Lecture Notes

Fundamentals of Control Systems

Instructor: Dr. Huynh Thai Hoang
Department of Automatic Control
Faculty of Electrical & Electronics Engineering
Ho Chi Minh City University of Technology
Email: [email protected]
[email protected]
H
Homepage: www4.hcmut.edu.vn/~hthoang/
4h t d / hth /

16 November 2011 © H. T. Hoang - www4.hcmut.edu.vn/~hthoang/ 1

 Chapter 7

MATHEMATICAL MODEL OF
DISCRETE TIME CONTROL SYSTEMS

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 2

 Content

‘ I t d ti to
Introduction t discrete-time
di t ti system
t
‘ The Z-transform
‘ Transfer function of discrete-time system
‘ State-space equation of discrete-time system

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 3

 Introduction to discrete
discrete--time systems

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 4

 Digital control systems

r(kT) u(kT)
(kT) uR(t) y(t)
Computer D/A Plant

yfb(kT)
A/D Sensor

‘ “Computer” = computational equipments based on

microprocessor technology (microprocessor,
(microprocessor
microcontroller, PC, DSP,…).
‘ Advantages
g of digital
g control system:
y
Ž Flexibility
Ž Easy to implement complex control algorithms
Ž Computer can control many plants at the same time.

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 5

 Discrete control systems

r(kT) Discrete u(kT) uR(t) y(t)

Hold Plant
processing

yfb(kT)
Sampling Sensor

‘ Discrete control systems are control systems which have

signals at several points being discrete signal.

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 6

 Sampling
‘ Sampling is the reduction of a continuous signal to a
discrete signal. * x(t) x (t)
‘ Mathematical expression T
describing the sampling process: x(t)
+∞
X * ( s ) = ∑ x ( kT )e −kTs t
k =0 0
x*(t)
‘ Shannon’s Theorem:
1 t
f = ≥ 2 fc
T 0

‘ If qquantization error is negligible,

g g , then A/D converters are
approximate the ideal samplers.
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 7
 Sampled--data hold
Sampled
‘ Sampled-data
Sampled data hold is the reconstruction of discrete signal
to a continuous signal. * x (t) xR (t)
ZOH
‘ Zero-order
Zero order hold (ZOH): keep
signal unchanged between x*(t)
two co
consecutive
secut ve sampling
sa p g
instants. t
0
‘ Transfer function of the ZOH.
ZOH xR(t)

1 − e −Ts
GZOH ( s ) = t
s 0

‘ If qquantization error is negligible,

g g , then D/A converters are
approximate the zero-order hold.
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 8
 The Z-transform

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 9

 Definition of the Z-
Z-transform
‘ Consider x(k),
x(k) k=0,1,2,…
k=0 1 2 being a discrete signal.
signal The Z-
Z
transform of x(k) is defined as:
+∞
X ( z ) = Z {x (k )} = ∑ x ( k ) z −k

k = −∞
where:
− z = eTs (s is the Laplace variable, T is the sampling period)
− X(z)
( ) : Z-transform
s o oof x(k).
Z
Notation: x ( k ) ←⎯→ X ( z )
+∞
‘ If x(k) = 0, ∀ k < 0 then X ( z ) = Z {x ( k )} = ∑ x ( k ) z −k
k =0

‘ Region Of Convergence (ROC): set of z such that X(z) is finite.

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 10
 An interpretation of the Z-
Z-transform
‘ Suppose x(t) being a continuous signal,
signal sample x(t) at the
sampling periode T , we have a discrete signal x(k) = x(kT).
‘ The mathematic model of the process of sampling x(t)
+∞
X * ( s ) = ∑ x (kT )e −kTs (1)
k =0

‘ The Z-transform of the sequence x(k) = x(kT).

+∞
X ( z ) = ∑ x ( k ) z −k (2)
k =0

Due to z = e , the right hand-side of the expression (1)

Ts
‘
and ((2)) are identical. So pperformingg Z-transform of a signal
g
is equivalent to discretizing this signal.
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 11
 Properties of the Z-
Z-transform
Given x(k) and y(k) being two sequences which have the Z- Z
transforms:
Z {x ( k )} = X ( z ) Z {y ( k )} = Y ( z )
‘ Linearity: Z {ax ( k ) + by ( k )} = aX ( z ) + bY ( z )
‘ Ti shifting:
Time hif i Z {x ( k − k0 )} = z − k0
X ( z)

‘ Scale in Z-domain: Z {a k
x ( k )} = X ( a −1
z)

‘ Derivative in Z-domain: Z {kx ( k )} = − z

dX ( z )
dz
‘ Initial-value theorem: x (0) = lim X ( z )
z →∞

‘ Final-value theorem: x ( ∞) = lim(1 − z −1 ) X ( z )

z →1

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 12

 The Z-
Z-transform of basic discrete signals

‘ Di impulse:
Dirac i l δ(k)
⎧1 if k = 0
δ (k ) = ⎨ 1
⎩0 if k ≠ 0 k

Z {δ (k
( k )} = 1
0

‘ Step function: u(k)

⎧1 if k ≥ 0 1
u(k ) = ⎨
⎩0 if k < 0 k
0
Z {u( k )} =
z
z −1
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 13
 The Z-
Z-transform of basic discrete signals (cont’)
‘ Ramp function: r(k)
⎧kT if k ≥ 0
r(k ) = ⎨
⎩0 if k < 0 k
0
Z {u( k )} =
Tz
(z − 1)2
‘ Exponential function: x(k)
⎧e -akT if k ≥ 0
x(k ) = ⎨
⎩0 if k < 0 k
0
Z {x ( k )} =
z
z − e −aT
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 14
 Discrete transfer function

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 15

 Derive transfer function (TF) from difference equation
u(k) y(k)
Discrete system

‘ The input-output
Th i t t t relation l ti shiphi off a discrete
di t system
t can be b
described by the difference equation:
a0 y ( k + n ) + a1 y ( k + n − 1) + ... + an −1 y ( k + 1) + an y ( k ) =
b0u( k + m) + b1u( k + m − 1) + ... + bm−1u( k + 1) + bmu( k )
where n>m, n is the order of the system.

‘ Taking the Z-transform

Z transform the two sides of the above equation:
a0 z nY ( z ) + a1 z n −1Y ( z ) + ... + an −1 zY ( z ) + anY ( z ) =
b0 z mU ( z ) + b1 z m −1U ( z ) + ... + bm−1 zU ( z ) + bmU ( z )
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 16
 Derive TF from difference equation ((con’t
con’t))

‘ T ki the
Taking th ratio
ti Y(z)/U(z)
Y( )/U( ) to
t obtain
bt i the
th transfer
t f function:
f ti
Y ( z ) b0 z m + b1 z m −1 + ... + bm −1 z + bm
G( z) = =
U ( z ) a0 z n + a1 z n −1 + ... + an −1 z + an

‘ The above transfer function can be transformed into the

equivalent form:
− ( n −m ) −1 − m +1 −m
Y ( z) z [b0 + b1 z + ... + bm −1 z + bm z ]
G( z) = =
U ( z) a0 + a1 z −1 + ... + an −1 z −n +1 + an z −n

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 17

 Derive TF from difference equation _ Example
‘ Consider a system described by the difference equation.
equation
Derive its transfer function:
y ( k + 3) + 2 y ( k + 2) − 5 y ( k + 1) + 3 y ( k ) = 2u( k + 2) + u( k )

‘ Solution: Taking the Z

Z-transform
transform the difference equation:

z 3Y ( z ) + 2 z 2Y ( z ) − 5zY ( z ) + 3Y ( z ) = 2 z 2U ( z ) + U ( z )

Y ( z) 2z2 + 1
⇒ G( z) = = 3
U ( z ) z + 2 z 2 − 5z + 3

Y ( z) z −1 ( 2 + z −2 )
⇔ G( z) = =
U ( z ) 1 + 2 z −1 − 5z −2 + 3z −3
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 18
 Calculate transfer function from block diagram
R(s) Y(s)
+ GC(z) ZOH G(s)
− T

H(s)

Y ( z) GC ( z )G ( z )
‘ The closed-loop TF: Gk ( z ) = =
R ( z ) 1 + GC ( z )GH ( z )
where
GC (z ) : TF of the controller, derive from difference equation

⎧⎪ G ( s ) ⎫⎪ ⎧⎪ G ( s ) H ( s ) ⎫⎪
−1
G ( z ) = (1 − z )Z−1
⎨ ⎬ GH ( z ) = (1 − z )Z ⎨ ⎬
⎪⎩ s ⎪⎭ ⎪⎩ s ⎪⎭
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 19
 Calculate TF from block diagram – Example 1
‘ Find the closed-loop transfer function of the system:
R(s) Y(s)
+ ZOH G(s)
−
T = 0.5
3
G (s) =
s+2

Solution: G ( z ) = (1 − z −1 )Z ⎧ G(s) ⎫ −1 ⎧ 3 ⎫
⎨ ⎬ = (1 − z )Z ⎨ ⎬
⎩ s ⎭ ( s + 2) ⎭
⎩ s (s
−2×0.5
−1 3 z (1 − e )
= (1 − z )
2 ( z − 1)( z − e −2×
2×0.5
)

⇒ 0.948 ⎧ a ⎫ z (1 − e − aT )
G( z) =
z − 0.368 Z⎨ ⎬= − aT
⎩ s ( s + a ) ⎭ ( z − 1)( z − e )
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 20
 Calculate transfer function from block diagram – Example 1 (cont’)

‘ The closed-loop transfer function:

0.948
= z − 0.368
G( z)
Gk ( z ) =
1 + G ( z ) 1 + 0.948
z − 0.368

⇒ 0.948
Gk ( z ) =
z + 0.580

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 21

 Calculate TF from block diagram – Example 2
‘ Calculate the transfer function of the system:
R(s) Y(s)
+ ZOH G(s)
−
T = 0.5

H(s)

3e − s 1
Given that G ( s ) = H ( s) =
s+3 s +1
‘ Solution:
The
h closed-loop
l dl transfer
f function:
f i
G( z)
Gk ( z ) =
1 + GH ( z )
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 22
 Calculate TF from block diagram – Example 2 (cont’)

−1 ⎧ G( s) ⎫ 3e − s
• G ( z ) = (1 − z )Z ⎨ ⎬ G( s) =
⎩ s ⎭ ( s + 3)
−1 ⎧ 3 e −s
⎫
= (1 − z )Z ⎨ ⎬
⎩ s( s + 3) ⎭
−3×0.5
−1 −2 z (1 − e )
= (1 − z ) z
( z − 1)( z − e −3×0.5 )

⇒
0.777
G( z) = 2 ⎧ a ⎫ − aT
z (1 − e T )
z ( z − 0.223) Z⎨ ⎬= − aT
⎩ s ( s + a ) ⎭ ( z − 1)( z − e )
z = eTs = e0.5 s
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 23
 Calculate TF from block diagram – Example 2 (cont’)
⎧ G ( s) H ( s) ⎫
−1 3e − s
• GH ( z ) = (1 − z )Z ⎨ ⎬ G ( s) =
⎩ s ⎭ ( s + 3)
−s
−1 ⎧ 3 e ⎫
= (1 − z )Z ⎨ ⎬ H (s) =
1
⎩ s ( s + 3)( s + 1) ⎭ ( s + 1)
−1 −2 z ( Az + B)
= 3(1 − z ) z
( z − 1)( z − e −3×0.5 )( z − e −1×0.5 )
(1 − e −3×0.5 ) − 3(1 − e −0.5 )
A= = 0.0673
3(1 − 3) ⎧ 1 ⎫ z ( Az + B)
Z ⎨ ⎬ = − aT −bT
⎩ −3×0.5
s ( s + a )( s + b ) ⎭ ( z − 1)( z − e )( z − e )
3e −3×0.5 (1 − e −0.5 ) − e −0.5 (1 − e ) −aT
B= b(1 − e = 0).−0346
a(1 − e −bT )
3(1 − 3) A=
ab(b − a)
⇒
0 .202 z + 0. 104
ae−aT (1 − e −bT ) − be −bT (1 − e −aT )
GH ( z ) = 2
z ( z − 0.223)( zB −= 0.607) ab(b − a)
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 24
 Calculate TF from block diagram – Example 2 (cont’)

‘ The closed-loop transfer function:

0.777
Gk ( z ) =
G( z) z 2
( z − 0. 223)
=
1 + GH ( z ) 0.202 z + 0.104
1+ 2
z ( z − 0.223)( z − 0.607)

⇒
0.777( z − 0.607)
Gk ( z ) = 4
z − 0.83z 3 + 0.135z 2 + 0.202 z + 0.104
0.777
G( z) = 2
z ( z − 0.223)
0.202 z + 0.104
GH ( z ) = 2
G
z ( z − 0.223)( z − 0.607)

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 25

 Calculate TF from block diagram – Example 3

‘ C l l t the
Calculate th closed-loop
l dl t
transfer
f function
f ti off the
th system:
t
R(s) e(k) u(k) Y(s)
+ GC(z) ZOH G(s)
− T=0.2

H(s)

−0. 2 s
Given that: G ( s ) = 5e
H ( s ) = 0.1
s2
The controller is described by the difference equation:
u( k ) = 10e( k ) − 2e( k − 1)

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 26

 Calculate TF from block diagram – Example 3 (cont’)

‘ Solution:
S l ti
The closed-loop transfer function:
GC ( z )G ( z )
Gk ( z ) =
1 + GC ( z )GH ( z )
‘ The TF of the controller is calculated from the difference
equation:
u( k ) = 10e( k ) − 2e( k − 1)
⇒ U ( z ) = 10 E ( z ) − 2 z −1E ( z )

U ( z)
⇒ GC ( z ) = = 10 − 2 z −1
E ( z)
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 27
 Calculate TF from block diagram – Example 3 (cont’)
−1 ⎧ G ( s) ⎫ −0.2 s
• G ( z ) = (1 − z )Z ⎨ ⎬ G ( s) =
5e
⎩ s ⎭ s2
−0.2 s
−1 ⎧ 5e
5e ⎫ −1 −1 ( 0. 2) 2
z ( z + 1)
= (1 − z )Z ⎨ 3 ⎬ = 5(1 − z ) z
⎩ s ⎭ 2 ( z − 1 ) 3

⇒
0.1( z + 1)
G( z) =
z ( z − 1) 2
−1⎧ G (s) H ( s) ⎫
• GH ( z ) = (1 − z )Z ⎨ ⎬ H ( s ) = 0.1
⎩ s ⎭
−1 ⎧ G ( s) ⎫
= 0.1(1 − z )Z ⎨ ⎬ ⎧ 1 ⎫ T z ( z + 1)
2
⎩ s ⎭ Z ⎨ 3⎬=
⎩ s ⎭ 2( z − 1)
3

⇒
0.01( z + 1)
GH ( z ) =
z ( z − 1) 2 z = eTs = e0.2 s
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 28
 Calculate TF from block diagram – Example 3 (cont’)

‘ The closed-loop transfer function:

⎡10 z − 2 ⎤ ⎡ 0.1( z + 1) ⎤
⎢ ⎥ .⎢ 2 ⎥
GC ( z )G ( z ) ⎣ z ⎦⎣ z ( z − 1) ⎦
Gk ( z ) = =
1 + GC ( z )GH ( z ) ⎡10 z − 2 ⎤ ⎡ 0.01( z + 1) ⎤
1+ ⎢ ⎥.⎢ 2 ⎥
⎣ z ⎦⎣ z ( z − 1 ) ⎦

⇒ z 2 + 0.8 z − 0.2
Gk ( z ) = 4 −1
z − 2 z + 1.1z + 0.08 z − 0.02 C
3 2 G ( z ) = 10 − 2 z
0.1( z + 1)
G( z) =
z ( z − 1) 2
0.01( z + 1)
GH ( z ) =
z ( z − 1) 2
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 29
 State--space model of discrete system
State

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 30

 The discrete state space (SS) equation
‘ The state-space model of a discrete system is a set of first-
order difference equations of the form:
⎧ x( k + 1) = Ad x( k ) + Bd r ( k )
⎨
⎩ y ( k ) = Cd x( k )
where:
h

⎡ x1 ( k ) ⎤ ⎡ a11 a12 K a1n ⎤ ⎡ b1 ⎤

⎢ x (k )⎥ ⎢a a K a ⎥ ⎢b ⎥
x(k ) = ⎢ 2 ⎥ Ad = ⎢ 21 22 2n ⎥
Bd = ⎢ 2 ⎥
⎢ M ⎥ ⎢ M M M ⎥ ⎢M⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎣ n ⎦
x (k ) ⎣ n1 n 2
a a K a nn ⎦ ⎣bn ⎦
C d = [c1 c 2 K c n ]
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 31
 Derive SS equation from difference equation
‘ Case 1: The right-hand side of the difference equation
does not involve the differences of the input:
a0 y ( k + n) + a1 y ( k + n − 1) + ... + an −1 y ( k + 1) + an y ( k ) = b0u ( k )

‘ Define the state variables:

Ž The first state variable is the x1 ( k ) = y ( k )
output off the
h system; x2 ( k ) = x1 ( k + 1)
Ž The ith state variable (i=2..n) x3 ( k ) = x2 ( k + 1)
is set to be one sample time-
time
M
advanced of the (i−1)th state
variable. xn ( k ) = xn −1 ( k + 1)

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 32

 Derive SS equation from difference equation
Case 1 (cont
(cont’))
⎧ x( k + 1) = Ad x( k ) + Bd r ( k )
‘ The state equations:
q ⎨
⎩ y ( k ) = Cd x( k )
where:
⎡ 0 1 0 K 0 ⎤ ⎡0⎤
⎡ x1 ( k ) ⎤ ⎢ 0 0 1 K 0 ⎥ ⎢ 0 ⎥
⎢ x (k )⎥ ⎢ ⎥ ⎢ ⎥
x( k ) = ⎢ 2 ⎥ Ad = ⎢ M M M M ⎥ B =⎢ M ⎥
⎢ M ⎥ ⎢ ⎥ d ⎢0⎥
⎢ ⎥ ⎢ 0a 0 0 K 1 ⎥ ⎢b ⎥
⎣ xn ( k ) ⎦ ⎢ − n − an −1 − an −2 K − a1 ⎥ ⎢ 0⎥
⎢⎣ a0 a0 a0 a0 ⎥⎦ ⎣⎢ a0 ⎥⎦

Cd = [1 0 K 0 0]
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 33
 Derive SS equation from difference equation – Case 1 example

‘ Write the state equations of the system described by:

2 y ( k + 3) + y ( k + 2) + 5 y ( k + 1) + 4 y ( k ) = 3u ( k )
⎧ x1 ( k ) = y ( k )
⎪
‘ Define the state variables: ⎨ x2 ( k ) = x1 ( k + 1)
⎪ x (k ) = x ( k + 1)
⎩ 3 2

⎧ x( k + 1) = Ad x( k ) + Bd r ( k )
‘ The state equations: ⎨
⎩ y ( k ) = Cd x( k )
where: 0⎡0⎤
⎡0 0 ⎤ Bd = 0 = ⎢ 0 ⎥
b0 ⎢ ⎥
0 1 0 1
Ad = 0 0 1 =⎢ 0 0 1 ⎥ ⎢⎣1.5⎥⎦
a3 a2 a1 ⎢ ⎥ a0
− − − ⎢⎣ − 2 − 2.5 − 0.5⎥⎦
a0 a0 a0 Cd = [1 0 0]
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 34
 Derive SS equation from difference equation
‘ Case 2: The right-hand side of the difference equation
involve the differences of the input:
a0 y ( k + n) + a1 y ( k + n − 1) + ... + an −1 y ( k + 1) + an y ( k ) =
b0u ( k + n − 1) + b1u ( k + n − 2) + ... + bn −2u ( k + 1) + bn −1u ( k )
‘ Define
D fi ththe state
t t variable:
i bl
Ž The first state variable is the
x1 ( k ) = y ( k )
output of the system;
x2 ( k ) = x1 (k + 1) − β1r ( k )
Ž The ith state variable (i=2..n)
is set to be one sample time- x3 ( k ) = x2 ( k + 1) − β 2 r ( k )
advanced of the (i−1)th state M
variable minus a quantity xn ( k ) = xn −1 ( k + 1) − β n −1r ( k )
proportional
i l to the
h input
i
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 35
 Derive SS equation from difference equation

Case 2 (cont
(cont’))
⎧ x( k + 1) = Ad x( k ) + Bd r ( k )
‘ The state equation:q ⎨
⎩ y ( k ) = Cd x( k )
where:
⎡ 0 1 0 K 0 ⎤
⎡ β1 ⎤
⎡ x1 ( k ) ⎤ ⎢ 0 0 1 K 0 ⎥ ⎢β ⎥
⎢ x (k )⎥ ⎢ ⎥ ⎢ 2⎥
x(k ) = ⎢ 2 ⎥ A =⎢ M M M M ⎥ B =⎢ M ⎥
⎢ M ⎥ d ⎢ ⎥ d
0 0 0 K 1 ⎢ ⎥
⎢ ⎥ ⎢ a ⎥ ⎢ β n −1 ⎥
⎣ n ⎦
x ( k ) a a
⎢ − n − n −1 − n −2 K − 1 ⎥ a
⎢⎣ β n ⎥⎦
⎣⎢ 0
a a 0 a 0 a 0⎥⎦

Cd = [1 0 K 0 0]
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 36
 Derive SS equation from difference equation

Case
C 2 (cont’)
( t’)
The coefficient βi in the vector Bd are defined as:
b0
β1 =
a0
b1 − a1β1
β2 =
a0
b2 − a1β 2 − a2 β1
β3 =
a0
M
bn −1 − a1β n −1 − a2 β n −2 − K − an −1β1
βn =
a0
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 37
 Derive SS equation from difference equation – Case 2 example

‘ Write the state equations of the system described by:

2 y ( k + 3) + y ( k + 2) + 5 y ( k + 1) + 4 y ( k ) = u( k + 2) + 3u( k )
⎧ x1 ( k ) = y ( k )
⎪
‘ Define the state variables: ⎨ x2 ( k ) = x1 ( k + 1) − β1r ( k )
⎪ x ( k ) = x ( k + 1) − β r ( k )
⎩ 3 2 2

⎧ x( k + 1) = Ad x( k ) + Bd u( k )
‘ The state equations: ⎨
⎩ y ( k ) = Cd x( k )
where: ⎡ β1 ⎤
0 1 0 ⎡0 1 0 ⎤ Bd = ⎢ β 2 ⎥
⎢ ⎥
Ad = 0 0 1 =⎢ 0 0 1 ⎥ ⎣⎢ β 3 ⎥⎦
a3 a2 a1 ⎢ ⎥
− − − ⎢⎣ − 2 − 2.5 − 0.5⎥⎦
a0 a0 a0 Cd = [1 0 0]
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 38
 Derive SS from difference equation – Case 2 example (cont’)

‘ ffi i t βi in
Th coefficient
The i the
th vector
t Bd are calculated
l l t d as:
⎧ b0 1
⎪ β1 = a = 2 = 0.5
⎪ 0
⎪ b1 − a1β1 0 − 1 × 0.5
⎨β 2 = = = −0.25
⎪ a0 2
⎪ β = b2 − a1β 2 − a2 β1 = 3 − 1 × ( −0.25) − 5 × 0.5 = 0.375
⎪ 3
⎩ a0 2

⎡ 0.5 ⎤
⇒ Bd = ⎢ − 0.25⎥
⎢ ⎥
⎢⎣ 0.375 ⎥⎦

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 39

 Formulation of SS from block diagram
r(t) e(t) e(kT) eR(t) y(t)
+ ZOH G(s)
−
T

‘ Step 1: Write the state space equations of the open-loop

continuous system:
⎧ x& (t ) = Ax(t ) + BeR (t )
eR(t) y(t) ⎨
G(s) ⎩ y (t ) = Cx(t )
‘ Step 2: Calculate the transient matrix:
Φ (t ) = L [Φ ( s )]
−1

Φ ( s ) = (sI − A)
-1
where
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 40
 Formulation of SS equations from block diagram (cont’)

‘ Step 3: Discretizing the open-loop

open loop continuous SS equation:
e(kT) y(kT)
ZOH G(s)

⎧ Ad = Φ (T )
⎧ x[( k + 1)T ] = Ad x( kT ) + Bd eR ( kT ) ⎪⎪ T
⎨ with ⎨ Bd = ∫ Φ (τ ) Bdτ
⎩ y ( kT ) = Cd x( kT ) ⎪ 0
⎪⎩Cd = C
‘ Step 4: Write the closed-loop discrete state equations
( hi h has
(which h input
i t signal
i l r(kT))
(kT))
⎧ x[( k + 1)T ] = [Ad − Bd Cd ]x( kT ) + Bd r ( kT )
⎨
⎩ y ( kT ) = Cd x( kT )
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 41
 Formulation of SS equations from block diagram – Example

‘ Formulate
l theh SS equations
i d
describing
ibi the
h system:

r(t) e(t) e(kT) eR(t) 1 x2 1 x1 K y(t)

+ ZOH
− s+a s
T

where a = 2, T = 0.5, K = 10

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 42

 Formulation of state equations from block diagram – Example (cont’)

‘ S l ti
Solution:
eR(t) 1 x2 1 x1 10 y(t)
‘ Step 1:
s+2 s
X 2( s )
X 1 (s) = ⇒ sX 1 ( s ) = X 2( s ) ⇒ x&1 (t ) = x 2 (t )
s
E R ( s) ⇒ ⇒ x&2 (t ) = −2 x2 (t ) +e R (t )
X 2 ( s) = ( s + 2) X 2 ( s ) = E R ( s )
s+2

⎡ x&1 (t ) ⎤ ⎡0 1 ⎤ ⎡ x1 (t ) ⎤ ⎡0⎤
⎢ x& (t )⎥ = ⎢0 − 2⎥ ⎢ x (t )⎥ + ⎢1⎥ eR (t )
⎣ 2 ⎦ 1 ⎣ 4243⎦ ⎣ 2 ⎦ ⎣{⎦
⇒ A B
⎡ x1 (t ) ⎤
y (t ) = 10 x1 (t ) = [10 0]⎢
123 ⎣ x2 (t )⎥⎦
C
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 43
 Formulation of state equations from block diagram – Example (cont’)

‘ Step 2: Calculate the transient matrix

−1 −1
⎛ ⎡1 0⎤ ⎡0 1 ⎤ ⎞ ⎛ ⎡s − 1 ⎤ ⎞
Φ ( s ) = (sI − A) = ⎜⎜ s ⎢ −⎢ ⎟ = ⎜⎢ ⎟
-1
⎥ ⎥ ⎟ ⎜ ⎥ ⎟
⎝ ⎣0 1 ⎦ ⎣0 − 2 ⎦ ⎠ ⎝ ⎣0 s + 2 ⎦ ⎠
⎡1 1 ⎤
1 ⎡ s + 2 1⎤ ⎢ s s ( s + 2) ⎥
= ⎢ ⎥ =⎢ ⎥
s ( s + 2) ⎣ 0 s⎦ ⎢ 1 ⎥
0
⎢⎣ s + 2 ⎥⎦
⎧⎡ 1 1 ⎤ ⎫ ⎡ −1 ⎧ 1 ⎫ −1 ⎧ 1 ⎫⎤
⎪ ⎢ ⎥ ⎪ ⎢ L ⎨ ⎬ L ⎨ ⎬⎥
⎪ ⎪ ⎩ s ( s + 2) ⎭ ⎥
Φ (t ) = L −1[Φ ( s )] = L −1 ⎨⎢ s s ( s + 2) ⎥ ⎬ = ⎢ ⎩ ⎭
s
⎪⎢ 0 1 ⎥⎪ ⎢ −1 ⎧ 1 ⎫ ⎥
0 L ⎨ ⎬ ⎥
⎪⎩⎢⎣ s + 2 ⎥⎦ ⎪⎭ ⎢⎣ ⎩s + 2⎭ ⎦
⎡ 1 − 2t ⎤
⇒ ⎢ 1 (1 − e )⎥
Φ (t ) = 2
⎢0 −2t ⎥
⎣ e ⎦
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 44
 Formulation of state equations from block diagram – Example (cont’)

‘ open loop ⎧ x[(k + 1)T ] = Ad x(kT ) + Bd eR (kT )

Step 3: Discretizing the open-loop
⎨
continuous state equations: ⎩ y (kT ) = Cd x(kT )

⎡ 1 −2t ⎤ ⎡ 1 − 2×0.5 ⎤
1 (1 − e ) 1 (1 − e )⎥ ⎡1 0.316⎤
Ad = Φ (T ) = ⎢ 2 ⎥ = ⎢ 2 =⎢ ⎥
⎢0 − 2t ⎥ ⎢ − 2×0.5 ⎥ ⎣ 0 0 . 368 ⎦
⎣ e ⎦ t =T ⎣0 e ⎦

T ⎧⎡
1 − 2τ ⎤ 0 ⎫ T ⎧⎡ 1 − 2τ ⎤ ⎫
T
⎪ 1 (1 − e )⎥ ⎡ ⎤ ⎪ ⎪⎢ (1 − e )⎥ ⎪
Bd = ∫ Φ (τ ) Bdτ = ∫ ⎨⎢ 2 ⎢ ⎥ d τ ⎬ = ∫⎨ 2 dτ ⎬
0 ⎪ ⎢0 − 2τ ⎥ ⎣1⎦ ⎪ 0 ⎪⎢ e −2τ ⎥ ⎪
⎩⎣ e ⎦ ⎭ ⎩⎣ ⎦ ⎭
0

T
⎡⎛ τ e − 2τ
⎞⎤ ⎡⎛ 0.5 e −2×0.5 1 ⎞⎤
⎢⎜⎜ + 2 ⎟⎟⎥ ⎢⎜⎜ + − 2 ⎟⎟⎥
= ⎢⎝ 2 2 ⎠⎥ = ⎢⎝ 2 2 2
2 ⎠⎥ = ⎡0.092⎤
⎢0.316⎥
⎢ e −2τ ⎥ ⎢ e − 2×0.5
1 ⎥ ⎣ ⎦
⎢ − ⎥ ⎢ − + ⎥
⎣ 2 ⎦0 ⎣ 2 2 ⎦

Cd = C = [10 0]

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 45

 Formulation of state equations from block diagram – Example (cont’)

‘ Step 4: The closed-loop discrete state equations:

⎧ x[( k + 1)T ] = [Ad − Bd Cd ]x( kT ) + Bd r ( kT )
⎨
⎩ y ( kT ) = Cd x( kT )

⎡1 0.316⎤ ⎡0.092⎤ ⎡ 0.080 0.316⎤

where [Ad − Bd Cd ] = ⎢ ⎥ −⎢ ⎥ [10 0] = ⎢ ⎥
⎣ 0 0 . 368 ⎦ ⎣ 0 . 316 ⎦ ⎣ − 3 . 160 0 . 368 ⎦

‘ Conclusion: The closed-loop

closed loop state equation is:
⎡ x1 ( k + 1) ⎤ ⎡ 0.080 0.316⎤ ⎡ x1 ( k ) ⎤ ⎡0.092⎤
⎢ x ( k + 1)⎥ = ⎢ − 3.160 0.368⎥ ⎢ x ( k )⎥ + ⎢0.316⎥ r ( k )
⎣ 2 ⎦ ⎣ ⎦⎣ 2 ⎦ ⎣ ⎦
⎡ x1 (k ) ⎤
y (k ) = [10 0].⎢ ⎥
⎣ 2 ⎦
x ( k )
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 46
 Calculate transfer function from state equation

‘ Gi
Given th state
the t t equation
ti
⎧ x( k + 1) = Ad x( k ) + Bd u( k )
⎨
⎩ y ( k ) = Cd x( k )

‘ Th corresponding
The di transfer
f function
f i is:
i
Y ( z)
G( z) = = Cd ( zI − Ad ) −1 Bd
U ( z)

16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 47

 Calculate transfer function from state equation - Example

‘ Calculate the TF of the system described by the SS equation:

⎧ x(k + 1) = Ad x(k ) + Bd u (k )
⎨
⎩ y ( k ) = C d x( k )
⎡ 1⎤ ⎡0 ⎤
Cd = [1 0]
0
Ad = ⎢ ⎥ Bd = ⎢ ⎥
⎣ − 0 . 7 − 0 . 1⎦ ⎣ 2⎦
‘ Solution: The transfer function is:
G ( z ) = Cd ( zI − Ad ) −1 Bd
−1
⎛ ⎡ 1 0⎤ ⎡ 0 1⎤ ⎞ ⎡0⎤
= [1 0]⎜⎜ z ⎢ ⎥ −⎢ ⎥ ⎟ ⎢ ⎥
⎟ 2
⎝ ⎣ 0 1⎦ ⎣ − 0 .7 − 0 . 1⎦ ⎠ ⎣ ⎦
⇒ G( z) = 2
z 2 + 0.1z + 0.7
16 November 2011 © H. T. Hoàng - www4.hcmut.edu.vn/~hthoang/ 48