇1 (𝑠)
= 32
−𝑠 𝑠 + 𝑠 + 2 Θ' (𝑠)
#
0
1
𝑠# + 𝑠 − 𝑇 (𝑠)
32 1
−𝑠 0 −𝑠𝑇1 (𝑠) M! 5 '+
Θ' 𝑠 = ⇒ Θ' 𝑠 = ⇒ 𝐺( 𝑠 = =
𝑠# + 𝑠 −𝑠 32[ 2𝑠 + + 2𝑠 ' + 3𝑠 # + 2𝑠 − 𝑠 # ] N$ 5 ;9 5! )5# )5)+
−𝑠 𝑠# + 𝑠 + 2 58
Modeling of Electromechanical Systems
Ø By the term of electromagnetic systems, we mean an
armature controlled DC motor (under a fixed
electromagnetic field) with a load.
Ø In general,
Input: 𝑉< (𝑠)
Output: ΘO 𝑠 𝑜𝑟 ΩO (𝑠)
𝐾< : 𝑚𝑜𝑡𝑜𝑟 𝑡𝑜𝑟𝑞𝑢𝑒 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 ∶ 𝑁. 𝑚/𝐴 𝐾. : 𝑚𝑜𝑡𝑜𝑟 𝑏𝑎𝑐𝑘 𝑒𝑚𝑓 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 ∶ 𝑉. 𝑠 /𝑟𝑎𝑑
ℒ 𝑑𝜃g (𝑡) ℒ
1 𝑇O 𝑡 = 𝐾< 𝐼< t ⇒ 𝑇O 𝑠 = 𝐾< 𝐼< 𝑠 2 𝑉. 𝑡 = 𝐾. ⇒ 𝑉. 𝑠 = 𝐾. 𝑠Θg 𝑠
𝑑𝑡
ℒ
𝑉. 𝑡 = 𝐾. 𝜔g (𝑡) ⇒ 𝑉. 𝑠 = 𝐾. Ωg 𝑠
v These two equations belong to the motor and relates the electric part to the mechanical part.
v To derive the transfer function of the motor, two parts must be considered separately as electrical and mechanical part
59
i) Electrical part: 𝑉G) + 𝑉H) + 𝑉I 𝑠 = 𝑉J 𝑠
(s 𝐿J + 𝑅J ) 𝐼J 𝑠 + 𝑉I (𝑠) = 𝑉J 𝑠 (*)
𝑙𝑜𝑎𝑑
If the system is moved to armature rod part:
?
𝑁<
≡
𝐽K = 𝐽J + 𝐽G
𝑁?
? ⟹ 𝑇K 𝑠 = 𝐽K 𝑠 ? + 𝑏K 𝑠 ΘK 𝑠 (∗∗)
𝑁<
𝑏K = 𝑏J + 𝑏G
𝑁?
a((,)
Ø To obtain 𝐺 𝑠 = , the equations belonging to the motor must be taken into account to relate the
\)(,)
mechanical part to the electrical part. If (*) is considered with (1) and (2)
60
𝑇g 𝑠
𝑠𝐿- + 𝑅- 𝐼- 𝑠 + 𝑉. 𝑠 = 𝑉- 𝑠 ⟹ 𝑠𝐿- + 𝑅- + 𝐾. 𝑠Θg 𝑠 = 𝑉- (𝑠)
𝐾-
If (**) is considered in that equation:
(𝐽g 𝑠 # + 𝑏g 𝑠)
𝑠𝐿- + 𝑅- Θg 𝑠 + 𝐾. 𝑠Θg 𝑠 = 𝑉- 𝑠
𝐾-
Θg (𝑠) 𝐾-
𝐺 𝑠 = =
𝑉- (𝑠) 𝐽g 𝐿- 𝑠 ' + 𝑏g 𝐿- + 𝐽g 𝑅- 𝑠 # + (𝑏g 𝑅- + 𝐾. 𝐾- )𝑠
𝑉- (𝑠) 1
𝐾 Θg (𝑠)
𝑠 ( +J3𝑠 + 4 (Third order System)
𝐽K 𝐿J 𝑠 A + 𝑏K 𝐿J + 𝐽K 𝑅J 𝑠 ? + (𝑏K 𝑅J + 𝐾I 𝐾J )𝑠
𝑉- (𝑠) 𝐾-
1 Ωg (𝑠) (Second order System)
(
𝑠 + 3𝑠 + 4
𝐽g 𝐿- 𝑠 # + 𝑏g 𝐿- + 𝐽g 𝑅- 𝑠 + (𝑏g 𝑅- + 𝐾. 𝐾- ) 61