2020 IEEE 4th International Conference on Intelligent Energy and Power Systems (IEPS)

Optimal Placement of Actuator for Vibration

Suppression Based on Intelligent PID Controller

Muhamad Sukri Hadi Nadzirah Mohd Mokhtar Intan Zaurah Mat Darus
Faculty of Mechanical Engineering Faculty of Civil Engineering School of Mechanical Engineering
Universiti Teknologi MARA Technology Universiti Teknologi Malaysia
Selangor, Malaysia Universiti Malaysia Pahang Johor, Malaysia
[email protected] Pahang, Malaysia [email protected]
[email protected]

Abstract — Many attempts have been proposed by the Nevertheless, this strategy is usually not applicable.
previous researchers due to reduce the undesired vibration by Passive control strategy has also been approached by applying
considering several control strategies. The simplest way in passive materials like vibration damper and dynamic observer.
reducing the vibration is to build more rigid structure so that However, this method is only applicable for high frequencies
less vibration will be produced. Nevertheless, this strategy is range, but not working well for low frequency range.
usually not applicable since the structures are need high power
Moreover, in meeting the demand of engineering applications,
consumption and limitation in operation speed. Furthermore, it
becomes a growing trend among the industries to use light many industries are putting efforts to keep the structural
weight of mechanical structure known as flexible plate. weight as low as possible, thus making the passive solutions
However, the critical problem faced by the industries is nonviable [8].
vibration on the structure that can lead to structural damage. In fact, an active vibration control (AVC) method is more
Hence, this research presents the optimal placement of actuator
efficient, reliable and flexible in controlling the unwanted
and sensor on the experimental rig for vibration cancelation of
the flexible plate structure based on intelligent PID controller. vibration of flexible structure. The potential of active vibration
The PID controller tuned by artificial bee colony (ABC) has been received remarkable attention from the researchers
algorithm was used to control the undesired vibration on the due to many applications that demand for effective vibration
structure. The robustness of developed controller was validated suppression especially in the aerospace structure, flexible
by varying the position of actuator on the experimental rig. It robotic arms and micro mechanical systems. AVC is a method
was indicated that point A2 leads to the good attenuation level to reduce the amplitude of the unwanted vibration by
by achieving highest attenuation value at the first mode of introducing the secondary sources of vibration to the
vibration with 28.83 dB which is equivalent to 21.62 % dynamical system [9].
attenuation, after the introduction of vibration control.
The key to achieve high attenuation level at the first mode
Keywords — flexible plate, active vibration control, artificial of vibration is by determining the exact position of actuators
bee colony, PID controller, placement actuator on the flexible plate structure. Therefore, this study presents
the effect of location of an actuator on the horizontal flexible
I. INTRODUCTION plate structure using active vibration control technique by
In recent times, the characteristics offered by the flexible employing artificial bee colony algorithm (ABC) for
plate structure such as lightweight, faster response, less power parameter estimation of PID controller.
consumption and less bulky design have received significant
considerations by the industries to apply its advantages into II. PID CONTROLLER TUNED BY ABC ALGORITHM
their engineering applications [1-4]. Nevertheless, the ABC algorithm is a swarm based on metaheuristic
vibration of the flexible plate structure is a critical problem algorithm that was introduced by Karaboga in 2005 for
faced by the industries, particularly in the airport baggage optimizing numerical problems and inspired by intelligent
transport conveyor, micro hand surgery system and foraging behavior of honey bees. In this research, ABC
semiconductor manufacturing industry which have a light algorithm was used to find the optimum value of PID
weight characteristic and relatively low damping for the parameters due to achieve the best attenuation level in
fundamental and initial models [5]. suppressing the unwanted vibration of the horizontal flexible
This drawback is often a limiting factor in the structure plate system. Details about artificial bee colony algorithm has
performance that can lead to the instability, fatigue and been described in this publication [5]. The objective function
structural damages [6]. Moreover, the frequency associated of the optimization used to tune the PID controller is
with this structure is commonly low which makes the nodes formulated based on the mean square error (MSE) of
vibration control become an important issue for the light horizontal flexible plate system. The PID-ABC controller is
flexible structures [7]. Many attempts have been proposed by implemented in the closed-loop controller using
the researchers in order to reduce the vibration by considering MATLAB/Simulink environment software. Fig. 1 shows the
several control strategies. The simplest strategy in reducing schematic diagram of PID-ABC controller.
the vibration is to build more rigid structure so that less The procedure of ABC algorithm in tuning the PID
vibration will be produced. parameters can be summarized as follows:

978-0-7381-0568-0/20/$31.00 ©2020 IEEE 237

Authorized licensed use limited to: UNIVERSITI TEKNOLOGI MARA. Downloaded on November 14,2023 at 04:54:08 UTC from IEEE Xplore. Restrictions apply.
 2020 IEEE 4th International Conference on Intelligent Energy and Power Systems (IEPS)

a) Selection of initial set of food source positions algorithm was tested on the experimental rig and the
performances of the controllers was evaluated. A sinusoidal
The algorithm is initialized by generating the random signal at the frequency of 27 Hz and amplitude of 1V was
number of bees at certain number of iterations. The initial employed as the disturbance signal to excite the experimental
value of iteration is started at small value due to achieve the rig while conducting the experiment. The controller was
faster optimization. Every PID parameters were assigned to utilised in a real-time computer control system using
the three food sources in the search space. Each food sources MATLAB/Simulink with sampling rate of 0.003 s. Details of
in the search space will be represented the value of KP, KI, the experimental setup can be found in [4]. Table I illustrates
and KD. PID controller parameters tuned by ABC algorithm which will
be validated experimentally. On the other hand, Fig. 2 shows
b) Evaluation for objective function of every bees the schematic diagram of the controller implemented in the
Every bee in the search space were assigned with the experiment. Next, Fig. 3 presents the block diagram of the
objective function which evaluated by MSE. controller implemented by MATLAB/Simulink. Fig. 4
describes the experimental setup used in this study.
c) Updating the best food sources of the bees
TABLE I. PID PARAMETER OF THE CONTROLLER
Employed bees are started to search the food sources
within the neighborhood food source and determine their Controller KP KI KD
nectar amount produced. The onlooker’s bee is selected the
PID-ABC 8.8632 0.2857 0.0046
best food sources which exploited by the employed bees and
the amount of nectar is determined. Then, the scouts are
explored a new food source for employed bees to be explored.
The best food source is memorized by the bees. The new best
food source is used to determine the optimal value of KP, KI,
and KD.
d) Termination strategy
The new set of PID parameters were submitted to the
control algorithm to determine the new objective function
value. The procedures (b) and (c) were repeated until the
algorithm meet the optimal value of KP, KI, and KD. Fig. 2. The schematic diagram of the controller used in the experiment.

Fig. 3. Block diagram of PID based controller using MATLAB/Simulink.

Fig. 1. The schematic diagram of PID controller tuned by ABC algorithm.

III. EXPERIMENTAL SETUP

As a validation of the developed simulated controller in
[5], the optimal value of PID parameters tuned by ABC Fig. 4. The Experimental setup used in the experiment.

238
Authorized licensed use limited to: UNIVERSITI TEKNOLOGI MARA. Downloaded on November 14,2023 at 04:54:08 UTC from IEEE Xplore. Restrictions apply.
 2020 IEEE 4th International Conference on Intelligent Energy and Power Systems (IEPS)

IV. PLACEMENT OF ACTUATOR As mentioned by Tavakolpour (2010), the flexible plate

This section demonstrates the effect of PID based nodal line is located at the middle of experimental rig, as
controller tuned by ABC algorithm in term of controller represented by the red line in Fig. 4. Hence, the location of the
performance by varying the position of actuator in the actuator needs to avert this line so that the actuator’s
experimental rig. Fig. 5 describes the location of sensors and performance can be preserved. The selection of actuator
actuators used during the experiment. This location was location at the high amplitude of frequency response will
considered based on the guidelines established in [6] and [7]. result in high vibration attenuation for the first mode of
The horizontal flexible plate experimental rig was divided into vibration [6]. Based on the impact test that has been
20 sections at x-axis and 20 sections at y-axis. Besides, the conducted, the high amplitude of frequency response was
primary disturbance is located at ∆x = 0.12m, ∆y = 0.12m, located at the first mode of vibration, which was at the middle
of experimental rig [14]. However, this location was averted
detection point is located at ∆x = 0.18m, ∆y = 0.18m and to sustain the actuator’s performance. Thus, several points
observation point is located at ∆x = 0.30m, ∆y = 0.30m. On the nearby the high amplitude of frequency response location
other hand, the location of actuator is varied into four points, were selected to investigate an ideal actuator position,
where point A1 is located at ∆x = 0.24m, ∆y = 0.36m, point A2 whereby the produced force will lead to maximum attenuation
is located at ∆x = 0.36m, ∆y = 0.36m, point A3 is located at of the vibration.
∆x = 0.24m, ∆y = 0.24m and point A4 is located at
V. RESULT AND DISCUSSION
∆x = 0.36m, ∆y = 0.24m.
The performance of the actuator at point A1, A2, A3 and
A4 were assessed in this section. Fig. 6 shows the time
response for horizontal flexible plate system at point A1, A2,
A3 and A4, respectively.

Fig. 5. Non-collocated actuator-sensor control technique. (a)

Based on previous research, collocated actuator-sensor

would give a good attenuation of vibration for flexible
structure system. However, due to the mechanical and
environmental constraints in the industries, it is impractical to
install the sensor and actuator at the same position [6]. In the
real-world applications related to this research such as surgical
hand immobilizer and airport transportation conveyor, it is
impossible to employ the collocation strategies of sensor and
actuator pairs for the active vibration control system. The
upper space of the application should be avoided from any
sensor and actuator in order to lend a hand on the surgical hand
immobilizer and transport the baggage on the conveyor
system. So, the sensor and actuator should be located at the (b)
lower space of application for this implementation purpose.
Thus, non-collocated actuator-sensor control technique was
chosen in this research in order to suppress the unwanted
vibration of the horizontal flexible plate experimentally [10].
Other than that, the location of actuator was selected based
on the corresponding first mode shape function of horizontal
flexible plate, which is the dominant mode in the system [11].
The main rule in placing the actuator is the actuator must be
located far enough from the point where the mode shape
function is changing sign or at the plate nodal lines. This is
crucial as to prevent the force produced by the actuator
experiencing the deterioration which is indirectly lead to
uncontrollable condition [12, 13].
(c)

239
Authorized licensed use limited to: UNIVERSITI TEKNOLOGI MARA. Downloaded on November 14,2023 at 04:54:08 UTC from IEEE Xplore. Restrictions apply.
 2020 IEEE 4th International Conference on Intelligent Energy and Power Systems (IEPS)

(d)
(d)
Fig. 7. Experimental frequency response for different actuator positions; (a)
Fig. 6. Experimental time response for different actuator positions; (a) Point A1; (b) Point A2; (c) Point A3; (d) Point A4
Point A1; (b) Point A2; (c) Point A3; (d) Point A4
It can be seen that the placement of actuator at point A2
Figs. 6 (a)-(d) present the time response of the controller
leads to the good attenuation level at the first mode of
at different actuator locations on the experimental rig. It could
vibration, while point A3 gives the lowest attenuation level in
be observed that different actuator location will result in a
this investigation. Other than that, point A2 has shown
different attenuation at the first mode of vibration. As can be
superiority by achieving the highest attenuation level at the
seen in the figure, point A2 has successfully reduced the
first mode of vibration with 23.83 dB. The attenuation value
vibration signal in time response to the lowest value as
has been reduced from 110.2 dB to 86.37 dB, which is
compared with the other points. Meanwhile for point A1, A3
equivalent to 21.62 % attenuation, after the introduction of
and A4, the time response produced at these locations are quite
vibration control. On the other hand, point A3 has shown
similar where the acceleration signal achieved at each points
inferiority by achieving the lowest attenuation level at the first
are only separated with small differences in values. Next,
mode of vibration with only 12.90 dB. The attenuation value
Figs. 7 (a)-(d) describe the frequency response of the
has been reduced from 117.1 dB to 104.2 dB, which is
attenuation level achieved at each point.2
equivalent to 11.02 % attenuation, after the introduction of
vibration control. The details of the controller performance
achieved at each location are as described in Table II.
Here, it is acknowledged that point A2 is the best position
to install an actuator to achieve a good vibration suppression
of the horizontal flexible plate. This investigation has led to
the significant effect of actuator location in vibration
suppression. A proper selection of actuator position will result
in a good vibration attenuation value, while inappropriate
actuator location will result in a bad vibration suppression of
the horizontal flexible plate system. As stated by Mohd Yatim
(a)
(2016), improper selection of the actuator location may reduce
the system damping performance and make it unstable, which
is indirectly leading to an uncontrollable system [11].

TABLE II. THE CONTROLLER PERFORMANCE AT DIFFFERENT

POSITION OF ACTUATOR

Percentage of
Position of Attenuation Level
Reduction
Actuator
1st mode 1st mode

Point A1 15.70 dB 13.13 %

Point A2 23.83 dB 21.62 %

Point A3 12.90 dB 11.02 %

(b)
Point A4 18.10 dB 15.12 %

VI. CONCLUSION
This paper was presented the effect of location of an
actuator on the horizontal flexible plate structure using active
vibration control technique by employing artificial bee colony
algorithm (ABC) for parameter estimation of PID controller.
The effectiveness of the developed controller in suppressing
the undesired vibration on the structure were assessed and
analysed in this study. The analysis was conducted by varying
the location of actuator on the experimental rig to obtain the
(c) best actuator placement for high vibration suppression

240
Authorized licensed use limited to: UNIVERSITI TEKNOLOGI MARA. Downloaded on November 14,2023 at 04:54:08 UTC from IEEE Xplore. Restrictions apply.
 2020 IEEE 4th International Conference on Intelligent Energy and Power Systems (IEPS)

purpose. Several points were chosen on the experimental rig Climbing and Walking Robots and the Support Technologies for
in placing the magnetic shaker actuator based on non- Mobile Machines, Shah Alam, Malaysia, August 2019.
collocated strategies of sensor and actuator pairs. This is [4] M. S. Hadi and I. Z. M. Darus, “Modelling and control of horizontal
flexible plate using particle swarm optimization,” International Journal
because it is impractical to employ with the collocated of Engineering & Technology, vol. 7, pp. 13-19, 2018.
strategies of sensor and actuator in the industrial applications. [5] M. S. Hadi and I. Z. M. Darus, “Modelling of horizontal flexible plate
Other than that, the number of rules have been followed to structure using artificial bee colony algorithm,” International Journal
of Engineering & Technology, vol. 7, pp. 415-419, 2018.
obtain the right location of actuator, such as the actuator
[6] A. R. Tavakolpour, “Mechatronic design of an intelligent active
location must be placed far away from the plate nodal line, vibration control for flexible structures ,” Ph. D. Thesis Department of
location must be averted from the mode of vibration of the Applied Mechanics, Faculty of Mechanical Engineering, Universiti
system, actuator placement is at high amplitude of frequency Teknologi Malaysia, 2010.
response in order to achieve maximum vibration reduction. [7] M. S. Saad, “Evolutionary optimization and real-time self-tuning active
Based on the conducted experimental study, it shows that vibration control of a flexible beam system,” Ph. D. Thesis Department
point A2 is the best location to place the magnetic shaker of Applied Mechanics, Faculty of Mechanical Engineering, Universiti
Teknologi Malaysia, 2014.
actuator. Here, it is revealed that proper selection of actuator
location is crucial to obtain high attenuation vibration, because [8] M. Verma, V. Lafarga and C. Collette, “Perfect collocation using self-
sensing electromagnetic actuator: Application to vibration control of
improper selection of actuator location may affect the flexible structures,” Sensors and Actuators A: Physical, vol. 313, pp.
controller performance. 112210.
[9] C. Q. Zhi, L. Cheng and M. Z. Xian, “Experimental study on active
ACKNOWLEDGMENT vibration control for a kind of two-link flexible manipulator,”
The authors would like to express their gratitude to Mechanical Systems and Signal Processing, vol. 118, pp. 623-644.
Minister of Higher Education Malaysia (MoHE), Universiti [10] O. A. G. Perez, G. S. Navarro and J. F. P. Solis, “Flexible-link robots
with combined trajectory tracking and vibration control,” Applied
Teknologi MARA (UiTM), Universiti Teknologi Malaysia Mathematical Modelling, vol. 70, pp. 285-298.
(UTM) and Universiti Malaysia Pahang (UMP) for funding [11] H. M. Yatim, “Evolutionary algorithms for active vibration control of
the research and providing facilities to conduct this research. flexible manipulator,” Ph. D. Thesis Department of Applied
This research is supported using FRGS-RACER Grant: Mechanics, Faculty of Mechanical Engineering, Universiti Teknologi
RACER/1/12019/TK03/UITM//1 Malaysia, 2016.
[12] D. Sun, J. K. Mills, J. Shan and S. K. Tso, “A PZT actuator control of
REFERENCES a single-link flexible manipulator based in linear velocity feedback and
actuator placement,” Mechatronics, vol. 14, pp. 381-401, 2004.
[1] T. A. Z. Rahman, A. As’arry and N. A. A. Jalil, “Active vibration
control of a flexible beam structure using chaotic fractal search [13] R. Dutta, R. Ganguli and V. Mani, “Swarm intelligence algorithms for
algorithm,” Procedia Engineering, vol. 170, pp. 299-306, 2017. integrated optimization of piezoelectric actuator and sensor placement
and feedback gains,” Smart Materials and Structures, vol. 20, pp.
[2] E. Omidi, S. N. Mahmoodi and W. S. Shepard “Multi positive feedback
105018-105032, 2011.
control method for active vibration suppression in flexible structures,”
Mechatronics, vol. 33, pp. 23-33, 2016. [14] M. S. Hadi, “Evolutionary swarm algorithm for modelling and control
of horizontal flexible plate structures,” Ph. D. Thesis Department of
[3] M. S. Hadi, H. M.Yatim, M. H. A. Talib, A. Jamali and I. Z. M. Darus,
Applied Mechanics, Faculty of Mechanical Engineering, Universiti
“Fuzzy-pid based controler for active vibration control of nonlinear
Teknologi Malaysia, 2017.
dynamic systems,” Proceedings of 22nd International Conference On

241
Authorized licensed use limited to: UNIVERSITI TEKNOLOGI MARA. Downloaded on November 14,2023 at 04:54:08 UTC from IEEE Xplore. Restrictions apply.