Antenna Azimuth Position Control with Classical

PID and Fuzzy Logic Controllers

H. Ibrahim Okumus E. Sahin O. Akyazi
Dept. of Electrical and Electronics Surmene Abdullah Kanca VHS, Surmene Abdullah Kanca VHS,
Engineering, Karadeniz Technical Karadeniz Technical University, Karadeniz Technical University,
University, Trabzon/Turkey Trabzon/Turkey Trabzon/Turkey
okumus@ktu.edu.tr esahin@ktu.edu.tr oakyazi@ktu.edu.tr

Abstract— In this study, an antenna azimuth position control As seen in Fig. 2, position direction is given by a
system is controlled by using a Proportional-Integral- potenatiometer manualy to antenna. Also, a potantiometer
Derivative (PID) controller and a fuzzy logic controller is used to get the real position of the antenna in the
(FLC) designed in Matlab/Simulink environment. In order to
obtain the best system response with FLC, different types of
system. The potantiometer converts the angle of the
fuzzy rules and membership functions are tested. System rotation to a voltage value. Then, differential amplifier
responses with proposed controller and PID controller are calculates the difference between input and output signal
compared to each other and commented according to these in order to get error signal. A power amplifier is used to
results. increase the amplitude of control signal in order to get a
suitable voltage value for the motor. In the system, it is
Keywords- Azimuth antenna control, FLC, Various fuzzy
rules and membership functions, PID. assumed that an armature controlled DC servo motor
having a fixed field is used [2]. Thanks to the gears and
motor, the position of the antenna is tried out to be
I. INTRODUCTION
controlled. Schematic representation of the antenna
The position control of antenna through the usage of azimuth position contol system is shown in Fig. 3 [1,2].
gears and feedback potentiometers is performed. Antenna
azimuth position control system is currently available
together with some control methods, some of which are H2
optimal and PID controls are some of these. Besides these
methods, a fuzzy logic controller designed in
Matlab/Simulink environment is used for controlling the
position of an antenna in a spherical coordinate system
called azimuth.
System concept of an antenna azimuth position control
system is shown in Fig. 1.[1,2]. Fig. 3. Schematic represantation of the antenna azimuth position control
system.

Also, functional block diagram of the antenna azimuth

position control system is given in Fig. 4 [1].

Fig. 1. System concept of an antenna azimuth position control system.

A detailed layout of the antenna azimuth position

control system given as a system concept is shown in Fig.
2 [1]. Fig. 4. Functional block diagram of the antenna azimuth position
control system.

II. SYSTEM MODELING IN MATLAB/SIMULINK

As expressed in previous section, an antenna azimuth
position control system consists of two potentiometers
which are used as input and output transducer, a
preamplifier, a power amplifier, a motor and a load.
Detailed block diagram of the antenna azimuth position
control system is shown in Fig. 5 [1].
Fig. 2. A detailed layout of the antenna azimuth position control
system.
978-1-4673-1448-0/12/$31.00 ©2012 IEEE

Fig. 5. Detailed block diagram of the antenna azimuth position control
system.
In the Table I, ''K'' represents the preamplifier block
and it is a gain value. The preamplifier gain ''K'' is
calculated by using Routh-Hurwitz criterion. According to
this criterion ''K'' is obtained between the range of 0 and
262.3 to get a stable system response [1]. In this study,
preamplifier gain ''K'' is selected as 100 in the limited
range. Simulink representation of the antenna azimuth
position control system is shown in Fig. 6.

The transfer function of motor and load block shown

in (1).
Φ m (s) Km
= (1)
Ea ( s ) s ( s + am )
The dampening and inertial components of the antenna
are adjusted with the help of gear ratios as seen in (2). Fig. 6. Simulink representation of the antenna azimuth position control
system.
N1 25 (2)
Kg = = = 0 .1
N2 250 If 10 turns are applied to the system as an input signal,
closed- loop system response without controller is shown
In (2), N1 and N2 are the gear teethes shown in Fig 1. in Fig. 7.
Calculations of dampening and inertial components are
given in (3) and (4).

J = J a + J L ( K g ) 2 = 0.02 + 1(0.1) 2 = 0.03 (3)

D m = D a + D L ( K g ) 2 = 0 . 01 + 1( 0 . 1) 2 = 0 . 02 (4)

In the above equations, Ja is the motor inertial constant

(kg-m2), JL is the load inertial constant (kg-m2), Da is the
motor dampening constant (N-m s/rad) and DL is the load
dampening constant (N-m s/rad).The pole and zero of the Fig. 7. Closed -loop system response of antenna azimuth position
motor and load block diagram is calculated as seen in (5) control system.
and (6).
As seen in Fig. 7, system has a percentage overshoot
DmRa + KbKt (0.02)(8) + (0.5)(0.5) about 35% and a settling time about 5 seconds. Also, little
am = = ≅1.71 (5) oscillation is observed in the system response. By using a
JRa (0.03)(8) PID controller and FLC, it is aimed to obtain a better
system response and results of these controllers are
Kt 0.5 compared.
Km = = ≅ 2.083 (6)
JRa (0.03)(8) III. DESIGN PROCESS OF THE FUZZY LOGIC
In these equations, Ra is the motor resistance (ohm), Kb CONTROLLER
and Kt are back EMF constant (V-s/rad) and motor torque In literature, a fuzzy logic controller is commonly
constant (N-m/A) respectively. Block diagram parameters defined as a system that makes decisions on behalf of
of preamplifier, power amplifier and gears are given in human. In this case, expert of human will be presented in
Table I [1]. the format of fuzzy rules. Then, these fuzzy rules make
decisions as a human does. A fuzzy logic controller
TABLE I. consists of three sections, which are fuzzification, rule
BLOCK DIAGRAM PARAMETERS OF THE ANTENNA AZIMUTH POSITION
base and defuzzification sections as shown in Fig. 8. [3-7].
CONTROL SYSTEM

Parameters Configurations
Kpot 0.318
K -
K1 100
a 100
Kg 0.1
Fig. 8. Basic structure of fuzzy logic controller.
 As seen in Fig. 8, the input values of the fuzzy logic TABLE II.
3X3 RULES FOR THE FUZZY LOGIC CONTROLLER
controller are error (e) and its change (∆e). These two
values are as defined in (1) and (2).
e(k ) = r(k ) − y(k ) (7)

Δe(k ) = e(k ) − e(k − 1) (8)

Respectively, r(k), y(k) and k are expressed as

reference to input, actual output and iteration counter.
Two inputs of the fuzzy logic controller are crisp values.
In fuzzification process, these crisp values are
converted into fuzzy membership values. To get fuzzy
membership values, error and its change are divided into
five fuzzy subsets. These are negative (N), zero (Z) and
positive (P). Triangular membership functions are used to
represent fuzzy subsets as shown in Fig. 9.
After fuzzification and rule based sections, by using
center of area method, fuzzified change of the control
signal is converted into the crisp values in order to get the
control signal in defuzzification section [8-13].
In the center of area method, maximum membership
degrees of the corresponding fuzzy subset are multiplied
by fuzzy membership values. Then, sum of these products
is divided by fuzzy membership values and change of
control signal is obtained. The general expression of the
center of area method is given in (7).
n

∑μ (Δu )⋅ Δu
i i i
for i=1,2,..,n (13)
ΔU (k ) = i=1
n

∑μ (Δu )
i =1
i i
Fig. 9. Triangular membership functions of each fuzzy subset.
A generalized overlooked view of the fuzzy logic
These fuzzy membership values are used in rule table to
controller is shown in Fig. 10. In this figure, subsets of
activate the related rules [3-7]. The mathematical
inputs are error; change of error and change of control
formulation of triangular membership function is shown in
signal are limited between positive one and negative one.
(3).
But, three gain coefficients are added to fuzzy logic
⎛ ⎛ x − x1 x2 − x ⎞ ⎞ controller to change the limits of different type of fuzzy
AΔ = max⎜⎜ min⎜⎜ , ⎟⎟,0⎟ (9) logic membership functions. Depending on the system,
⎟
⎝ ⎝ xT − x1 x2 − xT ⎠ ⎠ these coefficients used to adjust the limit membership
In this study, performances of different type functions may vary [3-15].
membership functions such as trapezoid, bell, Gaussian,
Cauchy, sigmoid or sinusoid are compared with each other
in order to obtain the best system response. The
mathematical representation of some of these membership
functions are shown below.
2
1 ⎛ x − xT ⎞
− ⎜ ⎟
AG = e 2⎝ W ⎠
(Gaussian) (10)

1
AC = 2m
(Cauchy) (11)
⎛ x − xT ⎞
1+ ⎜ ⎟
⎝ d ⎠
1 (Bell) (12)
AB = 2m
x − xT
1+
d Fig. 10. Generalized overlooked view of the fuzzy logic controller.

A rule based consists of data about the related system.

In this project, a nine, twenty-five and forty-nine rules
fuzzy logic controller are designed and tested to get the
best system response with proposed controller. One of
these rules utilized in making decisions are as defined in
Table II.
IV. APPLICATION AND RESULTS
In order to obtain the best system response with fuzzy
logic controller, firstly different type fuzzy membership
functions are performed. Simulink model representation of
antenna azimuth position control with different types
fuzzy membership functions are shown in Fig. 11.
 Fig. 13. System model with various fuzzy rules tables.

System responses with different number of rules of

fuzzy logic controller are shown in Fig. 14.

Fig. 11. System model with different type fuzzy membership functions.

System responses with triangle, trapezoid, Cauchy,

Gaussian and bell type fuzzy membership functions are
shown in Fig. 12. 10 turns is applied to the system as an
input.

Fig. 14. System model with various fuzzy rule tables.

As seen in Fig. 14, no overshoot is seen with different

number of fuzzy rule tables. Due to reduced number of
calculation, the fastest system response is obtained by
using 3x3 fuzzy rule table.
After obtaining the best antenna azimuth position
control response with triangular membership function and
3x3 fuzzy rule table, the performance of fuzzy logic
controller is compared with a classical PID controller.
Fig. 12. System responses with different type fuzzy membership Parameters of the PID controller are chosen as Kp=6,
functions. Ki=0.1 and Kd=1.5 respectively.
System block diagram model with PID controller and
As seen in Fig. 12, the fastest system response without FLC are shown in Fig. 15.
steady-state error is obtained by using triangular fuzzy
membership function. Since triangular fuzzy membership
function has sharper and more sensitive transitions
between its subsets compared to other membership
functions, it is expected result to get the best answer with
triangular type fuzzy membership function.
After determining the membership function type, 3x3,
5x5 and 7x7 fuzzy rule tables are tested by using
triangular membership function. System block diagram
with 3x3, 5x5 and 7x7 fuzzy rule tables are shown in Fig.
13. Fig. 15. Matlab/Simulink model representation of the system with FLC
and PID controller.

Antenna azimuth position control system responses

with FLC and PID controller are shown in Fig. 16. As an
input to the system 10 turns are applied via potentiometer.

Fig. 16. System responses with FLC and PID controller.
As seen in Fig. 16, by using FLC a better system
response is obtained.
Performances of these two controllers are also
compared with variable input values that are changed from
0 turn to 180 turns and 180 turns to 0 turn by increasing
and decreasing 10 turns to potentiometer. Comparison of
the performances with variable input turns are shown in
Fig. 17.
Fig. 17. System response of FLC and PID controller with variable
inputs.
As seen Fig. 17, with the variable input values, FLC
gives better results than PID controller.
V. CONCLUSION
In this study, antenna azimuth position system is
controlled with FLC and PID controllers designed in
Matlab/Simulink environment. In order to get the best
system response with FLC, various types of fuzzy
membership functions and various numbers of fuzzy rules
are tested. After obtaining the best FLC, performance
results of FLC are compared with a classical PID
controller. As a result, it is observed that FLC gives the
most convenient response to the desired system response.
REFERENCES
[1] J. Wiley &Sons, Inc. Nise, '' Control System Engineering '', 2000.
[2] Xuan L., Estrada J. and Digiacomandrea J.,''Antenna Azimuth
Position Control System Analysis and Controller Implementation'',
term project, 2009.
[3] I. H. Altas and A. M. Sharaf, “A Generalized Direct Approach for
Designing Fuzzy Logic Controllers in Matlab/Simulink GUI
Environment”, Accepted for publication in International Journal of
Information Technology and Intelligent Computing, Int. J. IT&IC
no.4 vol.1.
[4] İ. H. Altaş., “Bulanık Mantık : Bulanıklılık Kavramı”, Enerji
Elektrik Elektromekanik-3e, Temmuz 1999, Sayı 62, syf 80-85,
Bileşim Yayıncılık A.Ş., İstanbul
[5] İ.H. Altaş, “Bulanık Mantık: Bulanık Denetim, Enerji”, Elektrik,
Elektronik-3e, Sayı 64, sayfa 76-81, 1999
[6] İ.H. Altaş, “Bulanık Mantık Denetleyici: Matlab/Simulink Ortamı
için Bir Modelleme”, Otomasyon Dergisi, Bileşim Yayınları, Mart
2007, Sayfalar: 58-62
[7] C. Elmas, M.A. Akcayol, T. Yigit, “ Fuzzy PI Controller For Speed
Control of Switched Reluctance Motor “ J.Fac.Eng. Arch. Gazi
Univercity, Vol 22, No 1, 65-72, 2007
[8] M. Ali. Usta, Ö. Akyazı, İ. H. Altaş, “Design and Performance of
Solar Tracking System with Fuzzy Logic Controller Used Different
Membership Functions”, 7th International Conference on Electrical
and Electronics Engineering, ELECO 2011,1-4 December 2011,
Bursa, Turkey
[9] Ömür Akyazı, Murat Küçükali, Adem Sefa Akpınar, “ Sürekli
Mıknatıslı Senkron Motorun Hız Denetiminin Farklı Bulanık Üyelik
Fonksiyonları Kullanılarak Gerçekleştirilmesi”, Elektrik-Elektronik
Bilgisayar Sempozyumu, 5-7 Ekim 2011/Elazığ, Sayfa 311-315.
[10] Ö.Akyazı, M. A. Usta, İ. H. Altaş,”Farklı Bulanık Üyelik
Fonksiyonları Kullanarak Kıyıcı Beslemeli Doğru Akım Motorunun
Hız Denetiminin Gerçeklenmesi”, TOK’11 Otomatik Kontrol Türk
Milli Komitesi Otomatik Kontrol Ulusal Toplantısı, 14-16 Eylül
2011, İzmir, Türkiye. Sayfalar 66-70.
[11] Ö. Akyazı, H. Zenk, A. S. Akpınar, “Farklı Bulanık Üyelik
Fonksiyonları Kullanarak Sürekli Mıknatıslı DA Motorunun Hız
Denetiminin Gerçeklenmesi”, 6th International Advanced
Technologies Symposium (IATS’11)”, 16-18 Mayıs 2011,
Elazığ/Turkey, Sayfa 163-168.
[12] M. A. Usta, Ö.Akyazı, İ. H. Altaş, “Design and Performance of
Solar Tracking System with Fuzzy Logic Controller”, 6th
International Advanced Technologies Symposium (IATS’11)”, 16-
18 Mayıs 2011, Elazığ/Turkey, Sayfa 331-336.
[13] Ö.Akyazı, M. A. Usta, A. S. Akpınar, “Kapalı Ortam Sıcaklık ve
Nem Denetiminin Farklı Bulanık Üyelik Fonksiyonları Kullanılarak
Gerçekleştirilmesi”, 6th International Advanced Technologies
Symposium (IATS’11)”, 16-18 Mayıs 2011, Elazığ/Turkey, Sayfa
158-162.
[14] Mustafa E. Sahin, Ismail H. Altas, H. Ibrahim Okumus, “Design and
Analysis of Fuzzy Controller for a Single Holding Tank”,
International Symposium on INnovations in Intelligent SysTems and
Applications (INISTA 2010), 21-24 June, Kayseri/Turkey, Sayfa
250-254.
[15] E. Ozkop, H. I. Okumus, I. H. Altas, “A Fuzzy Logic Controlled
Electronic Differential for a Direct Wheel Drive Electric Vehicle”,
ICEM’08, XVIII, International Conference On Electrical Machines,
6th-9th of September 2008, Vilamoura – Algarve, Portugal.