You are on page 1of 6

Proceedings of the 2006 American Control Conference ThB11.

5
Minneapolis, Minnesota, USA, June 14-16, 2006

Fractional Order PID Control of A DC-Motor with Elastic Shaft:


A Case Study
Dingyü Xue and Chunna Zhao YangQuan Chen
Faculty of Information Science and Engineering CSOIS, Dept. of Electrical and Computer Engineering
Northeastern University Utah State University
Shenyang 110004, China Logan, UT 84322-4160, USA
xuedingyu@ise.neu.edu.cn; chunnazhao@163.com yqchen@ece.usu.edu

Abstract— In this paper, a fractional order PID controller of fractional order control in modeling and control design
is investigated for a position servomechanism control system have motivated renewed interest in various applications of
considering actuator saturation and the shaft torsional flexi- fractional order control [3], [4]. Some MATLAB tools of
bility. For actually implementation, we introduced a modified
approximation method to realize the designed fractional order the fractional order dynamic system modeling, control and
PID controller. Numerous simulation comparisons presented in filtering can be found in [5]. Reference [6] gives a fractional
this paper indicate that, the fractional order PID controller, order PID controller by minimizing the integral of the
if properly designed and implemented, will outperform the squared errors. Some numerical examples of the fractional
conventional integer order PID controller. order PID control were presented in [7], [8]. In [9], a PIα
Index Terms— Fractional order calculus, fractional order
control, PID control, servomechanism, elastic shaft, robustness. controller was designed to ensure that the closed-loop system
is robust to gain variations and the step responses exhibit
an iso-damping property. For speed control of two-inertia
I. I NTRODUCTION systems, some experimental results were presented in [10]
There is an increasing interest in dynamic systems of non- by using a fractional order PID controller. A comparative
integer orders. Extending derivatives and integrals from inte- introduction of four fractional order controllers can be found
ger orders to noninteger orders has a firm and long standing in [11]. It is also believed that FO calculus will be the right
theoretical foundation. For example, Leibniz mentioned this tool for biomimetic control [12].

concept in a letter to L Hospital over three hundred years The major contributions of this paper include 1) A newly
ago and the earliest more or less systematic studies have been modified approximate realization method for fractional or-
made in the beginning and middle of the 19th century by Li- der derivative; 2) An extensive simulation study using a
ouville, Riemann and Holmgren [1]. In the literature, people benchmark system with code available online to make our
often use the term “fractional order calculus”, or “fractional results reproducible1; 3) Show for the first time that, under
order dynamic system” where “fractional” actually means the same optimization condition, the best FO PID controller
“non-integer”. outperforms the best IO PID controller. Moreover, for the
Clearly, for closed-loop control systems, there are four first time, we show that the achieved robustness using FO
situations. They are 1) IO (integer order) plant with IO PID is robust against the approximate error in FO controller
controller; 2) IO plant with FO (fractional order) controller; realization.
3) FO plant with IO controller and 4) FO plant with FO The remaining part of this paper is organized as follows.
controller. In this paper, we focus on using FO-PID controller In Sec. II, the benchmark position servomechanism system
for an IO plant - “DC-Motor with elastic shaft”, a benchmark is introduced with detail model and model parameters. In
system from [2]. Sec. III, the fractional order PID controller and its math-
Intuitively, with noninteger order controllers for integer ematical foundation are presented. Section IV presents a
order plants, there is a better flexibility in adjusting the new modified finite dimensional realization method for FO
gain and phase characteristics than using IO controllers. This derivatives. In Sec. V, extensive simulation investigations
flexibility makes FO control a powerful tool in designing of the position servomechanism controlled by optimal IO
robust control system with less controller parameters to PID/PI controllers and optimal FO PID/PI controllers are
tune. The key point is that using few tuning knobs, FO presented to illustrate the superior robustness achieved by
controller achieves similar robustness achievable by using using fractional order controller. Finally, conclusions are
very high-order IO controllers. Since the tradeoff between drawn in Sec. VI.
the stability and other control specifications always exists, II. T HE B ENCHMARK P OSITION S ERVO S YSTEM
introducing fractional order control makes it more straight-
forward to achieve a better tradeoff. The possible advantages The Benchmark position servomechanism system is a
“DC-motor with elastic shaft” from [2]. To make this paper
The work done in Northeastern University was supported by the Key self-containing, in this section, we re-state from [2] the whole
Laboratory Project and the work by YangQuan Chen was supported in part
by the VPR’s TCO Technology Bridging Grant of Utah State University.
model used in our simulation, as depicted in Fig. 1 where
Corresponding author: YangQuan Chen. Tel. (435)7970148; Fax:
(435)7507390. 1 http://mechatronics.ece.usu.edu/foc/code/acc06.zip

1-4244-0210-7/06/$20.00 ©2006 IEEE 3182


TABLE I
we can see the benchmark system consists of a DC motor, S ERVOMECHANISM S YSTEM ’ S PARAMETERS
a gearbox, an elastic shaft and a load.
R
Value Value
θM gearbox Symbol Symbol
(SI units) (SI units)
V JM load θL
T i
) kθ 1280.2 ρ 20
βM ρ JL
)
kT 10 βM 0.1
motor model βL
Fig. 1. The benchmark position servomechanism model JM 0.5 βL 25

The benchmark system in Fig. 1 can be represented by the JL 50JM R 20


following differential equations:
kθ  θ M  βL
ω̇L = − θL − − ωL (1) III. F RACTIONAL O RDER PID C ONTROLLER : A N
JL ρ JL I NTRODUCTION
kT  V − kT ωM  βM ωM kθ  θM  To study the fractional order controllers, the starting
ω̇M = − + θL − (2)
JM R JM ρJM ρ point is of course the fractional order differential equations
using fractional calculus. A commonly used definition of the
where V is the applied voltage, T is the torque acting on the
fractional differointegral is the Riemann-Liouville definition
load, ωL = θ̇L is the load’s angular velocity, ωM = θ̇M is m  t
the motor shaft’s angular velocity, kθ and kT the torsional α 1 d f (τ )
D
a t f (t) = dτ (6)
rigidity and motor constant, JM and JL the motor and Γ(m − α) dt a (t − τ )1−(m−α)
nominal load inertia, βM and βL the motor viscous friction
for m − 1 < α < m where Γ(·) is the well-known
coefficient and load viscous friction coefficient, ρ the gear
Euler’s gamma function. An alternative definition, based on
ratio and R the armature resistance.
the concept of fractional differentiation, is the Grünwald-
Defining the state variables as xp = [θL ωL θM ωM ]T ,
Letnikov definition given by
the above model can be converted to an LTI state-space form:
⎡ ⎤ (t−a)/h
1  Γ(α + k)
0 1 0 0 α
f (t − kh). (7)
a Dt f (t) = lim
⎢ kθ βL kθ ⎥ h→0 Γ(α)hα Γ(k + 1)
⎢ − − 0 ⎥ k=0
⎢ JL JL ρJL ⎥
ẋp = ⎢ 0 ⎥ xp One can observe that by introducing notion of the fractional
⎢ 0 0 1 ⎥
⎣ kθ kθ βM + kT /R ⎦
2 order operator a Dtα f (t), the differentiator and integrator can
0 − 2 − be unified.
ρJM ρ JM JM
⎡ ⎤ Another useful tool is the Laplace transform. It is shown
0 in [13] that the Laplace transform of an nth (n ∈ R+ )
⎢ 0 ⎥ of a signal x(t) relaxed at t = 0 is given by:
⎢ ⎥ derivative
 
+⎢ 0 ⎥V (3) L Dn x(t) = sn X(s). So, a fractional order differential
⎣ kT ⎦
equation, provided both the signals u(t) and y(t) are relaxed
RJM at t = 0, can be expressed in a transfer function form
θL = [1 0 0 0] xp (4) a1 sα1 + a2 sα2 + · · · + amA sαmA
G(s) = (8)

b 1 s β 1 + b 2 s β 2 + · · · + b m B s β mB

T = kθ 0 − 0 xp . (5)
ρ where (am , bm ) ∈ R2 , (αm , βm ) ∈ R+ 2
, ∀(m ∈ N ).
The most common form of a fractional order PID con-
The only measurement available for feedback is θL . The troller is the PIλ Dμ controller [14], involving an integrator
load’s angular position must be set at a desired value by of order λ and a differentiator of order μ where λ and μ
adjusting the applied voltage V . The elastic shaft has a finite can be any real numbers. The transfer function of such a
shear strength, so the torque T must stay within specified controller has the form
limits. From an input/output viewpoint, the plant has a single
KI
input V , which is manipulated by the controller. It has two Gc (s) = KP + λ + KD sμ , (λ, μ > 0). (9)
outputs, one measured and fed back to the controller θL and s
one unmeasured T . The integrator term is sλ , that is to say, on a semi-logarithmic
Parameters of the experimental position servomechanism plane, there is a line having slope −20λdB . /dec.
system are shown in Table. I. The control signal u(t) can then be expressed in the time
The designed controller must set the load’s angular posi- domain as
tion θL at a given value. The elastic shaft has a finite shear
u(t) = KP e(t) + KI D−λ e(t) + KD Dμ e(t). (10)
strength, so the torque, T , must stay within specified limits
|T | ≤ 78.5Nm. Also, the applied voltage must stay within Clearly, selecting λ = 1 and μ = 1, a classical PID
the range |V | ≤ 220V [2]. controller can be recovered. The selections of λ = 1, μ = 0,

3183
and λ = 0, μ = 1 respectively corresponds conventional PI poles. One is −bωh /d, which is negative because ωh , b, d >
& PD controllers. All these classical types of PID controllers 0. The other two are roots of
are the special cases of the fractional PIλ Dμ controller
given by (9). It can be expected that the PIλ Dμ controller d(1 − α)s2 + asωh + dα = 0 (16)
may enhance the systems control performance. One of the whose real parts are negative as 0 < α < 1.
most important advantages of the PIλ Dμ controller is the Based on the well known zig-zag line approximation idea
better control of dynamical systems, which are described by in Bode plot, let
fractional order mathematical models. Another advantage lies ⎛ s ⎞α
in the fact that the PIλ Dμ controllers are less sensitive to 1+ d N
⎜ ⎟  
changes of parameters of a controlled system [14]. This is ⎜ b ωb ⎟ 1 + s/ωk
due to the two extra degrees of freedom to better adjust the ⎝ s ⎠ = Nlim→∞ 1 + s/ωk
(17)
1+ b k=−N
dynamical properties of a fractional order control system. d ωh
However, all these claimed benefits were not systematically 
demonstrated in the literature. In this paper, from practical where −ωk and −ωk are zero and pole of rang k
application point of view, we attempt to illustrate the benefits α−2k α+2k
in a reproducible manner.  dωb 2N +1 bωh 2N +1
ωk = , ωk = . (18)
It was pointed out in [15] that a band-limit implementation b d
of fractional order controller is important in practice, and
Hence
the finite dimensional approximation of the fractional order

N 
controller should be done in a proper range of frequencies ds2 + bsωh s + ωk
of practical interest. This is true since the fractional order sα ≈ K (19)
d(1 − α)s2 + bsωh + dα s + ωk
controller in theory has an infinite memory and some sort of k=−N
approximation using finite memory must be done. In the next where
section, we will present a modified approximation scheme α N
dωb ωk
whose performance is better than Oustaloup’s method [16]. K=  (20)
b ω
k=−N k
IV. A M ODIFIED A PPROXIMATE R EALIZATION M ETHOD
In this paper, we used b = 10 and d = 9.
Here we introduce a new approximate realization method The procedures for the modified approximation can be
for fractional derivative in the frequency range of interest briefly summarized in the following:
[ωb , ωh ]. Our proposed method here gives a better approxi-
• Given the frequency range [ωb , ωh ] and N
mation than Oustaloup’s method in both low frequency and 
• Based on the fractional order α, calculate ωk and ωk
high frequency parts.
according to (18)
Let ⎛ s ⎞α • Compute K from (20)
1+ d
⎜ ⎟
b ωb ⎟
• Obtain the approximate rational transfer function from
sα ≈ ⎜
⎝ s ⎠ (11) (19) to replace sα
1+ b
d ωh V. C OMPARATIVE S IMULATIONS
where 0 < α < 1, s = jω, b > 0, d > 0. Thus A. Best IO PID vs. Best FO PID
α α
bs −ds2 + d Simulations of position servomechanism controlled by
sα ≈ 1+ 2 . (12) normal PID controller and fractional order PID controller
dωb ds + bsωh
are carried out based on the parameters setting in Table.
Then, within [ωb , ωh ], using Taylor series expansion I with maximum output torque limitation ±78.5Nm. We
⎛ s ⎞α used constrained optimization routine to search for the best
1 + controller parameters. Two optimization criteria are used.
(dωb )α b−α ⎜ d ⎟
b ωb ⎟
sα ≈   ⎜ One is ITAE (integral of time-weighted absolute error) and
⎝ s ⎠
1 + αp(s) + α(α−1) 2 p2 (s) + · · · 1+ b another one is ISE (integral of squared error), where the
d ωh constraint is |T | < 78.5Nm. The reference signal is the unit
(13) step function.
where For the optimally searched IO PID using ITAE,
−ds2 + d
p(s) = 2 . (14) 21.13
ds + bsωh Gc1 (s) = 41.94 + − 8.26s; (21)
Truncating the Taylor series to the first order term, then s
⎛ s ⎞α For optimally searched IO PID using ISE,
α 1+
dωb ds2 + bsωh ⎜ d
b ωb ⎟
⎟ 10.65
α
s ≈ ⎜ Gc2 (s) = 110.09 + + 30.97s (22)
b d(1 − α)s2 + bsωh + dα ⎝ 1 + s ⎠ . s
b Fig. 2 shows the responses to unit step of the angular position
d ωh
(15) controlled by two integer order PID controllers Gc1 (s) and
Note that (15) is stable iff all the poles are on the left half Gc2 (s), respectively with the Bode plots of the open-loop
s-plane. It is easy to observe that the expression (15) has 3 controlled system shown in the same figure.

3184
Bode Diagram
Step Response
100
50 8

Magnitude (dB)
1.5 50 40
6
0 30
4
1 20
Amplitude

−50
2
−100 10
360
0.5 0 0
1.5 1.5

Phase (deg)
180
1.3 1.5 1.3 1.5
0 1.1 1.3 1.1 1.3
0.9 1.1 0.9 1.1
0.7 0.9 0.7 0.9
0 −180 0.7 0.7
0.5 0.5 0.5 0.5
−360
−2 −1 0 1 2
0 20 40 60 10 10 10 10 10
Time (sec) Frequency (rad/sec)
(a) ITAE(λ,μ) (b) ISE(λ,μ)
(a) Step responses (b) Bode plots
Fig. 4. Searching the best fractional orders (N = 4)
Fig. 2. Best IO PID Controllers. Solid line: ITAE; Broken line: ISE

in general, we can qualitatively tell that, the integer case


Now let us look at the best FO PID controllers. As the λ = 1 and μ = 1 is not optimal. In other words, the optimal
first attempt, let us first fix λ = 0.5 and μ = 0.6. Doing case most like corresponds to noninteger case. Moreover, we
the numerical search, we get the best ITAE of 2.22 and the can tell that, in this benchmark system, we prefer low order
corresponding fractional order PID controller is integral and lower order derivative.
0.01 C. Which N Is Good Enough?
− 31.6s0.6
Gc3 (s) = 135.12 +
(23)
s0.7 In implementing FO PID, we need to decide what the finite
while the best ISE is 0.87 and the corresponding fractional order is for the finite order approximation. In our case, we
order controller is need to decide the N . Let us first repeat the Fig. 4 using
91.95 N = 6.
Gc (s) = 61.57 + 0.5 + 2.33s0.6 . (24)
s
80 10
The step responses are compared in Fig. 3 with correspond- 60
8

ing Bode plots. 40


6

4
Step Response Bode Diagram
20 2
100
0 0
Magnitude (dB)

50 1.5 1.5
1.2 1.3 1.3
1.5 1.5
0 1.1 1.3 1.1 1.3
1 0.9 1.1 0.9 1.1
−50 0.7 0.9 0.7 0.9
0.7 0.7
Amplitude

0.8
−100
0.5 0.5 0.5 0.5

0.6 −150
360
0.4 (a) ITAE(λ,μ) (b) ISE(λ,μ)
Phase (deg)

180
0.2 0

0 −180 Fig. 5. Searching the best fractional orders (N = 6)


−0.2 −360
−2 −1 0 1 2
0 10 20 30 40 50 60 10 10 10 10 10
Time (sec) Frequency (rad/sec)

We conclude that, the difference between Fig. 4 and Fig. 5


(a) Step responses (b) Bode plots
is very small. To illustrate, let us run two examples with an
Fig. 3. Best FO PID Controllers. Solid line: ITAE; Broken line: ISE emphasis on the effect of different approximation order N .
Fig. 6 suggest that the changes in N will not contribute to
The observation is clear. The best FO PID performs better the differences in this benchmark problem.
than the best IO PID. This is not surprising but this may Step Response Step Response

not be fair since FO PID has two more extra parameters in 1.2 1.2

optimal search. 1 1

0.8 0.8
Amplitude

Amplitude

B. How To Decide λ and μ? 0.6 0.6

0.4 0.4
In the last section, we got a flavor that FO PID performs 0.2 0.2

better in side by side comparison. We simply fixed λ = 0.5 0 0

and μ = 0.6. But in reality, how to decide these two magic −0.2
0 10 20 30 40 50 60
−0.2
0 10 20 30 40 50 60
orders? To our best knowledge, this research question is not Time (sec) Time (sec)

well answered in the literature. In this paper, we only show (a) N = 2 (b) N = 6
a brutal force search result to partially justify why we fixed Fig. 6. Step responses comparisons with different N ’s
λ = 0.5 and μ = 0.6.
Here, we build two tables of optimal ITAE and ISE,
respectively, with respect to λ and μ which are enumerated D. Robustness Against Load Variations
from 0.5 to 1.5 with step of 0.1. In other words, we did In the last subsection, we have already seen the robust-
2 × 11 × 11 optimal searches. These two tables are visualized ness with respect to N . Here, the position servomechanism
in Fig. 4. control system is controlled by the best fractional order PID
Note that, in this investigation, we used the approximate controller (23) and the best IO PID controller (22) when load
order N = 4. It is unfortunate to observe that there is no changes ±50%. The results are summarized in Fig. 7. The
definite relationship can be established in Fig. 4. However, robustness against load variations is clearly seen from Fig. 7.

3185
Step Response Step Response However, if we use ISE, we should choose λ = 0.2. After
1.2 1.2

1 1
search, the optimal fractional order PI controller becomes
0.8 0.8
198.26
Gc (s) = −48.38 +
Amplitude

Amplitude
. (28)
s0.2
0.6 0.6

0.4 0.4

0.2 0.2
Fig. 10 summarizes the comparison of step responses and
0 0
Bode plots.
−0.2 −0.2
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Time (sec) Time (sec)
Step Response Bode Diagram

(a) +50% load variation (b) −50% load variation 1.4 50

Magnitude (dB)
1.2 0

Fig. 7. Step responses comparison with N = 4 (Dotted: Best PID; Solid 1


−50

Amplitude
line: Best FO PID) 0.8
−100

−150
360
0.6

Phase (deg)
180
0.4
0

E. FO PI Controllers 0.2 −180

0 −360 −2 −1 0 1 2
0 10 20 30 40 50 60 10 10 10 10 10
It will be interesting to check if we can see the case Time (sec) Frequency (rad/sec)

that “the best FO PI works better than the best IO PI”. We (a) Step responses (b) Bode plots
repeat what we performed for PID controllers and summarize
Fig. 10. Best FO PI Controllers. Solid line: ITAE; Broken line: ISE
briefly in the following.
Under the ITAE criterion, the following optimal IO PI
controller is searched The observation is again clear. The best FO PI performs
better than the best IO PI. This is again not surprising but
0.14
Gc (s) = 107.35 + (25) this may not be fair since FO PI has an extra parameter in
s optimal search.
and by the ISE criterion, When load increases ±50%, the corresponding results are
3.36 summarized in Fig. 11. Again the robustness against load
. Gc (s) = 106.82 + (26) variations can be observed.
s
Step Response Step Response
Fig. 8 shows the step responses and the Bode plots. 1.4 1.4

1.2 1.2
Step Response Bode Diagram
1 1
1.4 50
Amplitude

Amplitude
Magnitude (dB)

0 0.8 0.8
1.2

−50 0.6 0.6


1
Amplitude

−100 0.4 0.4


0.8
−150 0.2 0.2
0.6 −90
Phase (deg)

−180 0 0
0.4 0 10 20 30 40 50 60 0 10 20 30 40 50 60
Time (sec) Time (sec)
0.2 −270

0 −360
(a) +50% load variation (b) −50% load variation
−2 −1 0 1 2
0 10 20 30 40 50 60 10 10 10 10 10
Time (sec) Frequency (rad/sec)

Fig. 11. Step responses comparison with N = 4 (Dashed line: Best IO


(a) Step responses (b) Bode plots PI; Solid line: Best FO PI
Fig. 8. Best IO PI Controllers. Solid line: ITAE; Broken line: ISE

Similar to Fig. 4, we can draw Fig. 9 which provides the F. Robustness to Mechanical Nonlinearities
basis for λ selection. Using square wave as the reference input signal (period
12 1.05
T = 40 sec.) and adding Coulomb friction 0.1, the output
11 responses of the controlled angular position are shown in
10 1
Fig. 12. Similarly, we simulated the case with backlash, with
9
8 0.95
the deadband width of 0.5. The symmetrical square wave
7 position tracking responses are compared in Fig. 13. Finally,
6 0.9

5
we checked the case of deadzone with its parameter set as
4 0.85 ±0.5. In this case, the responses are compared in Fig. 14.
3
2 0.8
So, we can observe that best FO PID controllers perform
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
much better than their IO counterparts.
(a) ITAE(λ) (b) ISE(λ)
G. Robustness to Elasticity Parameter Change
Fig. 9. Searching the best fractional order (N = 4)
Finally, we are interested in checking the robustness with
respect to the changes of the elasticity parameter kθ in
From Fig. 9, if we use ITAE, we should choose λ = 0.05. Table 1. When kθ varies ±50%, the corresponding results are
Then, the optimal fractional order PI controller is summarized in Figs. 15 and 16, respectively. We can again
72.3 observe that the best FO controllers perform better than the
Gc (s) = 39.82 + . (27) best IO controllers.
s0.05
3186
1.5 1.5 Step Response Step Response

1 1 1
1

0.5 0.5 0.8


0.8

Amplitude

Amplitude
0.6
0 0 0.6
0.4
−0.5 −0.5 0.4
0.2

−1 −1 0 0.2

−1.5 −1.5 −0.2


0 20 40 60
0
0 20 40 60 80 100 0 20 40 60 80 100 Time (sec) 0 20 Time (sec) 40 60

(a) best PID (b) best PI (a) best PID (b) best PI

Fig. 12. Responses comparison with Coulomb friction (Dashed line: Best Fig. 16. Responses comparison when kθ decreases 50% (Dashed line:
IO Controllers; Solid line: Best FO Controllers Best IO Controllers; Solid line: Best FO Controllers

1.5 1.5

1 1
the control performance in any noticeable amount.
With the rapid development of computer performances, the
0.5 0.5
realization of fractional order control systems also became
0 0
possible and much easier than before. Despite fractional
−0.5 −0.5
order control’s promising aspects in modeling and control
−1 −1
design, fractional order control research is still at its primary
−1.5
0 20 40 60 80 100
−1.5
0 20 40 60 80 100 stage. The notable future research is to develop tuning rules
for FO PID and in particular on tuning the fractional orders.
(a) best PID (b) best PI
R EFERENCES
Fig. 13. Responses comparison with backlash (Dashed line: Best IO
Controllers; Solid line: Best FO Controllers [1] K. B. Oldham and J. Spanier, The Fractional Calculus, Academic
1.5 1.5 Press, New York and London; 1974.
[2] Mathworks Inc., Model Predictive Control Toolbox User’s Guide. [On-
1 1
line]: http://www.mathworks.com/access/helpdesk/
0.5 0.5 help/toolbox/mpc/
[3] J. A. Tenreiro Machado, Theory analysis and design of fractional-
0 0
order digital control systems, Journal Systems Analysis Modeling
−0.5 −0.5 Simulation, Vol. 27, 1997, pp. 107-122.
−1 −1
[4] A. Oustaloup, J. Sabatier and X. Moreau, From fractal robustness to
the CRONE approach, ESAIM, Vol. 5, 1998, pp. 177-192.
−1.5
0 20 40 60 80 100
−1.5
0 20 40 60 80 100
[5] Dingyü Xue and YangQuan Chen, Advanced Mathematic Problem
Solution Using MATLAB, Beijing: Tsinghua University Press; 2004.
(a) best PID (b) best PI [6] Z.F. Lv, Time-domain simulation and design of SISO feedback con-
trol systems, Doctoral Dissertation, National Cheng Kung University,
Fig. 14. Responses comparison with deadzone (Dashed line: Best IO Taiwan, China, 2004.
Controllers; Solid line: Best FO Controllers [7] R. Caponetto, L. Fortuna, and D. Porto, “A new tuning strategy for a
Step Response Step Response non integer order PID controller,” IFAC2004, Bordeaux, France, 2004.
[8] Chunna Zhao, Dingyü Xue and Yangquan Chen, “A fractional order
1 1 PID tuning algorithm for a class of fractional order planst,” Proc. of
0.8
the IEEE International Conference on Mechatronics and Automation,
0.8 Niagara Falls, Canada, 2005,pp. 216-221.
Amplitude

Amplitude

0.6
0.6 [9] C. A. Monje, B. M. Vinagre, Y. Q. Chen, V. Feliu, P. Lanusse and
0.4
0.4
J. Sabatier, “Proposals for fractional PIλ Dμ tuning”, The First IFAC
0.2 Symposium on Fractional Differentiation and its Applications 2004,
0 0.2 Bordeaux, France, July 19-20, 2004.
−0.2 Time (sec) 0 Time (sec)
[10] C. B. Ma and Y. Hori, “Design of fractional order PID controller for
0 20 40 60 0 20 40 60
robust two-inertia speed control to torque saturation and load inertia
(a) best PID (b) best PI variation,” IPEMC 2003, Xi’an, China,2003.
[11] Dingyü Xue and YangQuan Chen, “A comparative introduction of four
fractional order controllers,” Proceedings of the 4th World Congress on
Fig. 15. Responses comparison when kθ increases 50% (Dashed line: Best Intelligent Control and Automation, Shanghai, China, 2002,pp. 3228-
IO Controllers; Solid line: Best FO Controllers 3235.
[12] YangQuan Chen, Dingyü Xue and Huifang Dou. “Fractional cal-
culus and biomimetic control”. In Proc. of the First IEEE Int.
VI. C ONCLUSIONS Conf. on Robotics and Biomimetics (RoBio04), pages robio2004 347,
In this paper, a fractional order PID controller is ex- Shengyang, China, August 2004. IEEE.
[13] I. Podlubny, Fractional-order Systems and Fractional-order Con-
amined on a benchmark position servomechanism control trollers, The Academy of Sciences Institute of Experimental Physics,
system considering actuator saturation and the shaft tor- UEF-03-94, Kosice, Slovak Republic, 1994.
sional flexibility. Using numerical optimization, numerous [14] I. Podlubny, Fractional-order systems and PIλ Dμ -controllers, IEEE
Trans. Automatic Control, vol.44,1999,pp.208-214.
simulation comparisons presented in this paper indicate that, [15] A. Oustaloup, F. Levron, B. Mathieu, and F.M. Nanot, “Frequency-
the fractional order PID controller, if properly designed band complex noninteger differentiator: characterization and synthe-
and implemented, will outperform the conventional integer sis”, IEEE Trans. Circuits Syst. I, vol. 47, 2000,pp.25-39.
[16] Xue Dingyü, Chunna Zhao, and YangQuan Chen, “A modified ap-
order PID controller. It was shown that the best FO PID proximation method of fractional order systems”, submitted to IEEE
works better than IO PID. For actually implementation, we International Conference on Mechatronics and Automation, 2006,
introduced a modified approximation method to realize the Luoyang, China.
designed fractional order PID controller. We used simulation
to illustrate that the order the approximation does not affect

3187

You might also like