Journal of Al Azhar University Engineering Sector

Vol. 12, No. 42, January, 2017, 25-36

COMPARATIVE STUDY OF DIFFERENT CONTROL TECHNIQUES

APPLIED ON 6 DOF ROBOT ARM
Samer Al-Marzoqi1, Ahmad El-Bayoumy Mansour2
1
Research department faculty of engineering Al-Azhar University
2
Department of electrical engineering, Al-Azhar University

ABSTRACT:
Most of the industrial controllers used today for 6 DOF robot arm utilize PID controllers.
Using PID controller without tuning method leads to response with unsuitable time-domain
specifications (maximum-overshot, settling-time, rise-time, …). Ziegler-Nichols method was
introduced as the conventional method for tuning the PID controller. By using this method,
the enhancement of the response time-domain specifications is not sufficient. Recently, fuzzy
logic with and without PID-controller, and genetic algorithm with PID-controller are also
introduced to obtain suitable time-domain specifications of the response. This paper will
introduce a new comparative study for these four tuning methods applied on 6 DOF robot arm
in view of the response time domain specifications and the steady state error. Simulation
results are obtained using the Mat lap/Simulink. The results show that the Fuzzy with PID
controller responds better than the others in case of change in load and system disturbances.

Keywords: Robot Arm; Forward kinematics; Inverse Kinematics; PID controller; Fuzzy
controller; Genetic Algorithm and Fuzzy PID controller.

1. INTRODUCTION
In recent years, industrial and commercial systems take the advantages of robot technology.
Concentrated on control of robotic systems, large number of control researches and numerous
control applications were presented during the last years [1]. Robot manipulator is one of the
interested fields in industrial, educational and medical applications. It works in unpredictable,
hazard and inhospitable circumstances which human cannot reach. For example, working in
chemical or nuclear reactors is very dangerous, while when a robot is used instead of human it
involves no risk to human life. Therefore, modeling and analysis of the robot manipulators
and applying control techniques are very important before using them in these circumstances
to work with high accuracy. The essential problem is to study the robot manipulator problem
from two sides. The first one is the mathematical modeling of the manipulator and the
actuators, which includes an analysis for the forward kinematic, the inverse kinematic and the
modeling of the dc- motor [2, 3, 4] because it is an important issue in a robot manipulator.
The second problem is the control of the robot manipulator. The motivation of the control
technique is the usage of the high precision performance of the robot manipulators in
complicated and hazardous environments. Various controllers have been designed and applied
in the robot manipulator. Designs of various conventional controllers applied on some simple
systems have been introduced in [5]. Due to its simplicity of designing and implementation,
the conventional PID-controller may be the most widely used in the industrial and
commercial applications for the early decades [6]. However, robot manipulator is classified as
nonlinear system. Hence, most of the industrial controllers used today for robot arm utilize
PID controllers together with tuning method to can obtain response with suitable time-domain
specifications. Ziegler-Nichols method was introduced as the conventional method for tuning
 COMPARATIVE STUDY OF DIFFERENT CONTROL TECHNIQUES APPLIED ON 6 DOF ROBOT ARM

the PID controller, but the enhancement of the response time-domain specifications by using
this method is not sufficient. Recently, fuzzy logic with PID-controller [21] or without PID-
controller [7, 8] and genetic algorithm with PID-controller [9] are also introduced to obtain
suitable time-domain specifications of the response. This paper deals with controlling 6 DOF
robot arm such that the arm can follow any predefined path. The paper will introduce a new
comparative study for these four tuning methods applied on 6 DOF robot arm in view of the
response time domain specifications and the steady state error. The results show that the
Fuzzy with PID controller is the best method in view of the time domain specifications of the
response in case of change in load and system disturbances. The paper is organized as
follows: chapter 2 describes the forward kinematic model and the inverse kinematic model is
presented in chapter 3. Chapter 4 presents the mathematical model and the transfer function of
DC motor. Chapter 5 describes the structure of PID controller. Chapter 6 covers the different
tuning control techniques (Ziegler- Nichols Method, Fuzzy Controller without PID-controller,
Fuzzy-PID Controller and Genetic-PID controller). Chapter 7 presents the results and chapter
8 introduces the conclusion of the paper.

2. FORWARD KINEMATIC MODEL

The forward kinematic model is to determine the position and orientation of the end-effector,
given the values for the joint variables of the robot. The joint variables are the angles between
the links in the case of revolute or rotational joints, and the link extension in the case of
prismatic or sliding joints while it is possible to carry out all of the analysis using an arbitrary
frame attached to each link. In this convention, each homogeneous transformation is
represented as a product of six basic transformations. The following equation explains the
forward kinematic problem [10, 11].
A ( , ,…,) = [x, y, z, ] where , and are the input variables, [x, y, z] are the desired
position and is the desired rotation.
1 t   R ,  0 1  t 
A( )   
0 1  0 1 0 1 
Where:
is the direction vector.
is the position vector of point on the line.
so the rotations about this line are:
 R ,  1  R t 
 0 
 1 
The axes of the joints in their home position as
shown in Figure 1 is given by table 1.

Table 1. The axes of the joints in their home position

Joint V T
J1 K 0
J2 I 0
J3 I K
J4 K i
J5 I
J6 K i
 COMPARATIVE STUDY OF DIFFERENT CONTROL TECHNIQUES APPLIED ON 6 DOF ROBOT ARM

cos1  sin 1 0 0 1 0 0 0
 sin  cos1 0 0 0 cos  sin  0
A1 (1 )   1
A2 ( 2 )   2 2 
 0 0 1 0 0 sin  2 cos 2 0
   
 0 0 0 1 0 0 0 1
1 0 0 0  cos 4  sin  4 0 1  cos 4 D3 
0 cos  sin  
L2 sin  3   sin  cos 4 0  sin  4 D3 
A3 ( 3 )   3 3
A4 ( 4 )   4

0 sin  3 cos 3 0   0 0 1 0 
   
0 0 0 1   0 0 0 1 
1 0 0 0  cos 6  sin  6 0 1  cos 6 D3 
0 cos  sin  ( L2  D4 ) sin  5   sin  cos 6 0  sin  6 D3 
A5 ( 5 )   5 5
A6 ( 6 )   6

0 sin  5 cos 5 ( L2  D4 )1  cos 5   0 0 1 0 

   
0 0 0 1   0 0 0 1 
So the kinematic matrix is:
K( , , , )= [10]
3. Inverse Kinematic Model
The inverse kinematic is to find the joint variables of the robot manipulator for a given position and
orientation of the end effectors. The problem of the inverse kinematics (IK) is more difficult than the
forward kinematics problem [11,12,19]. It can be mathematically expressed as:
= (x, y, z, α, γ, φ), where k =1… I, is the joint angles and (x, y, z, α, γ, φ) represents the position
and orientation. The inverse kinematics can be solved for robot manipulator according to the following
steps:
1. Equate the general transformation matrix to the final transformation matrix of the robot
manipulator[11].
 11  12  13 x 
  22  23 y 
H i0   21   A1 A2  Ai (1)
 31  32  33 z 
 
0 0 0 1
2. For the both matrices define:
a) The elements that contain one joint variable.
b) Pairs of elements, which contain only one joint variable.
c) Elements, or combinations of elements, contain more than one joint variable.
3. After defining these elements, equate it to the corresponding elements in the other matrix to
form equations, and then solve these equations to find the values of joint variables.
4. Repeat step (3) to identify all elements in the two matrices.
5. In the case of inaccuracy, solutions look for another one.
6. If there is more joint variable to be found, multiply equation (1) by the inverse of a matrix
for the specified links.
7. Repeat steps (2) through (6) until solution to all joint variables have been found.
8. If there is no solution to the joint variable in terms of an element transformation matrix, it
means that the arm cannot achieve the specified position and orientation; the position is
outside the robot manipulator workspace.
 COMPARATIVE STUDY OF DIFFERENT CONTROL TECHNIQUES APPLIED ON 6 DOF ROBOT ARM

4. DC motor model

In armature control of separately excited DC motors, the voltage applied to the armature of the motor
is adjusted without changing the voltage applied to the field. Figure 2 shows a separately excited DC
motor equivalent model. The relation between the rotor shaft speed and the applied armature
voltage is represented by the transfer function:

= (2)
Since the relation between the position and speed is:

(3)
Thus the transfer function between the shaft position and the armature voltage at no load is [5]:

= (4)

Figure 3. Shows the DC motor model built in Simulink. Input ports are armature voltage (Va) and
load torque (Tload) and the output ports are angular speed in (w) and position (teta). For DC motor
with the following parameters: Moment of inertia = 0.000052 Kg.m2, Friction coefficient =
0.01 N.ms, Back EMF constant = 0.235 V/ms-1, Torque constant = 0.235 Nm/A, Electric
resistance = 2 ohm and Electric inductance = 0.23 H, the transfer function of the DC motor of
the first DOF considered is defined as:

= (5)

Figure 3: DC Motor Block Diagram Model

 COMPARATIVE STUDY OF DIFFERENT CONTROL TECHNIQUES APPLIED ON 6 DOF ROBOT ARM

5. PID CONTROLLER
PID controllers are widely used in industrial control applications due to their simple structures, comprehensible
control algorithms and low costs. Figure 4 shows the schematic model of a control system with a PID controller.
The mathematical model of the PID-controller with as input and as output is written as:

= (6)
or

= (7)
Where:
=Proportional gain
=integral gain
=derivative gain
=integral time
=derivative time
If the controller is digital, then the derivative term may be replaced with a backward difference and the integral
term may be replaced with a sum. For a small constant sampling time equation (7) can be approximated for
the digital controller as [5]:

= (8)

6. TUNING METHODS
Position of DC motor can be controlled by using one of following different tuning methods:-
6.1 PID Controller (Ziegler- Nichols Method).
6.2 Genetic PID controller.
6.3 Fuzzy Controller.
6.4 Fuzzy-PID Controller.
6.1 Ziegler-Nichols Method
PID controllers are usually tuned using Ziegler- Nichols frequency method. In this method and
is set initially and then using the proportional action only, the value of is increased from 0 to
ultimate value slowly until the process start to oscillate. The gain when this occurs is and the period
of the oscillation is .The advantage of this method of being very easy to use. Ziegler-Nichols (Z-N)
suggested that the value of , and can be found out by using the formula shown in
Table3[13, 22, 23].
 COMPARATIVE STUDY OF DIFFERENT CONTROL TECHNIQUES APPLIED ON 6 DOF ROBOT ARM

Table 3: Ziegler-Nichols rule

Controller
P 0.5
PI 0.45 /1.2
PID 0.6 /2 /8
6.2 Tuning of PID controller using genetic algorithm approach
Genetic Algorithms (GAs) are a stochastic global search method that mimics the process of natural evolution. It
is one of the methods used for optimization. The computational systems keep on improving and this makes GA
useful for optimization. The genetic algorithm starts with no knowledge of the correct solution and depends
entirely on responses from its environment and evolution operators such as reproduction, crossover and mutation
to arrive at the best solution. GA starts with an initial chromosome & check for the fitness value. The fittest
chromosomes are taken as parents further they are reproduced, crossed over & mutated. The offspring is checked
for the value of the fitness. The implementation of the tuning procedure through genetic algorithms starts with
the definition of the chromosome representation. The gains of , and are real numbers and characterize
the individual to be evaluated. [14, 18] GA can be applied to the tuning of PID position controller gains to ensure
optimal control performance at nominal operating conditions. The block diagram for the entire system is given in
Figure 5 and also the genetic algorithm parameters chosen for the tuning purpose are shown Table 4.

Figure 5: Block Diagram of the Entire System

Fitness Function (based on model reference approach):
Minimize J
Where:

J= (9)
system response.
model response.
= 500 Sample

= (10)
= 21, 26, 31 and 36 Sample
Note .Each 100 sample equal 1sec (sampling time =0.01sec).
Table 4: Parameters of GA
GA property Value/Method
Population Size 80
Maximum number of Generations 100
Performance Index/Fitness function Mean Square Error
Selection Method Normalized Geometric Selection
Probability of Selection 0.05
Crossover Method Scattering
Mutation Method Uniform Mutation
Mutation Probability 0.01
6.3 Fuzzy Logic Controller:
 COMPARATIVE STUDY OF DIFFERENT CONTROL TECHNIQUES APPLIED ON 6 DOF ROBOT ARM

Fuzzy logic is expressed by means of the human language. Based on fuzzy logic, a fuzzy controller converts a
linguistic control strategy into an automatic control strategy, and fuzzy rules are constructed by expert
experience or knowledge database. First, set the error e (t) and the error variation de (t) of the angular velocity to
be the input variables of the fuzzy logic controller. The control voltage u (t) is the output variable of the fuzzy
logic controller. The linguistic variables are defined as {NB, NM, ZR, PM, PB}, where NB means negative big,
NS means negative medium, ZR means zero, PM means positive medium and PB means positive big [15, 16]
and [17, 24, 25]. The fuzzy rules are summarized in Table 5.The fuzzy inference mechanism in this study
follows as:
Table 5: Fuzzy Rules
𝑒/ e NB NM ZR PM PB
NB NB NB NM NM ZR
NM NB NM NM ZR PM
ZR NM NM ZR PM PM
PM NM ZR PM PM PB
PB ZR PM PM PB PB

Figure 6: Membership function for inputs e and e

Here max-min type decomposition is used and the final output for system is calculated by
using center of area gravity method.
=MAX [ ,]
Where:
= membership function of
= membership function of
= membership function of
j is an index of every membership function of fuzzy set, m is the number of rules and is the
inference result. Fuzzy output u (t) shown in Figure 7. Can be calculated by the center of
gravity defuzzification as:

(11)
 COMPARATIVE STUDY OF DIFFERENT CONTROL TECHNIQUES APPLIED ON 6 DOF ROBOT ARM

Figure 7: Membership function for output u

NB stands for negative big with numerical equivalent of 0, NM stands for negative medium with numerical
equivalent value of 0.25, ZR stands for zero with numerical equivalent value of 0.5, PM stands for positive
medium with a numerical equivalent of 0.75, and PB stands for positive big with numerical equivalent of 1.

6.4 FUZZY PID CONTROL

Fuzzy-PID (FPID) provides good performance as compared to standalone Fuzzy. Simple rule base are
used for Fuzzy controller while FPID uses different rule base for proportional, integral and derivative
gains to make response faster [12]. Fuzzy-PID controllers are arranged into two sorts: the direct action
fuzzy control and the fuzzy supervisory control. The direct action sort replaces the PID control with a
feedback control loop to compute the action through fuzzy reasoning where the control actions are
resolved directly by means of a fuzzy inference. These sorts of fuzzy controllers are also called PID-
like controllers [19, 21]. On the other hand, the fuzzy PID type attempts to give nonlinear action for the
controller output utilizing fuzzy reasoning where the PID gains are tuned based on a fuzzy inference
system rather than the traditional methodologies. The design process of the fuzzy controller is
described as follows:
 Define the input and output variables of FLC. In this work, there are two inputs of FLC, the
error e (t) and its rate of change of error𝑒 (t) and three outputs, and are respectively as
shown in Figure 8.

Figure 8: Fuzzy PID control structure

 Fuzzify the input and output variables by defining the fuzzy sets and membership
functions. and the output of FLC has the following membership function as shown in
Figure 9 for the three outputs , and .
 Outline the inference mechanism rule to get the input-output relation. This work
utilized Mamdani (max-min) inference mechanism as shown in Figure 10.

 Defuzzify the output variable. Here, the center of gravity (COG) method, the most
frequently utilized method, is utilized. The control activity is[20]:
 COMPARATIVE STUDY OF DIFFERENT CONTROL TECHNIQUES APPLIED ON 6 DOF ROBOT ARM

Figure 9: PID parameters membership functions ( , and )

Figure 10: Fuzzy inference block

7. Result AND Discussion
Simulation result in Figure 11 and Figure 12 shows the output response of the proposed
controllers using the step as input signal. The tow figures show the performance of using the
different control approach. In addition, they show the effectiveness of the Fuzzy PID
controller for rejection disturbance. If the load torque with -0.7 N.m is applied on the desired
angle. The obtained result shows the effect of the disturbance on the output response after 0.5
second and the efficacy of the Fuzzy PID controller for tuning PID parameters and
eliminating the disturbance.
 COMPARATIVE STUDY OF DIFFERENT CONTROL TECHNIQUES APPLIED ON 6 DOF ROBOT ARM

Figure 11: Output responses for step input using proposed controllers without disturbance

Figure 12: Output responses for step input using proposed controllers with
disturbance
Clearly, the fuzzy PID control achieved better performance than other tuning methods in
terms of time response. The above figures show the effect of small disturbance after 0.5
second and effectiveness of the fuzzy PID control in eliminating the presence of disturbances.
Performance of proposed controllers is summarized in Table 9.
Table 9: Time-Domain Specifications Parameter Result
Time-Domain Specifications Parameter Characteristic Controller type
Steady state Settling Rising Overshoot (
error (SSE) time Time OS% )
0.009 0.25 0.0905 - PID controller(Ziegler-
Nichols)
0.007 0.16 0.068 - Fuzzy logic control
0.005 0.13 0.041 0.05 Genetic PID control
0.001 0.08 0.028 0.02 Fuzzy PID control
 COMPARATIVE STUDY OF DIFFERENT CONTROL TECHNIQUES APPLIED ON 6 DOF ROBOT ARM

8. CONCLUSION
Robot manipulators have become increasingly important in the field of flexible
automation. There are so many Controllers to control the actuator accurately and
robustly. In this paper, DC linear servo motor is used as actuator. Appling a control
technique is important to guarantee high efficiency and lower error for the motion of
the robot. This paper has introduced a new comparative study for four controlling
methods (Ziegler-Nichols method, fuzzy logic PID-controller with and without PID-
controller, and genetic algorithm with PID-controller) applied on 6 DOF robot arm in
view of the response time domain specifications and the steady state error. The 6 DOF
robot arm is considered such that the arm can follow any predefined path. Simulation
results are obtained using the Mat lap/Simulink. The results show that the Fuzzy PID
Controller can achieve better accuracy and has less or no deviation from the trajectory
compared to the other controllers.

REFERENCES
[1] Jong-Hwan Kim, Hyun Myung, “Robot Intelligence Technology and
Applications,” 4th Edition, 2016.
[2] Y. Yang, W. Wang, D.-J. Yu and G. Ding, “A Fuzzy Parameters Adaptive PID
Controller Design of Digital Positional Servo System,” IEEE Proceeding of the First
International Conference on Machine Learning and Cybernetics., Vol. 1, pp. 310-
314, Nov. 2015
[3] Asan M. K., Saravanakumar. G., Valamarthi. K.,Devaraj. D., Radhakrishnan. T.
K., Real‐Coded Genetic Algorithm for System Identification and Tuning of a
Modified Model Reference Adaptive Controller for a Hybrid Tank System, Applied
Mathematical Modelling, 37, pp. 3829‐3847, 2015.
[4] Wong, C. C., Lin, B. C., Lee, S. A. and Tsai, C. H., “GA-Based Fuzzy System
Design in FPGA for an Omni-Directional Mobile Robot,” Journal of Intelligent &
Robotic Systems, Vol. 44, pp. 327_347 (2015).
[5] K. Ogata “Modern Control Engineering, 4th Edition”, Dorling Kindersley Pvt.
Ltd., India (2009).
[6] Y. Yang, W. Wang, D.-J. Yu and G. Ding, “A Fuzzy Parameters Adaptive PID
Controller Design of Digital Positional Servo System,” IEEE Proceeding of the First
International Conference on Machine Learning and Cybernetics. Vol. 1, pp. 310-
314, Nov. 2002.
[7] M.W. Spong, S. Hutchinson and M. Vidyasagar, Robot Modeling and Control, 1st
Edition, Jon Wiley & Sons, Inc, 2005.
[8] J.J. Crage, Introduction to Robotics Mechanics and Control, 3rd Edition, Prentice
Hall, 2005.
[9] K. H. Kalil, Nonlinear Systems, 3rd Edition, Prentice-Hall, 2002.
[10] Indranil Pan, Saptarshi Das and Amitava Gupta, “Tuning of an optimal fuzzy
PID controller with stochastic algorithms for networked control systems with random
time delay”, ISA Transactions, Volume 50, Issue 1, pp. 28-36, January 2011.
[11] Baogang Hu, George K. I. Mann and Raymond G. Gosine, “New methodology
for analytical and optimal design of fuzzy PID controllers”, IEEE Transactions on
Fuzzy Systems, Volume 7, Issue 5, pp. 521-539, October 1999.
[12] Zhi-Wei Woo, Hung-Yuan Chung and Jin-Jye Lin, “A PID type fuzzy
controller with self-tuning scaling factors”, Fuzzy Sets and systems, Volume 115,
Issue 2, pp. 321-326, October 2000.
[13] Q.Wang, P Spronck and R Tracht, "An Overview Of Genetic Algorithms
Applied To Control Engineering Problems", Proceedings of the Second International
Conference on Machine Learning And Cybernetics, 2003
[14] P. I. Corke, "A simple and systematic approach to assigning Denavit-
Hartenberg parameters," IEEE Transactions on Robotics, vol. 23, pp. 590-594, June
2007.
 COMPARATIVE STUDY OF DIFFERENT CONTROL TECHNIQUES APPLIED ON 6 DOF ROBOT ARM

[15] R. P. Paul and H. Zhang, "Computationally e_cient kinematics for

manipulators with spherical wrists," Int. J. Robot. Res, vol. 5, pp. 32-44, 2012
[16] Jolly Shah, S.S.Rattan, B.C.Nakra, Kinematic Analysis of 2-DOF Planer Robot
Using Artificial Neural Network, World Academy of Science, Engineering and
Technology, Vol. 81, pp. 282-285, 2011
[17] S. A. Mazhari and S. Kumar, “PUMA 560 Optimal Trajectory Control using
Genetic Algorithm, Simulated Annealing and Generalized Pattern Search
Techniques,” International Journal of Electrical, Computer and Systems Engineering,
Vol. 2, No. 1, pp. 71-80, March 2008.
[18] Atul Kumar Dewangan, Sashai Shukla, Vinod Yadu “Speed Control of a
Separately Excited DC Motor Using Fuzzy Logic Control Based on Matlab
Simulation Program” International Journal of Scientific & Technology Research
Volume 1, Issue 2, ISSN 2277-8616 pp. 52 – 54 , March 2012.
[19] L. Wang, M. Tian and Y. Gao, “Fuzzy Self-adapting PID Control of PMSM
Servo System,” IEEE International Electric Machines & Drives Conference. Vol. 1,
pp. 860–863, May 2007.
[20] A new method for position control of a 2-DOF robot arm using neuro – fuzzy
controller, Indian Journal of Science and Technology, Vol. 51, pp. 2253-2257, 2012
[21] M. D. Youns, S. M. Attya and A. I. Abdulla, "Position Control of Robot Arm
Using Genetic Algorithm Based PID Controller", Al-Rafidain Engineering, Mosul,
Iraq, Vol.21,No. 6, pp. 19-30, December 2013.
[22] S. Yadegar an Azura binti Che Soh ,"Design Stable Robust Intelligent
Nonlinear Controller for 6- DOF Serial Links Robot Manipulator", International
Journal of Intelligent Systems and Applications (IJISA), MECS, pp.19-38, July 2014.
[23] Ch. R. Kumar, K. R. Sudha, D. V. Pushpalatha and Ch. V. N. Raja , "Fuzzy C-
Means Controller for a PUMA-560 Robot Manipulator", IEEE Workshop on
Computational Intelligence: Theories, Applications and Future Directions, IIT
Kanpur, India, pp. 57-62 , July 2013.
[24] Manafeddin Namazov, “DC motor position control using fuzzy proportional-
derivative controllers with different Defuzzification methods” An Official Journal of
Turkish Fuzzy Systems Association Vol.1, No.1, pp. 36-54, July 2010.
[25] Y. Li, K. Ang, G. Chong. PID Control System Analysis and Design: Problems,
Remedies and Future Directions. IEEE Control Systems Magazine Vol.1, No.1, pp.
32–41, February 2006.
[26] J. Han, “From PID to active disturbance rejection control,” IEEE Trans. Ind.
Electron., vol. 56, no. 3, pp. 900–906, March 2009.
[27] Mendel, J. M., Fuzzy logic systems for engineering: A tutorial. Proceedings of
the IEEE, vol. 83, no. 3, pp. 345–377, March 1995.
[28] A. Costa, A. De Gloria, F. Giudici, and N. Olivieri, “Fuzzy logic
microcontroller,” IEEE Micro, vol. 17, no. 1, pp. 66-74, November 1997.