AHMAD ALHASSAN et al: CLOSED-LOOP SCHEMES FOR POSITION AND SWAY CONTROL OF A GANTRY …

Closed-loop Schemes for Position and Sway Control of a Gantry Crane System

Ahmad Alhassan, Kumeresan A. Danapalasingam*, Muhammad Shehu, Auwalu M. Abdullahi, Auwal Shehu
Department of Control and Mechatronics Engineering
Faculty of Electrical Engineering, Universiti Teknologi Malaysia
Johor, Malaysia
*Corresponding Author / E-mail: [email protected], TEL: +607-5557128

Abstract—This paper presents the investigation into the performance of Lyapunov pole placement (LPP), linear quadratic
regulator (LQR) and proportional-integral-derivative (PID) control schemes for payload sway control and trolley position tracking
of a gantry crane system. A 2D gantry crane system is considered. The nonlinear model of the system is derived using the
Lagrangian energy equation and then linearized using Taylor’s series expansion. To investigate the performances of the designed
controllers, a unit step input as a reference perturbation is applied to the controllers. MATLAB simulation results of the responses
are analysed in time domain. The response time specifications of the trolley position, level of payload sway reduction, and
robustness to parameter variation and uncertainties are used to assess the performances of the controllers.

Keywords-closed-loop; gantry crane; Lagrange; linearization; LQR; Lyapunov; PID, simulation, Taylor series
the system and it is insensitive to modelling errors.
I. INTRODUCTION However, it leads to dissipation of energy or even burn out
Cranes are the most widely used tools to transport of the system (chattering) [13]. More so, Fuzzy logic
various types of goods efficiently and reliably from one controller was also developed to stabilized the responses of
point to another [1]. Gantry crane, tower crane and boom the GCS in [14],[15]. Intelligent control offers ease of
crane are the three major types of cranes used today [2]. execution and efficient control due to its ability to treat
Due to its cost effectiveness and ease of operation, gantry inaccurate model. However, significant parameter variation
crane system (GCS) is the most preferred crane system in affects its performance [13]. In addition, input shaping (IS)
the industries, shipping yards, mining sites, power plants, technique was proposed for vibration control of flexible
warehouses etc [3][4]. manipulator in [16]. As an open loop control, IS it’s simple
However, GCS are prone to vibration and deflection of to design and cost effective as it does not require feedback
the payload during operation and or in the presence of control or additional sensors. It only requires estimated
external disturbance (obstacle). This lead to inaccurate natural frequency and damping ratio of the system [17].
positioning of the load, delay in task completion and even a Conversely, small disturbance or variation in the system
damage to the system or the operating environment [2],[5]. parameters significantly affects its performance. Linear
Interestingly, to improve the system throughput, guarantee quadratic regulator (LQR) was also proposed in [18] for
safety of the environment and minimize maintenance cost balancing and control of an inverted pendulum. LQR
due to system failure, many researchers have engaged in modern control uses the ideas of weighting matrices to
developing the mathematical model of the system for achieved optimal conditions for the states and the control
precise dynamic analysis and effective control [1]. input. As a closed loop optimal controller, LQR is effective
The dynamic behaviour of the non linear GCS using and robust to uncertainties [19][20].
varying system parameters; trolley mass, payload mass and In this paper, an investigation into the performance of
cable length was presented in [1],[3][6]. It was observed Lyapunov pole placement (LPP), linear quadratic regulator
that, payload oscillation and trolley displacement are highly (LQR) and proportional-integral-derivative (PID) control
dependent on those parameters. In order to improve the schemes for payload sway control and trolley position
performance of the GCS, many control strategies were tracking of a gantry crane system is presented. A 2D gantry
presented. A simple state feedback controller (SFB) using crane system is considered. The non linear model of the
Ackerman’s formula was presented in [8]. The main issue system is derived using the Lagrangian energy equation and
with SFB is that the states of the system must be measurable then linearized using Taylor’s series expansion. To
and the gains depend on the accuracy of the model. Optimal investigate the performances of the designed controllers, a
PID controller was proposed using weight summation unit step input as a reference perturbation is applied to the
approach in [7]. Due to its simplicity of design and controllers. MATLAB simulation results of the responses
implementation, PID is applicable to many industrial are analysed in time domain. The response time
applications. Adaptive controller was also designed using specifications of the trolley position, level of payload sway
corrective control parameter in the presence of uncertainties reduction and robustness to parameter variation are used to
in [8][9]. Sliding mode control (SMC) was also developed assess the performances of the controllers. The robustness of
by assuming constant cable length in [10],[11]. This scheme the controllers is assessed by changing the payload mass,
was also improved by incorporating real time analysis using cable length and a sine wave input disturbance.
variable cable length in [12]. SMC retains the stability of

DOI 10.5013/IJSSST.a.17.32.28 28.1 ISSN: 1473-804x online, 1473-8031 print

 AHMAD ALHASSAN et al: CLOSED-LOOP SCHEMES FOR POSITION AND SWAY CONTROL OF A GANTRY …

The paper is organised as follows. In section II,

description of a gantry crane system is discussed. Section III
presents the mathematical modelling of the system. In
section IV, the derived nonlinear model is linearized using
Taylor’s series expansion and presented in state space form.
Section V describes the proposed control schemes. The
implementation and discussion of the proposed controllers is
discussed in section VI. Conclusion is finally presented in
section VII.
II. DESCRIPTION OF A GANTRY CRANE SYSTEM
The gantry crane system considered in this work is
shown in Fig. 1. The trolley slides along the horizontal jib.
The jib is supported by a pair of legs. A payload is
suspended via a suspension cable to a trolley. The schematic Fig. 1. Gantry crane system (GCS)
diagram of GCS is shown in Fig. 2. The applied force, F,
causes the trolley of mass, m1 to move a distance, x . This
motion leads to a deflection angle,  of the payload. The
trolley carries the load of mass, m 2 via am of hoisting cable
of length, l. The values of the system parameters used for
this study are tabulated in Table I[4]
III. MODELLLING OF NONLINEAR GANTRY CRANE
This section presents the nonlinear mathematical
modeling of a nonlinear gantry crane using the Lagrangian
energy equations.
To simplify the model derivation, the following
assumptions were adopted: (i) the force, F is considered as Fig. 2. Schematic diagram of GCS
the input to the system (ii) the mass of the hoisting cable is
TABLE I. DESCRIPTION OF PARAMETERS
neglected (iii) external disturbances are neglected (iv) the
effect of friction on the trolley is also neglected. The Parameters Value (unit)
Lagrange’s equations are given as [4]
Trolley mass (m1) 2.0 Kg

  Payload mass (m2) 0.5 Kg

d  L    L  Q i , i  1, 2,  , n (1)
dt   q  qi Cable length (l) 0.5 m
 i 
Acceleration due to gravity (g) 9.81 m/s2
LT P (2)
v 22  x 2  l 2 2  2 xl
 cos (5)
where P and T are respectively the total potential and kinetic
energy, n is total number of independent generalized
coordinate and Qi is non conservative generalized forces. 1
T2  m 2 ( x 2  l 2 2  2 xl
 cos ) (6)
The kinetic energy of the trolley given as: 2

1 (3) Therefore, (2) becomes

T1  m1 x 2
2
1 1
The kinetic energy of the payload is given as
L
2

m 2 x 2  l 2 2  2 xl 
 cos  m1 x 2  m 2 glcos
2
(7)

1 (4)
Therefore, the non-linear model of gantry crane system can
T2  m 2 v 22 be summarized as:
2

The velocity analysis of fig. 2 using cosine rule gives x  m 2 lcos  m 2 l 2 sin  F
( m1  m 2 )  (8)

m 2 l 2  m 2 lxcos
   m 2 glsin  0 (9)

DOI 10.5013/IJSSST.a.17.32.28 28.2 ISSN: 1473-804x online, 1473-8031 print

 AHMAD ALHASSAN et al: CLOSED-LOOP SCHEMES FOR POSITION AND SWAY CONTROL OF A GANTRY …

IV. LINEARIZATION OF THE NON LINEAR MODEL 0 1 0 0  0 

This section discusses the linearization process of the  m1g   1 
derived non linear model. Assuming a very small deflection 0 0
m2
0
, m 
A  0 
1 b   0 
angle of the payload (  <<1o). The nonlinear terms; sine
2
 0 0
 
and cosine can be linearized about an operating point using   g(m1  m2 )   1 
Tailor’s series expansion 0 0 0
 lm 2 
 lm2 
( x  x0 ) 2
f  x   f  x0    x  x0  f '
 x0   f ''
 x0    (10) C  1 0 1 0 (18)
2!
( x  x0 ) n
 f n
 x0  A pre-requisite to designing a control strategy is to
n! investigate the system's stability. In this case, the system has
Thus, for a small deflection angle of the payload, two poles located at the origin and two conjugate imaginary
poles (0, 0, +5j, -5j). For this reason, the response of GCS is
undesirable and behaves as an undamped oscillator. The
  0;  2  0; f  sin    ; f  cos   1 (11) system controllability was investigated by determining the
rank of the controllability matrix Gc as
Hence, the derived equation (8) and (9) of the non-linear
model can be approximately linearized as:  1 981
0 2
0
400
1 981 
x  m 2 l  F
( m1  m 2 )  (12)  0 0 

GC  b Ab A2b A3b   2
0
 400
981

m 2 l   m 2 lx  m 2 gl  0
2 1 0
(13)  40 
 981 
1 0 0 
Also, the desired responses; payload oscillation and the  40 
trolley displacement, can be re-arranged as:
G C  96 . 25  0 (19)
m g 1
x 2 
 F (14)
m1 m1 Thus, the controllability matrix is nonsingular; full rank and
hence, the system is controllable.
 ( m1  m 2 ) g 1
   F (15) A. Lyapunov pole placement scheme
m1l m1l
In this section, pole placement control based on
Lyapunov approach is presented. By assuming a measurable
V. CONTROL DESIGN state vector, x  [ x x   ] , a control law u   kx can be
implemented to the system. The Lyapunov function is given
In this section, three control strategies were designed as
based on Lyapunov pole placement, LQR and PID schemes.
The crane system is presented in the general form for state-
vector equation as follows [21]: AT  TF  bk (20)

x  Ax  bu (16)  1  [ k
k  kT k2 k3 k4 ] (21)
1

y  Cx  Du (17) The poles of the closed loop system can be selected along
the negative s- plane arbitrarily. In this case, fourth order
where A represents the system state matrix, b is the output conjugate poles are chosen as  1.5  0.5 j and  1  j . The
matrix, C represents the output matrix and y is the system matrix F of (20) can be formed from the assigned Eigen
output. By using the linear model of equation (14) and (15), values in a block diagonal form as
the state-vector equation matrices A, b and C can be
presented as:
1.5 0.5 0 0
0.5 1.5 0 0 
F  
 0 0 1 1  , k  [1 0 1 0] (22)
 
 0 0 1 1

By substituting for A,b, k , and F in (20) yields

DOI 10.5013/IJSSST.a.17.32.28 28.3 ISSN: 1473-804x online, 1473-8031 print

 AHMAD ALHASSAN et al: CLOSED-LOOP SCHEMES FOR POSITION AND SWAY CONTROL OF A GANTRY …

 14 73 168 144 1 0 0 0 0 0 0 0 0 0 0 0  0

 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0  0
   
 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0  0
   
 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0  0
 0 0 0 0 14 73 168 144 2.453 0 0 0 0 0 0 0  0.5
   
 0 0 0 0 1 0 0 0 0 2.453 0 0 0 0 0 0  0
 0 0 0 0 0 1 0 0 0 0 2.453 0 0 0 0 0  0
   
 0 0 0 0 0 0 1 0 0 0 0 2.453 0 0 0 0  0
 0 T  
0 0 0 0 0 0 0 14 73 0 144 1 0 0 0 0
   
 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0  0
   
 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0  0
 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1  0
   
 0 0 0 0 0 0 0 0  24.53 0 0 0 14 0 168 144    1
 0 0 0 0 0 0 0 0 0  24.53 0 0 1 0 0 0  0
   
 0 0 0 0 0 0 0 0 0 0  24.53 0 0 1 0 0  0
 0 0 0 0 0 0 0 0 0 0 0  24.53 0 0 1 0  0
  

(23)

By solving for T in (23) and substituting the value in (21) On the other hand, selecting elements of Q to be
yields the controller gains as relatively small compared to the elements of R will result in
costly control action and the system state x will not return or
k  [0.51 1.12 14.28  4.44] (24) converge fast to the origin.
The control law U=-Kx that minimizes the performance
index function J is called Kalman’s gain. For a LTI system
The general block diagram of a state feedback control is with cost function J, the optimal regulator is always a linear
shown in Fig. 3. To obtain zero steady state error, a gain is control law. For the closed-loop system, the system takes
added to the reference signal for the position tracking as the following form [22]
1
k in    0.51 (25)
 K    R 1   B   P 
T
(27)
C ( A  bk )  1 b

x ( t )  ( A  BK ) x ( t ) (28)
B. Linear Quadratic Regulator (LQR)
LQR control is a common approach employed in the 
control of hub angle and position of crane system [20]. The J   ( x T Qx  (  Kx ) T R (  Kx ) dt (29)
structure of LQR is given in Fig. 3. The design of LQR 0
control requires a linear state-vector model. Hence,
linearized state space model of equation (18) was utilized. The matrix P is obtained by solving the algebraic Riccati
The method involves obtaining a control law U=-Kx that equation given as
derives the system state to the origin (i.e. to zero) at the
same time minimizing the performance index function, J
 A   P    P  A   Q    P  B  R   B   P   0
1
with minimal control effort given in [23] as
T T
(30)

J   ( x ( t ) T Qx ( t )  U ( t ) T RU ( t )) dt (26) The closed-loop form of equation (28) is always stable if
0 matrix P is positive definite. The state penalty matrix Q and
the control effort penalty matrix R were selected as
where Q is a symmetrical positive semi-definite matrix
called state penalty matrix and R is the positive definite  0.75 0 0 0
symmetrical matrix known as control action penalty. For  0 1 0 0 
single input system, R reduces to a single number. Thus, J Q , R  1/ 40 (31)
represents weighted energy cost of the state and control.  0 0 1 0
To design LQR, the penalty weighting matrices Q and R  
are selected such that; If the elements of Q are relatively  0 0 0 1
large compared to that of R, then heavy penalty is applied
Using MATLAB command; lqr(A,b,Q,R), the control
to the deviations of the state x from the origin in comparison
gain was calculated as
to the deviations of the control action from zero.

DOI 10.5013/IJSSST.a.17.32.28 28.4 ISSN: 1473-804x online, 1473-8031 print

 AHMAD ALHASSAN et al: CLOSED-LOOP SCHEMES FOR POSITION AND SWAY CONTROL OF A GANTRY …

k  [0.51 1.12 14.28  4.44] , k i  0.22 (32) VI. IMPLEMENTATION AND RESULT
In this section, implementation and discussion of the
results is presented. To study the dynamics of the proposed
C. Proportional-Integral-Derivative control (PID) controllers, a unit step input is applied to the system. This is
In this section, PID control is presented. The general sufficient to make the GCS moves and then stop at the
representation of PID is given as desired position based on the given parameters (exact) of
Table I. LQR and LPP were implemented based on the
obtained control gains whereas the PID was tuned by trial

d
uc (t )  K p e(t )  Ki e(t )dt  K d e(t ) (33) and error. For a better performance of the PID, double PID is
dt utilized. One PID for position tracking and the other for
sway control as shown in Fig. 5. The performance of the
where u(t) is the control signal, Ki, Kp and Kd are LPP, LQR and PID control for the position tracking and
respectively the integral, proportional and derivative gains. payload sway control are respectively shown in Fig. 6-7.
e(t) is the undesired error calculated by taking the difference
Time response specifications and level of sway reduction
between the actual signal and the output response. PID can
be tuned manually or automatically to meet the desired were used to assess the control performances. For the sway
response based on the three available features. The control, mean absolute error (MAE) of the payload sway is
proportional gain is responsible for the system response. utilized. Small MAE means less sway and hence, the better
However, faster response leads to steady state error. The the performance of the controller. It can be observed that
integral gain takes care of the steady error. The derivative LPP gives a better sway reduction as compares to LQR and
feature reduces overshoot. Thus, if those features are not PID whereas PID provide better position tracking as
tuned properly, it may affect the closed loop stability of the compared to LPP and LQR in terms of settling time (ST) and
system. The block diagram of a PID control scheme is rise time (RT) and steady state (SS) error.
shown in Fig. 4. The signal is applied to the gantry crane.
The resulting responses will be feedback for comparison
with the reference input. The controller is tuned to make this
error zero.

LPP/LQR controller

Crane
Ki + ‐ ‐ System
Reference Output
‐ ‐
Signal response

Fig. 5. A simulink block for double PID controller

K 1.4
LPP
LQR
1.2 PID
Reference
Fig. 3. Block diagram of LQR/LPP control
1
Trolley position (m)

Kp .e(t) 0.8

t + Output 0.6
Reference + e(t) + u(t) Crane

Response
Ki e(t)
System
- 0 + 0.4

e(t) 0.2
Kd
t
PID
0
0 1 2 3 4 5 6 7 8 9 10
Time (s)
Fig. 4. Block diagram of PID controller Fig. 6. Trolley position the exact parameters.

DOI 10.5013/IJSSST.a.17.32.28 28.5 ISSN: 1473-804x online, 1473-8031 print

 AHMAD ALHASSAN et al: CLOSED-LOOP SCHEMES FOR POSITION AND SWAY CONTROL OF A GANTRY …

To investigate the robustness of the designed controllers, 0.06

the payload mass was increased to 1.0kg. Later, the cable LPP
length was also changed to 1.0m. The results were obtained 0.04 LQR
based on those values separately as shown in Fig 8-11. This PID

acts as a parameter variation procedure. It can also be 0.02

noticed that the variation of the parameters affects the

Payload oscillation (rad)

0
performance of LLP significantly as compared to PID and
LQR. To further test the external disturbance rejection -0.02
capability of the controllers, a continuous sine wave
disturbance (SWD) of 0.5 peak to peak magnitudes is -0.04

introduced to the system. This makes the LPP controller

unstable as shown in Fig. 12-13. However, the performances -0.06

of PID demonstrated that the controller effectively rejects -0.08

external disturbance and variation of the system parameters
while LQR was affected slightly. Table II summarized the -0.1

position tracking and sway control performance of the

designed controllers. -0.12
0 1 2 3 4 5 6 7 8 9 10
Time (s)
0.06 Fig. 9. Payload sway for 1.0kg trolley mass.
LPP
0.04 LQR 1.4
PID PP
LQR
0.02
1.2 PID
Reference
Payload oscillation (rad)

0
1

-0.02
Payload position (m)

0.8
-0.04

-0.06 0.6

-0.08
0.4

-0.1
0.2

-0.12
0 1 2 3 4 5 6 7 8 9 10
Time (s) 0
0 1 2 3 4 5 6 7 8 9 10

Fig. 7. Payload sway for the exact parameters. Time (s)

Fig.10. Trolley position for 1.0m cable length.

1.5
LPP 0.08
LQR LPP
PID 0.06 LQR
Reference PID
0.04
Payload oscillation (rad)

1 0.02
Trolley position (m)

-0.02

-0.04
0.5
-0.06

-0.08

-0.1

0 -0.12
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10

Time (s) Time (s)

Fig. 11. Payload sway for 1.0m cable length.

Fig. 8. Trolley position for 1.0kg trolley mass.

DOI 10.5013/IJSSST.a.17.32.28 28.6 ISSN: 1473-804x online, 1473-8031 print

AHMAD ALHASSAN et al: CLOSED-LOOP SCHEMES FOR POSITION AND SWAY CONTROL OF A GANTRY …

VII. CONCLUSION
2 LPP
LQR In conclusion, this paper investigates the performance of
PID
1.8
Reference LLP, LQR and PID controllers for trolley position tracking
1.6 and payload sway suppression. The effectiveness of the
1.4
designed controllers have been assessed in terms of trolley
position tracking, level of payload sway reduction, and
Trolley position (m)

1.2
robustness to parameter variation and uncertainties. Without
1 an external disturbance, acceptable performances have been
achieved with all the controllers. A comparative analysis of
0.8
the results has shown that double PID and LQR provides
0.6 precise position tracking with fast response whereas LLP
0.4
gives better sway reduction. However, best disturbance
rejection was achieved using double PID compared to both
0.2
LQR and LLP.
0
0 5 10 15 20 25 30
Time (s) ACKNOWLEDGMENT
Fig. 12. Trolley position for a sine wave disturbance.
This work was supported by Universiti Teknologi
Malaysia (UTM), the Fundamental Research Grant Scheme
LPP
(R.J130000.7823.4F273) from the Ministry of Higher
0.1 LQR Education Malaysia and the eScienceFund
PID
Reference
(R.J130000.7923.4S100) from the Ministry of Science,
Technology and Innovation Malaysia.
0.05
Payload oscillation (rad)

REFERENCES
0 [1] N. Đ. Zrni, V. M. Ga, and S. M. Bo, “Dynamic responses of a gantry
crane system due to a moving body considered as moving oscillator,”
Arch. Civ. Mech. Eng., pp. 2–9, 2014.
[2] A. Masoud, “Dynamics and Control of Cranes : A Review,” Vib.
-0.05
Control, vol. 9, pp. 863–908, 2003.
[3] V. S. Renuka and A. T. Mathew, “Precise Modelling of a Gantry
Crane System Including Friction , 3D Angular Swing and Hoisting
-0.1 Cable Flexibility,” Int. J. Theor. Appl. Res. Mech. Eng., vol. 2, pp.
119–125, 2013.
0 5 10 15 20 25 30 [4] H. Izzuan, Z. Mohamed, J. J. Jamian, A. Faiz, and Z. Abidin,
Time (s) “Dynamic Behaviour of a Nonlinear Gantry Crane System,” Procedia
Fig. 13. Payload sway for a sine wave disturbance. Technol., vol. 11, no. Iceei, pp. 419–425, 2013.
[5] J. Yoon, S. Nation, W. Singhose, and J. E. Vaughan, “Control of
TABLE II. LEVEL OF SWAY AND RESPONSE SPECIFICATIONS Crane Payloads That Bounce During Hoisting,” IEEE Trans. Control
Syst. Technol., vol. 22, no. 3, pp. 1233–1238, 2014.
Controller RT (s) ST (s) SS error MAE [6] A. B. Alhassan, B. B. Muhammad, K. A. Danapalasingam, and Y.
Sam, “Optimal Analysis and Control of 2D Nonlinear Gantry Crane
Exact 3.5 7.0 - 0.0100 System,” IEEE Int. Conf. Smart Sensors Appl., pp. 30–35, 2015.
1.0kg 4.0 12 - 0.0079 [7] H. I. Jaafar and M. F. Sulaima, “Optimal PID Controller Parameters
LLP for Nonlinear Gantry Crane System via MOPSO Technique,” IEEE
1.0m 3.5 9 - 0.0109 Conf. Sustain. Util. Dev. Eng. Technol., pp. 86–91, 2013.
[8] C. S. Teo, K. K. Tan, S. Y. Lim, S. Huang, and E. B. Tay, “Dynamic
SWD 3.0 6.5 0.58 0.0411 modeling and adaptive control of a H-type gantry stage,”
Mechatronics, vol. 17, pp. 361–367, 2007.
Exact 3.0 7.0 - 0.0131 [9] N. Sun, Y. Fang, and H. Chen, “Adaptive antiswing control for cranes
1.0kg 3.2 7.0 - 0.0135 in the presence of rail length constraints and uncertainties,” 2015.
LQR [10] G. Bartolini, A. Pisano, and E. Usai, “Second-order sliding-mode
1.0m 3.0 7.0 - 0.0131 control of container cranes,” Automatica, vol. 38, no. 10, pp. 1783–
1790, 2002.
SWD 2.5 4.0 - 0.017 [11] Q. H. Ngo and K. Hong, “Sliding-Mode Antisway Control of an
Offshore Container Crane,” vol. 17, no. 2, pp. 201–209, 2012.
Exact 2.5 4.0 - 0.0139 [12] L. A. Tuan, S. Moon, W. G. Lee, and S. Lee, “Adaptive sliding mode
control of overhead cranes with varying cable length †,” J. Mech. Sci.
1.0kg 2.5 4.0 - 0.0139
PID Technol., vol. 27, no. 3, pp. 885–893, 2013.
1.0m 2.6 4.0 - 0.0113 [13] C. Tai and K. Andrew, “Review of Control and Sensor System of
Flexible Manipulator,” J. Intell. Robot Syst., pp. 187–213, 2014.
SWD 2.5 4.0 - 0.0138 [14] J. Jalani, “Robust Fuzzy Logic Controller for an Intelligent Gantry
Crane System,” First Int. Conf. Ind. Inf. Syst. ICIIS, Sri Lanka, pp.
497–502, August 2006.

DOI 10.5013/IJSSST.a.17.32.28 28.7 ISSN: 1473-804x online, 1473-8031 print

 AHMAD ALHASSAN et al: CLOSED-LOOP SCHEMES FOR POSITION AND SWAY CONTROL OF A GANTRY …

[15] P. Hyla, “The Crane Control Systems : A Survey,” IEEE, pp. 505– and R. E. Samin, “Feedback Control Schemes for Gantry Crane
509, 2012. System incorporating Payload,” IEEE Symp. Ind. Electron. Appl.
[16] Z. Mohamed, A. K. Chee, A. W. I. M. Hashim, M. O. Tokhi, S. H. M. (ISIEA2011), Langkawi, Malaysia, pp. 370–375, 2011.
Amin, and R. Mamat, “Techniques for vibration control of a flexible [20] K. Ogata, “Modern Control Engineering.” 5th ed., New York, USA:
robot manipulator,” Robotica, vol. 24, no. 4, pp. 499–511, 2006. Prentice Hall, pp. 793–816, 2010.
[17] M. Maged and G. Shehata, “Anti-sway control of a tower crane using [21] C.-T. Chen, “Linear system theory and design.” 3rd ed., New York,
inverse dynamics,” IEEE, 2014. USA: Oxford Univesity Press, pp. 255–257, 1999.
[18] E. Vinodh Kumar and J. Jerome, “Robust LQR controller design for [22] C.-C. Huang, “Solving Algebraic Riccati Equation for Singular
stabilizing and trajectory tracking of inverted pendulum,” Procedia System Based nn Matrix Sign Function,” Int. J. Innov. Comput. Inf.
Eng., vol. 64, pp. 169–178, 2013. Control, vol. 9, no. 7, pp. 2771–2788, 2013.
[19] M. A. Zawawi, W. M. S. W. Zamani, M. A. Ahmad, M. S. Saealal,

DOI 10.5013/IJSSST.a.17.32.28 28.8 ISSN: 1473-804x online, 1473-8031 print