Alexandria Engineering Journal (2021) 60, 4029–4040

H O S T E D BY
Alexandria University

Alexandria Engineering Journal

www.elsevier.com/locate/aej
www.sciencedirect.com

PSO technique applied to sensorless field-oriented

control PMSM drive with discretized RL-fractional
integral
Waleed Abd El Maguid Ahmed a,b,*, Mahmoud M. Adel c, M. Taha d, Amr A. Saleh c

a
Engineering Mathematics and Physics Department, Faculty of Engineering, Fayoum University, 63514 Fayoum, Egypt
b
Zewail City of Science and Technology, 12578 Giza, Egypt
c
Electrical Department, Faculty of Engineering, Fayoum University, 63514 Fayoum, Egypt
d
Electrical Department, Faculty of Engineering, Cairo University,12613 Giza, Egypt

Received 8 July 2020; revised 12 February 2021; accepted 22 February 2021

Available online 11 March 2021

KEYWORDS
Permanent magnet syn-
chronous motor;
Fractional calculus;
Field oriented control;
Model reference adaptive
system;
Sensorless Control;
Particle swarm optimization
Abstract This paper introduces a Fractional Order Field Oriented Control (FO-FOC) with sensor-
less Model Reference Adaptive System (MRAS) for the Permanent Magnet Synchronous Motor
(PMSM). In this work, the conventional integer controllers of the vector control method are
replaced with the Fractional Order PI (FOPIa) controllers. Particle Swarm Optimization technique
(PSO) is employed to tune the gain, integration parameter and the fractional order parameter for
each of the speed, current and MRAS controllers to get the optimum values. The presented method-
ology is tested using MATLAB Simulink simulation and the results show improved performance of
the field oriented control technique by the employment of the fractional order controllers instead of
the conventional integer at different operating points. Furthermore, the additional degree of free-
dom in the fractional controllers helps the PSO technique to get the optimum cost at a very low
number of iterations compared with that of the integer one.
Ó 2021 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Alexandria
University. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

1. Introduction precise description for the complex physical systems, hence,

scientists go to the definition of generalizing the derivatives
Fractional Calculus (FC) is a generalization of the common and integrals to fractional order instead of integer one in the
derivatives and integrals to fractional order [1,2,3]. The non- modelling of electric systems and many applications [4]. In
integer order of differentiation and integration leads to a more the last two decades, the FC has acquired a wide area of

* Corresponding author at: Engineering Mathematics and Physics Department, Faculty of Engineering, Fayoum University, 63514 Fayoum,
Egypt.
E-mail addresses: waa01@fayoum.edu.eg, wabdelmagied@zewailcity.edu.eg (W. Abd El Maguid Ahmed).
Peer review under responsibility of Faculty of Engineering, Alexandria University.
https://doi.org/10.1016/j.aej.2021.02.049
1110-0168 Ó 2021 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Alexandria University.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
4030
applications like tuberculosis infection model [5], physical
engineering, economics [6,7], electric machines [8], biology
[9,10], transformers, inductors, parasitic capacitances in tran-
sistors, kinetic phenomena in semiconductors, thermal effects
in electronic devices [11], bio-engineering [12], many electric
control systems [13], noise detection [14], passive symmetric
network [15], analog signal processing [16], fractional propor-
tional integral differential controllers (PID) [17,18], diffusion
process [19], viscoelasticity, control, relaxation processes [20],
fluid dynamics [21], gas dynamics [22] and position servo con-
trol system [23].
FC can provide automatic control systems, especially those
that are designed based on the mathematical models of the
process and those that are intensively using proportional-inte
gral–differential (PID) controllers, an additional dimension
for the controller design [8]. This dimension results from the
degree of freedom gained from the fractional integrator and
differentiator parameters that allow better tuning of the PID
controllers. Electrical machine drives are part of the process
control systems that can provide an enhanced design using
fractional controllers. Many definitions are introduced for
the fractional integration and differentiation like Grunwald-
Letnikov fractional differ-integral, Riemann-Liouville frac-
tional differ-integral and Caputo fractional differ-integral
[24,25]. While the Cauchy repeated integration formula is used
for the integration for integer order, Riemann-Liouville (RL)
fractional differ-integral formula is a generalized integration
formula. In this work, the RL formula is used.
Permanent magnet synchronous motors (PMSMs) are
widely employed in many electrical drives due to its distinct
performance. The most distinctive advantages of PMSM are
the great power density, high efficiency and the numerous tor-
que to current ratio in addition to the decreased copper losses
in PMSM because of the absence of the rotor current [26]. The
field rotor winding is replaced with Permanent Magnets (PM)
in order to get rid of the drawbacks of other machines through
brushes removal, elimination of slip-rings and the cancellation
of rotor copper losses; which leads to optimum performance
[27].
One of the most common control methods for the electrical
machine drives is the field oriented control (FOC), also known
as vector control, that improves the performance of PMSM
drive [28]. In FOC, the speed control of PMSM becomes too
easy due to the decoupling of the torque component from
the flux component [29]. FOC incorporates direct axis and
quadrature axis current controllers and speed proportional
integral (PI) controllers to achieve the required control of
the current vector thus obtaining precise speed/position con-
trol. For optimum performance, these controllers should be
designed precisely.
Fractional order PI controller is a generalization of the
common integral of the conventional PI controller to a frac-
tional order [16]. The extra degree of freedom acquired by
the fractional order of the integrator has proved the ability
to strengthen the performance in the control systems [16,17].
Fractional order PI controllers show superiority to their coun-
terpart integer PI controllers [18]. The fractional order con-
trollers have been employed in many applications like
position servomechanism [30], aerospace control systems [31],
time delay systems [32] and integrating processes [33].
In order to strengthen the performance of the FOC tech-
nique, FO-FOC is presented in this work taking the advantage
W. Abd El Maguid Ahmed et al.
of the increased degree of freedom in the design of the frac-
tional controllers used in the proposed technique.
For vector controlled electrical machine drives, precise
motor position measurement is very important. Motor posi-
tion can be measured using different position (speed) sensors
like encoders, resolvers or Hall Effect sensors. However, they
have drawbacks and are not applicable for some tough appli-
cations. The high cost of these sensors is one of the most pop-
ular disadvantages that encourage the researchers to exclude
the position [34]. An alternative practical solution considered
for the reduction of the cost of the machine drive is to estimate
the rotor position/speed through currents and voltage mea-
surements (which is considerable reduced cost) rather than
measuring it. Hence, sensorless control is widely utilized for
electrical machine drive systems.
Various sensorless control techniques of PMSM have been
introduced by the researchers [34,35]. Model reference adap-
tive system position/speed estimator is used in this paper
because of its combined benefits of accurate position estima-
tion and its ease of implementation [36]. This method funda-
mentally is based on three parts: the reference model,
adaptive model, and the adaptive mechanism that seeks mini-
mum error between the reference model and the adaptive one
[37]. The precision of the estimated speed and position calcu-
lated from the MRAS is strongly based on the perfect design
of the MRAS controller.
In the proposed FO-FOC PMSM drive, Fractional order
PI controllers (FOPIa) controllers substitute the conventional
speed and current controllers of the FOC control drive. Fast
response with better performance (decreased rise time and set-
tling time) is the main advantage obtained from the employ-
ment of the FOPIa will be shown later in the results. In fact,
the increased degree of freedom for the PI controllers design
considering the fractional order parameters is what supports
this enhanced performance.
Tuning the parameters (Kp, Ki, a) of the speed and current
controllers are responsible for optimizing the control response
through minimizing the overshoot, rise time and settling time.
The calculated estimated speed and position are certainly
affected by the parameters (Kp, Ki, a) of the MRAS controller.
Many techniques found in literature are used for tuning the
PID controllers, some of which are deterministic such as Zieg-
ler Nichols [38], others are heuristic using different optimiza-
tion techniques [39]. In this work; the four controllers’
parameters optimal values (Kp, Ki, a) are obtained by the
employment of the Particle swarm optimization (PSO) tech-
nique in order to get an optimal control response with the
accurate estimated position.
In [8], the authors present the applicability of the fractional
order integrals for the FOC controllers and MRAS speed esti-
mation of a field oriented controlled PMSM drive. The authors
in [8] presented only the results obtained from the application
of the fractional order system with the consideration of the
PSO for the tuning of the controllers’ parameters. In this work,
the authors present the application of the PSO for the opti-
mization of the parameters of the sensorless MRAS FOC
using RL-fractional integration as applied on a Beckhoff
model AM247L PMSM machine and compared it to the coun-
terpart integer FOC controllers and MRAS speed estimation
models. The results for the fractional order control and speed
estimation model are compared to that of the counterpart
integer model showing the superiority of the fractional order
PSO technique applied to sensorless field-oriented control PMSM drive with discretized RL-fractional integral 4031

FOC controllers and MRAS speed estimation models in terms

of better performance than their counterpart integer FOC con-
trollers and MRAS speed estimation and in terms of the addi-
tional degree of freedom of the fractional order that leads to
reduced number of iterations of the PSO to reach the optimal
parameters.
Section 2 introduces the Riemann-Liouville integration def-
inition. The sensorless PMSM FOC drive is illustrated in Sec-
tion 3, through the description of the PMSM model, FOC
technique and MRAS position/speed estimator. Section 4 is
dedicated to introduce the proposed fractional order field ori-
ented control technique (FO-FOC) through introducing the
design of the fractional PI controllers as well as the tuning
of the controllers’ parameters. Section 5 presents the simula-
tion results obtained for both the FO-FOC PMSM drive and
the integer (conventional) FOC PMSM drive with the discus-
sion of the results obtained. Finally, section 6 concludes the
FO-FOC technique and the advantages gained over the con-
ventional FOC.
Fig. 1a abc to dq Reference Frame Transformation.
2. Riemann-Liouville integration

The Cauchy repeated integration formula as in Eq. (2.1) is gen-

eralized by Riemann-Liouville replacing the gamma function
instead of the factorial. The resulting Riemann-Liouville for-
mula is as in Eq. (2.2) [24,40,41].
Z t
1
D
n f ð t Þ ¼ ðt
sÞn
1 fðsÞds; t > 0: ð2:1Þ
t
ðn
1Þ! a
Where n is a positive integer.
Z t
1
D
a f ð t Þ ¼ ðt
sÞa
1 fðsÞds; a 2 Rþ : ð2:2Þ
t
CðaÞ a

3. Sensorless PMSM FOC drive

3.1. Permanent magnet synchronous motor (PMSM) model
It is common for the analysis and control of electrical machi-
nes to transform the machine model (mathematical equations)
in the rotating reference frame. For field oriented control, the
speed of the rotating reference frame is selected to be syn-
chronous speed, which is the rotor speed for the PMSM as
shown in Fig. 1a). This section presents the dynamic model
of the PMSM in the dq0 reference frame by the equations
given below in [2] and shown in Fig. 1b).
kq ¼ Lq irq ; ð3:1Þ
Fig. 1b Permanent Magnet Synchronous Motor Equivalent
Circuit.
Where kq is the q-axis flux, kd is the d-axis flux, vrq is the q-
axis voltage, vrd is the d-axis voltage, irq is the q-axis current, ird
is the d-axis current, rs is the stator winding resistance, xr is the
rotor electrical angular speed,xm is the rotor mechanical angu-
lar speed, km is the rotor magnetic flux linkage, p is the time
differentiator, Te is the developed torque, Tl is the load torque,
B is the motor viscous friction, P is the number of poles, and j
is the motor inertia.

3.2. Field oriented control (FOC) of PMSM

kd ¼ Ld ird þ km ; ð3:2Þ

vrq ¼ ðrs þ p Lq Þirq þ xr Ld ird þ xr km ; ð3:3Þ The field oriented control is one of the most common control
structures of the PMSM. The stator current consists of the flux
vrd ¼ ðrs þ p Ld Þird þ xr Lq irq ; ð3:4Þ component and the torque component. By decoupling the sta-
tor current to its components through a proper transforma-
3 P tion, the PMSM control can be seen as the dc machines
Te ¼ ½km irq þ ðLd - Lq Þ irq ird ; ð3:5Þ control. From Eq. (3.5), if the d-axis current is removed by
2 2
making the reference value of the d-axis current equal to zero,
dxm the torque equation will be only a function of the q current
Te ¼ Tl þ B xm þ j : ð3:6Þ
dt component, so a very appropriate mechanism is proposed to
4032 W. Abd El Maguid Ahmed et al.

force the total stator current in the q-axis only. Hence; Eq.
(3.5) after assigning zero to the reference d-axis current
becomes as follows [4]
3 P
Te ¼ km irq : ð3:7Þ
2 2
As shown in Fig. 1c), the vector control method employs
three controllers; the speed controller, the iq controller and
the id current controller. It is very clear that the target value
of q-axis current controller is the output of the speed controller
and the reference value of the d-axis current controllers is
assigned as zero by default. The difference between the refer-
ence value of the q-axis current and the measured q-axis cur-
rent value is the input to the q-axis current controller. Then
the reference value of vq voltage is obtained from the output
of the iq controller. The id controller input is the difference
between the id reference value (zero by default) and the mea- Fig. 2a Illustration of Model Reference Adaptive System.
sured id current. Hence, the reference vd voltage is obtained
from the output of id controller. ence model and the adjustable model. The stator current calcu-
In this work, the three employed conventional integer con- lated by the reference model is independent on the estimated
trollers for speed, d-axis current and q-axis current are speed, but the stator current calculated by the adjustable
replaced with fractional order controllers and the selection of model depends on the estimated speed and position [42]. The
the controllers’ parameters is done using the PSO technique. output of the two models are passed to the adaptation tech-
A schematic diagram for the field oriented control technique nique. In this paper, the adaptation mechanism employed is
using fractional order controllers instead of the conventional the PI controller that tuned the estimated rotor speed to the
ones are shown on Fig. 1c). actual rotor speed. In this work, Fig. 1c) Shows that the vq
and vd which are inputs to MRAS algorithm are not measured
3.3. Model reference adaptive system (MRAS) algorithm using voltage sensors but they are calculated from the park
transformation of the reconstructed voltage (va ; vb ; vc ) accord-
MRAS is one of the most commonly used techniques in ing to [43]. Eq. (3.8), (3.9) and (3.10) calculate the recon-
encoderless control due to its ease of implementation and its structed voltage (va ; vb ; vc ) using the pulse width modulation
simplicity. In this method, two models for speed and position periods (TA, TB, TC) and the dc link voltage according to
estimation are employed; the reference model and the adjusta- [43], then these reconstructed voltages will transformed to dq
ble one [37]. Fig. 2a) illustrates the algorithm of the MRAS. frame to get the vq and vd which are inputs to MRAS algo-
In MRAS algorithm, a motor variable is utilized as a state rithm. The elimination of the voltage sensors will provide
variable. This motor variable can be stator current or power space, minimize the setup cost and reduce the measurement
[37]. This variable will be calculated by the reference model error due to the inaccuracy of the voltage sensor. Therefore,
and the adaptive model. In this work, the current model one this will lead to a robustness and more accurate estimation
is utilized. Hence, the stator current is obtained from the refer- of the motor speed and position.

Fig. 1c Proposed Fractional Order Field Oriented Control Scheme.

PSO technique applied to sensorless field-oriented control PMSM drive with discretized RL-fractional integral 4033

va ¼
Vdc
FSW ð2TA
TB
TC Þ; ð3:8Þ (FOPIa) which consequently improves the performance of
3 the PI controllers because of the additional degree freedom
vb ¼
Vdc
FSW ð
TA þ 2TB
TC Þ; ð3:9Þ (fractional order a) as will be shown in the results. To tune
3 the PI controllers’ parameters, a PSO algorithm is developed
Vdc considering the tuning of the three parameters of each con-
vc ¼ FSW ð
TA
TB þ 2TC Þ: ð3:10Þ
3 troller, namely the proportional gain Kp, the integral gain Ki
Where TA, TB, TC are the pulse width modulation intervals and the fractional order parameter of the integrator term (a).
and FSW is the switching frequency of the inverter.
The equations of the current model can be introduced 4.1. Design of fractional order PI controller
according to [44] as follows
  "
rs L #
xr Lqd  id 
" vd # Podlubny introduced the fractional order controller concept in
d id Ld Ld 1997. The PI conventional controller has two main parameters;
¼ þ vq : ð3:11Þ
dt iq
xr LLd
Lrs iq Lq

xr kLmq the proportional part and the integral part. The arbitrary non-
q q
integer order is presented to the integration portion. Conse-
Define id *, iq *, vd *, vq * as follows quently an extra degree of freedom is gained by the addition
km  of the fractional order to Kp , and Ki for each controller [46].
id ¼ id þ ; i ¼ iq ; ð3:12Þ To design a fractional order controller (PIa), the integration
Ld q
to an arbitrary non-integer order (fractional order) should be
rs first determined. In this work, the integration to a fractional
vd ¼ vd þ km ; vq ¼ vq : ð3:13Þ
Ld order is designed according to the numerical approximation
From Eq. (3.12) and (3.13) of Riemann-Liouville integrals as follows [46]
" # " #"  # 2 3 For interval [0, X] with xn ¼ nh; n ¼ 0; 1; 2 . . . . . . . . . N
vd
d id
Lrsd ^ r LLqd
x id where h = X/N,
þ 4 v 5:
Ld
¼ ð3:14Þ  X
dt iq
x ^ r LLdq
Lrsq iq q

x ^ r kLmq ha N
Lq
Ia yn ðhÞ ¼ Cn;N yn : ð4:1Þ
Cð2 þ aÞ n¼0
Hence the estimated speed equation can be presented in Eq.
(3.15) as follows [45] applying the quadrature weights (from a product trape-
  zoidal rule), then:
Ki
b r ¼ Kp þ
x 8 9
s >
< ð1 þ aÞNa
N1þa þ ðN
1Þ1þa if n ¼ 0 >
=
  
km   Cn;N ¼ if 0 < n < N ;
1þa 1þa 1þa
Lq b Ld b Ld Lq ð N
n þ 1Þ
2ð N
nÞ þ ð N
n
1Þ
> >
 id i q
i d iq
iq þ b
iq
b iq b
id
: :
1 if n ¼ N
;
Ld Lq Lq Lq Ld
ð4:2Þ
ð3:15Þ
A schematic block diagram of MRAS adaptive mechanism Ki
PIa ¼ Kp þ : ð4:3Þ
is as shown in Fig. 2b). sa
The schematic diagram of the fractional order controller is
4. Proposed Fractional Order Field Oriented Control (FO- illustrated in Fig. 2c).
FOC) technique
4.2. Speed, current and MRAS controllers tuning
In this proposed technique, the integer PI controllers employed
in the FOC technique as well as the MRAS (position/speed The optimization techniques play a key role in PI parameters
estimator) are replaced with fractional order PI controllers selection. Hence, the PSO technique is applied for both the
scheme with integer conventional controllers and the proposed
scheme with fractional order controllers with the same fitness
function and the same constraints for a fair comparison.
Two measures are considered to compare the two schemes,
first is the dynamic and steady state response of the controlled
variable, i.e. the rotor speed and the second criteria is the effi-
ciency of the optimization technique, i.e., the number of
iterations to get the optimum cost. As mentioned earlier, this

Fig. 2b Adaptive Mechanism of Model Reference Adaptive

System. Fig. 2c Fractional Order Controller Block Diagram FOPIa.
4034 W. Abd El Maguid Ahmed et al.

work utilizes the PSO algorithm for the optimum selection of For the first measure regarding the dynamic and steady
the controllers’ parameters. The PSO algorithm equations state of the controlled variable, the rotor speed, the fitness
are as illustrated in [47]. function for the speed controller is to achieve the speed target

Fig. 3 Matlab Simulink FO-FOC Model.

PSO technique applied to sensorless field-oriented control PMSM drive with discretized RL-fractional integral 4035

of the absolute error between the reference speed and the

Table 1 Beckhoff model AM247L PMSM Machine
actual speed.
Parameters. hX i
Machine Parameters Value MIN ðspeed controllerÞcost ¼ MIN xref
xact : ð4:4Þ
Magnetic pole pairs 3
The primary goal for the current controllers is to select the
Rated speed (r.p.m) 3000
Inertia (kg.m2) 0.000198
parameters to minimize the difference between controllers’ ref-
Viscous damping coefficient (N.m.s) 0.000106 erence and the measured currents. The fitness function for
Stator resistance Rs (O) 5.95 both the Id and the Iq current controllers is
flux linkageWf (wb) 0.2955556
MIN ðcurrent controllerÞcostd;q
hX i
¼ MIN Id;qref
Id;qact : ð4:5Þ

value over a wide range of operation with a low rise time, low The limitation applied here for speed and current con-
settling time and acceptable overshoot. Accordingly, the trollers is the constraint of an acceptable overshoot
desired function employed here is to decrease the summation (overshoot < 5%).

Fig. 4a The Best Cost per iteration for the conventional and the proposed technique at 3000 r.p.m.

Fig. 4b Speed response comparison between the conventional and the proposed technique at 3000 r.p.m.

Fig. 4c Zoom of Speed response for Fig. 4b. for the transient response.
4036 W. Abd El Maguid Ahmed et al.

Fig. 4d Position response comparison between the MRAS estimated and the actual position at 3000 r.p.m.

maxðspeedÞ < 0:05  reference speed: ð4:6Þ 5. Simulation results

The MRAS observer plays a significant role in the improve-
ment of the drive control. The estimated speed will be used for
the speed control in the vector control technique while the esti-
mated position is needed for the transformation (park trans-
formation) of the measured currents in the abc physical
frame to the dq synchronous reference frame. Consequently,
its controller’s parameters should be chosen accurately, such
that the estimated speed tracked the actual speed. The cost
function thus is aimed to reduce the error between the mea-
sured speed and the estimated speed.
MIN ðMRAS controllerÞcost
hX i eters for the conventional method and the proposed one, two
¼ MIN ðxmeasured
xestimated Þ : ð4:7Þ
The proposed technique was examined using Matlab M
File
and Simulink model on a Beckhoff model AM247L PMSM
machine, as shown in Fig. 3, with the parameters given in
Table 1, at different operating points; 3000 r.p.m, 1500 r.p.m
and 700 r.p.m. The PSO optimization technique is used to tune
the parameters for both the integer PI controllers for the con-
ventional FOC and the fractional order controllers for the pro-
posed FO-FOC with the same constraints and the same fitness
function to get a fair comparison between the conventional
method and the proposed one. After getting the tuning param-
comparisons are considered; the first comparison is the
dynamic response of the rotor speed for both techniques and
the second one is the number of iterations to get the best cost
for each technique.

Fig. 5a The Best Cost per iteration for the conventional and the proposed technique at 1500 r.p.m.

Fig. 5b Speed response comparison between the conventional and the proposed technique at 1500 r.p.m.
PSO technique applied to sensorless field-oriented control PMSM drive with discretized RL-fractional integral 4037

Fig. 5c Zoom of Speed response for Fig. 5b. for the transient response.

Fig. 5d Position response comparison between the MRAS estimated and the actual position at 1500 r.p.m.

The optimization technique was performed on 3000 r.p.m
(high operating point) for both the conventional method and
the proposed method. Fig. 4a) shows that the additional
degree of freedom of the proposed FO-FOC helps the PSO
technique to get the optimum cost after almost 20 iterations,
but the PSO technique takes almost 70 iterations to get the
optimum solution. Also, it can be easily observed from
Fig. 4a) that the optimum cost for the FO-FOC is lower than
that of the conventional integer FOC. Fig. 4b) and Fig. 4c)
illustrate that the speed response of the proposed fractional
field oriented control (FO-FOC) has a better rise time with
no overshoot compared with that of the integer conventional
field oriented control. Also, the fractional controller can
achieve the reference in a lower time (smaller settling time)
compared with that of the integer controller. Fig. 4d) shows
the tracking of the MRAS estimated position to the actual
position.
The optimization technique was applied at medium speed
(1500 r.p.m) for the integer and fractional controllers.
Fig. 5a) illustrates that the optimum cost for the fractional
controller is much lower than the optimum cost for the integer
controller. Fig. 5b) and Fig. 5c) show obviously that the speed
response of the FO-FOC reaches the reference speed at 0.005
sec, while the speed response of the conventional FOC was
not yet able to achieve the reference speed. Besides, there is
no overshoot in the speed response of the fractional controller,
but the speed response of the integer conventional controller
has many oscillations and overshoot. Fig. 5d) shows the per-
fect estimation of the position by the MRAS algorithm at med-
ium speed.

Fig. 6a The Best Cost per iteration for the conventional and the proposed technique at 700 r.p.m.
4038 W. Abd El Maguid Ahmed et al.

Fig. 6b Speed response comparison between the conventional and the proposed technique at 700 r.p.m.

Fig. 6c Zoom of Speed response for Fig. 6b) for the transient response.

Fig. 6d Position response comparison between the MRAS estimated and the actual position at 700 r.p.m.

Table 2 Comparison between the proposed FO-FOC and the conventional FOC.
PSO Optimum cost Settling time (sec) Rise time (sec)
Proposed FO-FOC (3000 r.p.m) 2148.7338 0.007 0.00535
Conventional FOC (3000 r.p.m) 2726.2753 0.0075 0.00571
Proposed FO-FOC (1500 r.p.m) 1925.7380 0.005 0.00259
Conventional FOC (1500 r.p.m) 4591.2711 0.02 0.00263
Proposed FO-FOC (700 r.p.m) 1410.9686 0.009 0.00153
Conventional FOC (700 r.p.m) 2656.7084 0.015 0.001416

Finally, the proposed method was examined at low speed clearly achieved an optimum cost for the fractional controller
(700 r.p.m) to complete the comparison at high, medium and lower than the optimum cost of the integer controller because
low operating points. Fig. 6a) shows that the PSO technique of the advantage of an additional degree of freedom of the
PSO technique applied to sensorless field-oriented control PMSM drive with discretized RL-fractional integral
fractional controller. Fig. 6b) and Fig. 6c) illustrate that the
speed response of the fractional controller achieves the refer-
ence speed at almost 0.025 sec; however, the speed response
of the integer controller still does not achieve the reference
speed. Fig. 6d) shows the robustness of the MRAS position
estimation even at low speed.
Table 2 presented a comparison between the proposed FO-
FOC response and the conventional FOC technique.
5. Conclusion
This paper proposed FO-FOC MRAS for the control of
PMSM. Applying the FO-FOC in the simulation work
strengthened the robustness of the control system, decreases
the overshoot and improves the rise time and the settling time.
The simulation results proved the good response of the MRAS
and vector control at different operation points. The addi-
tional degree of freedom in the proposed fractional order con-
troller enhanced the PSO technique in achieving the fitness
function and getting the optimum cost with lower number of
iterations compared with that of the conventional controllers.
Finally, the proposed technique not only improves the speed
response but also decreases the number of iterations of PSO
for achieving the fitness function, which makes it applicable
for online tuning in the future work.
Declaration of Competing Interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
References
[1] J.A.T. Machado, A.M. Lopes, Multidimensional scaling locus
of memristor and fractional order elements, Journal of
Advanced Research (2020).
[2] D. Vindel, S. Haghbin, A. Rabiei, O. Carlson, R. Ghorbani,
Field-oriented control of a PMSM drive system using the
dSPACE controller, in: 2012 IEEE International Electric
Vehicle Conference, 2012: pp. 1–5.
[3] Y. Wei, Y. Chen, S. Cheng, Y. Wang, Discussion on fractional
order derivatives, IFAC-PapersOnLine. 50 (2017) 7002–7006.
[4] J. Jacob, A. Chitra, Field oriented control of space vector
modulated multilevel inverter fed PMSM drive, Energy
Procedia. 117 (2017) 966–973.
[5] N.H. Sweilam, S.M. Al-Mekhlafi, D. Baleanu, Optimal control
for a fractional tuberculosis infection model including the
impact of diabetes and resistant strains, Journal of Advanced
Research. 17 (2019) 125–137.
[6] J. Song, Research on fast numerical calculation based on
fractional calculus model for macroeconomic growth, Chaos,
Solitons & Fractals. 130 (2020) 109440.
[7] T. Chen, D. Wang, Combined application of blockchain
technology in fractional calculus model of supply chain
financial system, Chaos, Solitons & Fractals. 131 (2020) 109461.
[8] M.M. Adel, W.A. Ahmed, M. Taha, A.A. Saleh, Enhanced
Sensorless Field Oriented Controlled PMSM Drive Using
Fractional Calculus and PSO Technique, in: 2020
International Symposium on Power Electronics, Electrical
Drives, Automation and Motion (SPEEDAM), 2020: pp. 5–10.
4039
[9] B. Ghanbari, S. Kumar, R. Kumar, A study of behaviour for
immune and tumor cells in immunogenetic tumour model with
non-singular fractional derivative, Chaos, Solitons & Fractals.
133 (2020) 109619.
[10] S. Kumar, R. Kumar, R.P. Agarwal, B. Samet, A study of
fractional Lotka-Volterra population model using Haar wavelet
and Adams-Bashforth-Moulton methods, Mathematical
Methods in the Applied Sciences. 43 (2020) 5564–5578.
[11] A. Dzieliński, D. Sierociuk, G. Sarwas, Some applications of
fractional order calculus, Bulletin of the Polish Academy of
Sciences: Technical Sciences. (2010) 583–592.
[12] R.L. Magin, Fractional calculus in bioengineering: A tool to
model complex dynamics, in: Proceedings of the 13th
International Carpathian Control Conference (ICCC), 2012:
pp. 464–469.
[13] S. Tufenkci, B. Senol, B.B. Alagoz, R. Matušů, Disturbance
rejection FOPID controller design in v-domain, Journal of
Advanced Research (2020).
[14] Q. Wang, J. Ma, S. Yu, L. Tan, Noise detection and image
denoising based on fractional calculus, Chaos, Solitons &
Fractals. 131 (2020) 109463.
[15] M.S. Semary, M.E. Fouda, H.N. Hassan, A.G. Radwan,
Realization of fractional-order capacitor based on passive
symmetric network, Journal of Advanced Research. 18 (2019)
147–159.
[16] A.S. Ali, A.G. Radwan, A.M. Soliman, Fractional order
Butterworth filter: active and passive realizations, IEEE
journal on emerging and selected topics in circuits and
systems. 3 (2013) 346–354.
[17] Y. Luo, Y.Q. Chen, C.Y. Wang, Y.G. Pi, Tuning fractional
order proportional integral controllers for fractional order
systems, Journal of Process Control. 20 (2010) 823–831.
[18] S.E. Hamamci, Stabilization using fractional-order PI and PID
controllers, Nonlinear Dynamics. 51 (2008) 329–343.
[19] S. Kumar, S. Ghosh, B. Samet, E.F.D. Goufo, An analysis for
heat equations arises in diffusion process using new Yang-
Abdel-Aty-Cattani fractional operator, Mathematical Methods
in the Applied Sciences. 43 (2020) 6062–6080.
[20] S. Kumar, A new analytical modelling for fractional telegraph
equation via Laplace transform, Applied Mathematical
Modelling. 38 (2014) 3154–3163.
[21] S. Kumar, A. Kumar, D. Baleanu, Two analytical methods for
time-fractional nonlinear coupled Boussinesq-Burger’s
equations arise in propagation of shallow water waves,
Nonlinear Dynamics. 85 (2016) 699–715.
[22] S. Kumar, M.M. Rashidi, New analytical method for gas
dynamics equation arising in shock fronts, Computer Physics
Communications. 185 (2014) 1947–1954.
[23] C.A. Monje, B.M. Vinagre, V. Feliu, Y. Chen, On auto-tuning
of fractional order PIk Dl controllers, IFAC Proceedings
Volumes. 39 (2006) 34–39.
[24] R. Gorenflo, F. Mainardi, Fractional calculus: integral and
differential equations of fractional order, arXiv preprint
arXiv0805.3823. (2008).
[25] A.L. Alaoui, E. Azroul, A.A. Hamou, Monotone Iterative
Technique for Nonlinear Periodic Time Fractional Parabolic
Problems, Advances in the Theory of Nonlinear Analysis and its
Application. 4 (3) 194–213.
[26] Y. Zhao, Position/Speed Sensorless Control for Permanent-
Magnet Synchronous Machines, Ph.D. dissertation,
Deptartment Electrical Engineering, University of Nebraska,
Lincoln, NE, 2014.
[27] W. Qiao, X. Tang, S. Zheng, Y. Xie, B. Song, Adaptive two-
degree-of-freedom PI for speed control of permanent magnet
synchronous motor based on fractional order GPC, ISA
transactions. 64 (2016) 303–313.
[28] K. Boby, A.M. Kottalil, N.P. Ananthamoorthy, Mathematical
modelling of pmsm vector control system based on SVPWM
4040
with pi controller using Matlab, International Journal of
Advanced Research in Electrical, Electronics and
Instrumentation Engineering. 2 (2013).
[29] L. Saihi, A. Bouhenna, M. Chenafa, A. Mansouri, A robust
sensorless SMC of PMSM based on sliding mode observer and
extended Kalman filter, in: 2015 4th International Conference
on Electrical Engineering (ICEE), 2015: pp. 1–4.
[30] C. Zhao, X. Zhang, The application of fractional order PID
controller to position servomechanism, in: 2008 7th World
Congress on Intelligent Control and Automation, 2008: pp.
3380–3383.
[31] M.A.S. Aboelela, M.F. Ahmed, H.T. Dorrah, Design of
aerospace control systems using fractional PID controller,
Journal of Advanced Research. 3 (2012) 225–232.
[32] A. Yüce, N. Tan, D.P. Atherton, Fractional order pi controller
design for time delay systems, IFAC-PapersOnLine. 49 (2016)
94–99.
[33] E. Cokmez, S. Atiç, F. Peker, I. Kaya, Fractional-order PI
controller design for integrating processes based on gain and
phase margin specifications, IFAC-PapersOnLine. 51 (2018)
751–756.
[34] H.M. Kojabadi, M. Ghribi, MRAS-based adaptive speed
estimator in PMSM drives, in: 9th IEEE International
Workshop on Advanced Motion Control. 2006., 2006: pp.
569–572.
[35] V. Repecho, D. Biel, A. Arias, Enhanced high frequency
injection algorithm for sensorless sliding mode control PMSM
drives, in: 2014 IEEE 11th International Multi-Conference on
Systems, Signals & Devices (SSD14), 2014: pp. 1–6.
[36] A.A. Abd Samat, D. Ishak, A.M. Omar, S. Iqbal, M.A. Razak,
A new speed sensorless field oriented controller for PMSM
based on MRAS and PSO, Journal of Electrical Systems. 12
(2016) 566–573.
[37] Y. Ouyang, Y. Dou, Speed Sensorless Control of PMSM Based
on MRAS Parameter Identification, in: 2018 21st International
Conference on Electrical Machines and Systems (ICEMS), 2018:
pp. 1618–1622.
W. Abd El Maguid Ahmed et al.
[38] B. Jaganathan, R. Sharanya, S.K. Devi, S.K. Sah, Ziegler-
Nichol’s method of online tuning of PMSM for improved
transient response, in, 2010 International Conference on Power,
Control and Embedded Systems (2010) 1–4.
[39] Ö. Aydoğdu, M. Korkmaz, Optimal design of a variable
coefficient fractional order PID controller by using heuristic
optimization algorithms, International Journal of Advanced
Computer Science and Applications(IJACSA) 10 (3) (2019).
[40] S. Kazem, S. Abbasbandy, S. Kumar, Fractional-order Legendre
functions for solving fractional-order differential equations,
Applied Mathematical Modelling. 37 (2013) 5498–5510.
[41] S. Kumar, D. Kumar, S. Abbasbandy, M.M. Rashidi,
Analytical solution of fractional Navier-Stokes equation by
using modified Laplace decomposition method, Ain Shams
Engineering Journal. 5 (2014) 569–574.
[42] . Yujie, C. Shaozhong, Model reference adaptive control system
simulation of permanent magnet synchronous motor, in: 2015
IEEE Advanced Information Technology, Electronic and
Automation Control Conference (IAEAC), 2015: pp. 498–502.
[43] M. El Azzaoui, H. Mahmoudi, Design and simulation of the
space vector modulation and applied to a load RL powered by a
voltage inverter, International Journal of Advanced Computer
Research. 6 (2016) 153.
[44] A. Mishra, V. Mahajan, P. Agarwal, S.P. Srivastava, MRAS
based estimation of speed in sensorless PMSM drive, in: 2012
IEEE Fifth Power India Conference, 2012: pp. 1–5.
[45] . Shahgholian, M.H. Rezaei, A. Etesami, M.R. Yousefi,
Simulation of speed sensor less control of PMSM based on
DTC method with MRAS, in: 2010 Conference Proceedings
IPEC, 2010: pp. 40–45.
[46] D. Marshiana, P. Thirusakthimurugan, Fractional order PI
controller for nonlinear systems, in: 2014 International
Conference on Control, Instrumentation, Communication and
Computational Technologies (ICCICCT), 2014: pp. 322–326.
[47] M.I. Solihin, L.F. Tack, M.L. Kean, Tuning of PID controller
using particle swarm optimization (PSO), in: Proceeding of the
International Conference on Advanced Science, Engineering
and Information Technology, 2011: pp. 458–461.