DATA ANALYTICS AND APPLIED MATHEMATICS (DAAM)

e-ISSN: 2773-4854
VOL. 2, ISSUE 2, 060 – 068
DOI: https://doi.org/10.15282/daam.v2i2.6830

ORIGINAL ARTICLE

A mathematical model of a brushed DC motor system

N. N. A. Rahman1,*, N. M. Yahya1
1Faculty of Manufacturing and Mechatronic Engineering Technology, College of Engineering Technology, 26600 Pekan, Pahang, Malaysia.

ABSTRACT – Mathematical model has been proposed for some system that involves a ARTICLE HISTORY
brushed DC motor and it is widely used in industry. Brushed DC motor ideals for applications with Received: 21/08/2021
a low- torque, manage to change pace or speed and it is widely used in many applications such Revised: 24/11/2021
as x-y table positioning system, conveyor systems and other system that required to use the Accepted: 02/12/2021
features that brushed DC motor have. Mathematical model of brushed DC motor in order to verify
the performance of the DC motor. In this paper, mathematical model of brushed DC motor will be KEYWORDS
derived from a brushed DC motor circuit that consist of two parts that are electrical and mechanical DC motor
part. To validate the functionality of mathematical model, the performance of the brushed DC motor PI-PD controller
Transient response
without any controller will be compared with the brushed DC motor with the presence of PI-PD
Conveyor system
controller that will be tuned by trial-and-error method. Performances of both brushed DC motor with Trial-and-error method
and without controller will be compared in terms of transient response which are, rise time, Tr,
settling time, Ts, steady state error, ess and lastly percentage overshoot. At the end of the study,
the brushed DC motor with PI-PD controller show a better performance compared to the brushed
DC motor without any controller.

INTRODUCTION
An electrical motor is responsible for the majority of the mechanical growth we see around us. Electrical motor is a
mechanism that will transforms an electrical energy into mechanical energy. There are two categories of motors known
as AC motors, operate alternating current and brushed DC motors, operate on direct current. Based on a study carried by
[1], brushed DC motors are used in a variety of control systems, including home electrical systems, cars, trains, as well as
process control. The mathematical model is well known to be critical in the design of a control system. Four differenttypes
of DC motor have been mentioned by [2] which are permanent magnet motor, DC series motor, DC shunt motor and
compound motor. Then, there are few models of DC motor that accurately reflect to the system’s action.
For a number of purposes, brushed DC motors are often preferred over AC motors. Brushed DC motors are ideal for
low-torque applications, capable to change the pace or speed and it have a wide variety of speed controls options,
including both below and above the rated speed. Other than that, it also has a large and powerful staring torque, the price of
the brushed DC motors are also more affordable and last advantages that provide by brushed DC motors is that
maintenance of brushed DC motors is simple and takes little or no time [3]. Brushed DC motors have been applied for
many applications such as conveyors, turntables and other system that required for variable speed and constant or low-
speed torque or accuracy in position [4].
In a system that involve the usage of brushed DC motors, mathematical model with the correct parameters of the model
is important because an accurate behavior cannot be provided by the mathematical model if the parameters are not correct
[1]. Mathematical model of brushed DC motor can be derived from a brushed DC motor circuit that consist of two parts
that is mechanical part that have a rotor while electrical part has input voltage, resistor and also inductor.
Basically, the main goal of this paper is to generate a correct mathematical model of brushed DC motor which will be
used in many systems. Then, to validate the functionality of brushed DC motor mathematical model, PI-PD controller will
be implemented in order to compare the performance of the brushed DC motor with PI-PD controller and DC motor
without using any controller. Mathematical model of brushed DC motor will be analyzed and derived again if the
performance of the brushed DC motor with PI-PD controller does not excellent. A great mathematical model of brushed
DC motor will provide the best performance especially with an existence of the controller.

METHODOLOGY
Mathematical model of brushed DC motor
In general, the brushed DC motor circuit will be used to generate the mathematical modelling. As explain before,
brushed DC motor circuit consist of two part which are electrical part and mechanical part. Below is the figure of the
brushed DC motor circuit.

*CORRESPONDING AUTHOR | N. N. A. Rahman |  [email protected] 60

© The Authors 2019. Published by Universiti Malaysia Pahang Publishing. This is an open access article under the CC BY-NC-ND license.
 Rahman et al. │ Data Analytics and Applied Mathematics │ Vol. 2, Issue 2 (2021)

Figure 1. Brushed DC motor circuit.

Based on the Figure 1 above, to develop a mathematical modelling, Kirchoff’s voltage law (KVL) can be applied for the
electrical part which consist of voltage source, armature resistance, armature inductance, armature current as well as the
back emf voltage.

With these parameters and the KVL, the equation for the electrical part can be generated as below.

dI a
Ea = Ra ⋅ I a + La + Eb (1)
dt

where

𝐼𝐼𝑎𝑎 = armature current (A)

𝐸𝐸𝑏𝑏 = back emf voltage (V)
𝐸𝐸𝑎𝑎 = voltage source (V)
𝑅𝑅𝑎𝑎 = armature resistance (Ω)
𝐿𝐿𝑎𝑎 = armature inductance (H)

Then, back emf which is 𝐸𝐸𝑏𝑏 is proportional to the angular velocity of the shaft, 𝜃𝜃 by a constant factor, 𝐾𝐾𝑏𝑏 that can be
expressed as below.
dθ
Eb = K b (2)
dt

From the mechanical part of DC motor circuit, it can be derived as

d 2θ dθ
Jm 2
+ Bm − Tm (3)
dt dt

From the equation above, 𝐽𝐽𝑚𝑚 can be defined as the rotor moment of inertia, 𝐵𝐵𝑚𝑚 is the frictional coefficient and 𝑇𝑇𝑚𝑚which
is the torque of motor.
T=
m Kt ⋅ I a (4)

For the torque of motor, it is proportional to only the armature current that is 𝐼𝐼𝑎𝑎 by a constant factor, 𝐾𝐾𝑡𝑡.

From equations that have been derived above, substitution method will be applied into equation 1 and equation 3. By
substitute equation 2 into equation 1, it will give a new equation which is equation 5.

61 journal.ump.edu.my/daam ◄
 Rahman et al. │ Data Analytics and Applied Mathematics │ Vol. 2, Issue 2 (2021)

dI a dθ
Ea = Ra ⋅ I a + La + Kb (5)
dt dt

Then, substitute equation 4 into equation 3 which will create the equation 6.

d 2θ dθ
Jm + Bm Kt I a
=⋅ (6)
dt dt

By implementing the Laplace transform, both equation 5 and equation 6 can be written as:
Ea ( s ) − K b sθ ( s ) =+
( Ra sLa ) I a ( s ) (7)

J m s 2θ ( s ) + Bm sθ ( s ) =
Kt I a ( s ) (8)

Next, substitute the equation 8 into equation 7, then the transfer function from the voltage source, 𝐸𝐸𝑎𝑎(𝑠𝑠) to the outputangle,
𝜃𝜃(𝑠𝑠), directly follows:

θ (s) Kt
= (9)
Ea ( s ) J m La s + ( J m Ra + Bm La ) s 2 + Ra Bm s + K t K b s
3

By assuming the value for each brushed DC motor parameters which have been stated below thus, the transfer
function, equation 10 of the brushed DC motor can be expressed as:

𝐽𝐽𝑚𝑚 = 0.093 𝐾𝐾𝐾𝐾. 𝑚𝑚2

𝐵𝐵𝑚𝑚 = 0.008 𝑁𝑁𝑁𝑁𝑁𝑁
𝐾𝐾𝑏𝑏 = 0.6 𝑉𝑉⁄𝑟𝑟𝑟𝑟𝑟𝑟 𝑠𝑠−1
𝐾𝐾𝑡𝑡 = 0.7274 𝑁𝑁𝑁𝑁⁄𝐴𝐴
𝑅𝑅𝑎𝑎 = 0.6 𝛺𝛺
𝐿𝐿𝑎𝑎 = 0.006 𝐻𝐻
θ (s) 0.7274
= (10)
Ea ( s ) 0.000558s 3 + 0.055848s 2 + 0.44124 s

Equation 10 above will be used in order to verify the performance of the brushed DC motor itself. It will give an
output response which by having a simulation in MATLAB/Simulink software. Different input values of parameters will
provide a different system response such as stable or unstable system. In a simple way, mathematical model that derive
from the brushed DC motor circuit will influence the system response.
PI-PD controller
PI-PD or proportional integral-proportional derivative controller is widely used other than PID controllers due to its
capability for disturbance rejection such as load disturbance, minimize the rise time, and settling time. The concept of this
controller is that the first controller, PI controller placed in forward path while the PD controller placed in the feedback
loop.

62 journal.ump.edu.my/daam ◄
 Rahman et al. │ Data Analytics and Applied Mathematics │ Vol. 2, Issue 2 (2021)

Figure 2. Block diagram of PI-PD controller.

Based on Figure 2 above, there will be three different types of gains that need to be implemented in the block diagram
which are two of proportional gain, Kp1, Kp2, integral gain, Ki and derivative gain, Kd. The overall of equations of PI-PD
that involve the parameters’ gain can be expressed as:

 dc ( t ) 
u (=
t ) K i ∫ e ( t ) dt + K c e ( t ) −  K p c ( t ) + K d  (11)
 dt 

The output of the equation, 𝑢𝑢(𝑡𝑡) which have been written above is fed into the plant that is the transfer function in
equation 10, yielding a new output, 𝑐𝑐(𝑡𝑡). The new output, 𝑐𝑐(𝑡𝑡) is then fedback to find the new error signal 𝑒𝑒(𝑡𝑡) by
comparing it to the reference. The controller computes an update to the control input based on this new error signal.
Tuning of PI-PD controller
In a controller, there are many ways to tune the controller's parameters, which is Kp, Ki, and Kd. For example, widely
used tuning techniques are the Ziegler-Nichols method, bio-inspired computation or known as swarm intelligence, subset
from artificial intelligence and also trial and error method. In this context, where it will compare a brushed DC motor
without a controller with a brushed DC motor with a PI-PD controller, the trial-and-error method will be implemented.
This step is important in order to validate the mathematical model of brushed DC motor that have been derived earlier is
correct.
Trial and error are a simple way to tune a controller parameter. Since little or no knowledge of the regulated process is
needed, this method is simple to use for process operators [5]. Gains’ parameter which are Kc, Kp, Ki and Kd can be tuned
by increase the proportional gain until the control loop’s output oscillates at a constant rate, then set the integral and
derivatives terms to zero. This proportional gain increase should be done in such a way that the system’s response time
improves while remaining stable.
Next, set the integral term once the P-response is quick enough to reduce the oscillations gradually. Adjust the I value
before the steady state error is reduced, but it may lead to the overshoot. Last step is to increase the derivative gain before
the device responds rapidly to its set point after the P and I parameters have been set to desired values with a minimal
steady state error. The overshoot of the controller response is minimized by increasing the derivative expression.
For tuning process, four different gains will be involved which are Kc, Kp, Ki and Kd. In a closed loop step response
system, there are few characteristics that will influence the dynamic structure. First will be the rise time, Tr, which is the
time it takes for the plant output to go from 10% to 90% of the final value. Then, times it will takes for the system to be
in its steady state, which is settling time, Ts.
Next characteristics is peak time, Tp where it is the time taken for the output to reach it maximum value. Furthermore,
percentage overshoot which can be explained as how much higher the maximum point than stable state is calibrated
against the stable state. Lastly, the steady state error, ess which can be defined as the difference between the input step
value with the final value.
Development of PI-PD controller for brushed DC motor
Development of the PI-PD controller involves the structure of block diagram, which can be created by using a
software, MATLAB/Simulink. The transfer function of the brushed DC motor, which had been derived and assumed, was
implemented in development process of the PI-PD controller. The presence of PI-PD controller with a brushed DC motor
can be used for comparing the differences with a brushed DC motor only without applying any controller concerning its
performance toward the system with reference to the characteristics which will manipulate the feedback.

63 journal.ump.edu.my/daam ◄
 Rahman et al. │ Data Analytics and Applied Mathematics │ Vol. 2, Issue 2 (2021)

Other than that, the types of controller’s parameter gains that will be allocated is important in order to have the desired
output or performance of the brushed DC motor with the presence of controller. Four gains will be implemented in this
controller, which are proportional gain, Kc and Kp, integral gain, Ki, and derivative gain, Kd. Each of the gains will have
their function and can affect the system response.
Kp and Kc is a measure of system stiffness, and it specifies the amount of restoring force to be applied to counter the
position error. At the same time, the static torque load on the system is related to integral gain, Ki. Value of Ki “pushes”
the system at the end of the step to zero positioning error, and it is referred to as integral due to it rises with the time at the
end of the step. Next is Kd that represents damping effects on the system, functioning to minimize overshoot and
oscillations with Kp [6].

Figure 3. Block diagram of PI-PD controller in MATLAB/Simulink.

However, each of the gains will influence the system response in different ways. Based on the research carried [7], it
stated that Kp will help in minimize the rise time and steady state error but increase the overshoot of the system. Ki will
minimize the rise time as Kp but increase the overshoot and settling time. Lastly, Kd will have a small change in rise time,it
can reduce overshoot and settling time.
Performance evaluation
To check the validity of the brushed DC motor that have been derived, comparison between brushed DC motor without
controller and brushed DC motor with the presence of the PI-PD controller will be carried out with the help of trial-and-
error method to optimize the controller’s gains parameter. Thus, evaluation that will be conducted based on some
characteristics that have been mentioned before in order to decide which brushed DC motor is best to use for accuracy
and fast response relating to the position context.
Characteristics that important to evaluate are rise time, Tr, settling time, Ts, peak time, Tp, the percentage overshoot for
each of the conditions of the motor, existence of controller or without the controller and lastly the steady state error, ess. A
good system response is a system which have lower value in Tr where it refers to the amount of time it takes for a signal
to shift from low value to high value and the percentages are usually between 10% and 90% of the step high or final value
of the system.
Settling time or Ts where indicate the time elapsed between the application of an ideal instantaneous step input and the
time where output of the system entered and stayed within a specified error band also should be lower. Furthermore, time
need by the system response to reach the first peak of the overshoot is one of the important characteristics that need to
focus on. Moreover, steady state error which the time it takes for the system response to reach back the final value and
lastly is the overshoot of the response when the system exceeds its target.

64 journal.ump.edu.my/daam ◄
 Rahman et al. │ Data Analytics and Applied Mathematics │ Vol. 2, Issue 2 (2021)

Example of rise time, Tr

Tr = 90% − 10%
Tr = 2.202 − 1.141
Tr = 1.061s

Figure 4. Example of result for brushed DC motor without controller from Simulink scope.

Based on Figure 4 above is an example from a system response of a brushed DC motor without using any controller.
To get the system response parameters, a toolbar of cursor measurements in Simulink simulation, it will portray two
sections of time. It will be used to get the Tr by undergoing some mathematical operation, which is a minus operation
between the value of the second time and from the first time. Secondly, for Ts, this method can be done using only one
value and time that portray is the time for settling time. It can be taken directly from the bilevel measurements toolbar for
percentage overshoot characteristics, or it can be calculated manually. However, there will be some calculations involved
to know the exact value that needs to be used before getting the value of time for each of the characteristics.

RESULTS AND DISCUSSION

Trial-and-Error tuning method
Gains’ parameters from trial-and-error method which are proportional integral, Kp, integral gain, Ki and derivative
gain, Kd can be tabulated as below. These parameters will be applied in the block diagram along with the transfer function
based on type of controller used. Table 1 below is the result of trial-and-error tuning method that have been obtained.

Table 1. Trial-and-error tuning method.

Gain’s value
Type of
controllers Kp Ki Kd
PI 6.0 5.0 -
PD 1.0 - 0.65

Simulink simulation result

Based on the simulation in the Simulink software, result that have been obtained between the combination of brushed DC
motor with PI-PD controller and brushed DC motor without controller will be discussed. Step input or the final value

65 journal.ump.edu.my/daam ◄
 Rahman et al. │ Data Analytics and Applied Mathematics │ Vol. 2, Issue 2 (2021)

that used during the simulation is 5m. Figure 5 below shows the combination graph between brushed DC motor with the
presence of PI-PD controller and brushed DC motor without any controller.
As we can see from the figure below, the performance of the brushed DC motor without using any controller are in
stable condition. The stable performance which portrays by brushed DC motor without any controller achieved due to the
precise derivation of the mathematical model that have been done in earlier stage before the simulation process. Then, to
enhance the efficiency of the brushed DC motor in terms of the time response, PI-PD controller have been implemented.
The performance of the brushed DC motor with PI-PD controller provides a fast response compared to the brushed DC
motor which does not apply any controller based on the graph below.
The comparison between both brushed DC motor can be evaluated in terms of the transient response characteristics
which are rise time, Tr, settling time, Ts, percentage overshoot and steady state error, ess. which will be discussed on the
next topic.

Figure 5. Combination graph of brushed DC motor without controller and brushed DC motor with PI-PD
controller.
Transient response of brushed DC motor
Transient response for each brushed DC motor with PI-PD controller and brushed DC motor without controller may
be calculated using the step response graph shown in Figure 5. These transients can be achieved by adding a step to the
combination of two brushed DC motors, both with and without controllers. Due to the correct derivation of the
mathematical model, the performances of brushed DC motor for both with and without controllers then will be compared
and discussed based on the transient response characteristics. Table 2 below is the result of the transient response of both
different brushed DC motor with and without controller.

Table 2. Transient response of brushed DC motor with and without controller.

Transient response
Type of controllers
Tr (s) Ts (s) % OS SSE, ess.
Brushed DC motor
1.061 3.054 0.503 0
without controller
Brushed DC motor with
0.307 1.591 0.504 0
PI-PD controller

66 journal.ump.edu.my/daam ◄
 Rahman et al. │ Data Analytics and Applied Mathematics │ Vol. 2, Issue 2 (2021)

From the step response Simulink simulation of combination graph in Figure 5, and transient response result for
different approach of brushed DC motor, it shows that the mathematical model of brushed DC motor that have been
derived from the DC motor circuit is correct since the performance of the brushed DC motor without using any controllersis
stable. Then to have a better and fast response, the brushed DC motor was combined with the PI-PD controller, thus
brushed DC motor with the presence of PI-PD controller have a better performances and fast response compared to the
brushed DC motor without any controller.
As shown in Table 2 above, brushed DC motor with PI-PD controller has the shortest rise time, Tr which are 0.307s
while time required by system response from initial value of 10% to hit 90% of steady state response, brushed DC motor
without using any controller has the longest time, 1.061s. Proposed brushed DC motor with controller has the shortest
time of Tr due to the high value of proportional and integral gain which help in reducing the Tr characteristics.
For settling time, Ts which act as the time needed to reach and stay 2% of steady state value, again brushed DC motor
with PI-PD controller required a short time to reach it compared to brushed DC motor without any controller which are
1.591s and 3.054s. This is due to the value of derivative gain that help to minimize the Ts characteristic.
Moreover, overshoot that defined as the system response that exceed its steady state value or final value, both brushed
DC motor have the smallest value. Proposed brushed DC motor with proposed controller have slightly 0.001% difference
than brushed DC motor that is not applied any controller, 0.504% and 0.503% and the system is still stable since the
overshoot had occurred is below than 10%. Last characteristic that was evaluated is the steady state error, ess, where the
time it takes for the system response to reach back the final value for both conditions of the brushed DC motor is zero.
This is due to the precise method that have been implemented in mathematical model derivation of the brushed DC motor
which contribute to the stable response for the both brushed DC motor with and without the PI-PD controller.
Therefore, after analyzing the graph of the Simulink simulation, it can be inferred that a correct mathematical model
for brushed DC motor is crucial in order for the brushed DC motor to have a stable response even though the response is
still in slow response. By consider the evaluation of brushed DC motor with PI-PD controller and brushed DC motor
without controller in transient response aspects, the brushed DC motor with PI-PD controller provide an outperform
performances as it required shortest time from initial value of 10% to hit 90% of steady state response, to reach and stayin
2% of steady state response, small value of overshoot which in the range of 10% that will make the system to be stableand
there is no steady state as the system is in stable condition.

CONCLUSION
A mathematical model of brushed DC motor has been developed based on the circuit of brushed DC motor which
consist of electrical and mechanical part by implementing some method such as Kirchoff’s voltage law, Laplace transform
and other mathematical methods.
Based on the mathematical model of brushed DC motor that have been generated, a simulation by using Simulink
software have been conducted to verify the functionality and accuracy of the mathematical model that have been derived.
The simulation involves two conditions of the brushed DC motor which are brushed DC motor with PI-PD controller and
brushed DC motor without any controller.
From the simulation that have been made, both conditions of the brushed DC motor portray a stable response. This
indicates that the mathematical model that have been derived is correct and accurate especially a constant response that
provided by the brushed DC motor without any controller. For the purpose of enhancing the performance of brushed DC
motor, PI-PD controller was implemented. To optimize the gains’ parameter of proposed controller, trial-and-error
method was used.
As for the brushed DC motor with PI-PD controller, it provides a better performance and fast response compared to
the brushed DC motor only. An evaluation has been made for both of these brushed DC motor in terms of the transient
response aspects. Brushed DC motor with proposed controller shows a faster rise time, Tr than brushed DC motor without
controller which is 0.307s. Then, the time taken for the response to stay in 2% of the steady state, Ts also shorter than
brushed DC motor without controller, 1.591s.
Thus, it can be concluded that the main point to have a good performance of brushed DC motor is a precise derivation
for mathematical model of brushed DC motor itself. A correct mathematical model will influence the response of the
brushed DC motor. Then, to enhance the performance in fast response aspects, controllers can be implemented.

ACKNOWLEDGEMENT
The authors would like to thank Universiti Malaysia Pahang for providing financial support under Post Graduate Research
Scheme (PGRS) (Grant No. PGRS210365) and Faculty of Manufacturing and Mechatronic Engineering Technology,
Universiti Malaysia Pahang for laboratory facilities.

REFERENCES
[1] A. A. A. Emhemed and R. Bin Mamat. “Modelling and simulation for Industrial DC Motor using Intelligent control,” Procedia
Eng., vol. 41, no. Iris, pp. 420–425, 2012.
[2] K. Shrivastava and P. M. D. Pawar. “A review on types of dc motors and the necessity of starter for its speed

67 journal.ump.edu.my/daam ◄
 Rahman et al. │ Data Analytics and Applied Mathematics │ Vol. 2, Issue 2 (2021)

regulation,” Int. J. Adv. Res. Comput. Commun. Eng., vol. 5, no. 4, pp. 61–63, 2016.
[3] CRBTech. (2016). Advantages and Disadvantages of Prototype [Online]. Available: http://crbtech.in/CAD-
CAMTraining/advantages-disadvantages-prototype/.
[4] J. Kimbrell. (2016). The DC Motor Advantage [Online]. Available: https://www.processingmagazine.com/pumps-motors-
drives/article/15586862/the-dc-motor-advantage.
[5] M. Steinbuch, R. Huisman, D. A. F. Trucks, and F. Veldpaus. (2004). Tuning of PID-Type Controllers: Literature Overview
[Online]. Available: https://pure.tue.nl/ws/files/4286492/625529.pdf.
[6] D. Collins. (2016). What are PID gains and feed-forward gains? [Online]. Available:
https://www.motioncontroltips.com/faq-what-are-pid-gains-and-feed-forward-gains/.
[7] Jenkins. (2014). Tuning for PID Controllers [Online]. Available:
http://faculty.mercer.edu/jenkins_he/documents/TuningforPIDControllers.pdf.

68 journal.ump.edu.my/daam ◄