?

Speed Control of DC Motor Using Fuzzy PID Controller

Nassim Messaadi ? , Abdelkader Amroun ? ,
Department of Instrumentation and Automation
University of Science and Technology Houari Boumediene
arXiv:2108.05450v1 [eess.SY] 11 Aug 2021

Abstract
In this project, we designed a DC motor whose speed can be controlled by a PID controller. The
proportional, integral and derivative gains (KP, KI, KD) of the PID controller are adjusted according
to Fuzzy logic.
First of all, the fuzzy logic controller is designed according to rules so that the systems is basically
robust. There are 25 rules for the auto-tuning of each parameter of the PID controller. The FLC
(fuzzy logic controller) has two inputs. The first is the motor speed error between the reference
(setpoint) and the actual speed. The second is the variation of the speed error (derivative of the
speed error). Secondly the output of the FLC is the parameters of the PID controller which are
used to control the speed of the DC motor.
The study shows that both the precise characters of PID controllers and the flexible characters of
fuzzy controllers are present in the fuzzy self-tuning PID controller. The fuzzy auto-tuning approach
implemented on a conventional PID structure was able to control the speed of the DC motor. It also
improved the dynamic and static response of the system. The comparison between the conventional
response and the fuzzy self-tuning response was performed based on the simulation result obtained by
MATLAB/SIMULINK. The simulation results show that the designed self-adaptive PID controller
achieves good dynamic behavior of the DC motor, perfect speed tracking with short rise and settling
times, zero overshoot and steady state error and thus gives better performance compared to the
conventional PID controller. We then model the fuzzy PID using simple code on Arduino IDE and
perform a practical experiment, to confirm our theorical results.
Keywords: Speed control, DC motor, PID Controller, Fuzzy logic, Fuzzy PID.

? Seniorthesis
? Nassim Messaadi :
Email address : nassim.messaadi@etu.usthb.dz
Address : Cité hospitalière Kouba, bâtiment 1, n°7, 16006 Algiers, Algeria
? Abdelkader Amroun :

Email address : abdelkader.amroun@etu.usthb.dz

Address : Villa 13 cité ARMAF, rostomia, Dely brahim, 16047 Algiers, Algeria

Preprint submitted to Elsevier August 2021

1. Introduction

It is well known that, despite the great number of new methodologies that are continuously
developed by researchers in the field of automatic control, the most adopted controllers for industrial
processes are still the Proportional-Integral-Derivative (PID) controllers.
This is motivated by the fact that, although they usually do not achieve optimal performance, they
can guarantee satisfactory results for a wide range of processes and, due to their simple structure,
they often represent the best solution from a quality/price point of view.
In this project, a new methodology for tuning PID controllers for processes with underdamped
response is proposed. It allows to determine the proportional, integral and derivative gains through
a fuzzy inference system.
The main objective of this methodology is to improve the performance of the system in terms of
setpoint tracking, i.e. to decrease the rise time, overshoot and response time.
Fuzzy logic has already been adopted in the improvement of PID control and it seems appropriate
to use it also in this context, since the design of the tuning procedure can be done in an intuitive
way, reflecting the typical experience of the operator.

2. Modeling and Simulation

2.1. Modeling of the DC motor

The electrical equivalent of a DC motor is illustrated below. It can be represented by a voltage
generator (V) in series with an inductance (L) and a resistor (R) and this in series with an induced
voltage (e) of opposite direction to our voltage generator. The induced voltage is generated by the
rotation of the electric coil through the fixed flux lines of the permanent magnets. This voltage is
called electromotive force.

Figure 1: Representation of a DC motor

In general, the torque generated by a DC motor is proportional to the armature current and the
magnetic field strength. Assume that the magnetic field is constant and therefore the motor torque
(T) is proportional to the armature current (i) alone by a constant factor (Kt) called the torque
constant, as shown in equation (1). This is called an armature-driven motor.

T = Kt × i (1)
The back electromotive force emf (e) is proportional to the angular velocity of the shaft (θ) by a
constant factor (Kb) called the electromotive force constant, as shown in Figure 1.

e = Kb × θ (2)

2
 We derive the following governing equations based on Newton’s second law in equation (3) and
Kirchhoff’s mesh law in equation (4).

J θ̈ + bθ̇ = Kt × i (3)

di
La × + Ra × i = V − Kb × θ̇ × b (4)
dt
(J) is the moment of inertia of the rotor, (b) is the viscous friction constant of the motor, (La) is
the electrical inductance, (Ra) is the electrical resistance, and (V) is the voltage source. By applying
the Laplace transform, the modeling equations can be expressed in terms of Laplace variables as
follows in terms of Laplace variables, as shown in equations (5) and (6).

s(J.s + b) × Θ(s) = Kt × I(s) (5)

s(La.s + Ra) × I(s) = V (s) − Kb × s.Θ(s) (6)

The open-loop transfer function below was obtained by eliminating I(s) in equations (7), where the
rotational speed is considered as the output and the armature voltage as the input.

Θ(s) Kt rad/sec
G(s) = = [ ] (7)
V (s) (j.s + b)(La.s + Ra) + Kt × Kb V

Parameters Value
Torque constant (Kt) 0.5 N.m/A
Electromotive force constant (Kb) 1.25 V/rad/s
Electrical resistance (Ra) 5Ω
Viscous friction constant of the motor (b) 0.008 N.m/rad/s
Electrical inductance (La) 0.2 H
Moment of inertia of the rotor (J) 0.1 kg.m²

Table 1: Parameters of the DC motor

Using the parameters of our engine which are in the table above in equation (7), we obtain the
following equation:

Θ(s) 0.5 rad/sec

G(s) = = [ ] (8)
V (s) (0.1s + 0.008)(0.2s + 5) + 0.5 × 1.25 V
We can then translate the previous equation by the Simulink diagram shown below :

Figure 2: Representation of a DC motor

3
2.2. Conventional controller modeling
2.2.1. PID controller
The function of the PID controller is mainly to adjust an appropriate proportional gain (KP),
integral gain (KI) and differential gain (KD) in order to achieve optimal control performance.
There are several methods for adjusting the PID parameters, in our case we will use the ”Manual
Tuning” method.
The manual Tuning of the PID controller is done by setting the reset time to its maximum value
and the rate to zero and increasing the gain until the loop oscillates at a constant amplitude.
First, we set the ki and kd values to zero. We increase kp until the loop output oscillates, then kp
should be set to about half that value for a ”quarter-amplitude decay” type response. Then ki is
increased until any offset is corrected within a sufficient time for the process. However, a value that
is too high will lead to instability.
Finally, Kd is increased, if necessary, until the loop is fast enough to reach its set point after a load
disturbance. However, increasing Kd too much will result in excessive response and overshoot. Fast
PID loop tuning usually results in a small overshoot to reach the setpoint faster; however, some
systems cannot accept overshoot, in which case an over-damped closed-loop system is required,
which will require tuning well below half the setting that was causing the oscillation.
The values obtained are: Kp = 30, ki=12 ,kd=1.

Figure 3: Simulink diagram of the PID controller

We put this controller in series with our motor, we then obtain the Simulink diagram below:

Figure 4: Simulink diagram of the DC motor in series with a PID controller

4
2.2.2. Fuzzy PID controller
Fuzzy logic is expressed using human language. Based on fuzzy logic, a fuzzy controller converts
a linguistic control strategy into a control strategy, and the fuzzy rules are constructed by the
experience of an expert or a database.
The error (e) and the change of the error (de) of the speed are the variable inputs of the fuzzy
logic controller. The variable output of the fuzzy logic controller adjusts the parameters of the PID
controller, then the PID controller calculates the control output.

Figure 5: Inputs and Outputs of the Fuzzy Controller

All membership functions have an asymmetric form with more concentration near the origin.
This allows for greater accuracy in the steady state. Figures (6 and 7) show the input values (error
and error change) that are applied in the fuzzy PID controller.

Figure 6: Membership function of the Error

Figure 7: Membership function of the Error change

Figures 8, 9 and 10 show the output values (Kp, ki, kd) that are applied to the PID controller.

5
 Figure 8: Membership function of the gain Kp

Figure 9: Membership function of the gain Ki

Figure 10: Membership function of the gain Kd

e/de NL NS ZE PS PL
NL PVL PVL PVL PVL PVL
NS PML PML PML PL PVL
ZE PVS PVS PS PMS PMS
PS PML PML PML PL PVL
PL PVL PVL PVL PVL PVL

Table 2: The rules applied in the PID controller for Kp

e/de NL NS ZE PS PL
NL PM PM PM PM PM
NS PMS PMS PMS PMS PMS
ZE PS PS PVS PS PS
PS PMS PMS PMS PMS PMS
PL PM PM PM PM PM

Table 3: The rules applied in the PID controller for Ki

6
 e/de NL NS ZE PS PL
NL PVS PMS PM PL PVL
NS PMS PML PL PVL PVL
ZE PM PL PL PVL PVL
PS PML PVL PVL PVL PVL
PL PVL PVL PVL PVL PVL

Table 4: The rules applied in the PID controller for Kd

The Simulink diagram of our MCC in series with the fuzzy PID controller is shown in the following figure:

Figure 11: Simulink diagram of the DC motor in series with a Fuzzy PID Controller

For example, the rules applied in the PID Fuzzy as shown in Figure 12, if the error ’e’ is equal to zero (NL) and
the variation of the error ’de’ is equal to zero (PL), then Kp equal to 30 (PVL), Ki equal to 12 (PM) and kd equal to
1 (PVL).

Figure 12: The Fuzzy PID rule viewer

7
2.3. Simulation
2.3.1. System without controller
Before performing the control, we must first see the motor output in rpm. In this case (figure 13), the step
response does not reach the desired value, the rise time is too high and the steady state error is very high.

Figure 13: DC motor output without controller for a setpoint of 2000rpm

2.3.2. System using a PID controller

In this case (figure 14), the step response reaches the desired value, very small rise time, appropriate settling time
and the steady state error is very low (insignificant).

Figure 14: DC motor output with a PID controller in series

8
2.3.3. System using a fuzzy PID controller
In this case (figure 15), Fuzzy Logic is used for setting the PID parameters. The response reaches the desired
value, no steady state error, zero overshoot, the settling and rise times are minimal.

Figure 15: DC output with Fuzzy PID controller

Below (figure 16) we find the graphs of the error (in red) and the derivative of the error (in blue), we can see that
the error cancels out very quickly, we can also see that the derivative of the error reaches a very large negative value
from the first times which is considerably large but this is compensated very quickly.

Figure 16: The error and the derivative of the error of the DC motor in series with a Fuzzy PID controller

9
2.4. Conclusion
From the table (5) we notice that :
- None of the three setups show overshoot.
- The rise time of the DC motor without controller is considerably larger than with controller and
that the PID controller offers better performance than the fuzzy PID controller in terms of rise time.
- The steady state error without controller is very large, the simple PID controller offers negligible
error and the fuzzy PID offers no error.
- The response time without controller is quite large, the latter improves in the case of the simple
PID but remains relatively slow compared to the fuzzy PID which is 4 times faster than the PID
and 10 times faster than the DC motor without controller.

Parameters DC motor whitout controller PID Fuzzy PID

Overshoot(%) 0 0 0
Rise time (10-90%) (s) 1.556 s 55.791 ms 170.761 ms
Steady state error 500 1.5 0
Settling time (2%) (s) 2.759 1.19 0.274

Table 5: Comparison between the step responses of each system

Figure 17: Comparison between the step responses of each system

From table (5) and figure (17) we conclude that:

- DC motor alone offers unreliable performance.
- The classical PID is faster in rise time than the fuzzy PID but it takes much longer to reach its
final value.
- The fuzzy PID has the best compromise between rise time and response time, it also offers a very
accurate output.

10
3. Practical experiment

3.1. Material used

- Arduino UNO - 4 pins fan

- Driver L293D - Electronic relay module (5V)
- Resistance 1K Ω - Capacitor 70µF
- LCD display unit (16x2) - Breadboard

Table 6: List of material

3.2. Diagram of the installation

The diagram of the realization is shown in figure (18) below:

Figure 18: Installation

Our system is controlled with an Arduino UNO which is connected to the computer to obtain
the information of the system via the Arduino IDE interface, as we can see our system is powered
in 12V, in order to be able to control the power supply we use an electronic relay module between
the power supply and the board of experimentation. To make our Arduino communicate with our
motor we use the L293D driver. The capacitor and the diode between the positive and negative
terminals are used for filtering and elimination of parasitic signals, the DC motor in this system is
included in the 4 pins fan, including a pin for the +12V, a pin for the ground, a pin linked to a
Tacheometer (Hall effect sensor, which generates two pulses per revolution) and finally a PWM pin
(pulse width modulation) which allows us to control our DC motor and vary its speed

11
3.3. The program
We modeled the Fuzzy logic controller established on matlab in a simple program on Arduino
IDE.
We modeled the membership functions of the inputs: the error and the change of the error in math-
ematical equation (We already have (e) and (de) so we project on the curves of our membership
functions to obtain the ordinate which varies between 0 and 1. This ordinate will then be projected
on the curves of the membership functions of our output (Kp,Ki,Kd).
The choice of the curves to consider will depend on condition If-Then which correspond to the rules
established previously, table (2,3,4).

The code can be found on this link : https://github.com/aegniss/FUZZY-PID.git

3.4. Experiment Results

Once the code is launched, with a setpoint of 2000 rpm, we obtain the results shown in figure
(19) ”the error is represented in blue and the output in green”:

Figure 19: Plot of the DC motor Error and Output in RPM

Figure 20: Evolution of error and speed as a function of time

In figure (19) and figure (20) we can see that the DC motor reaches the set point very quickly
(at exactly 120ms), this is confirmed by the error trace that we can see in blue which is cancelled

12
at the same time. We also notice that we have an overshoot before stabilization which is due to the
high current absorbed at startup by the MCC. These observed results have a better response time
than the fuzzy PID in theory but the latter presents a considerable overshoot at start point which
is a negative point.

4. Conclusion

Project objectives:
In this project, we designed a DC motor whose speed can be controlled by a PID controller. The
proportional, integral and derivative gains (KP, KI, KD) of the PID controller are adjusted using
the fuzzy logic method.
The study shows that both the precise characters of PID controllers and the flexible characters of
fuzzy controllers are present in the auto-tuning fuzzy PID controller.
The simulation results show that the designed self-tuning PID controller achieves the good dynamic
behavior of the DC motor, perfect speed tracking with short rise and settling times, zero overshoot,
steady state error null and thus gives better performance compared to the conventional PID con-
troller.

Difficulties encountered during the experiment :

- Finding the right tuning method for the PID controller.
- Calculation of the derivative of the error in very small time intervals when coding on Arduino.
- Choice of the appropriate speed sensor given the very short measurement time and the speed of
rotation of the DC motor (In our case the IR sensor was not precise enough during the measure-
ment).
- Difficulty to implementing fuzzy logic on PIC microcontroller.
- Presence of a speed peak due to the amount of current absorbed at the start of the DC motor.

Possible improvements to the project:

- An alternative starting method to avoid speed peaks during start-up.
- Possibility to vary the speed (the setpoint) during the operation of the motor.
- Use of a more advanced microcontroller than the Arduino UNO.
- Possibility of entering the setpoint from a keyboard.
- Tuning of the PID parameters using genetic algorithms.

13
5. References

[1] Ahmed Alhussain Ali Ahmed ; Samar Mohammed Abdullah ; Mohamed Omer Babiker Mo-
hamed Ali ; Abdulmajeed Abdulmonim Abdullatif Abdulrahman.
“DC Motor Control of Speed Using FUZZY-PID Controller”- A Project Submitted in Partial Ful-
fillment for the Requirements of the Degree of B.Sc. (Honor) In Electrical Engineering. .

[2] Guokun Xie et Kai Zheng and Yajuan Jia.

”Design of Fuzzy PID Temperature Control System” .

[3] K. Hachemi ; B. Mazari ; H. Oirkozek ; A. Al Jazi ; M. Laouer.

“Etude Comparative des Régulateurs PID et Flou : Autopilotage d’un Moteur Synchrone à Aimant
Permanent ” .

[4] Umesh Kumar Bansal et Rakesh Narvey.

”Speed Control of DC Motor Using Fuzzy PID Controller”.

[5] Alfatih Yassin Awad Babiker.

“DC Motor Speed Control Using Fuzzy Logic and Proportional-Integral-Derivative Controller”-Thèse
de Master d’ingénierie informatique et reseaux .

[6] F. Martin McNeill et Ellen Thro.

“THE FUZZY WORLD” .

[7] Antonio Visioli.

“FUZZY LOGIC BASED TUNING OF PID CONTROLLERS FOR PLANTS WITH UNDER-
DAMPED RESPONSE” .

[8] DEBASMITA PATTANAIK ; BONANI SAHU et DEVADUTTA SAMANTARAY.

“MICROCONTROLLER BASED IMPLEMENTATION OF A FUZZY KNOWLEDGE BASED
CONTROLLER” .

[9] “Modelling and Control of DC Motor”.

14