See discussions, stats, and author profiles for this publication at: [Link]

net/publication/361876334

System Identification for a Mathematical Model of DC Motor System

Conference Paper · June 2022

DOI: 10.1109/I2CACIS54679.2022.9815461

CITATIONS READS

3 764

2 authors:

Nur Naajihah Ab Rahman Nafrizuan Mat Yahya

5 PUBLICATIONS 8 CITATIONS
Universiti Malaysia Pahang
35 PUBLICATIONS 190 CITATIONS
SEE PROFILE
SEE PROFILE

All content following this page was uploaded by Nafrizuan Mat Yahya on 08 March 2023.

The user has requested enhancement of the downloaded file.

 System Identification for a Mathematical Model of
DC Motor System
Nur Naajihah Ab Rahman Nafrizuan Mat Yahya
Faculty of Manufacturing and Mechatronic Engineering Faculty of Manufacturing and Mechatronic Engineering
Technology Technology
Universiti Malaysia Pahang, 26600 Pekan Pahang, Malaysia
Universiti Malaysia Pahang, 26600 Pekan Pahang, Malaysia
nafrizuanmy@[Link]
nurnaajihahabrahman@[Link]

Abstract— This project compares the performance of a DC may be employed to produce an outstanding output response.
motor PI-PD controller with and without a pole-zero system The MATLAB/Simulink program has a toolbox called
identification method. Cohen-Coon tuning is used to fine tune system identification that may be used for this purpose.
the proposed controller's parameters. In addition, a small
number of tests were performed with varying pole and zero According to the authors of several papers in [10-11],
values. The DC motor's performance with a PI-PD controller simplified, system identification (SI) refers to the science of
was evaluated throughout these tests for transient response creating an accurate mathematical model of a system's core
characteristics. At the end of the study, it can be concluded that properties, which may then be used to numerous technical
system identification is required in order to enhance the domains. For SI to be successful, it must go through several
performance of the DC motor. By that, the DC motor with the stages, such as estimating the time or frequency domain
PI-PD controller performed better than other variants, with signals of the system's input-output signals; developing a
0.067s rising time, 1.748s settling time, 1.170s peak time, and competitor model; making and applying decisions and
42.143 percent overshoot with no steady-state error, despite strategies; and finally, approving the assessed model and
having only two poles and no zeros in the proposed controller. assessing its suitability for application needs, all of which
should be done with a different set of dummies [12].
Keywords— DC motor system; mathematical model;
MATLAB/Simulink; poles and zeros; system identification It's feasible to identify a dynamic system model using
input-output data from a DC motor controller, even though
I. INTRODUCTION most systems are sophisticated and their models cannot be
easily derived from physical laws, thanks to this system
Electrical motors are responsible for the vast bulk of the identification [13]. According to Rahim et al. [14], the use of
mechanical growth we see around us. It is a mechanism that system identification has grown beyond mathematics and
converts electrical energy into mechanical energy [1]. It is control system theory to include biological, engineering,
used to power a wide range of equipment in companies, financial, and operations research. Some research has used
businesses, the public sector, and homes, including wind system identification to obtain the mathematical model of a
turbines, compressors, and machine tools [2-3]. DC motors system, particularly one that involves a DC motor. According
are a type of electric motor that may be found in a number of to Naung et al., [15], the system identification toolbox is
control systems, such as in domestic electrical systems, appropriate for defining dynamic systems and provides
automobiles, trains, and process control systems [4]. With a specific black-box model architectures, both linear and
wide variety of control options, it is suitable for low-torque nonlinear.
applications and provides the flexibility to alter the pace or
speed as well as it has a fair price at the same easy Aside from that, there are other approaches for identifying
maintenance and management [5-6]. a system. Referring to [16], for example, conducted a study
on the speed management of a DC motor in which the authors
Due to the benefits, DC motors have also been used in a used the artificial neural networks (ANNs) model to identify
variety of applications such as conveyors, turntables, electric the system. This SI procedure has been completed in order to
traction, cranes, and other systems that required variable build a competent controller capable of tracking a needed
speed as well as continuous or low-speed torque or accuracy trajectory.
in a position [7].
Furthermore, Riba et al., [17] published research on
According to Chotai and Narwekar [8], linear control parameter identification onto DC-DC converters in steady-
schemes including PID, pole placement approach, Linear state and transient situations using white-box models. The
Quadratic Regulator (LQR), incremental motion control, and research uses DC-DC buck converters as well as boost
so on are frequently employed in the control of DC motors. switch-mode power consumption (SMPC) for parametric
Then, for a system using DC motors, a mathematical model identification, which was produced using experimental data
with the proper parameters is critical. They must, however, and a white-box model. Experimental and simulation data
be meticulously designed, tested, and deployed. Because the using white-box models with known parameter values show
mathematical model cannot provide accurate behavior if the that the suggested technique is feasible and accurate [17].
parameters are incorrect [4,9]. But if the DC motor There is some study on system identification or parameter
specifications are wrong, this model may not be the best identification, which is used in determining the parameters of
mathematical model for the system to overcome. Thus, a new a model of a specific system to use experimental data to
approach for estimating the transfer function of the DC motor

XXX-X-XXXX-XXXX-X/XX/$XX.00 ©20XX IEEE

replicate its dynamic behavior, and the same is true for this The rotor moment of inertia 𝐽𝑚 , the frictional coefficient
paper. 𝐵𝑚 , and the motor's torque 𝑇𝑚 , are all specified in the
equation above.
The major goal of this article is to implement a system
identification onto a created mathematical model obtained (4)
from the DC motor circuit to have an appropriate
mathematical model and a better DC motor response. Then, in The torque of the motor is only proportional to the
terms of transient reaction characteristics, a mathematical armature current, 𝐼𝑎 by a constant ratio, 𝐾𝑡
model derived from system identification in the presence of a The substitution approach will be applied to (1) and (2)
controller will be evaluated. It is possible to get the best generated from the preceding (3). By adding (2) into (1), we
performance out of a DC motor by modelling it obtain a new equation, (5).
mathematically.

II. METHODOLOGY (5)

A. Mathematical Model of DC Motor Then, substitute (4) into (3) which will create (6).
Mathematical modelling will be done in general with the
help of the DC motor. As previously stated, the DC motor (6)
system is divided into two parts: electrical and mechanical.
The DC motor circuit is depicted in the diagram below. Both (5) and (6) may be expressed using the Laplace
transform.
(7)
(8)
Next, substitute (8) into (7), then the transfer function
from the voltage source, to the output angle, ,
directly follows

(9)

Fig. 1. DC motor circuit When the DC motor parameters listed below are assumed,
the transfer function, (10) may be represented as follows:
Kirchoff's voltage law (KVL) may be used to represent
the electrical part based on Fig. 1, which comprises a voltage
source, armature resistance, armature inductance, armature
current, and the back emf voltage. The equation for the
electrical portion can be derived using these parameters and
the KVL as shown below.

(1)
(10)
Where,
The aforementioned (10) will be used to check the DC
= armature current (A) motor's performance. It will produce an output response after
= back emf voltage (V) running a simulation in the MATLAB/Simulink software.
= voltage source (V) Varied parameter input values will result in different system
= armature resistance (Ω) responses, such as a stable or unstable system. Simply put,
= armature inductance (H) the system response will be influenced by a mathematical
model derived from the DC motor circuit.
As a result, the back emf, 𝐸𝑏 which is directly
proportional to the shaft's angular velocity, 𝜃 by a constant
factor, 𝐾𝑏 may be stated as follows. B. System Identification
An exact mathematical model, or the best to fit with the
(2) DC motor, is required to have a large output response of the
DC motor, especially with the presence of a controller. That
The following may be deduced from the mechanical part being said, system identification is a key component that can
of electric motor: assist a system in producing a steady and superior output
response. As a result, mathematical models or transfer
functions produced from the DC motor circuit, (10) must deal
(3) with system identification.
In (10), a mathematical model will be connected with a
controller, the PI-PD controller. Tuning settings of the
mentioned controller using Cohen-Coon technique. The steps
for system identification will begin with measuring the output
signals when the system is loaded with an input signal, which
may be accomplished by using the block diagram of the DC
motor with PI-PD controller in MATLAB/Simulink software.

a) Input

Fig. 2. Block diagram of DC motor with PI-PD controller for system b) Output
identification
Fig. 4. Input and output signal
The input and output signals, as shown in Fig. 2., are then
exported to the MATLAB workspace. These export processes The identification method will produce varying accuracy
can be completed by using the Simulink software's 'To of model estimation based on these three different
Workspace' capabilities, and it is set as the input variable and assessments. The results of three distinct tests' system
the output variable. Then, as shown in Fig. 3., system identification methods will be reviewed in-depth in the
identification is recognised using the equation (10) while input following item.
and output signals are shown in Fig. 4. A variety of models,
such as transfer function models and nonlinear models, could C. Tuning method for 𝐾𝑝 , 𝐾𝑖 𝑎𝑛𝑑 𝐾𝑑
be used to anticipate the system model [18]. In this study, The 𝐾𝑝 , 𝐾𝑖 𝑎𝑛𝑑 𝐾𝑑 parameters of a controller can be
estimation in terms of the transfer function will be used.
tuned in a variety of ways. The most common tuning
strategies are the Ziegler-Nichols method, the Cohen-Coon
method, and bio-inspired computation as well as swarm
intelligence which is a type of artificial intelligence. The
Cohen-Coon approach will be used to compare the
performance of a DC motor and a suggested controller with
varied pole and zero values.
Cohen-coon method is characterized as an offline tuning
approach, which means that once the input has reached the
steady-state, step change can be added. The output can then
be monitored using the time constant as well as time delay,
and the response utilized to determine the initial control
parameters. Closed-loop systems can benefit from this tuning
method since the reaction time for closed-loop system is
Fig. 3. System identification faster [20].
As for Cohen-Coon method, three process characteristics
The input and output signals can be derived from the
have been applied, which are the process gain, dead time as
system identification procedure depicted in Fig. 3. while using
well as the time constant. These three characteristics can be
the time plot tools and the signal response depicted in Fig. 4.
assessed through the use of step test and analysis of the result.
below.
Fig. 5. shows how to find the process variable and the
Three distinct experiments were carried out in order to
output of the controller to get the three things listed above.
determine the optimal mathematical model that may be
As for the process gain, it can be calculated by dividing the
employed by the DC motor. These tests were carried out by
change in process variable (PV) with the change in controller
varying the values of poles and zeros. In terms of the poles, it
output (CO) in percentage form.
described as a value of s in the transfer function that equals
zero in the denominator, whereas a value of s that equals zero A tangent line needs to be developed from the maximum
in numerator is referred to as a zero of transfer function [19]. slope of the PV graph which will intersect with the original
Poles and zeros are frequencies when the denominator and level of PV (before the step change in CO). Therefore, the
numerator become zero. System stability and performance are dead time characteristics can be calculated. Dead time can be
analysed with poles and zeros.. defined as the time difference between the change in CO and
the intersection of the tangent line with the original PV level.
Referring to Fig. 3., there are three tests, the first of which
is a transfer function with poles two and zeros zero. The Next, after 𝑡𝑑 has been calculated, to have the value of the
second transfer function has two and one values for poles and time constant, 𝜏, 63% of the total change in PV needed to be
zeros, respectively, and the third transfer function has three computed. Then, the time needed for the PV to reach the level
and zero values for poles and zeros. can be found [21]. Lastly, by having these parameters, gains
for each of the controllers, which are 𝐾𝑝 , 𝐾𝑖 𝑎𝑛𝑑 𝐾𝑑 can be
developed as the Cohen-Coon tuning rules in Table I below.
 Static torque demand on the system is inversely
proportional to integral gain, Ki, at the same time. At the end
of the step, the value of Ki "pushes" the system to zero
positioning error, and it is called integral because it grows with
time. Following that comes 𝐾𝑑 , which depicts the system's
damping effects, which work to limit overshoot and
oscillations with 𝐾𝑝 [22]. However, each of the benefits will
have a different impact on the system's reaction. According to
Jenkins [23], 𝐾𝑝 will help reduce the rising time and steady-
state error while increasing the system's overshoot. 𝐾𝑖 's rise
time will be less than 𝐾𝑝 's, but its overshoot and settling time
will be greater. Last but not least, 𝐾𝑑 's rising time will be
Fig. 5. Step test for Cohen-Coon method slightly shifted in order to aid minimise overshoot and settling
time.
TABLE I. COHEN-COON TUNING RULES
The dynamic structure of a closed-loop step response
Cont Gain’s value system can only be affected by a few factors. 𝑇𝑟 , is the amount
roller 𝑲𝒑 𝑲𝒊 𝑲𝒅 of time it will take for the system reaction to grow from 10%
0.9 𝜏 𝜏 + 0.092𝑡𝑑 to 90% of its final value. There is a settling time known as 𝑇𝑠 ,
PI ( + 0.092) 3.33𝑡𝑑
𝑔𝑝 𝑡𝑑 𝜏 + 2.22𝑡𝑑 which is a reference to the time it takes for the system to
0.9 𝜏 𝜏 − 0.324𝑡𝑑
PD ( + 0.092) 0.27𝑡𝑑 stabilise. The peak time, 𝑇𝑝 , is the amount of time it takes for
𝑔𝑝 𝑡𝑑 𝜏 + 0.129𝑡𝑑
the output to reach its highest level. Because the highest point
D. Development of DC Motor with PI-PD controller is greater than the steady state, a percentage overshoot is
Using the above-mentioned Fig. 3., three alternative calibrated against this difference.
transfer functions provide varying degrees of precision that Finally, there is the steady-state error, 𝑒𝑠𝑠 , which is defined
match the estimation data. These accuracy results will be as the difference between the initial step value and the final
listed in Table II below, along with their mathematical model. value. In Fig. 6., block schematics of each mathematical
From the results shown in Table II, it can be seen that each model with PI-PD controller are shown.
of the mathematical models produced by estimating the value
of poles and zeros demonstrates that the identification is good
because it is nearly 100% accurate, resulting in a better
response for the DC motor with proposed controller
performance.
Then, in MATLAB/Simulink software, these
mathematical models, which comprise the original, generated
from the DC motor circuit, will be constructed with a PI-PD
controller. The parameters of proposed controller for each
mathematical model will be tweaked differently using a single
method, the Cohen-Coon tuning method. Having the desired
output or performance of the DC motor in the presence of a
suggested controller's settings is crucial.

TABLE II. MATHEMATICAL MODEL AND ACCURACY OF SYSTEM

IDENTIFICATION

System Identification
Mathematical model Accuracy Fig. 6. Block diagram of DC motor with PI-PD controller

1(P=2,
94.76% E. Performance Evaluation
Z=0)
2(P=2, Using the parameters listed above, an assessment will be
99.16% conducted to determine which of the four mathematical
Z=1)
3(P=3, models of DC motor with PI-PD controller performs best in
99.77% terms of accuracy and speed of reaction for the purpose of
Z=0)
positioning. Main characteristics that will be compared is 𝑇𝑟 ,
𝑇𝑠 , and 𝑒𝑠𝑠 while 𝑇𝑝 as well as the percentage overshoot is an
Proportional gain 𝐾𝑐 and 𝐾𝑝 , integral gain 𝐾𝑖 , and additional element. These transient response characteristics
derivative gain 𝐾𝑑 will be included in this controller. Each will be compared with existence studies regarding the DC
benefit has a function and affects how the system responds. motor also which are by Chong et al., [24] and Akpama et al.,
𝐾𝑝 and 𝐾𝑐 are system stiffness metrics that indicate the [25].
amount of restoring force required to correct for position
inaccuracy.
 III. RESULT AND DISCUSSION While other mathematical models took longer to achieve
and stay within 2% of steady-state value, 𝑇𝑠 , the first transfer
A. Simulink Result of System Identification function had a lower value for 𝑇𝑠 , which is the time necessary
The results achieved between four distinct mathematical to reach and stay within 2% of the steady-state value, 1.748s.
models of DC motor with PI-PD controller will be reviewed This occurred as a result of the simulation's usage of tiny
based on the simulation in the Simulink software about the values of derivative profits. The first transfer function's peak
system identification processes. The final value or step input time, 𝑇𝑝 , is 1.170s, whereas other mathematical models
used during the simulation is 5m. Fig. 7. depicts a combination required 1.166s, 1.172s, and 1.199s, respectively.
graph of mathematical models of DC motors with PI-PD
controllers produced from the identification procedure and TABLE III. TRANSIENT RESPONSE OF DC MOTOR WITH PI-PD
mathematical models of DC motors with PI-PD controllers CONTROLLER FOR EACH MATHEMATICAL MODEL
derived from the DC motor circuit and assumptions of DC
motor parameters. Transient response
Tr (s) Ts (s) Tp (s) %OS
As shown in the figure below, the performance of the DC 1 (P=2, Z=0) 0.067 1.748 1.170 0 42.143
motor with PI-PD controller, which has two poles and zero
2 (P=2, Z=1) 0.057 2.271 1.166 0 65.254
zeros, is good and stable. Although the accuracy is not as good
as in other mathematical models, this performance was 3 (P=3, Z=0) 0.078 2.471 1.172 0 65.833
accomplished due to the precision attained in the identification 4 (From DC
0.077 2.045 1.199 0 50.758
procedure with the given value of poles and zeros. When all motor circuit)
four different mathematical models are compared, the first Overshoot is defined as the system response that exceeds
transfer function delivers a faster response than the others. its steady-state value or final value. The first transfer function
The next topic will go over the evaluation in terms of of a DC motor with a PI-PD controller shows a small value of
transient response characteristics for each of the mathematical overshoot when compared to the others, which are 42.14%,
models of DC motors with PI-PD controllers. 65.25%, 65.83%, and 50.76%, respectively. The value of
overshoot was influenced by the suggested controller's gains
parameter, which is proportional and integral gains.
Overshoot will be greater if the value of proportional and
integral gains is greater.
The final characteristic that has been assessed is the a
system's steady-state error, which is the difference between
the expected and actual values, is zero for all previously
defined transfer functions. This occurred as a result of the
identification process's precision as well as the construction of
a mathematical model of a DC motor from the DC motor
circuit itself.
It is reasonable to deduce that an accurate mathematical
Fig. 7. Simulation result of the combined block diagram of DC motor with model of a DC motor is crucial for producing a reliable
PI-PD controller response, even if the response is still delayed, from the
Simulink simulation graph. The DC motor mathematical
B. Transient Response of DC Motor with PI-PD model with PI-PD controller obtained from the identification
controller process with two poles and no zero outperforms performance
because it required a short time from an initial value of 10%
The step response graph in Fig. 7. was used to calculate to hit 90% steady-state response, to reach and remain in 2%
the transient response for each mathematical model of a DC steady-state response, a small value of overshoot compared to
motor with a PI-PD controller. PI-PD controllers and DC other DC motor mathematical models.
motors were compared in depth in this investigation. A PI-PD
controller is used to measure the DC motor's transient
response, which is shown in Table III. IV. CONCLUSION
As shown in Fig. 7, the step response Simulink simulation System identification based on the poles and zeros idea has
of combination graph shows that the mathematical model of been established using several approaches such as Kirchhoff's
DC motor derived from the circuit of DC motor and voltage law, Laplace transform, and other mathematical
identification process with two poles and no zeros is the most methods applied to the mathematical model formed from the
stable output response compared to the other mathematical DC motor circuit. An precise mathematical model of a DC
models. motor and the usage of a controller, a PI-PD controller tuned
using the Cohen-Coon tuning approach, are required for
As shown in Table III, the second mathematical model has system identification in order to increase the performance of
the shortest rise time, 𝑇𝑟 , which is 0.057s, whereas the first, the DC motor.
third, and fourth mathematical models with PI-PD controller
required 0.067s, 0.078s, and 0.077s, respectively, for the According to the simulation results with varied values of
system response to hit from 10% to 90% of steady-state poles and zeros, the DC motor with PI-PD controller, which
response. Because of the high value of proportional and has two poles and zero zeros, is in a good and more stable
integral gains, the second transfer with two poles and one zero state. In terms of transient response, a detailed comparison of
has the shortest rising time. four alternative mathematical models of DC motor with PI-
PD controller was made. For the increased time, it takes
0.067s to go from 10% to 90% of the final and 1.748s to stay [11] A. Kefal, C. Maruccio, G. Quaranta, and E. Oterkus, “Modelling
in and maintain 2% of the steady-state, which is the smallest and parameter identification of electromechanical systems for
time compared to the others. It took 1.170s to reach the initial energy harvesting and sensing,” Mech. Syst. Signal Process., vol.
high point of the overshoot during the peak time. The 121, pp. 890–912, 2019.
mathematical model with two poles and no zero value has won
[12] A. A. M. Zahir, S. S. N. Alhady, W. A. F. W. Othman, and M. F.
a lower value of the overshoot, and there is no steady-state
Ahmad, “Genetic Algorithm Optimization of PID Controller for
error as it returns to the system's ultimate value.
Brushed DC Motor,” Intell. Manuf. Mechatronics, vol. 10, pp.
To summarise, the important point for having an efficient 427–437, 2018.
output response for a DC motor is a specific derivation for the [13] T. Sands, “Nonlinear-adaptive mathematical system
mathematical model and some identification method to identification,” Computation, vol. 5, no. 4, p. 47, 2017.
determine the optimal transfer function to be implemented.
[14] N. A. Rahim, M. N. Taib, and M. I. Yusof, “Nonlinear system
This is due to the fact that an excellent reaction can be
identification for a Dc motor using NARMAX approach,” in Asian
obtained and is influenced by a flawless mathematical model,
particularly when a controller is used. Conference on Sensors, 2003. AsiaSense 2003., 2003, pp. 305–
311.
[15] Y. Naung, A. Schagin, H. L. Oo, K. Z. Ye, and Z. M. Khaing,
ACKNOWLEDGMENT
“Implementation of data driven control system of direct current
The authors would like to thank Universiti Malaysia motor by using system identification process,” in 2018 IEEE
Pahang for providing financial support under UMP Internal
Conference of Russian Young Researchers in Electrical and
Grant (Grant No. RDU210370) and Faculty of Manufacturing
Electronic Engineering (ElConRus), 2018, pp. 1801–1804.
and Mechatronic Engineering Technology, Universiti
Malaysia Pahang for laboratory facilities. [16] D. L. Mute, K. Chaudhari, R. Khamari, and A. Singare, “System
Identification Using Neural Network Model for Speed Control of
Direct Current Motor,” Int. Res. J. Eng. Technol., vol. 02, no. 03,
REFERENCES
pp. 1316–1320, 2015.
[1] N. N. A. Rahman and N. M. Yahya, “A mathematical model of a [17] J.-R. Riba, M. Moreno-Eguilaz, S. Bogarra, and A. Garcia,
brushed DC motor system,” Data Anal. Appl. Math., vol. 2, no. 2, “Parameter Identification of DC-DC Converters under Steady-
pp. 60–68, 2021. State and Transient Conditions Based on White-Box Models,”
[2] M. Kushwah and P. A. Patra, “Tuning of PID Controller for Speed Electronics, vol. 7, no. 12, p. 393, Dec. 2018.
Control of DC Motor Using Soft Computing Techniques - A [18] W. J. Tang, Z. T. Liu, and Q. Wang, “DC motor speed control
Review,” Adv. Electron. Electr. Eng., vol. 4, no. 2, pp. 141–148, based on system identification and PID auto tuning,” 36th Chinese
2014. Control Conf. CCC, pp. 6420–6423, 2017.
[3] R. K. Achanta and V. K. Pamula, “DC Motor Speed Control using [19] K. Craig, “The significance of poles and zeros,” 2012.
PID Controller Tuned by Jaya Optimization Algorithm,” 2017 [Link]
IEEE Int. Conf. Power, Control. Signals Instrum. Eng., pp. 983– (accessed Aug. 08, 2021).
987, 2017. [20] P. Woolf et al, “PID Tuning via Classical Methods,” Engineering
[4] A. A. A. Emhemed and R. Bin Mamat, “Modelling and simulation LibreTexts, 2021.
for Industrial Direct Current Motor using Intelligent control,” [Link]
Procedia Eng., vol. 41, no. Iris, pp. 420–425, 2012. ngineering/Book%3A_Chemical_Process_Dynamics_and_Contr
[5] CRBTech, “Advantages and Disadvantages Of Prototype,” 2016. ols_(Woolf)/09%3A_Proportional-Integral-
[Link] Derivative_(PID)_Control/9.03%3A_PID_Tuning_via_Classical
prototype/ (accessed Apr. 04, 2021). _Methods (accessed Jun. 24, 2021).
[6] D. A. Barkas, G. C. Ioannidis, C. S. Psomopoulos, S. D. [21] J. Smuts, “Cohen Coon Tuning Rules,” Reflection of a Process
Kaminaris, and G. A. Vokas, “Brushed DC Motor Drives for Control Practitioner, 2011.
Industrial and Automobile Applications with Emphasis on Control [Link] (accessed May 16,
Techniques: A Comprehensive Review,” Electronics, vol. 9, no. 6, 2022).
p. 887, May 2020. [22] D. Collins, “What are PID gains and feed-forward gains?,” 2016.
[7] J. Kimbrell, “The Direct Current Motor Advantage,” 2016. [Link]
[Link] feed-forward-gains/ (accessed Sep. 17, 2021).
drives/article/15586862/the-dc-motor-advantage (accessed Apr. [23] Jenkins, “Tuning for PID Controllers,” 2014.
04, 2021). [Link]
[8] J. Chotai and K. Narwekar, “Modelling and Position Control of [Link] (accessed Mar. 24, 2021).
Brushed DC Motor,” 2017 Int. Conf. Adv. Comput. Commun. [24] S. H. Chong, T. Tze Ter, and V. Sakthivelu, “Positioning Control
Control, pp. 1–5, 2017. of Ball Screw System Driven by DC Motor,” Appl. Mech. Mater.,
[9] J. N. Rai, “Speed control of DC motor using Fuzzy Logic vol. 761, pp. 142–147, 2015.
Controller,” IOSR J. Electr. Electron. Eng., vol. 3, no. 6, pp. 41– [25] E. Akpama, E. Effiong, and R. Ezenwosu, “Simulink Design of a
48, 2012. DC Motor Control for Water Pump using Fuzzy Logic,” in 2nd
[10] G. F. Sirca Jr and H. Adeli, “System identification in structural International Conference on Electrical Power Engineering
engineering,” Sci. Iran., vol. 19, no. 6, pp. 1355–1364, 2012. (ICEPENG 2021), 2021, no. July, pp. 16–19.

View publication stats