Professional Documents
Culture Documents
2: P IDfrom
ID controller: controller:
points tofrom Figure
points
plane. 3.2: P ID con
to plane.
Fractional order
3.2.2 Bode’s
3.2.2 ideal
For a wide class of controlled D. Van we
objects loopideal
Bode’s
Nostrandtransfer
loopfunction
transfer
3.2.2 function
recommend Company, Bode’s ideal loop
the fractional Inc.,
P I n DNew
δ
York, 1945.
controller, which is a particular case of P I λ D δ controller, where λ = n, n ∈
Fractional-order systems
Fractional order systems and controllers Bode’s ideal loop transfer function
CHAPTER 3. FRACTIONAL-ORDER CONTROLLERS
CHAPTER 3. FRACTIONAL-ORDER CONTROLLERS
CHAPTER 3. FRACTIONAL-ORDER CONTROLLERS
14
14
W(s) Y(s)
K
G(s) = s!
W(s) Y(s)
K
G(s) = s!
W(s) + E(s) U(s) Y(s) CHAPTER 3. FRACTIONAL-ORDER CONTROLLERS 14
Gc(s) Gs(s) Figure 3.3: Bode’s ideal loop.
W(s) Y(s)
– G(s) =
K
Figure
(a) Open loop:
s ! 3.3: Bode’s ideal loop.
• Magnitude: constant slope of −α20dB/dec.;
(a) Open loop: • Crossover frequency: a function of K;
General characteristics
•Figure 3.3: Bode’s
of the Bode’s
ideal loop.
ideal line
• Phase: horizontal
slope of −α20dB/dec.;
transfer
of −α ; function: π
2
1 Magnitude: constant
Gs ( s ) = βn
• Nyquist: straight line at argument −α π2 .
+ an −1s βn −1 + ! + a1s β1 + a0 s β0
• Crossover frequency: a function of K;
an s Start (a) Open loop: (b) Closed loop:
!! "" • Phase: horizontal line of −α π2 ;
• Magnitude: constant slope of −α20dB/dec.;
Figure 3.2: P ID controller: from points to plane.
• Gain margin: Am = infinite;
βn
an D y (t ) + an−1D βn −1 β0
y (t ) + ! + a0 D y (t ) = u (t ) ! " • Nyquist: straight line at argument
• Crossover frequency: a function of K;
−α π2 . ! "
• Phase margin: constant : Φm = π 1 − α2 ;
63 / 90 • Phase:
(b)horizontal line of −α π2 ;
Closed loop: • Step response: y(t) = Ktα Eα,α+1 (−Ktα ) ,
• Gain margin: AmFor a wide class of controlled objects we recommend the fractiona
Back • Nyquist: straight line at argument −α π2 . where Ea,b (z) is Mittag-Leffler function of two parameters [43].
= infinite;
Full screen (b) Closed loop: ! " λ δ
• Phase margin: controller,
constant : Φmwhich
= π 1 −is α
2
;a particular case of P I D controller, where λ =
Close • Gain margin:• AStep
3.2.3 Illustrative example
m = infinite;
N and
response: y(t)
! =
α
"δ
Kt ∈
α
E R.
α,α+1 Integer
(−Kt α
), order integrator is important for steady-sta
We can illustrate the fractional order control properties by an example. As-
End • Phase margin: where
constant : Φ = π 1 − ;
Ea,b (z) is Mittag-Leffler
m 2 function of two parameters [43].
cancellation but on the other hand the fractional integral is also impo
• Step response: y(t) = Ktα Eα,α+1 (−Ktα ) ,
suming that the transfer function of the DC motor is
2 where Ea,b (z) is Mittag-Leffler function of two parameters [43]. Km 5
obtaining a Bode’s ideal loop transfer function response
3.2.3 Illustrative example
G(s) =
Js(s + 1)
, (3.9) with consta
Km
G(s) = 3.2.2
m
Bode’s ideal loop transfer function
transfer, function that gives this phase
Js(s + 1)
(3.9)
margin is
G(s) = , (3.9)
Js(s + 1) 1
with J being the payload inertia. Our specification is the constant Go (s) = √
phase . (3.10)
with J beingmargin
the payload inertia. Our specification
Bode suggested an ideal shape of the loop transfer function in his
is the constant phase
s3s
independent of the payload changes.
margin independent of the payload changes. design of feedback amplifiers in 1945. Ideal loop transfer function has
Assume that we would like to have a closed loop system that is insensitive
Assume that we would like to have a closed loop system that is insensitiveo
to gain variations with a constant phase margin of 60 . Bode’s ideal loop
to gain variations with a constant phase margin of 60o. Bode’s ideal loop ! " α
transfer
transfer function function
that gives that margin
this phase gives this
is phase margin is s
L(s) = ,
1
G (s) = √ .
o
1
Go (s) = √ . (3.10) (3.10) ωgc
s3s s3s
where ωgc is desired crossover frequency and α is slope of the idea
characteristic.
Phase margin is Φm = π(1 + α/2) for all values of the gain. The am
margin Am is infinity. The constant phase margin 60o , 45o and 30o co
• Gain margin:
## Am = infinite;
22 ! "
• Phase margin: constant : Φm = π 1 − α2 ;
##
• Step response:
""
22 y(t) = Ktα Eα,α+1 (−Ktα ) ,
where Ea,b (z) is Mittag-Leffler function of two parameters [43].
##
log
log !!
PI!Dµ controllers
suming that the Illustrative
transfer function of the DC motor is
3.2.3 Bode’s ideal loop transfer function: example
example
Kmproperties by an example. As-
We can illustrate the fractional order control
The
suming that the transfer G(s) of=theofDC
function
transfer function a DC ,is
motormotor is (3.9)
Js(s + 1) Fractional-order PID controllers:
G(s) =
Km
, J is payload inertia
(3.9) from points to plane
with J being the payload inertia. Our+specification
Js(s 1) is the constant
phase CHAPTER 3. FRACTIONAL-ORDER CONTROLLERS 13
margin independent of the payload changes.
with J being the payload inertia. Our specification is the constant phase
Assume that we would like to have a closed loop
Assume that
margin we would
independent like
of the to have
payload a closed
changes.
gainloop systemwith thata is insensitive
Assumesystem that
that we would is
likeinsensitive to loop
to have a closed variations
system thatois insensitive
to gain variations with
constant
to gain variations
a constant
withphase
phase
margin
a constant
margin
phaseofmargin
60 . Bode’s
o of 60
of 60o. ideal
. Bode’s
Bode’sloop
ideal loop
ideal loop
transfer function that
transfer
CHAPTER
transfer function gives
3. this
that function phase
FRACTIONAL-ORDER
gives this phase margin
that margin
gives this isphase margin is
CONTROLLERS
is 15
PI!Dµ controllers
#
! " !log ! " ! ωgc
Bode’s ideal loop transfer function: example
log
Since Go (s) = C(s)G(s), J we can find1 the controller transfer
1 function in the
C(s) = s2/3 + 1/3 = K s2/3 + 1/3 , (3.11) where ωgc is desired crossover frequency and α is slope of the ideal cut-off
following form Km s s
Figure
Figure 3.4:!plots
3.4: Bode Bodeof plots" of function
transfer transfer
! function
(3.8). " (3.8).
which is a particular case Jof the2/3 1 function of
transfer 2/3the P1I λ D δ controller (3.1), characteristic.
C(s) = s + 1/3 = K s + 1/3 , (3.11)
Kcontroller s controller transfer s function in the Phase margin is Φm = π(1 + α/2) for all values of the gain. The amplitude
Since
where
Since GK(s)==J/K
Sinceo Go (s) =mC(s)G(s), , we
is thewe
C(s)G(s), m
the
cancontroller
constant.
can find the transfer
find the controller function
transfer is in the
function o
which
The is
following a particular
phase
form margincase of the
of the transfer function
controlled system with I λ D δ controller
of theaPforward (3.1),
loop controller margin Am is infinity. The constant phaseComparison
margin 60of 45o and
, unit 30o correspond
step responses
following form ! " ! "
Gwhere
c (s) is
K = J/Km is the J controller
2/3 ! 1
constant. "2/3 1! " to the slopes α = −1.33, −1.5 and −1.66.of the fractional-order “reality” and
C(s)
The phase margin = s +
J controlled = K1 s + , (3.11)
1 controller its integer-order “model”
C(s)Kmof=Φthe
m = arg[C(jω
2/3
ss1/3 + 0system
)G(jω with
0=)] K a forward
+s1/3
π,s2/3 + loop , (3.12) (3.11) The Nyquist curve for ideal Bode transfer function is simply a straight line
Gc (s) is Km s1/3 s1/3
which
where is ω
a 0particular case of the
is the crossover transfer function of the P I λ D δ controller (3.1),
frequency. through the origin with arg(L(jω)) = απ/2 (see e.g. [1, 39]).
Φm = this
is is
of arg[C(jω a 0particular case of fractional PID (3.1),
)G(jω0 )] + π, (3.12)
which
where K=
The is a particular
J/K m is phase
obtained case
the controller
margin the transfer
constant. function of the P I λ D δ controller
where ωK0 is the crossover frequency. #constant. $ Bode’s transfer function (3.6) can be used as a reference system in the
The phase =margin mofisthethecontrolled system
where J/K
The obtained phase margin is 60 : controller 1with o a forward loop 4 πcontroller
π following form:
Gc (s)ΦisThe
m =obtained phase margin
arg[C(jω)G(jω)] is arg = π −a forward
The phase margin of +the π= controlled
# (jω)4/3 system
$
+ π with
32
= . loop
3
(3.13)
controller K
Φm = arg[C(jω0 )G(jω10 )] + π, 4π π(3.12) Gc (s) = α (0 < α < 2) (3.7)
G (s)
Φ
The constant
c mis= arg[C(jω)G(jω)] + π =
phase margin is not dependent arg + π =
4/3 of the payload
π − =
changes. (3.13)
and the s +K
where ω0 is the crossover frequency. (jω) 3 2 3
system gain K and phase curve Φm =isarg[C(jωa horizontal 0 )G(jωline0at)] + π,
−2π/3. (3.12) K
The constant
obtained phase margin
margin isis not dependent of the payload changes and the
Step response:
The
Step response
where
system ωgain0 is K
of
theand closed control
crossover
phase
# loop can
frequency.
curve is a
$ be expressed as:
horizontal
1 line at −2π/3.
4π π
Go (s) = α
s
(0 < α < 2), (3.8)
ΦmThe %
= arg[C(jω)G(jω)] = arg& is 4/3 + π = π − ' = . ( (3.13)
+1πmargin
Step obtained phase
y(t)response
= L−1 %of closed control loop
1+1/3 + 1) &
(jω)
= canEbe expressed
t1+1/3# 1+1/3, 2+1/3 $
as:1+1/3
3 2−t 3 , (3.14) where Gc is transfer function of closed loop and Go (s) is transfer function in
The constant phase s (s '
of the1 payload changes ( 4 πthe π
−1 margin is 1 not dependent 1+1/3 1+1/3 and open loop. General characteristics of Bode’s ideal transfer function are:
my(t) = L = targ E1+1/3, 2+1/3+ −t −, (3.14)
whereΦgain
system step=Kresponse
arg[C(jω)G(jω)]
and phase is
s (s independent
curve
1+1/3 is
+ 1)
+
a π = of the
horizontal payload
line
(jω) at inertia
−2π/3.
4/3
π =andπ α =
32
=
4/3. 3
. (3.13)
Step response of closed independent
control loop can be payload
expressed as: 8
where
The step response
constant %phaseis margin & is not of the
dependent inertia
of the and α = 4/3.
payload changes and the
1 ' 1+1/3 (
systemy(t) = L−1 K and
gain phase curve = t1+1/3is aE1+1/3, 2+1/3 −t
horizontal line at, −2π/3. (3.14)
s (s1+1/3 + 1)
Step response of closed control loop can be expressed as:
where step response is %
independent of the
& payload inertia and α = 4/3.
1 ' (
y(t) = L−1 = t1+1/3 E1+1/3, 2+1/3 −t1+1/3 , (3.14)
s (s1+1/3 + 1)
where step response is independent of the payload inertia and α = 4/3.
9
ss
being developed is the method of dominant roots [37], based on the given
stability measure and the damping ratio of the closed control loop. Assuming
that, the desired dominant roots are a pair of complex conjugate root as follows:
s1,2 = −σ ± jωd ,
designed for the damping ratio ζ and natural frequency ωn . The damping
constant (stability measure)
! is σ = ζωn and the damped natural frequency of
oscillation ωd = ωn 1 − ζ 2 . The design of parameters: Kp , Ti , λ, Td and δ
where w(t) is the desired value of closed control loop and y(t) is the real value
of closed control loop.
This method does not insure the desired stability measure of the closed
control loop. Measure of stability has to be checked out additionally. We can
use a frequency method described in [39].
• MATLAB command
• Display
+ + + + + +
1
- - - + + +
Y2n* (s)
Z2n -1(s)
Y2n -2 (s)
Z2n-3 (s)
Y2 (s)
can be expressed in the form: expansion, which is identical with the equation (6.1):
1
b1 (s) Z(s) = Z1 (s) +
1
G(s) ! a0 (s) + Y2 (s) +
CHAPTER 4. CONTINUED FRACTION EXPANSION (CFE) 22 Z3 (s) + 1
b2 (s)
a1 (s) + b3 (s) Y4 (s) +
1
a2 (s)+ a (s)+... + + + + + + .................................
3
CHAPTER 4. CONTINUED FRACTION EXPANSION (CFE) 19 - - -
1
+ + +
1
b1 (s) b2 (s) b3 (s) 1
Y2n−2 (s) +
= a0 (s) + ... (4.1) Y2n* (s) 1
Z2n−1 (s) +
a1 (s)+
+ a2 (s)+ a3 (s)+ Y2n (s)
G(s) Z2n -1(s)
- Similarly to the above considerations, we can obtain a continued fraction
where ai s and bi s are rational functions of the variable s, or are constants. Y2n -2 (s)
expansion of the transfer function of the other interesting type of a nested
H(s)
The application of the method
CHAPTER yields a rational
4. CONTINUED FRACTION function,
EXPANSION ! (CFE)which is19an
G(s), Z2n-3 (s) multiple-loop control system, depicted in Fig. 4.7:
4.3 control
CFE and
4.4 CFEsystems
stability of linear
and nested systems
multiple-loop
.................................
1
1
Y2n−2 (s) +
control
CHAPTER systems
4. CONTINUED FRACTIONbetween EXPANSION (CFE)frac- 20 1
Let us now establish an interesting new relationship continued Z2n−1 (s) +
Y2n (s)
CFEs and nested multiple loops CFEs and nested multiple loops
It is also knowntionsthat continuous
and nested fraction
multiple-loop expansions
control systems. can be used for investi-
Let us now establish an interesting new +relationship between continued frac-
gating
CHAPTER stabilityFRACTION
4. CONTINUED of
Welinear
first
tions systems.
recall
and the known
nested
EXPANSION Forfact
this,
that
multiple-loop
(CFE) thethecharacteristic
control
19 transfer
-
systems.
1
function polynomial Q(s)
R(s) of the control Similarly to the above considerations, we can obtain a continued fraction
expansion of the transfer function of the other interesting type of a nested
loop with
of the differential aWe
equation negative feedback
firstofrecall
the system
the shown
known in Fig.
should
fact that 4.1
thebe is *given
divided
transfer by [8]
inR(s)
function twoof parts, the
the control multiple-loop control system, depicted in Fig. 4.7:
Y (s) 2n
+ loop with a negative feedback shown in Fig. 4.1 is given by [8]
“even” part (containing
-
G(s) even powers of s) and G(s) the “odd” part (containing odd Z(s) =
1
(4.9)
R(s) =
Figure 4.2: Nested multiple-loop
G(s) . (4.4)
control system – level 1. 1
powers of s): H(s) R(s)1 =+ G(s)H(s) . (4.4) Z1 (s) +
CHAPTER 4. CONTINUED FRACTION EXPANSION (CFE)Y2 (s) + 23 1
1 + G(s)H(s)
+ + 1
Q(s) = m(s)
From (4.4) it immediately + n(s).
follows 1
that the- transfer function of the circuit Z3 (s) +
Figure 4.1: A control loop with From
a negative
(4.4)feedback. +
it immediately follows that the transfer function of the circuit + + + + + .................................
Y2n(s)
CHAPTER 4. CONTINUED shown in Fig. 4.2 is (CFE) - - - 1
FRACTION EXPANSION 20 + +
Then this two parts of the characteristic
CHAPTER 4. CONTINUED FRACTIONshown in Fig.(CFE)
EXPANSION 4.2 is 20 polynomial
Y * (s)are used for creating 2n
Y2n−2 (s) +
1
The rational function R(s) should be written in the form of a continuous Z2n -1(s) 1
1 1 11 Z2n−1 (s) +
fraction: its test function + - in1 the
+ form of Pa2nfraction, =(s) = in which
(s)P2n ∗ ∗ = =
the highest
Z2n -1(s) ,, power of(4.5)s is
(4.5) Y2n (s)
1 1 11 ·+Y12n
1 +(4.3) · (s)
Y2n (s) Y2n
Y(s)
R(s) = - 2n (s) Y2n -2 (s)
contained in bthe
1 s +
denominator:
Y * (s)
2n
1
Y2n* (s) where Y 1 Y(s) ∗ = Y ∗ (s) +"1. 4.3: Nested multiple-loop
where b2 sY +2n (s) =2n 2n (s) + 2n1.Figure # control system – level 2. Z2n-3 (s)
. m(s)
. . . . . . .system
Figure 4.2: Nested multiple-loop control . . . . – level 1. n(s)
R(s) =
Figure 4.2: Nested multiple-loop control system – level 1. or R(s) = .
1 Using the equations (4.4) and (4.5) we obtain the transfer function of the
Y2n-4 (s)
+
+
-
- 1
1
+
bn−1
+ s + bsystem
+ ns
n(s)
shown in Fig. 4.3:
m(s) Nested loop of type II
If bk > 0, k = 1, . . . , n, then theY2nsystem
* (s) is +stable. If some bk is negative, 1
Z 1 (s)
then the system is unstable. Y2n* (s) Q2n−1 (s) = Z2n−1 (s) + P2n (s) = Z2n−1 (s) + . (4.6)
Y2n (s) Figure 4.7: Nested multiple-loop control system of the second type.
Considering the continued fraction Z (s) (4.3) as a tool for designing a corre-
Z2n2n-1-1(s)
sponding LC circuit, we can conclude that stability of a Combining linear systemthe equations (4.4) and (4.5) we find the transfer function of
is equiv-
Figure 4.3: Nested multiple-loop control system – level Both types of nested multiple-loop systems, presented in this section, can
alent to realizability
Figure 4.3: of its test
Nested function control
multiple-loop R(s) with the
systemthe help2.2.ofmultiple-loop
–nested
level only passive system shown in Fig. 4.4:
be used for simulations and realizations of arbitrary transcendental transfer
electric components.
Using the equations (4.4) and (4.5) we obtain theCHAPTER transfer function 4. of
CONTINUED
the Q2n−1 (s)
FRACTION 1
EXPANSION (CFE) 20 functions. For this, the transfer function should be developed in a continued
Using the equations (4.4) and (4.5) we obtain the transfer function P2n−2of(s)the = = fraction, which after truncation can be represented by a nested multiple-loop
system shown in Fig. 4.3: 1 + Q2n−1 (s)Y2n−2(s) 1
system shown in Fig. 4.3: Y2n−2 (s) + system shown in Fig. 4.6 or Fig. 4.7.
4.4 CFE and nested multiple-loop 1 +
1
Q2n−1 (s)
Q2n−1 (s) = Z2n−1 (s) + P2n (s) = Z2n−1 (s) + 1 . (4.6)
Q2n−1 (s) = Z2n−1 (s) + P2n (s) = Z2n−1 (s) + Y2n (s). (4.6) - 1
control systems Y2n (s) =
CHAPTER 4. CONTINUED FRACTION EXPANSION 1 (CFE) 20
(4.7)
4.5 CFE and rational and irrational numbers
Combining the equations (4.4) and (4.5) we find the transfer function of Y2n−2 (s) + Y2n* (s)
Combining the equations (4.4) and (4.5) we find the transfer function of 1
the
Letthe nested
usnested multiple-loop
now establish system shown
an interesting new in Fig. 4.4:
relationship between continued frac- Z2n−1 (s) +
multiple-loop system shown in Fig. 4.4: + Every rational and irrational number may be written in CFE form. Basically
tions and nested multiple-loopQ2n−1 control 1 Y2n (s)
(s) systems. 1 -
Figure 4.2: Nested multiple-loop control system – level 1. the process of finding a continued fraction development consists of two steps:
Q2n−1 (s) 1
2n−2 (s) =
We firstPPrecall
2n−2 (s)the= known
1 + Qfact that2n−2
2n−1 (s)Y
=
the(s)transfer
= function R(s) 1 of the control
1 + Q2n−1 (s)Y2n−2 (s) Y2n−2 (s) + The1 transfer function of the system Y2n* (s)
shown in Fig. 4.5 is then given by the if the fraction m/n is greater than 1, then divide. Otherwise, write the fraction
loop with a negative feedback shown in Fig. 4.1 Y2n−2 is (s) + Qby
given (s)
[8] + + m/n as 1/(n/m) and proceed with the first step. Continue until a numerator
relationship
2n−1
Q2n−1 (s) 1
=
=
1
1
G(s)
1
(4.7)
Figure (4.7)
-
4.2: Nested multiple-loop control system + – level 1. CHAPTER 5. APPROXIMATION OF FRACTIONAL OPERATOR
of 1 is obtained. Continued fractions are often used to get good rational
CHAPTER 5. APPROXIMATION OF FRACTIONAL OPERATOR 25 25
Y2n−2R(s)
(s) += 1 . Q2n−3 (s)(4.4) = Z2n−3 (s) + P 2n−2 (s) approximations for real numbers. Let us consider the numbers written in the
CHAPTER 5. APPROXIMATION OF FRACTIONAL OPERATOR 25
Y2n−2 (s) + 1 + G(s)H(s)
Z (s) + 1
1 Y2n* (s)
CHAPTER 4. CONTINUED FRACTION + 1 form: a0 + b1 /(a1 + b2 /(a2 + b3 /(a3 + . . . ))). In ”simple” continued fractions,
Z2n−1 (s) +EXPANSION
2n−1
Y2n (s)
(CFE) 21
= Z (s) 1 +
+
(4.8)
Y2n (s)
transfer function of the circuit 2n−3 - + 1 all the bi , ∀i, are 1 and the number can be re-written as [a0 ;5.
a1 , aAPPROXIMATION
2 , a3 , . . . ].
From (4.4) it immediately follows that the
The transfer function of the system shown in Fig. 4.5 is then given by the Z2nY-12n−2
(s) (s) + nominator polynomials. nominator Probably,
polynomials. this Probably,
fact wasthis
CHAPTER noted fact forwas the first
OF FRACTIONAL OPERATOR
noted time
for the firstintime[17],in [17], 25
shownThe transfer
in Fig.
relationship
4.2 function
is of the system shown in Fig. 4.5 is then given by the Y2n* (s)
Z (s) +
1 nominator polynomials.
Example 4.5.1.
where Probably,
theappeared:
following this fact
idea was noted
appeared: for the
a CHAPTER
dense 5.first
interlacing timeofinsimple
APPROXIMATION [17], OF poles and
FRACTIONAL OPERATOR 25
relationship 2n−1
Y2n (s) where where
the following
the
The CFE idea
following
for π, which
idea
gives the ”best”aapproximation
appeared:
dense
nominatorinterlacing
polynomials.
of a given order, is of simple
Probably, this fact waspoles
noted for andthe first time in [17],
Q2n−3 (s) P
(s) = 2nZ(s)
= Z2n−3 =(s)+
1
+ P2n−2∗(s) =
1
, Figure 4.3: (4.5)
Nested multiple-loop
Z2n -1(s) control system – level 2. [3, 7, 15, zeros along
1, 292, 1, 1, 1, 2,a1,line
3, 1, in the1,a1,swhere
14, 2,
dense
plane
2, 2, 2,the
interlacing
2,is,...]. inThesome
following very
of simple
way,
idealarge
appeared: equivalentpolesinterlacing
a dense toanda branch of wascut;poles and
simple
Q2n−3 2n−3 (s)
1 +P12n−2
· Y2n (s)(s) Y12n (s)
1
Continuing this process, we obtain the transfer function of the nested multiple- zeros along a
zeros along line
term 292 in
a line
and
the
meansin α the
sthat
s
, 0the
plane
sα plane
<following
is, in
< 1,convergent
viewed
some
is, in iszeros
some
asa good
an
way,
way,
approximation
operator,
along a line
equivalent
nominator
equivalent
[13].
has
in the s aplane
to
polynomials. a
tois,ainidea
branch branch
cut
some
branch
Probably,
along
way,cut;
thiscut;
the
fact
negative
equivalent
noted for the first time in [17],
tointerlacing
a branch cut;
= Z2n−3 (s) + (4.8) where the following appeared: a dense of simple poles and
= Z (s) + loop
1 control system
(4.8) shown in Fig.
(4.5)4.6 we in the form of a2. continued fraction
where Y2n (s) = Y2n ∗
(s) ++1. 2n−3+ Y2n−2 (s) ++
1
1 Using the
1
Figure 4.3: Nested
equations multiple-loop
(4.4) and control system
obtain the– transfer
level function of and sαand
the , 0 <sαα
, 0< < 1,α< viewed
1,axis
real viewed as
for an
as an
argumentsoperator,
355 and
operator,
[3, 7, 16] = of= s
has
on
sα , 0 <
has(−π, aa. .α.branch
branch
< 1, viewed
π) but is
zeros cut
alongcut
as along
otherwise along
aanline
operator,
inthe
free
the the
has
of
a negative
branch cut along the negative
snegative
poles
plane is, inand
somezeros.
way, equivalent to a branch cut;
- - Y2n−2 (s) + Z+2n−1system (s) + 1shown in Fig. 4.3: [3, 7, 15, 1] = 3.14159292
113 real axis for arguments αof s on (−π, π) but is otherwise free of poles and zeros.
Z2n−1 (s) + Y2n (s) and s free
, 0 < ofα <poles
1, viewed
CHAPTER 4. CONTINUED FRACTION EXPANSION
Y * (s)
(CFE)
Using
Y 2n (s) 21 (4.4) and (4.5) we obtain the transfer function of the
the equations real axis
realfor
axisarguments
for arguments of sof on s on (−π, (−π,π)π)but but isis otherwise
otherwise free of poles andasand
an operator, has a branch cut along the negative
zeros. zeros.
Continuing this process, we obtain the2ntransfer function system shownmultiple- 1 real axis for arguments of s on (−π, π) but is otherwise free of poles and zeros.
of the nested Qin2n−1
Fig.(s)
4.3:
= Z2n−1 (s) + P2n (s) = Z2n−1 (s) + . (4.6)
Z1 Z3 Z 2n -3 Z 2n -1
A devices or a circuit exhibiting fractional-order
Z1 Z3 Z 2n -3 Z 2n -1
behaviour is called a fractance. Figure 6.2: Finite ladder circuit.
• domino ladder circuit network, Let us consider the circuit depicted in Fig. 6.2, where Z2k−1 (s) and Y2k (s),
k = 1, . . . , n, are given impedances of the circuit elements. The resulting
• a tree structure of electrical elements, impedance Z(s) of the entire circuit can be found easily, if we consider it in
Figure 6.2: Finite ladder circuit.
the right-to-left direction:
• transmission line circuit Z(s) =ladder
Domino
Figure 6.2: Finite ladder
1 circuit.
Z1 (s)+ lattice networks can approximate fractional operator more
1
(6.1)
Y2 (s) +
effectively than the lumpedZ3networks [10, 30]. 1
(s) +
Domino ladder Let lattice
us considernetworks
the circuit depicted
can in Fig. 6.2, where1fractional
Y4 (s)approximate
+ Z2k−1 (s) and Yoperator 2k (s), more
k = 1, . . . , n, are given impedances of the . . . .circuit
. . . . . . . . .elements.
. . . . . . . . . . . . The
. . . . . .resulting
..
Design of fractances can be using a truncated CFE, effectively than the lumped
impedance Z(s) of the networks [10,
entire circuit can30].
1
be found easily, if1we consider it in
Y2n−2 (s) +
1
the right-to-left
Let us consider the circuit direction:
depicted in Fig. 6.2,Zwhere 2n−1 (s) + Z2k−1 (s) and Y2k (s),
which gives a rational approximation. k = 1, . . . , n, Z(s)
are =given 1
Y (s)
2n
(s)in transmision
Z(s) = kkZ=a line
.Z . in. ,series (6.7)
a b
Let us consider the circuit depicted oneZ hasin Fig. to 6.3,
obtain where aand ZYof
continuous
2k−1 (s) =fraction
and YY2k (s),
of(s), = 1,b . . line,
representation n. which of the canfunction Z(s), !15
Magnitude [dB]
(Za ) and the same in shunt (ZbImpedance ), we get a2k−1 such kind 2k transmision be used also as a
Z Z Z
1 3
structure
5
depicted in Fig. 6.4.
2n-1
!20
Z(s) = = s + (6.6)
Take
Z(s) Y !
as a transmission line Y
or Y
a symetrical
Y Y
impedance Y
domino
Y Y
ladder (Z
isZ by O. Heaviside a ) and
lattice the same
network in
as shunt
well. (Z b ), we get a structure depicted in Fig. 6.4. !30
3
as 1
1 2 3 4 5 6 2n-1 2n
Z(s) =model Za .Z R
. real −1/2 R − (6.7)
If we substitute = aofbresistance (s) Z =R
(s)and a by O.s Heaviside
capacitance |+ Zb = in1/sC, then thebe expressed as Frequency [rad/sec]
CHAPTER
impedance 6. GENERAL
is the same APPROACH TO FRACTANCE R in−1/2 DEVICES R −1/2 −jπ/4 34 !50
In the case if we assume elements in transmision = line s series
Phase [deg]
From "this expansion itZ(s)
follows that
Z(s) Zb
If"we substitute aC resistance =
CZaω = Re and |of as=jω .
capacitance Z(6.8)
b = 1/sC, then the
!60
Then
Zb Zb Zb Zb Zb Zb Zb
(Za ) and the same in Za shuntZa(Zb ), we get aR one Zahas to obtain
structure depicted
R a continuous
in Fig. fraction
6.4. representation the function Z(s),
one has to obtain Z(s)
Za
a continuous
= s−1/2 1 =impedance
fraction −1/2 −jπ/41of the function Z(s),
representation
ω(s) eis 4− |s=jω 3 !70
!3 + 4s for the 2 analogue realization 1Z(s) = in sthe 2s +form = 1 of ωthe first Cauer’s Frequency [rad/sec]
2s 1 e |s=jω . (6.8)
6.4 Tree structure circuit Z(s) = LC Za .Zcircuit 2s + C − C
s +
canonic b. [20] we have 1 to choose 1 (6.7) the following
− s+ 12 values 3 of coils and Figure 7.5: Bode plots of the I 1/2 controller where half order integral was
Let us consider the total impedance of the fractance circuit as shown in Fig. 6.5, 12 3 − s
If we substitute a resistance
which has a recursive Zacapacitors:
structure with = combination
the R and aofcapacitancetwo impedances ZZ a b
= 1/sC, then − s the 2 approximated by transmission line depicted in Fig. 7.2.
ber of stage in the recursive self infinity (binary) structure on the impedance
be realized using passive electric
characteristics [31]. with the 6.3
components.
help of active However, Transmission lines circuit
they cannamely
components,
CHAPTER 6. GENERAL APPROACH TO FRACTANCE DEVICES 35
be realized operating amplifiers. We can
AZ(s)
similarwith the
Ytree structure
Y 1 help
Y ofY active
as a fractal Ymodel ofYcomponents,
2 realize
rough
3
interface
Y this Ynamely
between
4 by two
5 negative operating
inpedance
6 amplifiers.
converterWe
2n-1 whichcan was described in previous
2n
materials of very Zdifferent conductivities, e.g.
Z an electrode and Zan electrolyte
Z
realize this by negative
2 Let us inpedance
consider
subsection.
4 the
was studied in [21]. There was described a slightly different structure of tree
circuit
converter depicted
Z which 6 was
Z in Fig.
described Z 6.3, in where
previous
2n Z Z2k−1 (s) and Y2k (s), k = 1 3 5 2n-1
subsection. 1, . . . , n, are given impedances of the circuit elements. This structure is known
Figure 6.3: General structure of transmissionZ(s)
line.
as a transmission line or a symetricallines
6.3 ZaTransmission dominocircuit
ladder lattice network as well. Y1 Y2 Y3 Y4 Y5 Y6 Y 2n-1 Y 2n
Za Za Z2 Z4 Z6 Z 2n
6.3 The Za
resultinglines
Transmission impedance Z(s) of the entire circuit can be expressed as a CFE
circuit
described Let by (6.1)us consider with the circuit depicted
symetrical distributionin Fig. 6.3, where Z2k−1
of elements, (s) and
Z2k−1 (s) Y=2k (s), k=
Z2k (s)
Z(s) Figure 6.3: General structure of transmission line.
Let
Zb us Zb consider
Zb the
Zb Zbcircuit 1, . .=
Zb depicted .Zb, n, are
YZain Zb given
Fig. 6.3,impedances
where Z2k−1of(s) theand
circuit elements.
Y2k (s), k = This structure is known
General Approach to Fractances General Approach to Fractances
Za Za
and Y2k−1
Za
(s) 2k (s), k = 1, . . . , n.
1, . . . , n, are given impedances as a transmission
of the circuit Za line or Za
elements. a symetrical domino ladder lattice network as well.
the Thissamestructure
Za
In the case if we assume elements is known Za
in transmision line in series
CHAPTER 7. REALIZATION OF FRACTIONAL CONTROLLERS 41
CHAPTER 7. REALIZATION OF FRACTIONAL CONTROLLERS 42
(ZFigure
a ) and
6.4:
astheTransmission
same line circuit
in 6.5,
shunt (Z=b ),composed
we of structure
two inpedance Za and in
Zb .Fig. 6.4.
Let us consider the total
In the impedance
case ifofwe theassume
fractance circuit
the same shown in Fig.
elements in transmision
Z(s) Zaget.Zba. in
line series depicted (6.7)
0.4
0.2
0.2
If we substitute
and Zb . Impedance is derived as geometrical mean6.4
Zb model
of Zaof Tree
aand
real structure
resistance
Zb ,cable = circuit
Za(solved R and a Heaviside
capacitance Zb =can 1/sC, then the !0.4 !0.2
Impedance
Za of such kind of transmision line,linewhich canby beO.used also asina 1887), be expressed as !0.6
!0.4
!
Take
!0.8 0 0.005 0.01 0.015 0.02 0.025 0.03
! 0 0.005 0.01 0.015 0.02 0.025 0.03
CHAPTER 7. REALIZATION OFTime
FRACTIONAL CONTROLLERS 44
impedance is usbyconsider
[sec]
CHAPTER 7. REALIZATION OFTime
FRACTIONAL
[sec] CONTROLLERS 43
Za. Let the total impedance
in 1887),ofZ(s) the fractance
model of realZ(s) cable= line
Zb (solved
Za .Z b O. Heaviside
"structure can
" be Zacircuit
= expressed .Zb . as asshown in Fig. 6.5, (6.7) 1
1
which has a !
recursive with the combination of two impedances Z 0.5
If
andwe Z(s) = Za .Z
Zbsubstitute .derived (6.7)
Z(s) = aisbresistance
. Impedance Z=a = R and
s as geometrical ω −1/2 aofcapacitance
eZ−jπ/4 Z = 1/sC, then
a and| Zb , . b
the
sine
0
(6.8) 0
If we substitute a resistance Za Za
If we substitute a resistance
Za
= R and a capacitance
Zb
impedance
Z a = R b = 1/sC,C
Zis
and a
then
capacitance " Z b
!C
=
Z(s) = Za .Z" 1/sC,b.
then the
s=jω !0.5
unit step !0.5
!1
0 0.005 0.01 0.015 0.02 0.025 0.03
impedance " is
R −1/2 R −1/2 −jπ/4
Zb " Za Z(s) = s = ω e |s=jω . (6.8) 0.6
1/2
Figure 7.6: Time response of the I 1/2 controller to unit-step input where half
Figure 7.7: Time response of the I controller to sin input where half order
Then
integral0.4was approximated by transmission line depicted in Fig. 7.2.
0.2
Z(s) = s = |s=jω , order integral was approximated by transmission line depicted in Fig. 7.2.
sIf−1/2
we substitute ω resistance a = R and a capacitance Zb = 1/sC, then
e−jπ/4 |Zs=jω
0.2
C
the impedance shows C the fractance characteristics as (see also (6.8)) !0.2 !0.2
which is a fractional order integral with absolute value of impedance propor- " " !0.4
0 0.005 0.01 0.015 0.02 0.025 0.03
!0.4
0 0.005 0.01 0.015
Time [sec]
0.02 0.025 0.03
tional toFigure
ω −1/26.5: A phase
self-similar treeiscircuit composed of independent
two inpedance Z and Z b.
Time [sec]
and angle constant −π/4, ofa the frequency.R −1/2 R −1/2 −jπ/4 1
Z(s) = s = ω e |s=jω , 1
0.5
Input signal: E [V]
circuit, where capacitance C was the same at every stage and resistance R
Fig. 6.5.increased
Therefore, it seems to be important to study the effect of the num-
saw ramp
0
0
by the ratio a at every stage of branching.which
The resulting impedance
is a fractional order integral with absolute value of impedance propor- !0.5
ber of stage in the recursive self infinity (binary) structure on the impedance !0.5
phase
characteristics [31].angle is constant !"/4, independent of the frequency
Z(s) had the form of a continued fraction expansion. tional to ω −1/2 and phase angle is constant −π/4, independent of the frequency. !1
0 0.005 0.01 0.015
Time [sec]
0.02 0.025 0.03
!1
0 0.005 0.01 0.015
Time [sec]
0.02 0.025 0.03
•• A fractional
If ! orderthe
can be found, system
commensurate order
CHAPTER 7. REALIZATION OF FRACTIONAL CONTROLLERS 40
Figure 7.1: Analogue fractional-order integrator.
For example:
CHAPTER 7. REALIZATION OF FRACTIONAL CONTROLLERS 39
Z (s)
7.2 Realization of fractional I λ controller 1µF 1k !
22k ! 22k ! 22k ! 22k ! 22k ! 22k ! 22k ! 22k ! 22k ! 22k ! 22k ! 22k !
Z F (s) F
22k ! 22k ! 22k ! 22k ! 22k ! 22k ! 22k ! 22k ! 22k ! 22k ! 22k ! 22k !
can be written
It as bethat
is obvious as thus the original order
of the P I λ D µ controller, (if Kp = 0 and Td =1µF
• system
is transmission
a particularline
casecircuit. 0). The con-
1k !
1k !
Figure 7.2: RC
Figure 7.2: RC transmission
1k !
troller line
1k !1k !
transmission
1k !
1k !
wascircuit.
realized in three forms, namely by: the symetrical domino ladder or
1k !line (see
1k !
1k !Fig. 7.2)
1k !
1k ! Z F (s) ladder (see Fig. 7.3) 1µF
for n=6, the finite domino 1k !
1µF
can !=1/6,
written
C
where
1µF 1µF 1µF 1µFZ
1µFconnected
impedance
were 1µF to feedback
1µF in operational amplifier (Fig. 7.1).
1µF 1µF 1µF 1µF 1µFF
1µF 1µF
Z F (s)
Z F (s) It should be noted that the described methods work for arbitrary orders, but
the circuit elements with computed values are not usually available. Because
Figure 7.4: RC binary tree circuit.
of this, in our experiment we proposed and realized the integrator with order
Figure 7.3: RC domino
λ =ladder ladder circuit.
0.5. circuit.
Figure 7.3: RC domino !5
beofrealized alsoanalogue
using the methods described in [31, 33, 38, 54], which do not
we have Ti = 1.4374. The transfer function
order I λ controller is:
the realized fractional-
!20 ©Dingyu Xue, NEU, PR China
order I λ controller is: involve explicit rational approximations.
1.4374 s−0.5.
where
C(s) = −0.5 (7.2) !25
C(s) = 1.4374 s In .the case, if we will use (7.2)
identical resistors (R-series)
!30 and identical ca-
λ
Adjustment of the integration constant
pacitors T of
(C-shunt)
i the fractional-order
in the Ifractances,
λ I con-
then the behaviour
!35 of the circuit will
Adjustment of the integration constant Ti of the fractional-order con-
troller (or half order integrator) depicted in Fig. 7.1 was done by resistor Ri .
troller (or half order integrator) depictedbe in asFig.a 7.1
half-order
was doneintegrator/differentiator.
by resistor Ri .
If we change the resistor Ri , the integration constant changes the value in the
We used the!40
resistor
10 values R = 10
2 3 4
10 FOC tutorial III 15
If we change the resistor Ri , the integration constant
(Rj = changes
R, j = the1, . .value thethe capacitor values C = 1µF , (Cj = C, j = Frequency [rad/sec]
. , n) inand
required interval.
1kΩ, ©Dingyu Xue, NEU, PR China
required interval. 1, . . . , n).
100 For better measurement
±10 V. results we used two operational amplifiers
!30
For measuremets we used frequency Hz and amplitude
For measuremets we used frequency 100 Hz and amplitude
TL081CN in inverting±10connection.
V. !40 FOC tutorial III 15
The resistors R1 and R2 are R1 = R2 = 22kΩ. The integration constant Ti
!50
©Dingyu Xue, NEU, PR China
Phase [deg]