Professional Documents
Culture Documents
1.Introduction
Currently,elasticjointmanipulatorshavereceivedagreat
deal of attention due to their light weight, high
manoeuvrability, flexibility, high power efficiency and
large number of applications. However, controlling such
systems still faces numerous challenges that need to be
addressed. The control issue of the flexible joint is to
designthecontrollersothatthelinkoftherobotcantrack
aprescribedtrajectorypreciselywithminimumvibration
to the link. In order to achieve these objectives, various
methods using different techniques have been proposed.
Yim [1], Oh and Lee [2] proposed an adaptive output
feedbackcontrollerbasedonabacksteppingdesign.This
technique is proposed in order to deal with parametric
uncertainty in a flexible joint. The relevant work also
being done by Ghorbel et al. [3]. Lin and Yuan [4] and
Spong et al. [5] introduced a nonlinear control approach
using the feedback linearization technique and the
integralmanifoldtechniquerespectively.Arobustcontrol
design was reported by Tomei [6] by using simple PD
1 Mohd Ashraf Ahmad, Mohd Zaidi Mohd Tumari and Ahmad Nor Kasruddin Nasir:
Composite Fuzzy Logic Control Approach to a Flexible Joint Manipulator
www.intechopen.com
ARTICLE
www.intechopen.com
Int J Adv Robotic Sy, 2013, Vol. 10, 58:2013
control and Yeon and Park [7] by applying robust H
control. An openloop optimal control approach for
generating the optimal trajectory of the elastic robotic
armsinpointtopointmotionisproposedbyKorayemet
al. [8]. This method is based on Pontryagins minimum
principle which considers the mobility of the elastic
manipulator.
Inrecentyears,toolsofcomputationalintelligencesuch
as artificial neural networks and fuzzy logic controllers
have been credited in various applications as powerful
tools capable of providing robust controllers for
mathematicallyilldefinedsystems.Thishasledtorecent
advancesintheareaofintelligentcontrol[9,10].Various
neural network models have been applied in the control
of flexible joint manipulators which have led to
satisfactory performances [11]. H. Chaoui et al. [12] used
a sliding mode control approach that learns the systems
dynamics through a feedforward neural network. A
timedelay neurofuzzy network was suggested in [13],
where a linear observer was used to estimate the joint
velocitysignalsandeliminatedtheneedtomeasurethem
explicitly. Subudhi et al. [14] presented a hybrid
architecturecomposedofaneuralnetworktocontrolthe
slow dynamic subsystem and an H to control the fast
subsystem. A feedback linearization technique using a
TakagiSugeno neurofuzzy engine was adopted in [15].
LinandChen[16]proposeacombinedrigidmodelbased
the computed torque and fuzzy control of flexiblejoint
manipulators.
Figure1.Rotaryflexiblejointmanipulator
Figure2.Descriptionoftheflexiblejointmanipulatorsystem
3.TheModellingoftheFlexibleJointManipulator
Thissectionprovidesabriefdescriptiononthemodelling
oftheFJMsystemasabasisofasimulationenvironment
for the development and assessment of the proposed
control techniques. The EulerLagrange formulation is
considered in characterizing the dynamic behaviour of
thesystem.
,
x Ax Bu
y Cx
= +
=
(1)
where the vector
T
x u o u o ( =
2
2
0 0 1 0
0 0 0 1
0 0
0 0
0 0 , 1 0 0 0 .
stiff m g t m g eq m
eq eq m
stiff eq arm m g t m g eq m
eq arm eq m
m g t g m g t g
eq m eq m
K K K K B R
A
J J R
K (J J ) K K K B R
J J J R
K K K K
B C
J R J R
(
(
(
(
+
(
=
(
(
(
+ +
(
(
(
( ( = =
(
(2)
In equation (1), the input u is the input voltage of the
servomotor
m
V which determines the FJM base
movement.Inthisstudy,thevaluesoftheparametersare
definedinTable1.
TipangleerrorE={NMNSZEPSPM},
TipangleerrorrateV={NMNSZEPSPM},
VoltageU={NMNSZEPSPM},
where,Rn,n=1,2,Nmax,isthenthfuzzyrule,Ei,Vjand
Ukfori,j,k=1,2,,5,aretheprimaryfuzzysets.
Figure3.PDtypefuzzylogiccontrolstructure
No. Rules
1. If( e isNM)and( e isZE)then(uisPM)
2. If( e isNS)and( e isZE)then(uisPS)
3. If( e isNS)and( e isPS)then(uisZE)
4. If( e isZE)and( e isNM)then(uisPM)
5. If( e isZE)and( e isNS)then(uisPS)
6. If( e isZE)and( e isZE)then(uisZE)
7. If( e isZE)and( e isPS)then(uisNS)
8. If( e isZE)and( e isPM)then(uisNM)
9. If( e isPS)and( e isNS)then(uisZE)
10. If( e isPS)and( e isZE)then(uisNS)
11. If( e isPM)and( e isZE)then(uisNM)
Table2.LinguisticRulesofthePDtypeFuzzyLogicController
4.2PDtypefuzzylogicwithanoncollocatedfuzzylogic
controller(PDFLCFLC)
AcombinationofPDtypefuzzylogicandanoncollocated
fuzzylogiccontrolschemeforthecontrolofthetipangular
position and the vibration suppression of the system
respectively is presentedin thissection. The use ofa non
collocated control system, where the deflection angle of the
flexible joint manipulator is controlled by measuring its
angle,canbeappliedtoimprovetheoverallperformanceas
more reliable output measurement is obtained. The control
structurecomprisestwofeedbackloops:(1)Thehubangleas
aninputtothePDtypeFLCfortipangularpositioncontrol.
(2) The deflection angle as an input to a separate non
collocatedcontrollawforvibrationcontrol.Thesetwoloops
are then summed together to give a torque input to the
system. A block diagram of the composite fuzzy logic
controlschemeisshowninFigure4.
Fortipangularpositioncontrol,thePDtypeFLCstrategy
developedintheprevioussectionisadoptedwhereasfor
the vibration control loop, the deflection angle feedback
through a noncollocated fuzzy logic control scheme is
utilized. In designing the noncollocated fuzzy logic
control,basictriangularandtrapezoidalformsarechosen
fortheinputandoutput membershipfunctions.Figure5
shows the membership functions of the fuzzy logic
controllerforvibrationcontrol.ItconsistsofNegativeBig
(NB), Negative Small (NS), Zero (Z), Positive Small (PS)
and Positive Big (PB), as shown in the diagram. The
universesofdiscoursesofthedeflectionangle,deflection
angle rate and input voltage are from 1.2 to 1.2 rad,
0.015to0.015rad/sand447to447Vrespectively.
m m g t g
m
V K K
R
(3)
Figure4.PDtypeFLCwithanoncollocatedfuzzylogiccontrol
structure
4 Int J Adv Robotic Sy, 2013, Vol. 10, 58:2013 www.intechopen.com
Figure5.Membershipfunctionsoftheinputandoutputsignals
Deflectionangle
rate
Deflectionangle
PB PS Z NS NB
PB PB PB PB NB NB
PS PB PS PS NS NB
Z PB PS Z NS NB
NS PB PS NS NS NB
NB PB PB NB NB NB
Table3.Linguisticrulesfornoncollocatedfuzzylogiccontrol
Figure6.Rulesgenerationbasedonthemotioncondition
Meanwhile,ifthejointrotatesinaclockwisedirection,as
shown in Figure 6(b), and the link deflects in an anti
clockwise direction, the torque should be imposed in an
anticlockwise direction to suppress the deflection
motion.Inthecasewherethereisnodeflection,notorque
shouldbeapplied.Furthermore,theproposedfuzzylogic
control adopts the wellknown Mamdani minmax
inferenceandcentreofarea(COA)methods.
4.3PDtypefuzzylogicwithanoncollocatedPIDController
(PDFLCPID)
AcombinationofaPDtypeFLCandnoncollocatedPID
control scheme for the control of the rigid body motion
and vibration suppression of the system is presented in
this section. The control structure of the PDFLCPID is
not very different from the PDFLCFLC and a block
diagramofthecontrolschemeisshowninFigure7.
Overshoot10%
Settlingtime0.75sec
Risetime0.55sec
Figure7.PDtypeFLCwithanoncollocatedPIDcontrol
structure
4.4PDtypefuzzylogicwithinputshaping(PDFLCIS)
A control structure for the control of the rigid body
motionanddeflectionanglereductionoftheflexiblejoint
manipulator based on a PDtype FLC and inputshaping
scheme is proposed in this section. The positive input
shapers are proposed and designed based on the
propertiesofthesystem.Inthisstudy,theinputshaping
control scheme is developed using a ZeroVibration
-400 -300 -200 -100 0 100 200 300 400
0
0.5
1
Input voltage (V)
NB NS Z PS PB
D
e
g
r
e
e
o
f
m
e
m
b
e
r
s
h
i
p
-0.015 -0.01 -0.005 0 0.005 0.01 0.015
0
0.5
1
Deflection angle rate (rad/s)
D
e
g
r
e
e
o
f
m
e
m
b
e
r
s
h
i
p
NB NS Z PS PB
1 0.8 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 -1
0
0.5
1
Deflection angle (rad)
NB NS Z PS PB
D
e
g
r
e
e
o
f
m
e
m
b
e
r
s
h
i
p
5 Mohd Ashraf Ahmad, Mohd Zaidi Mohd Tumari and Ahmad Nor Kasruddin Nasir:
Composite Fuzzy Logic Control Approach to a Flexible Joint Manipulator
www.intechopen.com
DerivativeDerivative (ZVDD) inputshaping technique
[21]. Previous experimental studies with a flexible
manipulator have shown that significant vibration
reduction and robustness is achieved using a ZVDD
technique[24].AblockdiagramofthePDtypeFLCwith
aninputshapingcontroltechniqueisshowninFigure8.
Theinputshapingmethodinvolvesconvolvingadesired
command with a sequence of impulses known as input
shapers. The design objectives are to determine the
amplitudeandtimelocationoftheimpulsesbasedonthe
natural frequencies and damping ratios of the system.
Thepositiveinputshapershavebeenusedinmostinput
shaping schemes. The requirement of positive amplitude
for the impulses seeks to avoid the problem of large
amplitudeimpulses.Inthiscase,eachindividualimpulse
must be less than one in order to satisfy the unity
magnitude constraint. In addition, the robustness of the
inputshaper to errors in the natural frequencies of the
systemcanbeincreasedbysolvingthederivativesofthe
system vibration equation. This yields a positive ZVDD
shaperwithaparameteras:
t1=0,t2=
d
,t3=
2
d
,t4=
3
d
1
2 3
1
1 3 3
A
H H H
,
2
2 3
3
1 3 3
H
A
H H H
2
3
2 3
3
1 3 3
H
A
H H H
,
3
4
2 3
1 3 3
H
A
H H H
(4)
where:
2
1
H e
,
n
and represent the natural frequency and damping
ratiorespectively.Fortheimpulses,tjandAjarethetime
locationandamplitudeofimpulsejrespectively.
Figure8.PDtypeFLCwithaninputshapingcontrolstructure
5.ImplementationandResults
In this section, the proposed control schemes are
implemented and tested within the simulation
environment of the flexible joint manipulator and the
corresponding results are presented. The manipulator is
required to follow a trajectory of 50. System responses
namelythetipangularpositionandthedeflectionangle
areobserved.Toinvestigatethevibrationofthesystemin
the frequency domain, the power spectral density (PSD)
of the deflection angle response is obtained. The
performances of the control schemes are assessed in
terms of vibration suppression, trajectory tracking and
time response specifications. Finally, a comparative
assessment of the performance of the control schemes is
presentedanddiscussed.
(a)Tipangularposition
(b)Deflectionangle
(c)Powerspectraldensity
Figure9.ResponseoftheflexiblejointmanipulatorusingPD
FLC
2
1
n d
0 0.5 1.0 1.5 2.0 2.5 3.0
0
10
20
30
40
50
60
Time (s)
T
i
p
a
n
g
u
l
a
r
p
o
s
i
t
i
o
n
(
d
e
g
)
0 0.5 1.0 1.5 2.0 2.5 3.0
-20
-15
-10
-5
0
5
10
15
20
Time (s)
D
e
f
l
e
c
t
i
o
n
a
n
g
l
e
(
d
e
g
)
0 2 4 6 8 10 12 14 16 18
-100
-80
-60
-40
-20
0
20
40
Frequency (Hz)
P
o
w
e
r
S
p
e
c
t
r
a
l
D
e
n
s
i
t
y
(
d
B
)
8.83 Hz
2.94 Hz
14.52 Hz
6 Int J Adv Robotic Sy, 2013, Vol. 10, 58:2013 www.intechopen.com
Figure 9 shows the responses of the flexible joint
manipulator to the reference input trajectory using the
PDFLCinatimedomainandafrequencydomain(PSD).
These results were considered as the system response
under rigid body motion control and will be used to
evaluate the performance of the noncollocated fuzzy
logic control and inputshaping scheme. The steadystate
tip angular trajectory of 50 for the flexible joint
manipulator was achieved within the rise and settling
times and an overshoot of 0.222 s, 0.565 s and 1.78 %
respectively. It is noted that the manipulator reaches the
required position within 1 s, with little overshoot.
However,anoticeableamountofvibrationoccursduring
the movement of the manipulator. It is noted from the
deflectionangleresponsethatthevibrationofthesystem
settles within 3 s with a maximum residual of 15.
Moreover, and from the PSD of the deflection angle
response,thevibrationsattheflexiblejointaredominated
by the first three vibration modes, which are obtained as
2.94Hz,8.83Hzand14.52Hzwithamagnitudeof32.38
dB,12.08dBand26.52dBrespectively.
(a)Tipangularposition
(b)Deflectionangle
(c)Powerspectraldensity
Figure10.Responseoftheflexiblejointmanipulatorusing
compositePDFLCFLC,PDFLCPIDandPDFLCIS
0 0.5 1.0 1.5 2.0 2.5 3.0
0
10
20
30
40
50
60
Time (s)
T
i
p
a
n
g
u
l
a
r
p
o
s
i
t
i
o
n
(
d
e
g
)
PD-FLC-FLC
PD-FLC-PID
PD-FLC-IS
0 0.5 1.0 1.5 2.0 2.5 3.0
-15
-10
-5
0
5
10
Time (s)
D
e
f
l
e
c
t
i
o
n
a
n
g
l
e
(
d
e
g
)
PD-FLC-FLC
PD-FLC-PID
PD-FLC-IS
0 2 4 6 8 10 12 14 16 18
-100
-80
-60
-40
-20
0
20
40
Frequency (Hz)
P
o
w
e
r
S
p
e
c
t
r
a
l
D
e
n
s
i
t
y
(
d
B
)
PD-FLC-FLC
PD-FLC-PID
PD-FLC-IS
2.94 Hz
8.83 Hz
14.52 Hz
7 Mohd Ashraf Ahmad, Mohd Zaidi Mohd Tumari and Ahmad Nor Kasruddin Nasir:
Composite Fuzzy Logic Control Approach to a Flexible Joint Manipulator
www.intechopen.com
Table 4 and Figure 11 summarize the magnitude of the
vibration and the level of the vibration reduction of
deflection, respectively, for the proposed composite control
schemes.Itisseenthatahighperformanceinthereduction
of the vibration of the system is achieved using both non
collocatedfuzzylogicandPIDcontrolascomparedtofeed
forward control based on an inputshaping technique. This
isobserved and comparedtothe PDFLC for the first three
modes of vibration. In addition, as demonstrated in the tip
angulartrajectoryresponse,aslightlyfasterresponsewitha
higher overshoot is obtained using both noncollocated
fuzzy logic and PID control scheme as compared to the
inputshaping technique. Comparisons of the specifications
of the tip angular trajectory responses are summarized in
Table5.Inaddition,andasdemonstratedinthetipangular
trajectoryresponsewithPDFLCFLCandPDFLCPID,the
minimumphasebehaviourofthemanipulatorisunaffected.
Intermsofthedesigncomplexity,theimplementationofthe
inputshaper technique is much easier as compared to both
noncollocatedcontrolsasalargeamountofdesigneffortis
required in order to determine the controller parameters,
such as the range of the membership function in the fuzzy
logic design and kp, ki and kd gains in the PID design.
However, the inputshaper strategy is very sensitive to the
system parameter variation and could not compensate for
anyeffectofdisturbanceascomparedtothenoncollocated
control.
Figure11.Levelofthevibrationreductionofthedeflectionangle
6.Conclusions
The development of composite fuzzy logic schemes for
the trajectory tracking and vibration suppression of a
flexiblejointmanipulatorhasbeenpresented.Thecontrol
schemes have been developed based on a PDtype FLC
with noncollocated fuzzy logic control, a PDtype FLC
withnoncollocatedPIDcontrolandaPDtypeFLCwith
an inputshaping scheme. The proposed composite
control schemes have been implemented and tested on a
flexible joint manipulator. The performances of the
control schemes have been evaluated in terms of input
tracking capability and vibration suppression for the
resonance modes of the manipulator. An acceptable
performance in input tracking and vibration control has
been achieved with the proposed control strategies. A
comparative assessment of the control schemes has
shown that the PDFLC with noncollocated strategies
performs better than the PDFLC with a feedforward
technique in respect of the deflection angle reduction of
the elastic joint. Moreover, in terms of the speed of
responses, PDFLC with noncollocated PID and FLC
results in a faster settling time response with high
overshoot as compared to PDFLC with a feedforward
scheme. The work thus developed and reported in this
paper forms the basis of the design and development of
composite control schemes for the input tracking and
vibration suppression of multidegree of freedom robotic
armswithflexiblelinksandjoints.
7.Acknowledgments
This work was supported by the Faculty of Electrical &
Electronics Engineering, University of Malaysia Pahang,
and especially the Control & Instrumentation (COINS)
ResearchGroup.
8.References
[1] Yim, W. (2001) Adaptive Control of a Flexible Joint
Manipulator. Proc. 2001 IEEE, International Robotics
&Automation,Seoul,Korea,pp.34413446.
[2] Oh,J.H.andLee,J.S.(1997)ControlofFlexibleJoint
Robot System by Backstepping Design Approach.
Proc.ofIEEEInternationalConferenceonRobotics&
Automation,pp.34353440.
[3] Ghorbel, F., Hung, J.Y. and Spong, M.W. (1989)
Adaptive Control of Flexible Joint Manipulators.
ControlSystemsMagazine.9:913.
[4] Lin,L.C.andYuan,K.(2007)ControlofFlexibleJoint
Robots via External LinearizationApproach. Journal
ofRoboticSystems.1(1):122.
[5] Spong, M. W., Khorasani, K. and Kokotovic, P. V.
(1987) An Integral Manifold Approach to the
Feedback Control of Flexible Joint Robots. IEEE
JournalofRoboticsandAutomation.3(4):291300.
[6] Tomei, P. (1991) A Simple PD Controller for Robots
with Elastic Joints. IEEE Trans. on Automatic
Control.36(10):12081213.
0 5 10 15 20 25 30 35
1
2
3
Level of vibration reduction (dB)
M
o
d
e
o
f
f
r
e
q
u
e
n
c
i
e
s
PD-FLC-FLC
PD-FLC-PID
PD-FLC-IS
8 Int J Adv Robotic Sy, 2013, Vol. 10, 58:2013 www.intechopen.com
[7] Yeon, J. S. and Park, J. H. (2008) Practical Robust
Control for Flexible Joint Robot Manipulators. Proc.
2008 IEEE International Conference on Robotic and
Automation,Pasadena,CA,USA,pp.33773382.
[8] Korayem, M.H., Nohooji, H.R. and Nikoobin, A.
(2011) Path planning of mobile elastic robotic arms
byindirectapproachofoptimalcontrol.International
JournalofAdvanceRoboticSystems.8(1):1020.
[9] Ahmad, M.A., Raja Ismail, R.M.T., Ramli, M.S.,
Zawawi, M.A., Hambali, N. and Abd Ghani, N.M.
(2009) Vibration Control of Flexible Joint
Manipulator using Input Shaping with PDtype
Fuzzy Logic Control. Proceedings of IEEE
International Symposium on Industrial Electronics,
Seoul,pp.11841189.
[10] Ahmad,M.A.,Nasir,A.N.K.,Hambali,N.andIshak,
H. (2008) Vibration and Input Tracking Control of
Flexible Manipulator using Hybrid Fuzzy Logic
Controller. Proceedings of the IEEE International
Conference of Mechatronics and Automation,
Kagawa,pp.593598.
[11] Chaoui, H., Gueaieb, W., Yagoub, M. and Sicard, P.
(2006) Hybrid neural fuzzy sliding mode control of
flexiblejoint manipulators with unknown dynamics.
Proceedings of the 32nd Annual Conference of the
IEEE Industrial Electronics Society (IECON2006),
pp.40824087.
[12] Chaoui, H., Sicard, P. and Lakhsasi, A. (2004)
Reference model supervisory loop for neural
network based adaptive control of a flexible joint
with hard nonlinearities. IEEE Canadian Conference
on Electrical and Computer Engineering. 4: 2029
2034.
[13] Hui, D., Fuchun, S. and Zengqi, S. (2002) Observer
based adaptive controller design of flexible
manipulators using timedelay neurofuzzy
networks. J. Intell. Robot. Syst.: Theory and
Applications.34(4):453466.
[14] Subudhi, B. and Morris, A. (2006) Singular
perturbation based neuroh innity control scheme
for a manipulator with exible links and joints.
Robotica.24(2):151161.
[15] Park, C.W. (2004) Robust stable fuzzy control via
fuzzy modeling and feedback linearization with its
applications to controlling uncertain singlelink
flexible joint manipulators. J. Intell. Robot. Syst.:
TheoryandApplications.39(2):131147.
[16] Lin, L. C. and Chen, C. C. (1995) Rigid modelbased
fuzzy control of flexiblejoint manipulators. J. of
IntelligentandRoboticSystems.13:107126.
[17] Tokhi, M.O. and Azad, A.K.M. (1996) Control of
flexible manipulator systems. Proceedings IMechEI:
Journal of Systems and Control Engineering. 210:
113130.
[18] Ahmad, M.A., Suid, M.H., Ramli, M.S., Zawawi,
M.A., Ismail, R.M.T.R. (2010) PD Fuzzy Logic with
noncollocated PID approach for vibrationcontrol of
flexible joint manipulator. Proceedings of 6th
International Colloquium on Signal Processing and
ItsApplications,Mallaca,Malaysia,pp.15.
[19] Ahmad, M.A. (2008) Vibration and Input Tracking
Control of Flexible Manipulator Using LQR with
NonCollocated PID Controller. Proceedings of the
2
nd
UKSIM European Symposium on Computer
Modeling and Simulation, Liverpool, United
Kingdom,pp.4045.
[20] Shchuka,A.andGoldenberg,A.A.(1989)Tipcontrol
of a singlelink flexible arm using feedforward
technique.MechanicalMachineTheory.24:439455.
[21] Mohamed,Z.andAhmad,M.A.(2008)HybridInput
ShapingandFeedbackControlSchemesofaFlexible
Robot Manipulator. Proceedings of the 17
th
World
Congress The International Federation of Automatic
Control,Seoul,Korea,pp.1171411719.
[22] Rattan, K.S., Feliu, V. and Brown, H.B. (1990) Tip
position control of flexible arms. Proceedings of the
IEEEConferenceonDecisionandControl,Honolulu,
pp.18031808.
[23] Quanser Student Handout, Rotary Flexible Joint
Module.http://www.quanser.com.
[24]Mohamed,Z.andTokhi,M.O.(2002)Vibrationcontrol
of a singlelink flexible manipulator using command
shaping techniques. Proceedings IMechEI: Journal of
SystemsandControlEngineering.216:191210.
9 Mohd Ashraf Ahmad, Mohd Zaidi Mohd Tumari and Ahmad Nor Kasruddin Nasir:
Composite Fuzzy Logic Control Approach to a Flexible Joint Manipulator
www.intechopen.com