Fuzzy Logic Based Adaptive Parameter Estimation System in Moving
Measurement Systems
1
ZUHAL ER, 2
BARIŞ GÖKÇE, 3
SALIH METIN YURTER
1Maritime Faculty, Istanbul Technical University (ITU),
Istanbul, TÜRKIYE
2Deparmentt of Mechatronic Engineering. Faculty of Engineering, Necmettin Erbakan University,
Konya, TÜRKIYE
3Zenopix Technology Industry and Trade Limited Company, Necmettin Erbakan University,
Ankara, TÜRKIYE
Abstract: -Fast and accurate weighing of the weighing systems used in industrial filling systems is of great
importance in terms of increasing production capacity and maintaining product quality. In facilities that grind
and process grain, machines and equipment are positioned horizontally and vertically on steel structures.
Since these machines continuously perform grinding, transferring, filling, and emptying operations, they
create continuous vibration in the mechanical systems they are connected to. Moving weighing systems are
significantly affected by these mechanical systems. When the impact effect of pneumatic valves-controlled
covers in moving weighing systems is added to these structural mechanical vibrations, there are significant
waits and delays in weighing systems that measure performance. For this reason, in a performance
measurement system in a flour mill, the measurement interval increases as the amount of weighing increases.
For example, in a moving weighing system that performs 50 kg performance weighing, the measurement
interval can increase up to 15 seconds, which is quite long. In this study, an applied study has been conducted
to increase the weighing performance in moving weighing systems and to minimize the measurement interval.
The data collection process in the study focuses on two main components: load cell data and IMU data. Thus,
it is aimed to overcome the difficulties of traditional methods used in weighing systems, which are generally
observed to be insufficient to combat slow and noisy data. The analysis techniques used in this study are
Kalman Filtering, Dynamic Q and R Matrix Updates, Comparative Analysis and Statistical Analysis. The
Kalman filter was used for the integration of Load cell and IMU data and was applied to filter out noise and
oscillations in the weighing data and make more accurate weight estimates. The results obtained showed that
the dynamic Kalman filtering method can provide faster and more accurate weighing results compared to
traditional methods, with error rates varying between 0.4% and 1% for different combinations of Q and R
values in measurements made on the scale. Dynamic Kalman filtering method effectively filters oscillatory
and noisy load cell signals, with error rates of 0.7% to 1% for Q=0.02 and R=17 parameters, and error rates of
0.4% to 0.7% for Q=0.07 and R=13 parameters. was able to obtain more accurate weight estimates. This study
has shown that the dynamic Kalman filtering method is a potential method that can be used in industrial filling
systems. This method can contribute to increasing production capacity and maintaining product quality by
providing faster and more accurate weighing results. In this respect, the research has a unique contribution.
This method provides a revolutionary development in industrial weighing systems and fills an important gap
in the literature.
Key-Words: -Kalman filter, Fuzzy logic, Dynamic measurement, Inertial measurement unit, Load cell
Received: October 2, 2023. Revised: August 11, 2024. Accepted: September 4, 2024. Published: October 7, 2024.
1. Introduction
Mobile measurement systems are systems used to
collect accurate and reliable data in dynamic
environmental conditions. These systems are widely
used in various application areas, for example,
robotics, vehicle control, aircraft, marine vessels and
mobile sensor networks. Fuzzy logic-based adaptive
parameters are used to increase the performance of
such systems. In today's modern industrial systems,
weighing is a critical process control phase. In raw
material and related sectors, the need for fast and
precise weighing is increasing day by day. The grain
industry is also at the forefront of these sectors [1],
[2]. Active weighing systems are systems used in the
grain industry to weigh products such as wheat, flour
and bran during production, without interrupting
production. Thanks to the efficiency systems with
International Journal of Electrical Engineering and Computer Science
DOI: 10.37394/232027.2024.6.15
Zuhal Er, Bariş Gökçe, Salih Metin Yurter
E-ISSN: 2769-2507
126
Volume 6, 2024
advanced technological infrastructure, production-
related information such as capacity and efficiency
values can be easily accessed at any time while
production continues. In this context, it is critical that
the measured values exactly match the real values.
Even the slightest error in measured values can lead
to huge financial losses in mass production factories.
The use of the dynamic weighing method in the grain
industry and related production systems provides an
increase in the amount of product weighed per unit
time due to the products being in motion. In this way,
time and economy are saved [1], [3], [4]. In the
dynamic weighing method, in order to reach the
desired weighing speeds, the products must be
weighed without stopping; However, mechanical
vibrations and environmental harmonics may cause
distortions in the measurement signal that vary
depending on the speed of the moving system and the
weight of the object to be measured [5]-[7], [17],
[18], [27]-[30].
Load Cells are generally used as weight sensors in
weighing systems. The combination of load cells
with an oscillatory response due to their nature and
the low-frequency disturbance caused by vibrations
in the system results in a noisy measurement signal
[6], [19], [31]. It is very difficult to separate these
disturbing effects, which occur from mechanical and
structural effects, from the measurement signal. To
correct the system response, filtering of the
measurement signal is generally used [5]. Generally
speaking, when we look at the literature, passive
vibration reduction approaches are used in
engineering applications when the general
characteristics of vibration are known, but over time,
both structural changes and modifications on the
system may cause problems with low-frequency
vibrations [8], [9], [20]-[32].
With the advancement of technology, new types
of actuators and sensors are emerging, and with the
cheaper computing technology, active vibration
control has become applicable to many problems
[10], [11], [32]. While different numerical models are
designed to create active vibration control algorithms,
it is important that the dynamic properties of the
structure to be measured must be preserved
throughout the control or measurement process. This
enables the adjustment of some controllers used to
reduce or characterize vibration, such as positive
position feedback (PPF) [9], [11], while controls to
be made with algorithms such as linear quadratic
(LQ) [10], [12] or model estimation. [13], [14] enable
results to be obtained in different ways.
In industrial filling systems, especially in grain
processing facilities, high-precision and high-speed
weighing operations are critical for production
efficiency and product quality. The continuous
movement of machines and equipment in these
facilities causes significant vibrations in the
associated structural systems. These vibrations
negatively affect the measurement accuracy of
moving weighing systems, leading to delays and
erroneous results in weighing processes. The impact
effect of pneumatic valves, in particular, further
exacerbates these negative effects. This study aims to
investigate ways to improve weighing performance
and reduce measurement times in moving weighing
systems in grain processing facilities. To this end,
dynamic parameters were determined using the
Kalman filtering method with load cell and Inertial
Measurement Unit (IMU) data. The primary
objective of the study is to develop a model to obtain
more accurate and faster weighing results, despite the
adverse effects of vibrations and external factors
2. Materıals and Methods
In this study, MPU6050 was used as the IMU and
STM32F407V series 168 MHz processor was used as
the microcontroller. Figure 1 shows the
microcontroller used in the study. Through this
control card, the UART communication output was
connected to the computer with a UART-USB
converter and the module was made ready for data
transfer.
Fig. 1. Data collection card used
The data collection module connection is
seen in figure 2. The Control Card is placed in two
different orientation positions on the mobile grain
weighing system. The reason for using two different
sensor modules was that the sensitivity of the
acceleration movement in the Z axis would be low,
and the two cards were placed at a 90-degree angle to
each other. Thus, more accurate data was obtained.
Both cards were placed in the chamber within the
measurement system, thus creating a data collection
environment that could absorb all the noise that the
weighing system would be exposed to.
International Journal of Electrical Engineering and Computer Science
DOI: 10.37394/232027.2024.6.15
Zuhal Er, Bariş Gökçe, Salih Metin Yurter
E-ISSN: 2769-2507
127
Volume 6, 2024
Fig. 2. Control system connected to active weighing
system
3. Results and Dıscussıons
Figure 3 shows the acceleration values on the X
axis from the data taken in the measurement
environment. From the amplitude value given here, it
can be seen that periodic accelerations or noise occur.
Fig. 3. X-axis acceleration value of the mobile
weighing system
Figure 4 shows the acceleration amplitudes of the
Y axis to which the control system is connected.
When these values are compared with the X-axis
acceleration values in the previous graph, it is seen
that the oscillation is not much in the Y direction.
This direction is actually 90 degrees perpendicular to
the pneumatic valves. In this way, limitation in one
direction affects the vibration.
Fig. 4. Y-axis acceleration value of the mobile
weighing system
Figure 5 shows the Z axis acceleration value of
the mobile weighing system. Here, it has been
observed that in normal cases, no vibration is
observed or the noise in the Z direction is very low
due to the gravitational effect on the Z axis, but it has
been observed that it has serious effects on the Z axis
due to the knocking during opening and closing of
the covers connected to the pneumatic pistons. This
situation affects the weighing process extremely
negatively.
Fig. 5. Z axis acceleration value of the mobile
weighing system
These noises in the Z axis are the noises created
by the system's moving elements (pneumatic
systems, etc.) and the released elements in the system
(joints, etc.). These noises completely change the
measurement characteristics. Therefore, it was
concluded that it would be more accurate to
determine the parameters by taking into account the
effects of these elements in order for the selected
filters to be adaptive.
Noise was eliminated by applying a Kalman filter
on the Z-axis speed and position displacement values
in the active weighing system. After removing the
noise, RMS values were obtained for speed and
position change. Accordingly, the RMS values of
speed and position change are RMSpeed = 0.51;
RMSposition = 0.46.
Figure 6 shows the Z axis speed change value
over rms value graph in the active weighing system.
When this graph is compared with the previous
graph, the noise effect of moving and free elements
on the measurement between measurement periods is
clearly seen. If you pay attention, the noise
characteristics in each period are not the same.
Therefore, if the noises within a 120-second
measurement are taken as reference, the filter
parameters can be determined as accurately as
possible. As can be seen here, position changes occur
in both directions of the axis and position changes are
very high due to moving free elements during
measurement. Here too, taking displacement values
above the RMS values enabled the grouping of the
data's measurement periods.
Fig. 6. Z axis Speed and Position Change Value in the
active weighing system is above the RMS Value
Within the scope of this study, adaptive
adjustment of the fuzzy logic-based measurement
International Journal of Electrical Engineering and Computer Science
DOI: 10.37394/232027.2024.6.15
Zuhal Er, Bariş Gökçe, Salih Metin Yurter
E-ISSN: 2769-2507
128
Volume 6, 2024
noise covariance matrix R or process noise
covariance matrix Q was provided. Adaptively
specifying these parameters improves Kalman filter
performance and prevents the filter from biasing
when R or Q are uncertain.
In dynamic weighing systems, mechanical covers
create delays in weighings due to physical constraints
during stops and starts. These physical constraints
and time delays impose certain acceptable time
intervals for measurements. For example, in grain
performance weighing systems, weighing intervals
vary between 5 and 15 seconds. Generally, the first 4-
5 seconds of this interval is due to mechanical
constraints.
This fuzzy logic based adaptive Kalman filter was
tested on a digitally developed weighing system. It
has been observed that the fuzzy logic based adaptive
Kalman filter gives much better performance at
acceptable phase shift.
An algorithm using fuzzy logic principles has
been created to adaptively adjust the Q and R matrix
of the noise covariance matrix. Speed and position
data derived from acceleration data received from the
IMU sensor were used in the fuzzy logic design.
Speed and position data here play a key role in
determining noise during measurement. The obtained
average speed and position data were used as fuzzy
logic input set membership function data. At this
stage, minimum and maximum speeds and position
displacements are given in Table 1.
Table 1. Inimum and Maximum Values of
Speed and Position Data
Minimum
Maxsimum
Average
Speed
0
5
3,71
Position
0
13
8,88
At this stage, a fuzzy logic inference system with
two inputs and two outputs is designed. This system
was made for both Mamdani and Sugeno. Speed and
position data were determined as inputs, and Q and R
values were determined for the outputs. For
Mamadani, the “AND” method and minimum value
were determined for two entries, and the center of
gravity method was selected in the rinsing process.
Figure 7. A visual representation of the Mamdani
type fuzzy logic inference system is given. The
selected features are also shown on the image.
Fig. 7. Visual demonstration of mamdani type fuzzy
logic inference system
Speed and location data were used as input. In
Table 2, the velocity input membership function data
is given, and in Figure 8, the graphical representation
of the velocity input membership function data is
given. Here, the input data is divided into five fuzzy
sets and determined as smallest (vs), small (s),
medium (m), high (h) and highest (vh). Triangular
membership function was preferred. While
determining the minimum and maximum values of
the data, 0 and the maximum value were determined
by rounding the highest speed value read on the
weighing system to the upper integer.
Table 2. Velocity Input Membership
Function Data
Velocity Input Membership Function
Vs
0
0,75
1,50
S
0,25
1,5
2,5
M
1,5
2,5
3,5
H
2,5
3,5
4,75
Vh
3,5
4,25
5
Fig. 8. Graphical representation of velocity entry
membership function data
International Journal of Electrical Engineering and Computer Science
DOI: 10.37394/232027.2024.6.15
Zuhal Er, Bariş Gökçe, Salih Metin Yurter
E-ISSN: 2769-2507
129
Volume 6, 2024
Figure 9 shows the graphical representation of the
distance input membership function. As in the
velocity table, a triangular membership function was
used here and the data was determined by rounding
the data from 0 to the maximum displacement data to
the upper integer. Here, the creation of a triangular
membership function has been determined entirely
based on experience.
Fig. 9. Graphical representation of Q Parameter
output membership function
Figure 10 shows the graphical representation of
the Q parameter output membership function.
Determining the output of this parameter is based
entirely on observation. The lowest Q value
completely eliminates noise by making a very flat
prediction at the filter output and creates a noticeable
phase shift. The lowest acceptable Q value observed
on the scale is 0.001. This value is the minimum
value that can be used if there is very high speed and
very large positional displacement. If it is for the
lowest speed and smallest positional displacement,
the highest Q value that can be used is 0.1. Fuzzy
inference will make an inference in the meantime.
Fig. 10. Graphical representation of distance entry
membership function
Figure 11 shows the graphical representation of
the R parameter output membership function. As can
be seen from here, the change in R affects the output
with a linear change, not with an exponential change
like Q.
Fig. 11. R parameter output membership function
graphical representation
Table 3 gives the rule table defined for output.
While creating the rule table defined for the output,
the output definition according to the inputs was
determined according to the effect of 40% speed and
60% position change. Since this proportional
determination has a greater effect of spatial
displacement on noise than speed, output
memberships were determined according to this rule.
Table 3. Rules Definations
Velocity
Distance
Q
R
1
VS
VS
VH
S
2
VS
S
H
S
3
VS
M
H
M
4
VS
H
M
M
5
VS
VH
M
H
6
S
VS
VH
S
7
S
S
H
S
8
S
M
M
M
9
S
H
M
M
10
S
VH
S
H
11
M
VS
H
S
12
M
S
H
M
13
M
M
M
M
14
M
H
S
H
15
M
VH
S
H
16
H
VS
H
M
17
H
S
M
M
18
H
M
M
H
19
H
H
S
H
20
H
VH
VS
H
21
VH
VS
M
M
22
VH
S
M
M
23
VH
M
S
H
24
VH
H
S
H
25
VH
VH
VS
H
The distribution of these inferences made for
Mandani on the membership functions is given in
figures 12, 13, 14 and 15.
International Journal of Electrical Engineering and Computer Science
DOI: 10.37394/232027.2024.6.15
Zuhal Er, Bariş Gökçe, Salih Metin Yurter
E-ISSN: 2769-2507
130
Volume 6, 2024
Fig. 12. Mamdani Fuzzy Logic Inference Example
Input Speed = 0.5, Position = 1 for Q = 0.07 , R = 6
Fig. 13. Mamdani Fuzzy Logic Inference Example
Input Speed = 2.5, Position = 6.5 for Q = 0.028 , R
= 12
Fig. 14. Mamdani Fuzzy Logic Inference Example
Input Speed = 4,9, Position = 12.9 for Q = 0.005, R
= 16
Fig. 15. Mamdani Fuzzy Logic Inference Example for
Input Speed = 3.71 and Position = 8.8 , Q = 0.02
and R = 12
The same fuzzy logic inference was also made in
the Sugeno method, which gives linear output. The
distribution of these inferences made for Sugeno on
the membership functions is given in figures 16, 17,
18, and 19.
Fig. 16. Sugeno Fuzzy Logic Inference Example for
Input Speed=0.5 and Position=1, Q=0.1 R=2
Fig. 17. Sugeno Fuzzy Logic Inference Example for
Input Speed=2.5 and Position=6.5, Q=0.09 R=8
International Journal of Electrical Engineering and Computer Science
DOI: 10.37394/232027.2024.6.15
Zuhal Er, Bariş Gökçe, Salih Metin Yurter
E-ISSN: 2769-2507
131
Volume 6, 2024
Fig. 18. Sugeno Fuzzy Logic Inference Example for
Input Speed=3.7 and Position=8.8, Q=0.07 R=13
Fig. 19. ugeno Fuzzy Logic Inference Example for
Input Speed=4.9 and Position=12.9, Q=0.001 R=17
Figure 20 shows the Sugeno fuzzy logic software
interface made in Python.
Fig. 20. Mamdani Fuzzy Logic Software Made in
Python Programming Language
When the Table 4 is examined for the fuzzy
inference results obtained with Mamdani and
Sugeno, it is seen that the linear inference results
obtained from Sugeno are more meaningful and
closer to reality.
Table 4. Table Type Styles
Velocity
Position
Q
R
Input
Sugeno
1
0,5
1
0,1
2
2
2,5
6,5
0,09
8
3
4,9
12,9
0,001
21
4
3,71
8,8
0,07
13
Input
Mamdani
1
0,5
1
0,07
6
2
2,5
6,5
0,028
12
3
4,9
12,9
0,005
16
4
3,71
8,8
0,02
17
Figure 21 shows Mamdani fuzzy inference results
and original weighing data graph, final weighing
result graph using Kalman filter with parameter
defined for Q=0.02 and R=17. Here, it has been
observed that for the coefficients R = 17, Q = 0.02,
the error rate is 0.7% in 6 weighings, 0.9% in 7
weighings and 1% in 8 weighings. The reason for the
increase in the error rate can be attributed to the
increased phase lag between the raw data and the
filtered data as the system better suppresses noise. As
the phase shift increases, the error also increases
significantly, resulting in a longer weighing time. The
data obtained as a result of weighing were as shown
in Figures 22 and 23, respectively.
Fig. 21. R:17 Q:002 Number of Weighings:6
Fig. 22. R:17 Q:002 Number of Weighings:7
International Journal of Electrical Engineering and Computer Science
DOI: 10.37394/232027.2024.6.15
Zuhal Er, Bariş Gökçe, Salih Metin Yurter
E-ISSN: 2769-2507
132
Volume 6, 2024
Fig. 23. R:17 Q:002 Number of Weighings:8
Figure 24 shows the original weighing data
and graph of Sugeno fuzzy inference results, data
graph results using Kalman filter with parameter
defined for Q=0.07 and R=13. It has been observed
that the system works successfully with an error rate
of 0.4% in 6 weighings, 0.6% in 7 weighings and
0.7% in 8 weighings for the coefficients R = 13, Q =
0.07. The data obtained as a result of weighing were
as shown in Figures 25 and 26, respectively.
Fig. 24. R:13 Q:007 Number of Weighing:6
Fig. 25. R:13 Q:007 Number of Weighing:7
Fig. 26. R:13 Q:007 Number of Weighing:8
The available studies focusing on dynamic
parameter estimation with the Kalman filter method
using acceleration, velocity and position data are
presented in the Table 5.
Table 5. Comparison of This Study with the
Literature
Refrenc
es
Main
Theme
Application Area
[1], [2],
[5], [6],
[7],
[17],
[18],
[20],
[27],
[28],
[29],
[30],
Dynamic
Weighing
Continuous mass
measurement in
checkweighers, dynamic
compensation, dynamic
load identification
[1],
[21],
[22],
[23],
[24],
[25],
[26]
Kalman
Filter
attitude estimation, state
estimation, adaptive
filtering
[8], [9],
[10],
[11],
[12],
[13],
[14],
[15],
[16],
Mechanic
al
Vibration
LQG control of vibrations
in flexible structures,
vibration control of active
structures.
This
study
Kalman
filter,
mobile
measurem
ent, load
cell, IMU
sensor,
As can be seen above, this
study shares common
aspects with other studies
in the literature, but it
offers originality in its
practical application. By
eliminating environmental
and structural vibrations
International Journal of Electrical Engineering and Computer Science
DOI: 10.37394/232027.2024.6.15
Zuhal Er, Bariş Gökçe, Salih Metin Yurter
E-ISSN: 2769-2507
133
Volume 6, 2024
mobile
grain,
weighing
system,
Estimatio
n,
Measure
ment,
fuzzy
logic
inference
system,
membersh
ip
function,
accelerati
on,
velocity,
position
that occur during dynamic
grain weighing, it
accelerates the weighing
process. This is achieved by
estimating dynamic
parameters using the
Kalman filter method with
raw load cell data and
acceleration, speed, and
position data obtained from
the IMU sensor. This
approach allows for faster
weighing.
4. Conclusıons
Active farming systems produce very noisy
measurement results due to both the mechanical and
structural vibrations of their environment and various
external factors. This can lead to deviations in
measurement values and sometimes even serious
measurement errors. While measurements are made
relatively quickly in non-dynamic structural systems,
filtering measurements of vibrations on dynamic and
moving systems and vibrations on a dynamically
operating system is very important for the industrial
sector.
In this study, filters that can be used for
measurements of moving weighing systems and the
effects of these filters on measurement and
performance characteristics were investigated. The
results obtained allowed the development of a fuzzy
logic-based parameter extraction system to update the
coefficients of the active filters used.
In this study, Mamdani and Sugeno fuzzy inference
methods were combined with the Kalman filtering
technique to reduce measurement errors in moving
weighing systems. While both methods yielded
successful results to a certain extent, the Sugeno
method was observed to improve system
performance with lower error rates.
Mamdani Method: In the Mamdani method, although
the obtained results suppressed the noise in the
system better, they caused an increase in the error
rate due to the phase shift between the raw data and
the filtered data. This situation is undesirable,
especially in applications requiring high precision.
Sugeno Method: The Sugeno method, on the other
hand, was successful in both noise suppression and
minimizing phase shift, resulting in lower error rates.
This result indicates that the Sugeno method is more
suitable for such applications.
In conclusion, a fuzzy logic-based parameter update
system has been developed to increase the
measurement accuracy in moving weighing systems.
In this system, it has been observed that the Sugeno
method is more successful and provides significant
improvement when used with Kalman filtering. The
obtained results indicate that this method can also be
used in different industrial applications
Acknowledgment
We would like to thank the “Endüstriyel Elektrik
Elektronik San. ve Tic. A.Ş.” company for
contributing to the testing of the system and
collection of data.
Declaration of Generative AI and AI-Enabled
Technologies in the Writing Process
During the preparation of this study, only the
authors used ChatGPT4.0 mini for grammar and
language checking. After using this tool/service, the
authors reviewed and edited the content as necessary
and assumes full responsibility for the content of the
publication.
References:
[1] Nied´zwiecki, M. and Wasilewski A.,
"Application of Adaptive Filtering to Dynamic
Weighing of Vehicles", Control Engineering
Practice,4:635-644,1996.Lee, K.-F., Automatic
Speech Recognition: The Development of the
SPHINX SYSTEM, Kluwer Academic
Publishers, Boston, 1989.
[2] Mesut Y., Ismail Y., "Dinamik Agırlık Ölçümü
Için Filtre Tasarımı", 2017.
[3] Clement J., Novas N., “High Speed Intelligent
Classifier of Tomatoes by Colour, Size And
Weight”, Spanish Journal of Agricultural
Research,2,314-325,2012.
[4] Yabanova, I. and Yumurtacı, M., "Basic
Specification for IP Fast Reroute: Loop-Free
Alternates", RFC 5286, 2016.
[5] Yamazaki, T., Sakurai, Y., Ohnishi, H.,
Kobayashi, M., Kurosu, S., "Continuous Mass
Measurement ˙ In Checkweighers And Conveyor
Belt Scales", Proceedings of the 41st SICE
Annual Conference, Osaka, 5-7 Aug., 2002.
[6] Boschetti, G., Caracciolo, R., Richiedei, D.,
Trevisani, A., "Modelbased dynamic
compensation of load cell response in weighing
machines affected by environmental vibrations",
International Journal of Electrical Engineering and Computer Science
DOI: 10.37394/232027.2024.6.15
Zuhal Er, Bariş Gökçe, Salih Metin Yurter
E-ISSN: 2769-2507
134
Volume 6, 2024
Mechanical Systems and Signal Processing,
34(1-2), 116-130, 2013.
[7] Pietrzak, P., Meller, M., Nied´zwiecki, M.,
"Dynamic mass measurement in checkweighers
using a discrete timevariant low-pass filter",
Mechanical Systems and Signal Processing,
48(1-2), 67-76, 2014.
[8] A. Montazeri, J. Poshtan, and A. Choobdar,
“Performance and robust stability trade-off in
minimax LQG control of vibrations in flexible
structures,” Engineering Structures, vol. 31, no.
10, pp. 2407–2413, 2009.
[9] A. Preumont, "Vibration Control of Active
Structures", Kluwer Academic, 2nd edition,
2002.
[10] C. R. Fuller, S. J. Elliott, and P. A. Nelson,
“Active Control of Vibration”, Academic Press,
San Francisco, Calif, USA, 1st edition, 1996.
[11] H. Benaroya and M. L. Nagurka,
“Mechanical Vibration: Analysis, Uncertainities
and Control”, CRC Press/Taylor & Francis,
Boca Raton, Fla, USA, 3rd edition, 2010.
[12] N. Chandiramani and S. Purohit, “Semi-
active control using magnetorhelogical dampers
with output feedback and dis- tributed sensing,”
Shock and Vibration, vol. 19, no. 6, pp. 1427–
1443, 2012.
[13] M. Hassan, R. Dubay, C. Li, and R. Wang,
“Active vibration control of a flexible one-link
manipulator using a multivariable predictive
controller,” Mechatronics, vol. 17, no. 1, pp.
311–323, 2007.
[14] G. Taka´cs and B. Rohal, Predictive
Vibration Control: Effi- cient Constrained MPC
Vi-Bration Control for Lightly Damped
Mechanical Systems, Springer, London, UK,
2012.
[15] Z. Zhang, F. Hu, Z. Li, and H. Hua,
“Modeling and control of the vibration of two
beams coupled with fluid and active links,”
Shock and Vibration, vol. 19, no. 4, pp. 653–
668, 2012.
[16] G.Takács, T.Polóni, and B.Rohal’-Ilkiv,
“Adaptive Model Predictive Vibration Control
of a Cantilever Beam with Real-Time Parameter
Estimation”, Hindawi Publishing Corporation
Shock and Vibration, Volume 2014, pp. 1-15,
2014.
[17] Shinder, I. and Moldover, M. (2010),
Feasibility of an Accurate Dynamic Standard for
Water Flow, Flow Measurement and
Instrumentation,
https://tsapps.nist.gov/publication/get_pdf.cfm?p
ub_id=832344
[18] Shinder, I. and Moldover, M. (2009),
Dynamic Gravitational Standard for Liquid
Flow: Model and Measurements, Annual
International Symposium on Fluid Flow
Measurement, 7th CEESI, Anchorage, AK,
https://tsapps.nist.gov/publication/get_pdf.cfm?p
ub_id=903358
[19] Ştefănescu, D.M. (2020). Evolution of Strain
Gauge Force Transducers-Design, Fabrication,
Testing, Calibration and Databases. In:
Handbook of Force Transducers. Springer,
Cham. https://doi.org/10.1007/978-3-030-35322-
3_7
[20] Liu, R., Dobriban, E., Hou, Z. et al.
Dynamic Load Identification for Mechanical
Systems: A Review. Arch Computat Methods
Eng 29, 831–863 (2022).
https://doi.org/10.1007/s11831-021-09594-7
[21] Azam, Saeed Eftekhar, Eleni Chatzi, and
Costas Papadimitriou. "A dual Kalman filter
approach for state estimation via output-only
acceleration measurements." Mechanical
systems and signal processing 60 (2015): 866-
886.
[22] Park, S.; Gil, M.-S.; Im, H.; Moon, Y.-S.
Measurement Noise Recommendation for
Efficient Kalman Filtering over a Large Amount
of Sensor Data. Sensors 2019, 19, 1168.
https://doi.org/10.3390/s19051168
[23] Xinchen Li, Levent Guvenc, Bilin Aksun-
Guvenc, Vehicle State Estimation and
Prediction, Electrical Engineering and Systems
Science, Systems and Control,
arXiv:2304.11694,
https://doi.org/10.48550/arXiv.2304.11694
[24] H. Şahın, B. Gürkan and V. E. Ömürlü,
"Sensor Fusion Design by Extended and
Unscented Kalman Filter Approaches for
Position and Attitude Estimation," 2022
International Congress on Human-Computer
Interaction, Optimization and Robotic
Applications (HORA), Ankara, Turkey, 2022,
pp. 1-10, doi:
10.1109/HORA55278.2022.9799879.
[25] Deibe, Á.; Antón Nacimiento, J.A.;
Cardenal, J.; López Peña, F. A Kalman Filter for
Nonlinear Attitude Estimation Using Time
Variable Matrices and Quaternions. Sensors
2020, 20, 6731.
https://doi.org/10.3390/s20236731
[26] Gerald Cook; Feitian Zhang, "Application of
Kalman Filtering," in Mobile Robots:
Navigation, Control and Sensing, Surface
Robots and AUVs , IEEE, 2020, pp.133-169,
doi: 10.1002/9781119534839.ch5
[27] https://ch.mathworks.com/help/fusion/ug/tun
ing-kalman-filter-to-improve-state-
estimation.htmlhtps://www.nist.gov/pml/sensor-
science/fluid-metrology/dynamic-weighing-
method-unsteady-liquid-flow-measurements
[28] https://www.nar-robotics.com/dynamic-
weighing-technology-pioneering-solutions-for-
dynamic-industries/
[29] https://www.rdsmme.com/blog/onboard-
weighing-systems-for-the-arable-agricultural-
industry/
[30] https://www.camaweigh.com/blog/post/what
-is-the-difference-between-static-and-dynamic-
weighing/
[31] https://www.iqsdirectory.com/articles/load-
cell/types-of-load-cells.html
[32] Vivek Warke, Satish Kumar, Arunkumar
Bongale, Pooja Kamat, Ketan Kotecha,
Ganeshsree Selvachandran, Ajith Abraham,
Improving the useful life of tools using active
vibration control through data-driven
approaches: A systematic literature review,
Engineering Applications of Artificial
Intelligence, Volume 128, 2024, 107367,ISSN
0952-1976,
https://doi.org/10.1016/j.engappai.2023.107367
International Journal of Electrical Engineering and Computer Science
DOI: 10.37394/232027.2024.6.15
Zuhal Er, Bariş Gökçe, Salih Metin Yurter
E-ISSN: 2769-2507
135
Volume 6, 2024
Contribution of Individual Authors to the
Creation of a Scientific Article (Ghostwriting
Policy)
The authors equally contributed in the present
research, at all stages from the formulation of the
problem to the final findings and solution.
Sources of Funding for Research Presented in a
Scientific Article or Scientific Article Itself
No funding was received for conducting this study.
Conflict of Interest
The authors have no conflicts of interest to declare
that are relevant to the content of this article.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the
Creative Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en
_US
International Journal of Electrical Engineering and Computer Science
DOI: 10.37394/232027.2024.6.15
Zuhal Er, Bariş Gökçe, Salih Metin Yurter
E-ISSN: 2769-2507
136
Volume 6, 2024
