See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/271436688

Design of a fuzzy logic controller for a 2-DOF robot manipulator

Conference Paper · November 2013

DOI: 10.1109/ICECCO.2013.6718279

CITATIONS READS
6 835

3 authors:

Mustafa Şinasi Ayas Yahya Danayiyen

Karadeniz Technical University Karadeniz Technical University
46 PUBLICATIONS 524 CITATIONS 15 PUBLICATIONS 38 CITATIONS

SEE PROFILE SEE PROFILE

İsmail H. Altas
Karadeniz Technical University
194 PUBLICATIONS 2,613 CITATIONS

SEE PROFILE

All content following this page was uploaded by Yahya Danayiyen on 13 February 2023.

The user has requested enhancement of the downloaded file.

 DESIGN OF A FUZZY LOGIC CONTROLLER FOR A 2-DOF ROBOT MANIPULATOR

Mustafa Şinasi Ayas#1, Yahya Danayiyen#2, İsmail Hakkı Altaş#3

#
Electrical and Electronics Engineering Department, Karadeniz Technical University
1
[email protected]
2
[email protected]
3
[email protected]

ABSTRACT The performance of the controllers can be observed

by comparing the overshoot, settling time, rise time and
In this study, a 2-DOF robot manipulator is designed and steady-state error [7]. Since the performance indices are
controlled by modified fuzzy logic controller. In order to have based on the time response of the error signal, they include
a good position tracking a fuzzy logic controller, which is all information about the dynamic behavior of the controlled
convenient for nonlinear complex systems, is developed by system, they are used to measure the performances of
modifying the ones from previous works. The performance different controllers applied to the same system.
of the developed FLC is compared with the classical PID In this study, a two degree of freedom (2-DOF)
controller of which the parameters are tuned by means of robot manipulator shown in Fig. 1 is modeled by using
Integral Squared Error (ISE) performance measure. The Matlab/Simulink graphical user interface and the positions of
optimum results for both developed FLC and classical PID the links are controlled by developed FLC. The performances
controller are given in the result section. of the proposed FLC and classical PID controller is compared
using performance index called Integral Squared Error (ISE)
Index Terms— fuzzy logic, controller, robot which is a version of the integral-error based performance
manipulator, performance index. measure methods. The equation of the ISE is given as

1. INTRODUCTION ISE= ∫ ( )2 dt (1)

Robot arms or manipulators which are mechanical devices where e(t) is the error signal.
driven by electrical motors or actuators have been used in
many applications in industry such as painting, drilling, 2. 2-DOF ROBOT MANIPULATOR DYNAMICS
material handling and pick and place. Since the main aim is
to increase the productivity in these fields, the control of A 2-DOF robot manipulator is shown in Fig. 1. Since
manipulators come into prominence. mobility of this manipulator is two, two actuators are required
The most commonly used industrial controllers are in order to control the manipulator. The abbreviations of the
proportional-integral-derivative (PID) controllers of which parameters used in Fig. 1 and their equivalents are defined in
the first tuning method is proposed by Ziegler and Nichols Table 1.
[1]. Although advanced controllers are available, PID
controller have been preferred because of its simple and easy
to understand structure in industrial applications where the
performance is satisfactory. However, due to their constant
parameters, PID controllers are not able to handle the control
process of the systems with uncertainties and nonlinearities
as the adaptive controllers do. In order to achieve adaptivity
in the controllers, novel controllers such as Fuzzy-PID [2],
Neuro PID [3] and PSO-PID [4] controller have been
developed and used during last decades.
Since fuzzy logic deals with the uncertainty and
vagueness, it is superior to PID controller whose parameters
are tuned according to precise mathematical model and the
effectiveness of Fuzzy Logic Controller (FLC) has been
proven in non-linear complex systems [2, 5, 6]. Figure 1. The 2-DOF robot manipulator

978-1-4799-3343-3/13/$31.00 ©2013 IEEE 265 ICECCO 2013

 Table1. Symbols of the parameters KE = K1+K2 (8)
Mi Mass of link I (kg) P1= M1gL1 sinƟ1 (9)
Li Length of link I (m)
Ɵi Angle of link I (degree) P2= M2g[ L1 sinƟ1 + L2 sin(Ɵ1+ Ɵ2) ] (10)

g Gravitational constant PE = P1+P2 (11)

Ki Kinetic energy of link i
Ɫ = KE – PE (12)
Pi Potential energy of link i
Ɫ Ɫ
τƟi Necessary torque of link i τƟ1,2 = [ Ɵ ̇,
]- Ɵ ,
(13)

Using Lagrange equations, motion equations of the τƟ1 = M1L22 (Ɵ̈1+Ɵ̈2) + M2L1L2 cosƟ2(2Ɵ̈1+Ɵ̈2 )
2-DOF manipulator can be calculated. Equations 14 and 15
+(M1+M2) L12Ɵ̈1 - M2L1L2Ɵ̇22sinƟ2
show the necessary torque to be applied the actuators in order
to reach the required reference point. - 2 M2L1L2Ɵ̇1Ɵ̇2 sinƟ2 + M2gL2 cos(Ɵ1+ Ɵ2)
+ (M1+M2)gL1cosƟ1 (14)
x1=L1 cosƟ1 (2)
τƟ2 = M2L2 (Ɵ̈1+Ɵ̈2) + M2L1L2 Ɵ̈1cosƟ2
2

y1=L1 sinƟ1 (3)

+ M2L1L2Ɵ̇12sinƟ2 + M2gL2 cos(Ɵ1+ Ɵ2) (15)
x2=L1 cosƟ1 + L2 cos(Ɵ1+ Ɵ2) (4)
In this study, SimMechanics library of
y2=L1 sinƟ1 + L2 sin(Ɵ1+ Ɵ2) (5) Matlab/Simulink is used to develop 2-DOF robot
manipulator. Fig. 2 shows the physical part of the
K1= M1L12 Ɵ̇12 (6) manipulator which is move by means of DC motors.
K2= M2L12 Ɵ̇12 + M2L22 ( Ɵ̇1 + Ɵ̇2)2

+ M2L1L2 ( Ɵ̇12 +Ɵ̇1Ɵ̇2 ) cosƟ2 (7)

Figure 2. The 2-DOF robot modeled in Matlab/Simulink

3. FUZZY LOGIC CONTROLLER Fuzzification part convert crisp input values to

linguistic variables and determine the membership values of
Fuzzy logic controllers are convenient for nonlinear complex each input for each fuzzy set by means of membership
systems that have unknown or unmodeled dynamics. A fuzzy functions which are triangular type in this study.
logic controller consists of four main part called fuzzification, Rule base contains the necessary rules that will
rule base, inference system and defuzzification. Fig. 3 shows process according to input values. The rules are constituted
the basic structure of a FLC [8]. according to IF-THEN-ELSE structure. The rule base which
consists of nine rules is given in Table 2. The abbreviations
of parameters used in Table 2. Are defined as: N, Negative;
Rule Base
P, Positive; Z, Zero; e, error; Δe, error change; Δu, control
input change.

Table2. Rule Table

Fuzzification Inference Defuzzification e
Δu
N Z P
Figure 3. The basic structure of a FLC
N N N Z
Δe Z N Z P
P Z P P

266
 Inference system calculate the fuzzy result obtained Table3. The results obtained by trial error method
from the rule base according to the input values. Mamdani,
Theta 1
Max-min, max-dot, and Takagi-Sugeno-Kang are the
common methods used in industrial fuzzy logic controllers. PID Kp Ki Kd ISE
Mamdani fuzzy reasoning is used in this study. 23 7 0.8 4288
The output of the inference system which is a fuzzy FLC 4060
value is convert to the crisp output value using the methods Theta 2
such as center of area (COA) and center of gravity(COG) in PID 28 6 0.4 1586
defuzzification part. COA is used as defuzzification method FLC 1210
in this study.
The inside structure of the developed FLC is given
in Fig. 4. The details of the developed FLC is clarified in [8]. Theta 1
50

45

40
ref
35 FLC
PID

Position (degree)
30

25

20

15

10

0
0 1 2 3 4 5 6 7 8 9 10
time (sec)

Figure 6. The position waveform for theta 1.

Figure 4. The inside of the developed FLC block in

Matlan/Simulink Theta 2
35

4. RESULTS
30

The Simulink model of 2-DOF robot manipulator position PID

control system is shown in Fig.5. The controller parameters 25
ref
of both classical PID controllers and developed FLC is FLC
Position (degree)

determined by using ISE performance measure method. 20

Several simulation results are obtained by using trial and

error methods in order to obtain optimum result, minimum 15
ISE value.
The parameters of PID controllers for minimum ISE 10
value is given in Table 3. Fig.6 and Fig.7 shows the position
waveform of the links obtained using optimum parameters. 5
It is clear from Table 3., Fig.6. and Fig. 7. that FLC
gives better performance for both link. 0
0 1 2 3 4 5 6 7 8 9 1
Time (sec)

Figure 7. The position waveform for theta 2.

Figure 5. 2-DOF robot arm position control model

267
 5. CONCLUSION

In this study a 2-DOF manipulator has been designed in

Matlab/Simulink. The position control of the designed
manipulator is obtained using both classical PID controller
and the developed FLC. The optimal parameters of the
controllers are tuned using ISE performance measure method.
A comparison of the results is obtained by applying the both
controllers under the same operating conditions. It is quite
obvious that the developed FLC gives better result than the
classical PID controller.

6. REFERENCES

[1] J. G. Ziegler, and N.B. Nichols, "Optimum setting for automatic

controllers," Trans. ASME, vol. 64, pp. 759-768, 1942.
[2] R. Arulmozhiyal and R. Kandiban, “Design of Fuzzy PID
Controller for Brushless DC Motor”, International Conference on
Computer Communication and Informatics, Coimbatore, INDIA,
Jan. 10 – 12, 2012.
[3] F. Hong and Y. Xiaohong, “Design and Simulation of Neuron
PID Controller”, International Conference of Information
Technology, Computer Engineering and Management Sciences, vol.
1, pp.80-82, 2011.
[4] W. Dongsheng, Y. Qing and W.Dazhi, “A Novel PSO-PID
Controller Application to Bar Rolling Process”, Proceedings of the
30th Chinese Control Conference, Yantai, China, July 22-24, 2011.
[5] R.Arulmozhiyal and K.Baskaran, “Implementation of Fuzzy PI
Controller for Speed Control of Induction Motor Using FPGA”,
Journal of Power Electronics, Vol.10, No.1, Jan 2010, pp.65-71.
[6] P. Ananthababu and B. Amarendra Reddy; “Control of PMDC
Motor using Fuzzy PI Controller”; International conference on
“control, automation, communication and energy conservation, 4th-
6th June 2009.
[7] W. Tan, J. Liu, T. Chen, H. J. Marquez; “Comparison of some
well-known PID tuning formulas”; Computers and Chemical
Engineering, pp. 1416-1423, 2006.
[8]I.H.Altas and A.M. Sharaf, “A generalized direct approach for
designing fuzzy logic controllers in matlab/simulink gui
environment”, International Journal of Information Technology and
Intelligent Computing, Vol.1, No:4, 2007.

268

View publication stats