ISSN 2278-3091

Aaron Don M. Africa et al., International Journal ofVolume

Advanced8,Trends
No.4,inJuly – August
Computer 2019
Science and Engineering, 8(4), July- August 2019, 1427 - 1432
International Journal of Advanced Trends in Computer Science and Engineering
Available Online at http://www.warse.org/IJATCSE/static/pdf/file/ijatcse61842019.pdf
https://doi.org/10.30534/ijatcse/2019/61842019

Modeling in the Frequency Domain of Control Systems

Aaron Don M. Africa, Patrick Bernard T. Arevalo, Arsenic S. Publico, Mharela Angela A. Tan
Department of Electronics and Communications Engineering
De La Salle University, Manila
2401 Taft Ave., Malate, Manila 1004, Philippines
[email protected]

in the frequency domain, however, a more effective tool for

ABSTRACT this analysis is a spectrum analyzer. A spectrum analyzer
displays 2D graph of the power of the signal at the x-axis and
Modeling in the frequency domain for control systems are its frequency at the y-axis. The said graph is referred to as the
made up of many sub-topics such as modeling of frequency spectrum of a signal since it indicates how strong a
electric/electronic circuits, modeling of translational signal is at all values of frequencies. The simplest and most
mechanical systems, modeling of rotational mechanical common example in the frequency domain is the sine wave,
systems, modeling of rotational mechanical systems with which is a perfectly periodic signal. The frequency spectrum
gears, modeling of electromechanical systems, nonlinear of a 100-Hz sine wave is composed of only one frequency,
systems, linearization techniques, concepts of poles and zeros and so the frequency spectrum is illustrated below.
of a system, and the pole-map zero. These topics, as well as
other topics related to modeling in the frequency domain, will Several research studies which involve identification
be discussed in this paper. procedures are developed in the frequency domain. According
to one study, methods which are based on the frequency
Key words: Frequency Domain, Control Systems, domain use concepts such as storage modulus and loss factor.
Mechanical Systems, Rotational Systems. Sometimes, equations presented in the time domain must be
transformed to the frequency domain, the common tool for
this transformation is known as the Fourier transform [2].
1. INTRODUCTION
The approach of the frequency domain for linear systems is
Frequency Domain analysis studies signal with respect to where the control, processing, and communication dependent
frequency, rather than with respect to time. Information being on. This approach has been popular in the fields of science and
studies is plotted having frequency on the x-axis and the engineering. Key principles connected to frequency analysis
amplitude on the y-axis. The said plot displays how the energy is the Frequency Response Function (FRF), which is the basis
of the signal is distributed as a function of frequency. of the concepts such as Nyquist stability, Bode diagrams,
filter designs, and many more [3].
Time and frequency domain representations are both present
in electrical signals. Both voltage and current measurements 2. LITERATURE REVIEW
are expressed as functions of time in the time domain.
Furthermore, digital information is usually displayed by a In one particular study, the grey box system identification
voltage as a function of time as well as the signals being method was pertained to address the issue of fractional-order
measured on an oscilloscope. model identification from existing frequency domain data.
Due to linked expenses, this was the only possible and
On the other hand, frequency-domain representations are feasible method or approach for industrial cases of system
commonly used in linear system analysis. Most interference identification wherein the capability of conducting
in and source of signals are defined in the time domain, experiments involving time domain were limited. This study
however, the frequency domains offer more convenient evaluates the potential of the said approach and examines the
system behavior and signal transformations. The use of a model parameter estimation method depending on the existing
linear frequency response function in linear frequency domain frequency domain data. For this study, there are three
analysis allows the evaluation to have a fundamental basis in optimization algorithms used: Trust Region Reflective (TRR)
learning system configurations like the control bandwidth and nonlinear least-squares estimation method,
sensitivity function [1]. Levenberg-Marquardt (LM) nonlinear least-squares
estimation method, and the Nelder-Mead (NM) direct search
When performing a measurement of the frequency of a signal, simplex method. This research suggested a grey box model
it is also the analysis of the signal in the frequency domain.
identification method for fractional-order systems relying on
Some oscilloscopes are used in the analysis of periodic signals

1427
Aaron Don M. Africa et al., International Journal of Advanced Trends in Computer Science and Engineering, 8(4), July- August 2019, 1427 - 1432

the frequency domain data available. The said identification most basic, if not the most basic concept that is known in
approach was verified using an academic sample where a lot electrical circuits is the Resistance Inductance Capacitance
of optimization algorithms under the identification process circuit or the RLC circuit The RLC circuit is a big help in
were put to test. Based on the scope of this study, the Trust understanding behaviors that are evident in control systems.
Region Reflective (TRR) method has obtained the most RLC circuits may be considered as series or parallel the main
exemplary results among the different optimization difference of these two circuits are obvious from the names
algorithms. On the other hand, the Nelder-Mead simplex itself, series RLC circuits have a series connection for the
method had shown almost the same quality of performance, resistors, inductors, and capacitors, while for parallel RLC
therefore, it could generally be utilized to solve the problems circuits, resistors, inductors, and capacitors are connected in
where bounds are either hard to estimate or unknown. For parallel with respect to one another [7].
future research work, the researchers recommend the
assessment of the current approach based on real-life The next topic for modeling in the frequency domain is about
industrial information and re-evaluation of the translational mechanical systems. One of the first things to
Levenberg-Marquardt algorithm using the approach of know about translational mechanical systems is that it moves
coordinate transformation to permit the application of search along a straight line and that it is made up of three basic
parameter bounds [4]. elements, which are mass, spring, and dashpot or damper.
When a force is applied to a given translational mechanical
In another study, a frequency domain with a three-phase system, an opposing force is present which is caused by the
transformer model was implemented using nodal method. mass, elasticity, and the friction which is present in a given
The model was able to show and demonstrate effectiveness system. It is also known that since the force is applied and the
and ease through its involvement on basic distribution force which is present in the system which may be considered
networks. Also taken into account was the angular as opposing forces are of opposite directions, the algebraic
displacement parallel to winding connections, which verified sum of the forces in the system and is applied to the system is
and proved the versatility of the model. Furthermore, this zero.
paper also described the function of frequency-domain
transformer modeling processes for electromagnetic transient The first of the three is mass (figure 1), it may be defined as
analysis of networks. The said was dependent on the the body’s property which is able to store kinetic energy. It
utilization of admittance matrices which can be incorporated was also mentioned in the previous part that when a force is
with bigger networks later on. For the actual experiments, the applied to a body that has a mass, it will be opposed by a force
researchers have used the model to compute the transient over which is due to the mass of the system. It should also be noted
voltages which are results of switching and the Numerical that this force that opposes is proportional with respect to the
Laplace Transform (NLT) to obtain the time domain response. acceleration of a given body.
According to the researchers, the model is easy to modify in
order to try and consider winding connections and the
appropriate angle displacement inherent to them. The
accuracy and effectiveness of the model were evaluated
through having the results compared with EMTP-RV or
Electromagnetic Transients Program [5].

The study by Gocheva et.al. focused on proposing a novel Figure 1: Mass Force Diagram
approach with regard to using generalized nets in the
modeling of electronic circuits. Generalized nets are utilized The next is spring (figure 2), it may be defined as an element
in choosing components, connecting, modifying, and which is able to store potential energy. For a spring, when a
simulating electronic circuits. The proposed matrix calculus force is applied on a spring, it will be opposed by a force
was done with respect to time discretization. It takes into which is caused by the elasticity of a spring. This opposing
account the performance capacity of generalized nets as an aid force is proportional with respect to the displacement of the
in parallel processing. Systematization of electronic spring.
components is one of the most essential elements and
considerations of the proposed approach. Furthermore, it
provides opportunities to manage complex electronic circuits
using basic methods like matrix equations instead of solving
differential equations [6].

3. THEORETICAL CONSIDERATION

One of the important concepts in modeling in the frequency Figure 2: Spring Diagram
domain in control systems is electrical circuits. One of the

1428
Aaron Don M. Africa et al., International Journal of Advanced Trends in Computer Science and Engineering, 8(4), July- August 2019, 1427 - 1432

Lastly, dashpot can be described as having a torque applied to

Lastly, for translational mechanical systems is the dashpot, it it with an opposing torque which is caused by the rotational
may be explained by having a force applied to a dashpot and friction (figure 6) of the dashpot itself. This is also
the opposing force is caused by the friction of the dashpot proportional with respect to the velocity of the body [8].
(figure 3). This opposing force is proportional with respect to
the velocity of the body.

Figure 6: Rotation of B Diagram

Figure 3: Damper Diagram
The next topic which is under modeling in the frequency
Next for modeling in the frequency domain is rotational domain is also rotational mechanical systems, but this time
mechanical systems (figure 4), these systems move in a fixed with gears. If one thinks about it, it is not common that we see
axis and are made up of three elements, which are the moment or know of mechanical systems that have no gear trains in
of inertia, torsional spring, and dashpot. It should also be them. It should be noted that gears are responsible in allowing
noted that the application of torque to a rotational mechanical the trade-off of speed for torque. For practice problems and
system is opposed by opposing torque which is caused by the for the basics of gears, it is assumed that gears fit perfectly,
moment of inertia, elasticity, and friction of the system. but this is not what really happens since gears exhibit
Similar to the algebraic sum in translational mechanical backlash, this means that one of the gears will move through a
systems, the torque applied in rotational mechanical systems small angle at a time before it even meets or connects with the
and the opposing torque are in opposing directions, thus the teeth of the other gears. The image below shows two gears
torques acting on the system is zero. with their teeth connected to one another [9].

The first in rotational mechanical systems is the moment of This next topic is about modeling of electromechanical
inertia. It is similar to mass in translational mechanical systems, this system is made up of an electrical subsystem as
systems in a sense that it is the one which stores kinetic well as a mechanical subsystem, thus the name
energy. It should be noted that when a torque is applied on a electromechanical system. Some previous concepts like mass,
body with a moment of inertia, an opposing torque is caused elasticity, and damping are also related to electromechanical
by the moment of inertia in the body itself and that opposing systems. There are cases for some devices like motors and
torque is proportional with respect to the angular acceleration speakers where the mass is driven with the use of a force that
of the body. is generated by the electrical subsystem and there are also
times when in some devices like microphones, the motion of
the mass is the one which generates either a voltage or current
in the electrical subsystem. One of the most important
concepts in electromechanical systems is an electric motor.
There are various types of electric motors, however, the two
main categories that are considered in most systems are direct
Figure 4: Inertia Diagram current motors (DC) and alternating current motors (AC) and
under the DC motor category, the two considered are
The next is torsional spring (figure 5), again similar to the armature-controlled motor and field-controlled motor [10].
spring in translational mechanical systems, it is the one which
stores the potential energy. When a torque is applied to a The last topic to be discussed in this part is the concept of
torsional spring, an opposing torque which is caused by the poles and zeros. Poles and zeros are used in analyzing the
elasticity of torsional spring opposes it giving the torsional performance of a system as well as to check the stability of it.
spring a relationship of being proportional to the angular It may also be defined as it is the one responsible for
displacement of the torsional spring. controlling the system whether it will work or not. It should be
noted that most of the time, the numbers of poles and zeros are
equal with one another in a given system, but there are also
cases when the number of poles is greater than the number of
zeros.

Poles may be defined as the roots of the denominator of a

given transfer function. In the equation shown below, poles
are the roots of D(s) and may be evaluated by taking D(s) = 0
Figure 5: Rotation of K Diagram and s are solved. In obtaining the response for a system, the
location of poles is analyzed with the real and imaginary value
1429
Aaron Don M. Africa et al., International Journal of Advanced Trends in Computer Science and Engineering, 8(4), July- August 2019, 1427 - 1432

parts of each of the poles. The real part is the one which gives
the exponential, while the imaginary part is the one which
gives the sinusoidal values.
The output equation can then be defined as

Zeros may be defined as the roots of the nominator of a

transfer function. In the same equation that was given for and in order to obtain the transfer function, it is required to
poles, the zeros may be obtained through taking N(s) = 0 and s perform a Laplace transform which then yields
are solved. As mentioned earlier, zeros are less or equal to the
number of poles. Zeros also show that the output at those
given frequencies are zero [11].

When poles and zeros have been found, plotting them is

Shown below (figure 9) is the block diagram for the RLC
possible. Poles and zeros are plotted in the S-Plane (figure 7),
series circuit.
it is a complex plane that contains a real and an imaginary axis
which refers to the complex-valued variable z. When mapping
poles in the plane, it is denoted by “x”, while an “o” for zeros.
The figure below shows a sample of plotting the poles and
zeros [12].

Figure 9: RLC Series Circuit

For the parallel RLC circuit (figure 10), the equation is shown
by

Figure 7: The S plane

4. DESIGN CONSIDERATION

The series RLC circuit (figure 8) shown below can be used to Figure 10: RLC Parallel Circuit
analyze and understand the performance of the control system
[13].

and the output can be represented through the capacitor and

when combined will yield

The Laplace transform is then given by [14]

Figure 8: RLC Circuit

and the resulting transfer function will be
From the figure, the input of the system is given by

1430
Aaron Don M. Africa et al., International Journal of Advanced Trends in Computer Science and Engineering, 8(4), July- August 2019, 1427 - 1432

[5] G. Bilal, J. M. Villanueva-Ramirez, and P. Gomez,

The block diagram for the parallel RLC circuit is seen below "Frequency domain transformer model for
(figure 11) [14]. electromagnetic transient analysis of networks," 2017
North American Power Symposium (NAPS). pp. 1-5,
2017.
https://doi.org/10.1109/NAPS.2017.8107183
[6] P. V. Gocheva, N. L. Hinov and V. P. Gochev, "Modeling
of Electronic Circuits with Generalized Nets," 2018 IX
National Conference with International Participation
(ELECTRONICA). pp. 1-4, 2018.
https://doi.org/10.1109/ELECTRONICA.2018.8439168
[7] P. D. Love, J. Zhou, and J. Matthews, “Project controls
for electrical, instrumentation and control systems:
Enabling role of digital system information modelling,”
Automation in Construction, Vol. 103, pp. 202-212,
2019.
Figure 11: Parallel RLC Block Diagram
https://doi.org/10.1016/j.autcon.2019.03.010
[8] K.A. Lazopoulos, D. Karaoulanis, and Α.Κ. Lazopoulos,
5. CONCLUSION
“On fractional modelling of viscoelastic mechanical
Modeling in the frequency domain can help in analyzing the systems,” Mechanics Research Communications, Vol.
behavior of a certain system. For example, in a study entitled 98, pp. 54-56, 2019.
Grey Box Identification of Fractional-order System Models https://doi.org/10.1016/j.mechrescom.2017.08.006
from Frequency Domain Data by Aleksei Tepljakov, Eduard [9] W. Liang, X. Yang, and J. Yao, “Rotation effects of force
Petlenkov, and Juri Belikov in 2018, it was used in presenting transducer on the output of the build-up system,”
and processing the data using different algorithms. The Measurement, Vol. 138. pp. 659-671, 2019.
System can be programmed using the database pattern of https://doi.org/10.1016/j.measurement.2019.01.071
[10] A. Kefal, C. Maruccio, G. Quaranta, and E. Oterkus,
[15,16]. The researches from [17,18,19] can be used as a
“Modelling and parameter identification of
pattern to make the program codes. Using the frequency
domain data, is the most feasible approach for the type of electromechanical systems for energy harvesting and
research that they are conducting. For the researchers, it was sensing,” Mechanical Systems and Signal Processing,
considerably more economical compared to other approaches. Vol. 121, pp. 890-912, 2019.
The same case can be said for the other research. Modeling in https://doi.org/10.1016/j.ymssp.2018.10.042
the Frequency Domain not only provides a more detailed [11] M. Dorosti, R.H.B. Fey, M.F. Heertjes, M.M.J. van de
approach in understanding the system but also proposes a Wal, and H. Nijmeijer, “Iterative pole-zero finite element
mathematical approach to carefully interpret the system. model updating using generic parameters,”
Mechatronics, Vol. 55, pp. 180-193, 2018.
https://doi.org/10.1016/j.mechatronics.2018.06.012
[12] G. Blanco, “Poles of the complex zeta function of a plane
REFERENCES
curve,” Advances in Mathematics, Vol. 350, pp. 396-439,
2019.
[1] Y. Yoon, Z. Sun, and H. Du, “Inverse modeling https://doi.org/10.1016/j.aim.2019.04.048
approach for parametric frequency domain analysis of an [13] A. Yazdani, and R. Iravani, “Voltage–Sourced
electrohydraulic system,” Mechanical Systems and Converters in Power Systems–Modeling,” Control and
Signal Processing. Vol. 121, pp. 412-425, 2019. Applications, 2010.
https://doi.org/10.1016/j.ymssp.2018.11.041 https://doi.org/10.1002/9780470551578
[2] S. Zhang, H. Wang, J. Gao, and C. Xing, “Frequency [14] A. Ramirez, M. Abdel-Rahaman, and T. Noda,
domain point cloud registration based on the Fourier “Frequency-domain modeling of time-periodic switched
transform,” Journal of Visual Communication and Image electrical networks: A review,” Ain Shams Engineering
Representation, Vol. 61, pp. 170-177, 2019. Journal, 2017.
https://doi.org/10.1016/j.jvcir.2019.03.005 https://doi.org/10.1016/j.asej.2017.07.002
[3] R. Bayma, Y. Zhu, and Z. Lang, “The analysis of [15] A. Africa, C. Alcantara, M. Lagula, A. Latina Jr. and C.
nonlinear systems in the frequency domain using Te, “Mobile phone graphical user interface (GUI) for
nonlinear output frequency response functions,” appliance remote control: An SMS-Based electronic
Automatica, Vol. 94, pp. 452-457, 2018 appliance monitoring and control system.” International
https://doi.org/10.1016/j.automatica.2018.04.030 Journal of Advanced Trends in Computer Science and
[4] A. Tepljakov, E. Petlenkov and J. Belikov, "Grey Box Engineering (IJATCSE). Vol. 8, No. 3, pp. 487-494,
Identification of Fractional-order System Models from 2019.
Frequency Domain Data," 2018 41st International https://doi.org/10.30534/ijatcse/2019/23832019
Conference on Telecommunications and Signal [16] A. Africa, and C. Uy, “Development of a cost-efficient
Processing (TSP). pp. 1-4, 2018. waste bin management system with mobile monitoring
https://doi.org/10.1109/TSP.2018.8441247
1431
Aaron Don M. Africa et al., International Journal of Advanced Trends in Computer Science and Engineering, 8(4), July- August 2019, 1427 - 1432

and tracking.” International Journal of Advanced Trends

in Computer Science and Engineering (IJATCSE). Vol.
8, No. 2, pp. 319-327, 2019.
https://doi.org/10.30534/ijatcse/2019/35822019
[17] A. Africa, “A Rough Set Based Solar Powered Flood
Water Purification System with a Fuzzy Logic Model.”
ARPN Journal of Engineering and Applied Sciences.
Vol. 12, No. 3, pp.638-647, 2017.
[18] A. Africa and J. Velasco, “Development of a Urine Strip
Analyzer using Artificial Neural Network using an
Android Phone.” ARPN Journal of Engineering and
Applied Sciences. Vol. 12, No. 6, pp. 1706-1712, 2017.
[19] A. Africa, S. Bautista, F. Lardizabal, J. Patron, and A.
Santos, “Minimizing Passenger Congestion in Train
Stations through Radio Frequency Identification (RFID)
coupled with Database Monitoring System.” ARPN
Journal of Engineering and Applied Sciences. Vol. 12,
No. 9, pp. 2863-2869, 2017.

1432