International Journal of Scientific & Engineering Research, Volume 5, Issue 11, November-2014 1044

ISSN 2229-5518

Speed Control of DC Motor Using PID & Smart

Controller
Purushotam Kumar, Prabhakar Kumar Prabhat, Mithun Kumar, Dr. S.D. Choudhary

Abstract- The thesis describes about the concept of DC motor and Speed control separately excited DC motor. Motor speed is
controlled with PID controller and first system is checked without controller on loaded and unloaded condition then add PID controller
and system is tuned using its existing tuning methods. After it system is further tuned in order to get desired value with less steady state
error. And then the result is discussed. This paper describes about the basic concepts of Fuzzy Set and Fuzzy Logic, the speed control
with the help of Fuzzy controller. Fuzzy controller provides better control strategies than other controllers. Optimization of Fuzzy
controller with Simulink model describes in this paper and a new way for faster response and smooth output. The comparison of these
two controllers’ results is also showed. From the results it is proved that Fuzzy Controller is the best controller. Finally the MATLAB
Simulation is discussed.

Index Terms: DC Motor, PID Controller, Existing tuning, Steady State Error, Fuzzy logic, Fuzzy controller, Simulink, MATLAB.

1. CONCEPTS OF DC MOTOR
For A motor convert electrical energy into mechanical energy.
There are two types of motor: AC motor & DC motor. A
simple DC motor use electricity and magnetic field for
producing torque which rotate the motor. PMDC permanent

IJSER
magnet DC motor outperforms to AC motor because it
provide better speed control on high torque loads and use in
wide industrial application. The applied voltage describes the
speed of motor while current in the armature windings shows
the torque. If applied load increased in the shaft of motor then
in order to sustain its speed motor draw more current from Where, V A is the armature voltage. (In volt)
supply and if supply is not able to provide enough current E b is back emf the motor (In volt)
then motor speed will be affected. DC motor provides more I a is the armature current (In ampere)
effective results if chopping circuit is used. Low power DC R ais the armature resistance (In ohm)
motors usually use in lifting and transportation purposes as L a is the armature inductance (In Henry)
low power AC motors don’t have good torque capability. DC T m is the mechanical torque developed (In Nm)
motor used in railway engines, electric cars, elevators, robotic J m is moment of inertia (In kg/m²)
applications, car windows and wide verity of small appliances
B m is friction coefficient of the motor (In Nm/ (rad/sec))
and complex industrial mixing process where torque cannot
ω is angular velocity (In rad/sec)
be compromised.
We know that
2. SPEED CONTROL OF DC MOTOR ω = (V a -I R)/K aϕ …………….. (1)
Where, ϕ = Field flux per pole
The term speed control stand for intentional speed
Ka = Armature constant = PZ / 2πa
variation carried out manually or automatically DC motors
Where, P = No. of pole
are most suitable for wide range speed control and are there
for many adjustable speed drives. Specification of the dc motor:
Armature resistance (R a) = 0.5Ω
• Purushotam Kumar, Research Scholar, Department of Electrical
Engineering, Asansol Engineering College, WB, India E-mail: Armature inductance (L a) = 0.02 H
purushotamkumar88@gmail.com
• Prabhakar Kumar Prabhat is currently associated with the Armature voltage (V a) = 200
department of Electrical Engineering,DAVIET,Daltonganj,
Jharkhand, India, E-mail: pkp.nitp@gmail.com Mechanical inertia (j) = 0.1 Kg .m
• Mithun Kumar is currently associated with the department of Friction coefficient (B m) = 0.008 N.m/rad/sec
Electrical Engineering, DAVIET, Daltonganj, Jharkhand, India,
E-mail: mithunkism@gmail.com Back emf constant (k) = 1.25 V/rad/sec
• Dr. S.D. Choudhary is currently associated with the department
of Electronics & Communication Engineering, DAVIET, Rated speed = 1500 r.p.m
Daltonganj, Jharkhand, India, E-mail: suryadeo.bit@gmail.com Motor torque constant=N.m/A
IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 11, November-2014 1045
ISSN 2229-5518

Calculation:
Speed at full load when ω=157.07 rad/sec
E b=1.25x157.07=196.3 V

VA= E b+I R

200 = 196.3+I R
IA =7.325 Amps

3. SPEED CONTROL WITH PID

CONTROLLER
A proportional–integral–derivative controller (PID controller)
Fig: Three values of Kp , Ki and Kd
is widely used in industrial control systems. It is a generic
control loop feedback mechanism and used as feedback The integral term contribute error and duration of error
controller. PID working principle is that it calculates an error proportionally. Error sum gives offset that corrected
value from the processed measured value and the desired previously. The calculated error is multiplied by integral gain
reference point. The work of controller is to minimize the error and then added to controller output. It finally reduced the
by changing in the inputs of the system. If the system is not steady state error.
clearly known then applying PID controller provide the best
results if it is tuned properly by keeping parameters of the

IJSER
system according to the nature of system.

4. TUNING OF PID CONTROLLER BY

ZIEGLER-NICHOLS METHOD

Fig: Blok diagram of PID controller

The PID measurement depends upon three parameters which
is called the proportional, the integral and derivative part
which is called P, I and D part. P determine the reaction to
current error, I determine reaction to the sum of recently
appeared errors, D Determine reaction according to the rate
off error changing. The sum of all three parts contribute the
control mechanism such as speed control of a motor in which
P value depends upon current error, I on the accumulation of Fig: Simulink model of DC motor with PID controller
previous error and D predict future error based on the current
rate of change. As derivative action is sensitive to noise so As in this project the target is to control the speed so speed is
mostly the controllers are PI controller rather than PID as it is send back for checking the system in closed loop and tuned
not possible a system without disturbances. Integral part helps PID controller. The method used for tuning is Ziegler–Nichols
the system to reach onto its target value while P part increase method.
overshoots. The P term take the output to proportional of error
value. Its response can be adjusted by multiplying the error by According to Ziegler–Nichols method:
a constant Kp which is called proportional gain. If
proportional gain is large then it creates a high overshoot  Run the controller by taking only P value.
which unstable the system, while a small output change  Increase P value of the system until it self-oscillating
makes a small control action. with constant amplitude.
 Then take controller gain time period
IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 11, November-2014 1046
ISSN 2229-5518
Different values of K P , K I & o K d of PID controller for position
from graph analysis:-

IJSER CONCLUSSION:

Applying PID controller both system become marginally

stable. But settling time is 10 sec for angle of pendulum which

should be decreased. Too much damping is occurs for both

angle and position. So we can’t say that PID is good for

inverted pendulum.

5. SPEED CONTROL WITH FUZZY

CONTROLLER
Different values of k P ,k I & o K d of PID controller for angle from graph
analysis:- Controllers based on the fuzzy logic give the linguistic
strategies control conversion from expert knowledge in
automatic control strategies. The first fuzzy logic based
controllers application was done by Assilian and Mandani [2].
The recent fuzzy logic controller application [3] in water
quality control, train operation automatic system, elevators
control, nuclear reactor control and fuzzy computers shows an
efficient way for using the fuzzy control in complex process
which can be controlled by a skilful human being without
IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 11, November-2014 1047
ISSN 2229-5518
knowing its dynamic. The fuzzy logic controller consists of a others control system. While the others control system use
linguistic propositions and rules set, which defines individual difficult mathematical calculation to provide a model of
control actions. A fuzzy logic controller designed on the basis the controlled plant, it only uses simple mathematical
of the fuzzy logic is an approximate reasoning-based calculation to simulate the expert knowledge. Fuzzy set of
controller, which does not require exactly analytical models theory represent the human reasoning with knowledge that is
and is much closer in spirit to human thinking and natural almost impossible to represent in quantitative measures or for
language than the traditional logic system. Essentially, the that control plants that are hard to control or ill defined. Fuzzy
control strategies in the FLC are based on expert experience, inference system models the system using if-then rules. Fuzzy
so the fuzzy logic controller can be regarded as the simulation set theory proposed the membership function at range of
of a humanoid control model. When the designing a FLC, the numbers [0, 1] or False or true membership function. This
control strategies have to be regarded as the simulation of a theory provides the mathematical strength to check the
humanoid control rules pre-constructed by control result fails uncertainties connected with human thinking or reasoning.
to meet the system requirement due to a change in the outside Fuzzy logic is suitable for a model that is hard to control or
environment. The possible solution to this problem is that we non linear models. This system also provides control over
can adjust either the membership function of the fuzzy sets or MIMO systems and also allows decision making with
the control rules to achieve the control objective. incomplete information. Human reasoning can also be known
as multi valued imprecise. [21] The requirement for the
application of a FLC arises mainly in situations where:

6. The description of the technological process is

IJSER
available only in word form, not in analytical form.
7. It is not possible to identify the parameters of the
process with precision.
8. The description of the process is too complex and it is
more reasonable to express its description in plain
language words.
9. The controlled technological process has a "fuzzy"
character.
10. It is not possible to precisely define these conditions.

A fuzzy logic controller has four main components as shown

in Figure:

1. Fuzziffication
Fuzzy logic is a type of multi valued logic. It deals with
2. Inference engine
approximate reasoning rather than precise. Fuzzy logic
3. Rule base
derived from fuzzy set theory. Fuzzy logic was first proposed
4. Defuzzification
by LotfiZadeh in 1965. Fuzzy logic has currently used in
control theory, artificial intelligence systems specially to Fuzzification
control complex aircraft engines and control surfaces,
helicopter control, messile guidance, automatic transmission, The first step in designing a fuzzy controller is to decide
wheel slip control, auto focus cameras and washing machines, which state variables represent the system dynamic
railway engines for smooth drive and fuel consumption and performance must be taken as the input signal to the
many industrial processes. Fuzzy logic provide better results if controller. Fuzzy logic uses linguistic variables instead of
we compared it with PID controller. Fuzzy logic control is a numerical variables. The process of converting a numerical
control algorithm based on a linguistic control strategy, variable (real number or crisp variables) into a linguistic
which is derived from expert knowledge into an automatic variable (fuzzy number) is called fuzzification. This is
control strategy. The operation of a FLC is based on achieved with the different types of fuzzifiers. There are
qualitative knowledge about the system being controlled .It generally three types of fuzzifiers, which are used for the
doesn't need any difficult mathematical calculation like the fuzzification process; they are:
IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 11, November-2014 1048
ISSN 2229-5518
 Singleton fuzzifier abscissa of the center of gravity of the fuzzy set is calculated as
 Gaussian fuzzifier follows:
 Trapezoidal or triangular fuzzifier

Rule Base
Where x i is a point in the universe of the conclusion
A decision making logic which is, simulating a human (i=1, 2, 3….) and µ c (x i) is the membership value of the
decision process, inters fuzzy control action from the resulting conclusion set. For continuous sets summations are
knowledge of the control rules and linguistic variable replaced by integrals
definitions [9]. The rules are in “If Then" format and formally
the If side is called the conditions and then side is called the
conclusion. The computer is able to execute the rules and
compute a control signal depending on the measured inputs
error (e) and change in error (de). In a rule based controller the
control strategy is stored in a more or less natural language. A
rule base controller is easy to understand and easy to maintain
for a non- specialist end user and an equivalent controller
could be implemented using conventional techniques [14].

Inference engine Fig: Centre of Gravity

IJSER
Inference engine is defined as the Software code which Bisector of area (BOA)

processes the rules, cases, objects or other type of

The bisector of area (BOA) defuzzification method calculates
knowledge and expertise based on the facts of a given
the abscissa of the vertical line that divides the area of the
situation. When there is a problem to be solved that involves
resulting membership function into two equal areas. For
logic rather than fencing skills, we take a series of inference
discrete sets, is the abscissa x j that minimizes.
steps that may include deduction, association, recognition,
and decision making. An inference engine is an information
processing system (such as a computer program) that
systematically employs inference steps similar to that of a
human brain.
Here i max is the index of the largest abscissa x i max. BOA is a
Defuzzification computationally complex method.

The reverse of Fuzzification is called Defuzzification. The use Mean of maximum (MOM)
of Fuzzy Logic Controller (FLC) produces required output
In this method the crisp value is to choose the point with the
in a linguistic variable (fuzzy number). According to
highest membership. There may be several points in the
real world requirements, the linguistic variables have to
overall implied fuzzy set which have maximum membership
be transformed to crisp output. There are many
value. Therefore it is a common practice to calculate the mean
defuzzification methods but the most common methods are as
value of these points. This method is called mean of maximum
follows [11]:
(MOM) and the crisp value is calculated as follows:
 Center of gravity (COG)
 Bisector of area (BOA)
 Mean of maximum (MOM) Here I is the (crisp) set of indices i where µ c(x I)
reaches its maximum µmax cardinality (the number of
Center of gravity (COG)
members), and | I | is its Implementation of an FLC requires
For discrete sets COG is called center of gravity for the choice of four key factors:
singletons (COGS) where the crisp control value is the

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 11, November-2014 1049
ISSN 2229-5518
 Number of fuzzy sets that constitute 1. Centre of gravity (COG)
linguistic variables. 2. Centre of gravity method for singletons (COGS)
 Mapping of the measurements onto the 3. Bisector of area (BOA)
support sets. 4. Mean of Maxima (MOM)
 Control protocol that determines the 5. Left most maximum (LM), and right most maximum
controller behavior. (RM)
 Shape of membership functions.
Sugeno Fuzzy Inference
Mamdani fuzzy inference
Mamdani style is not computationally efficient as it find the
The most common method is used currently is fuzzy inference centroid of two dimensional shapes by integration of carrying
system. In 1975, Professor Ebrahim Mamdani of London function. Michio Sugeno proposed a new method to use single
University introduced first time fuzzy systems to control a spike, a singleton, as a membership function inputs. Its mean
steam engine and boiler combination. He applied a set of fuzzy set is at unity point at one particular point on the
fuzzy rules experienced human operators. The mamdani universe of discourse and zero at remaining area. This system
system usually done in four steps. [21] is almost same of Mamdani method but with the exception of
consequent change and instead of fuzzy set it use a
 Fuzzification of the inputs mathematical function as input variable. [21]
 Rule evaluation.
 Aggregation of the rules. Rules

IJSER
 Defuzzification.
The general rules for Dc motor speed control is that if motor
Fuzzification converts input data to degree of membership speed is less than desired speed then speed up the motor and
functions. In this process data is matched with condition of if motor speed is more than reference speed then slows it
rules and determined how well data is matched with rule at speed. There are nine possible conditions which motor can be
particular instance. Thus a degree of membership function is seen and nine possible regions are selected from which 25
developed. possible rules in fuzzy controller are written. In the process of
producing necessary output voltage with Fuzzy Logic
Controller the speed error should be minimized. The bigger
speed error causes the bigger controller input. In addition
changing of the error plays an important role to define
controller input.

Then in Rule-base block rules are written according to system

requirement. Fuzzy controller work on both MIMO and SISO.
In case of Dc motor there are two input variables Error and
Change in error are selected. This system is limited to single
loop control. Usually rules are in if, and, then form. In
inference engine aggregation is done in which degree of
fulfillment is calculated of the condition specifies by a rule. In
activation min of two aggregated value is selected and only
thickened part of singleton are activated. Its multiplication
result in slighter smooth control. Then all activated
conclusions are accumulated using max operation.

Defuzzification block converted resulting fuzzy set into a

number that is sent to the system and this number is actually
the control signal. There are seven defuzzification methods. Sugeno Fuzzy Inference

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 11, November-2014 1050
ISSN 2229-5518
In region one according to Fig. error is in positive reign while
error direction is going to negative reign, so output of
controller must be positive. In this case speed still not touch to
the reference point so increase in speed is required which is
directly proportional to the voltage of controller so increase in
voltage help the motor to got reference speed. In third reign, if
error value is negative large and change of error value is
negative large than output will be negative large. This
condition is corresponding to the reign 3 interval but this
result gives us crisp value for fuzzy control this crisp value
should be converted into linguistic form.

The surface in the next Fig. shows that motor touch the
reference speed in smooth way. According to this design
motor can reach to the maximum speed of 300 rad/sec.


output is NL
IJSER
If error is NL and change in error is NL then

If error is NL and change in error is NZ then

output is NL
If error is NL and change in error is ZZ then
Fig. Surface View

Fig. shows the simulink model of DC motor speed control

through fuzzy controller.

In fuzzy controller error and change in error is measure by

output is NZ
following formulas:
The output of controller plotted against the rules describes. Error (ek) = Wref – wm
Rules behavior can be checked by changing of error or change Change in error (ce) = e(k) – e(k-1)
in error point. From the surface an idea can be built that in a Back emf provides the error while change in error measured
certain case what will be the output of controller. that error goes in negative direction or positive direction. In
simulink model reference speed is selected 150 rad/sec.

Fig. Simulated Outputs (Sugeno Fuzzy Inference) Fig. shows the output of DC motor
IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 11, November-2014 1051
ISSN 2229-5518
In Fig. desired speed is 150 rad/sec. Motor achieve this speed Comparison between fuzzy PID and conventional PID
in less than 0.2 second and torque load is applied at 0.5 second controller:
but it did not have any effect on the speed and it remain  Self-tuned tuning PID controller is less compared to
constant. This is the advantage of fuzzy controller that in pole conventional PID controller.
placement and PID controller speed reduced for minor time  The three parameters "K P ", "K I ", "K D " of conventional
and then goes to reference point but in this case speed not PID control need to be constantly adjusted online in
reduced. order to achieve better control performance.
 Fuzzy self-tuning PID parameters controller can
Design of Fuzzy Rules automatically adjust PID parameters in accordance
Rule bases for tuning K P with the speed error and the rate of speed error-
change, so it has better self-adaptive capacity fuzzy
PID parameter controller has smaller overshoot and
less rising and settling time than conventional PID
controller and has better dynamic response
properties and steady-state properties.
 Steady state error in case of self tuned fuzzy PID is
less compared to conventional PID controller.

IJSER
Rule bases for tuning K I

Different values of K p ,K i & K d for fuzzy logic controller:

Sl. No. Kd Kp Ki

1 0.75 15 1

2 3.5 5 20

3 0.02 -3 15

4 10 2.5 0.8

Rule bases for tuning K D 5 55 4 -10.25

11. RESPONSE OF FUZZY LOGIC

CONTROLLER

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 11, November-2014 1052
ISSN 2229-5518
12. SIMULATION RESULTS

IJSER

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 11, November-2014 1053
ISSN 2229-5518
13. DISCUSSION AND CONCLUSION OF THE [7] Weiming Tang, Guanrong Chen and Rongde Lu, “A Modified
RESULT Fuzzy PI Controller for a Flexible-joint Robot Arm with
Uncertainties”, Fuzzy Set and System, 118 (2001) 109-119.
In this project we have studied about different method for [8] PavolFedor, Daniela Perduková, “A Simple Fuzzy Controller
speed control of DC motor. The steady state operation and its Structure,”ActaElectrotechnica ET Informatica No. 4, Vol. 5, 2005
[9]Maher M.F. AlgreerandYhyaR.M.Kuraz, “Design Fuzzy Self
various torque-speeds, torque-current characteristics of DC
Tuning of PID Controller for Chopper-Fed DC Motor
motor are studied. We have also studied basic definition and
drive.”Kuraz
terminology of fuzzy logic and fuzzy set. This project [10] BomedieneAlloua ,AbdellahLaoufBrahimGasbaoui and
introduces a design method of two inputs and three outputs ABdessalamAbderrahamani, “Neuro-Fuzzy DC Motor speed
self-tuning fuzzy PID controller and make use of MATLAB Control Using Particle Swarm Optimization,” Leonaro Electronic
fuzzy toolbox to design fuzzy controller. fuzzy controller Journal of Practices and Technologies ISSN,1583-1078.
adjusted the proportional, integral and derivate (K P , K I , K D ) [11]ManafeddinNamazov and OnurBasturk, “DC motor position
control using fuzzy proportional-derivative controllers with
gains of the PID controller according to speed error and
different defuzzification methods,” Turkish Journal of Fuzzy
change in speed error .From the simulation results it is Systems (eISSN: 1309–1190), Vol.1, No.1, pp. 36-54, 2010.
concluded that ,compared with the conventional PID [12] Wang Xiao-kan, Sun Zhong-liang, Wanglei, Feng Dong-qing,
controller, self-tuning PID controller has a better performance "Design and Research Based on Fuzzy PID-Parameters Self-
in both transient and steady state response. The FLC has better Tuning Controller with MATLAB," Advanced Computer Theory
dynamic response curve, shorter response time, small and Engineering, International Conference on, pp. 996-999, 2008
International Conference on Advanced Computer Theory and
overshoot, small steady state error (SSE), high steady precision
Engineering, 2008.
compared to the conventional PID controller.
[13] GaddamMallesham and AkulaRajani ,”AUTOMATIC

IJSER
TUNING OF PID CONTROLLER USING FUZZY LOGIC .” 8th
14. FUTURE SCOPE International Conference on DEVELOPMENT AND
APPLICATION SYSTEMS Suceava, Romania, May 25 – 27, 2006.
MATLAB simulation for speed control of DC motor has been [14]NurAzliza Ali, “Fuzzy logic controller for controlling DC
done which can be implemented in hardware to observe motor speedusing MATLAB applications”.
actual feasibility of the approach applied in this thesis. This [15]M.Chow and A. Menozzi ,”on the comparison of emerging
technique can be extended to other types of motors. The and conventional techniques for DC motor control” proc.IECON
parameters of PID controller can also be tuned by using ,PP.1008-1013,1992.
[16]A.course in fuzzy systems and control by Li-Xin Wang,
genetic algorithm (GA).
Prentice-Hall international, Inc.
[17]Ogata, K., Modern Control Engineering. Englewood Cliffs, NJ:
REFERENCES Prentice Hall, 2001.
[18]MATLAB and SIMULINK Version 2010a, the MathsworksInc,
[1]B.J. Chalmers, “Influence of saturation in brushless permanent
USA.
magnet drives.” IEE proc. B, Electr.PowerAppl, vol.139, no.1,
[19]P.S Bhimbhra,electrical machinery, New Delhi, Khanna
1992.
Publishers.
[2]C.T. Johnson and R.D. Lorenz, “Experimental identification of
[20] www.wikipedia.org
friction and its compensation in precise, position controlled
Books
mechanism.” IEEE Trans. Ind ,Applicat, vol.28, no.6, 1992.
[3]J. Zhang, N. Wang and S. Wang, “A developed method of
Gopal, M., Control Systems-Principles and Design; 3rd Edition, New
tuning PID controllers with fuzzy rules for integrating process,”
Delhi: McGraw-Hill, 2008.
Proceedings of the American Control Conference,Boston, 2004,
Ajit K. Mandal, Introduction to Control Engineering-Modeling,
pp.1109-1114.
Analysis and Design; 1st Edition, New Delhi, New Age
[4] K.H. Ang, G. Chong and Y. Li, “PID control system analysis,
International, 2008.
design and technology,” IEEE transaction on Control System
Franklin, G.F., J.D. Powel and A. Emami-Naeini; Feedback Control
Technology, Vol.13, No.4, 2005, pp. 559-576
of Dynamical Systems; 5th Edition, Upper Saddle River, NJ: Pearson
[5] H.X.Li and S.K.Tso, "Quantitative design and analysis of Fuzzy
Education 2005.
Proportional-Integral-Derivative Control- a Step towards Auto
Dorf, R.C. and R.H. Brishop; Modern Control System, 10th Edition,
tuning", International journal of system science, Vol.31, No.5,
Upper Saddle River, NJ: Pearson Education 2004.
2000, pp.545-553.
Nise, N.S.; Control System Engineering; 4th Edition Danvers, MA:
[6] Thana Pattaradej, Guanrong Chen and PitikhateSooraksa,
John Wiley, 2003.
"Design and Implementation of Fuzzy PID Control of a bicycle
Kuo, B.C. and F. Golnaraghi: Automatic Control System; 8th Edition
robot", Integrated computer-aided engineering, Vol.9, No.4, 2002.
Danvers, MA: John Wiley, 2003.

IJSER © 2014
http://www.ijser.org