Alexandria Engineering Journal 108 (2024) 518–538

Contents lists available at ScienceDirect

Alexandria Engineering Journal

journal homepage: www.elsevier.com/locate/aej

Performance enhancement of sensorless induction motor drive using

modified direct torque control techniques for traction application
S. Usha a , P. Geetha b , A. Geetha a , R. Palanisamy a , T.M. Thamizh Thentral a , Bidyut Mahato b ,
Nimay Chandra Giri c, * , Mohammed Alharbi d, *
a
Department of Electrical and Electronics Engineering, SRM Institute of Science and Technology, Kattankulathur, Tamil Nadu 603203, India
b
Department of Electrical and Electronics Engineering, Galgotias College of Engineering and Technology, Greater Noida, India
c
Department of Electronics and Communication Engineering, Centurion University of Technology and Management, Jatni, Odisha 752050, India
d
Department of Electrical Engineering, College of Engineering, King Saud University, Riyadh 11421, Saudi Arabia

A R T I C L E I N F O A B S T R A C T

Keywords: In recent scenarios, induction motors are used in many different fields due to their small size, advanced tech-
IM drive nology, and capacity to handle large amounts of power. One of the annoying downsides of conventional AC
Five phases traction drive systems is the engine’s speed and the uncertainty of the ripple content. The goal of the proposed
VSI inverter
work is to use various control techniques to analyze and enhance the performance of two five-phase induction
Direct torque control
Space vector modulation (SVM)
motors that are supplied in parallel via an inverter. The control techniques are scalar control, six-step voltage
Adaptive model referencing system traction type source inverter (VSI), space vector modulation (SVM) based VSI inverter, field-oriented control, model
application reference adaptive system-based field-oriented control (FOC), direct torque control (DTC), direct torque control
-SVM, model reference adaptive system based direct torque control-SVM for traction application. The modified
space vector modulation-based induction motor (IM) sensorless drive is realized together with a comparison of
several control strategies. It makes use of direct torque control approaches with a model reference adaptive
system (MRAS). An IM is designed, and performances are analyzed in high-speed and low-speed regions. Two
bogies sets were considered in the closed-loop scheme, tested under parameter variations and in low-speed areas.
The simulation result examined the most significant control methods for an IM fed by an inverter for electric
traction applications. The performance of the torque in terms of reduced torque ripple in the IM drive is
analyzed. The proposed direct torque control technique will reduce torque harmonic ripple by 100 % under the
steady-state condition. Under a change in a parameter, the torque harmonic ripple is decreased by 99 %. For the
IM drive, the hardware implementation of the advanced control approaches is examined.

1. Introduction difficult situations, SVM is selected for high-performance and precise

control applications. While weighing aspects like complexity, cost,
An IM provides several advantages, which include a clearly defined performance, and resilience, each strategy offers advantages and is
construction, known performance, and a limited capacity of determined appropriate for a variety of motor drive applications. A power boost
power. Sensor-type parallel linked induction motors needed an inverter converter, a single voltage power source inverter with a primary in-
configuration for every motor under conventional electric traction. The duction motor, and a fading combination rectifier are all components of
application-specific needs must be taken into consideration while traditional traction drives [1]. The alternating electric current motors
selecting the right control method for IM drives. Scalar control is best have countless and explicit improvements, mainly the structure is very
suited for straightforward, budget-conscious applications, whereas VSI simple and observable, well-grounded, and defined operation, the
is more versatile and works with a wider range of control schemes. While small-scale quantity of precise energy. Furthermore, the stable magnif-
MRAS-based techniques are favoured for adaptive, sensor-less control in icence and the process of enormous force changing/fluctuating

* Corresponding authors.
E-mail addresses: [email protected] (S. Usha), [email protected] (P. Geetha), [email protected] (A. Geetha), [email protected]
(R. Palanisamy), [email protected] (T.M.T. Thentral), [email protected] (B. Mahato), [email protected] (N.C. Giri), [email protected]
(M. Alharbi).

https://doi.org/10.1016/j.aej.2024.07.095
Received 1 June 2024; Received in revised form 22 July 2024; Accepted 23 July 2024
Available online 5 August 2024
1110-0168/© 2024 The Author(s). 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/).
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

wavelength machine together with its domination approach are used. In current equations, etc. A DC to AC converter coupled with a sensor for
its development by the traction drive, the alternating current drive such speed is commonly used to power an AC motor, although such a speed
as a locomotive grip force has come to more significant criteria. Right sensor may diminish system dependability and increase expenditures on
now, the IM is enforced in the grip force motor and has more drawbacks investment. Its execution is equally difficult [15,16].
such as reliability, speed variation, eminent torsion rippling, and control Many kinds of literature have proven that both hard, as well as soft
technique therefore it is not dynamic. Moreover, in the sensor-type switching techniques used in inverters, are comparable for analyzing the
parallel, combined induction motors in the current electric traction circulating bearing current and shaft voltages. The spikes occurring at a
drive system required the construction of an inverter associated with common-mode voltage (CMV) can likely be removed by moving the
each motor [2]. dead time over the phase pole. The authors [17] propose a carrier-based
The suggested arrangement of single inverter-fed polyphase induc- pulse width modulation (PWM) for reducing the CMV in three phases to
tion motors in parallel reduces the complexity and cost of the standard the polyphase. CMV produces is the cause of producing the circulating
way to create an energy-efficient traction drive. When compared to bearing current. CMV may also result in control equipment failures in
three-phase machinery, polyphase has various advantages. Scalar-type many industries. Many researchers have observed CMV has been the
controlling techniques were initially utilized to maintain the induction reason for the fault occurrence and bearing failure in motors [18].
motors, however, the methodology became difficult due to calculations. Circulating bearing current (CBC) is a significant issue in inverter-fed
The motor speed was measured using Tacho-generators, and the input parallel-connected traction motors. A 2-level DC to AC converter with a
feedback type system’s speed variation was calculated using the Lya- greater CMV will result in tedious deviations in voltage (dv/dt) along
punov formula. The system requires improved sophisticated controlling with the generation of higher frequency CBC. The bearing-to-CMV ratio
performance, precision, as well as torque power responsiveness [3]. is investigated for various machine ratings [19]. High common mode
Analyzing the characteristics of the machine and designing controls voltage generates a common mode current that leads to different issues
require accurate IM modelling. Speed, torque, and flux are examples of in the industries to damage motor bearings, which results in the faulty
system performance indicators that are impacted by a controller’s operation of machines as well as relay tripping. The major reason for
design. The first basic principle method, system mechanism framework those faults and bearing damage is because of common-mode voltage
verification methods, finite element component technique, comparable (CMV) [20,21]. The completion of common-mode voltage (CMV) pre-
magnetized circuit methodology, standard differential equations, pares the machine for bearing current elimination.
linear/nonlinear differential equation, a structure via lumped parameter The DTC techniques in IM drives are analyzed in low-speed regions,
and disrupted parameter, both static and dynamic modeling are among often in the region of 100 RPM. In terms of response time, many LSTM
the various forms of mathematical computational modeling [4]. The models have also been shown. Using the MATLAB/Simulink model, a
electrical machine notion uses a mathematical computational model comprehensive comparison of LSTM, fuzzy, and ANFIS-based DTC was
specified by variable differential equations. Partial differential equa- performed. The performance has been assessed in both constant and
tions can be used to describe the temperature distribution, magnetic transient situations [22]. The PMSM drive is analyzed for traction
field, and other characteristics of electrical equipment. One drawback of application. The generated high-frequency resistance may therefore be
employing partial solutions differential equations for research is that used to monitor the permanent magnet temperature. The method’s
they are more difficult to solve. [5]. nonintrusive character is demonstrated by the removal of extra sensors
To improve the accuracy as well as reliability of the traction motor, and the absence of any hardware changes to the existing system. Finally,
sensor-less speed control is required under low-speed performance. With experimental findings are used to validate the efficacy of the proposed
the proper usage of resources, it is possible to reduce the cost. Tem- method [23]. One common technique for managing the drive of a per-
perature variation under steady as well as transient conditions is pre- manent magnet synchronous motor (PMSM) is direct torque control
sent. Hence it is necessary to perform thermal analysis for both (DTC). One drawback of DTC is its high torque and flux ripples.
conditions to improve efficiency. Hence satisfactory speed deviation and Recently, several duty-cycle-controlled DTC strategies have been
the relevant speed control are not stated in the literature so far [6,7]. developed to increase the PMSM’s steady-state responsiveness. On the
Because of their dependability, durability, and low cost, induction mo- other hand, the present duty-cycle-controlled DTC solutions are so-
tors are the most often utilized kind of motor in industries. At first, a phisticated and result in greater switching frequency. This paper offers a
direct current (DC) motor was employed since it provides controllable unique duty-controlled DTC for decreasing torque and flux ripple
torque as well as speed. However, because of the presence of a without significantly increasing switching frequency or calculation
commutator in it, it requires regular servicing, which could be avoided overhead [24].
by employing an induction motor. The absence of a commutator which Two different voltage vectors (VVs) are used by the proposed DTC to
is present in an IM results in a reduced size, cost as well as a wider range regulate stator flux and torque during two different sampling intervals.
of applications. Unfortunately, the IM had an issue with speed control, Candidate VVs are selected using a look-up database, and their duty
which was difficult to achieve [8]. ratios are then calculated to get the stator flux and torque reference
The speed of the IM is controlled by advances in power electronic values in two sample intervals. As a result, stator-flux ripples and torque
converters [9,10], which outline various methods of speed control that are greatly reduced. In addition, the proposed method establishes the
make use of a proportional-integral (PI) controller. Inverters were now order of potential VVs to reduce the quantity of switches and, thus, the
employed to regulate the motor’s speed. The use of an inverter solved frequency of switching. The suggested approach is put to the test in an
the problem of speed control but also caused another difficulty with the experiment and contrasted with conventional DTC, an active duty-based
durability of an IM powered by an inverter [11,12]. Here, the reference DTC, and deadbeat DTC to confirm its effectiveness [25].
frame considered is a synchronous revolving rotor flux orientation and The researcher proposes a novel technique for direct torque control
so the traction machine is modeled using it. Accuracy with at least some (DTC) of a two-level inverter (2LI) at constant switching frequency,
of the information of a motor that is asynchronous was required for utilizing fractional-order sliding mode control (FOSMC) in conjunction
vector control without sensors and IM control approaches. Thermal and with space vector pulse width modulation (SVPWM). This suggested
motion analysis induction motors were analyzed in a motor solver [13, controller’s efficacy is contrasted with that of a traditional sliding mode
14]. controller (SMC) and a conventional proportional-integral controller
Generally, while performing the dynamic modeling of a machine, [26]. Fractional-order controllers (FOSMCs) provide an optimal and
speed versus torque characteristics, as well as the mechanical machine more resilient control system compared to the SMC. This method yields
equations, are considered. This modeling can able to offer flux linkages, the optimal performance and control efforts. These algorithms can be
differential voltage equations, the rotating rotor, stationary stator successfully applied since power converters are by their very nature

519
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

variable structure systems. Differentiators and integrators of fractional ⎡ ⎤

2π 4π 6π 8π
order are used by these controllers [27].
Vd ⎢ cos(ωt) cos(ωt − 5 ) cos(ωt − 5
) cos(ωt −
5
) cos(ωt − )⎥
5 ⎥
By combining the high resilience of the sliding-mode observer (SMO) =⎢
Vq ⎣ 2π 4π 6π 8π ⎦
with the flexibility of the model reference adaptive system (MRAS), the sin(ωt) sin(ωt − sin(ωt − ) sin(ωt − ) sin(ωt − )
sliding-mode-based MRAS can improve the speed estimate performance 5 5 5 5
of speed-sensorless IM (IM) drives. However, the analysis in this article (6)
shows that there is a mismatch problem between the fixed value of SMO The equation of direct and quadrature axis voltages is given by the
gains and the changing stability boundary, which changes depending on following.
the speed of IM operation. This results in a decrease in speed estimation
performance or even instability in the high-speed range [28]. This d
Vds = Rs ids + ψ − ωe ψqs (7)
problem is addressed by suggesting a discrete SMO, which increases the dt ds
effective operating range of the sliding-mode-based MRAS [29]. d
The new FO controller design process for the Voltage Source Inverter Vqs = Rs iqs + ψ + ωe ψds (8)
dt qs
(VSI) fed IM is presented in this brief. Using the chirp signal injection
approach with the Hardware-In-Loop (HIL) system for the laboratory d
Vdr = Rr idr + ψ − (ωe − ωr )ψqr (9)
prototype, the real FO VSI-fed IM transfer function model is identified dt dr
[30].
Contribution of the paper: The proposed work concentrates on the d
Vqr = Rr iqr + ψ + (ωe − ωr ) ψ dr (10)
different control techniques: scalar control, six-step VSI inverter, SVM- dt qr
based VSI inverter, FOC, Model reference adaptive system-based FOC, Equations for flux linkages in the d-q axis are given in the following
Direct torque control (DTC), Direct torque control (SVM), Direct torque
control (SVM) is based on a model reference adaptive system. The ψds = L1s ids + Lm, (ids + idr ) (11)
outcome of the simulation looked at the most important control strate-
gies for an IM that is fed by an inverter and used for electric traction.
ψdr = L1r idr + Lm, (ids + idr ) (12)
Two bogies setups were considered in the closed-loop scheme and tested ( )
ψ qs = L1s ids + Lm, iqs + iqr (13)
under parameter variation and low-speed region. The IM drive’s hard-
ware implementation of better control techniques is examined.
ψqr = L1r idr + Lm, (iqs + iqr ) (14)
The different control techniques for IM drive are described in Section
2. The simulated results and discussion are explained in Section 3. The
ψdm = Lm, (ids + iqr ) (15)
hardware implementation is examined in Section 4. The proposed work
is concluded in Section 5.
ψqm = Lm, (iqs + iqr ) (16)

2. Control techniques of IM drive The flux linkage in the d and q axis depends on the direct and
quadrature axis stator and rotor current.
The scalar control technique is adopted when specific performance is The equation for direct axis and quadrature axis currents in terms of
not required and for inexpensive applications. The method’s main flux and inductance are given by the following
disadvantage is less accuracy and torque response under parameter
variations due to the coupling effect, but it is simple to carry out. For
ψds (L1r + Lm ) − Lm ψdr
ids = (17)
more excellent dynamic performance, field-oriented control (FOC) is (L1s L1r + L1s Lm + L1r Lm )
adopted [31]. In recent times, direct torque control (DTC) holds the
ψqs (L1r + Lm ) − Lm ψqr
advantage of an electric propulsion mechanism that produces faster iqs = (18)
torque control and will not require large calculations [32]. The main (L1s L1r + L1s Lm + L1r Lm )
attitude of the DTC method is the direct selection of ideal inverter
ψds (L1r + Lm ) − Lm ψds
switching states. Sensor-less Direct -Vector- Control (DVC) method is idr = (19)
(L1s L1r + L1s Lm + L1r Lm )
used for AC locomotive applications widely [33]. The proposed work
concentrates on various control techniques that are applied on Induction
ψds (L1r + Lm ) − Lm ψds
motors powered by a voltage power source inverter (VSI), for propulsive idr = (20)
(L1s L1r + L1s Lm + L1r Lm )
motor applications. The equation of five-phase IM voltage is represented
by the following (19). The equation for electrical torque developed in the rotor of a quinary
motor is given by (19).
Va = Vm Sin (ωt) (1)
( ) Te = PLm (iqs idr − ids iqr ) (21)
2π
Vb = Vm sin ωt − (2) The equation of reference speed of the rotor is given by the Eq. (19).
5
∫
P
(
4π
) ωr = (Te − T1 )dt (22)
Vc = Vm sin ωt − (3) 2J
5
The IM is modeled based on the above design in the proposed work.
( )
6π
Ve = Vm sin ωt − (4) 2.1. Six-step voltage power source inverter
5
( )
8π The Six-step voltage power source inverter-fed induction motors
Vf = Vm sin ωt − (5) constitute a voltage /frequency regulator being exposed in Fig. 1a. The
5
stator frequency is obtained from the reference speed, and negotiating
The conversion of five-phase voltages towards the d-q axis is given by slip frequency. The flux location is computed and utilized to develop the
six-step pulse required by the polyphase inverter [34]. The major benefit
of this technique is its easiness and implementation over the other

520
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Fig. 1. (a) Six-step voltage source inverters fed five five-phase motors in parallel (b) Six-step voltage source inverters control techniques.

methods. Due to the coupling effect among flux and torque, IM drive / /
becomes more unstable and establishes a slower dynamic response. The Vm (R1 + Rʹ2 s) (25)
Voltage/Frequency ratio is maintained within a range of nominal speed. The phase current fundamental is
The motor runs in a weaker flux when operated above the minimum √̅̅̅̅̅̅̅̅̅̅̅̅̅̅
speed. Beyond minimal speed, the motor activates in the weaker flux Im = (Imag2 + I2t ) (26)
mode; the voltage of the motor is sustained constant, but the frequency
is raised proportionally to the reference speed [35]. While reversing the Based on the above basic equation the six-step inverter-fed type IM is
speed, a small suspension is needed the zero speed to let air gap flux fade analyzed.
to zero. The six-step inverter-fed AC motor drive is simulated and the The motor operating conditions are comparable to those of a voltage
speed and torque performance are analyzed. Fig. 1b shows how the source system. Normally, this will only occur if the motor is linked with
standard Sinusoidal pulse width modulation approach is utilized to the power source, however, in the present situation, the voltage power
regulate the switches in the inverter to control the motor’s speed. The source waveforms contain some harmonics as well as vary in a particular
equation is described below (17) frequency. Torque is then created by the fundamental element of the
The relation between voltage and current is described by the firing motor’s current in induction motors, and then the harmonic components
angle are ancillary to the regular operation of the motor. Similarly, to a motor
that is linked with the mains, thus the rated current consumed by the
Vc = 1.35XVs XCos(Alpha) + 1.35XVs X{Cos(Alpha − 30) motor is directly proportional to its parameters, notably its magnetiza-
(23)
− Cos(Alpha) }XFN (Idc ) tion needs leakage reactance, load as well as resistance. This current’s
power factor absorbs its fundamental value. In this system, the inverter
Where,FN (Idc ) - load current is a minimum impedance source of power which enables the motor
The motor terminal voltage is given by parameters to control the current level along with phase angle [14].
√2
Vm = XVc (24)
√3 2.2. Modified space vector modulation-based voltage source inverter
The torque component is
Due to the internal coupling effect within torque and flux in the

521
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

machine, the dynamic response of this drive is slow. The sampling time
of the speed controller must be an integral multiple of the space vector
pulse width modulation (SVPWM) sampling time. The size of the
simulation step should be a hundred times less than the switching
period. The modified SVPWM techniques are illustrated in Fig. 2. The
SVPWM-based VSI fed IM drive is simulated and the speed and torque
performance is analyzed. SVPWM-based AC motor drive consists of a
feedback loop along with slip regulation. The output of the speed control
circuit is to be added with the measured speed of the rotor to derive
stator flux frequency. The required PWM signals are obtained by
computing reference voltage and stator frequency. Easy implementation
and simplicity portray the benefit of this method over the other vector- Fig. 3. Space vector modulation-based voltage source inverter fed five-phase
controlled drives and these are shown in Fig. 3. The equation is IM in parallel.
described below (19)
The space vector modulation-based voltage inverter is designed by integral controller in the speed control circuit. Motor line current at five
the following fundamental equations. The equation to find the times for junctions corresponds to torque and flux references and is calculated by
voltages is given by the FOC controller. A poly-phase current regulator is utilized to feed the
five currents to the motor. Motor speed control, current, and torque
T1 = T∗s M∗ Sin(60 − θ) (27)
signals are obtained at the side of the output. For even operations, the
T2 = T∗s M∗ Sin(60) (28) FOC method is used at high performance. Currents id (direct axis) and iq
(quadrature axis) are preferred over the phase currents. They provide
TO = Ts − T1 − T2 (29) flux and torque that are independent of one another in the motor. By
denotation, the iq current and flux of the stator are in phase with each
Where, M - modulation index, other and id is vertical to iq. The FOC-based VSI-fed induction drive
demonstrated in Fig. 4 has improved performance compared with the
M=
√3Vref
(30) traditional technique but the energy efficacy is not sufficient because of
E the speed sensor.
Based on the above basic equation space vector pulse width
modulation-based five-phase IM is analyzed. 2.4. Modified sensorless field-oriented control
Take a controller, whose sampling time is 20 µs, better simulation
outcomes are obtained for a step simulation time of 20 µs. However, the The speed sensor is no longer needed. Motor speeds, current, output,
time step cannot be higher than the controller time step. The speed
controller works based on space vector modulation which controls the
motor slip. The motor speed is added to the slip value calculated by the
controller to generate the necessary inverting frequency. The Voltage/
Frequency (V/F) ratio is maintained constant with the latter’s frequency
being used to derive the stipulated inverter voltage. The performance of
the voltage power source inverter-fed IM drive powered by SVPWM
enhanced over the standard technique.

2.3. Sensor-based field-oriented control

Sensor-based FOC method is carried out for the proposed induction

motors drive. The reference value of Flux and torque is compared to the
Fig. 4. Field-oriented sensor-based control power of a 5-phase IM drive.
actual values of the motor by using a comparator with a Proportional

Fig. 2. Space vector modulation techniques for five-phase induction motor.

522
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

and torque indications have been obtained. Thus, the proposed MRAS
J
∑
technique was adopted for adaptive motor control. The proposed MRAS Ok = ωkj Øj (x) (37)
method which is simple and more efficient is demonstrated in Fig. 5. j=1
This technique has two prototypes; one is an adjustable model and the
other is a voltage reference model. The Vref model does not need the ωkj (t + 1) = ωkj (t) + ηe (t)Øj (38)
measure to be found, whereas the adaptive model needs the rotor speed. ( )
In the conventional MRAS design scheme, the voltage reference model is Øj ωj x − cj
cj (t + 1) = cj (t) + ηe (t) (39)
associated with the output of adjustable prototypical and the error for σ2

driving adequate adjustable machinery, where output quantity is ⃦ ⃦2

required to be determined. − ⃦x − cj ⃦
Øj (x) = exp⌊ ⌋ (40)
This method is proceeded by supplying power to the IM with a PWM 2
σj
VSI (voltage source type inverter), that is constructed by a Universal
⃦ ⃦
Bridge Block. The references are compared and checked with the actual ηe(t)Øj ωj ⃦x − cj ⃦
values of the motor by using a comparator with a Proportional integral σ j (t + 1) = σj (t) + (41)
σ3
controller in the speed control circuit. Motor line current at five junc-
The equation for the speed adaptation technique is given by
tions corresponding to torque and flux references is calculated by the
∫
FOC controller. The five currents are sent to the motor through a poly- ( )
ωr = Kp eids ψ qr s − eiqs ψ dr s + K1 (eids ψ qr s − eiqs ψ dr s )dt
phase current regulator. Motor speed, current, and torque signals are
obtained at the side of the output. For even operations, the FOC method (42)
is used at high performance. Currents id (direct axis) and iq (quadrature
The modified neural network-based MRAS technique is developed to
axis) are preferred over the phase currents. Its speed is approximated by
achieve the sensorless IM drive for traction application. The modified
terminal currents and voltages on the basis MRAS technique as shown in
MRAS method is modeled to estimate the driving speed of an induction
Fig. 6a. Fig. 6a shows that the FOC-based voltage power source inverter
motor. The adaptive mechanism compares the predicted speed to the
powered by an induction drive outperforms the conventional method.
reference speed until the speed error deviation is nil (zero).
The equation of reference models is given by (28)
The five currents are sent to the motor through a poly-phase current
∫
− 1 s L regulator. At the output, signals for motor speed, current, and torque are
ψ̂ sdr = ψ − ωr ψ sqr + m isds (31)
Tr dr Tr acquired. For even operations, the FOC method is used at high perfor-
∫ mance. Currents id (direct axis) and iq (quadrature axis) are preferred
1 s L over the phase currents. Based on the MRAS (Model Referencing
ψ̂ sqr = ωr ψ sdr − ψ + m is (32)
Tr qr Tr qs Adaptive System) approach, its speed is estimated using terminal cur-
For modeling the reference model which is present in an IM drive and rents and voltages. As a result, the speed sensor is no longer required. At
executing the adaptive model reference scheme, the rotor flux present in the output, signals for motor speed, current, and torque are acquired.
the direct and quadrature axes is considered 28). The suggested MRAS approach is used for motor adaptive control. The
The equation for the adaptive model is given by field-oriented control techniques are depicted in Fig. 6b.
) As a result, in the proposed artificial neural network (ANN) based on
ψ dr (k) = w1 ψ ∗ dr (k − 1) −
∗
w2 ψ ∗ qr (k − 1) + w3 ids (k − 1 (33) MRAS, the standard model that is adaptable and is substituted using the
artificial neural network, as well as the PI controller is modified by ANN
ψ ∗ qr (k) = w1 ψ ∗ qr (k − 1) − w2 ψ ∗ dr (k − 1) + w3 iqs (k − 1) (34) in the adaptation process. Thus, the output of the two different methods
is correlated with each other until there are nil faults between the two
The calculated rotor flux in direct and quadrature axes is used to
methods. The sensorless induction drive characteristics of the FOC-
create the adaptive model which is present in an IM drive for executing
based voltage source inverter have been established Relative to the
the adaptive model reference scheme.
typical technique, however, the sensorless drive provides adequate ef-
The error signal is given by
ficiency in energy.
Ts Ts
W1 = (1 − ) , W2 = ωr Ts , W3 = Lm (35)
Tr Tr
2.5. Direct torque control
̂ sdr ψ sqr − ψ sdr ψ
ξ=ψ ̂ sqr (36)
This technique is carried out by feeding the IM with a PWM voltage
Where ψ
s
̂ dr &ψ ∗ qr are the fluxes calculated along the d-q axis source inverter (VSI) that is constructed by a Universal type Bridge
Block. A PI controller is used in the speed control circuit to generate
The equation for neural network depends on an Adaptive approach
torque and flux references for the DTC. It calculates the flux and torque
equation that is represented by the following
and then compares the same with the respective references. Inverter
switching pulses are generated by the output gained by comparators by
optimal switching. Motor speed, current, and torque signals are obtained
at the output side. Thus, the flux and torque calculator block are
inculcated to determine the electromagnetic torque and rotor flux
components (αβ). This calculator works depending on the equation
synthesis in the motor. The αβ vector square is utilized to determine the
αβ plane’s sector where the flux vector is present. The Torque and Flux
Hysteresis block consists of a three-tier hysteresis comparator for torque
monitoring and control and a two-tier hysteresis comparator needed for
flux regulation. The Switching table block consists of two peruse tables
that select a particular voltage vector in terms of the output of the
comparators of Torque and Flux Hysteresis. The machine’s initial flux is
also produced in the same block that is illustrated in Fig. 7a. The direct
Fig. 5. Model reference adaptive system and control. torque control techniques are represented in Fig. 7b. The equation is

523
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Fig. 6. (a) Sensorless field-oriented IM drive control (b) Field-oriented-control block diagram.

described below (25) speed, current, and torque are acquired. The IM drive shown in Fig. 8,
The electromagnetic torque is described as follows remodels a DTC with a modified adaptive system reference model.
3P( )
Te = . . λsd ∗ isq − λsq ∗ isd (43) 2.7. Proposed method of sensorless direct torque controlled IM drive
22
The torque is directly proportional to the stator flux A pulse width modulation VSI provides the motor in this manner.
2 Thus, the speed control circuit employs a PI controller to develop the
3 P |Vs | Rr /S
Te = . . . (44) torque and flux references required by the DTC-SVM controller. The
2 2 πf (Rs + Rr /S)2 + (Xis + Xir )2
DTC-SVM controller calculates five-phase motor line current references
to torque and flux and then feeds these currents to the motor by a poly-
|Vs |
Where, |λs | = (45) phase current regulator being shown in Fig. 9. Motor speed is deter-
2π f
mined from terminal currents and voltages using the MRAS method.
Hence, speed sensors are not required anymore also no longer based on
the hysteresis controller. The estimated speed, torque, and current sig-
2.6. Space vector modulation for direct torque control
nals are taken as an output from the block. The performance of the DTC-
based voltage power source inverter powered by a sensor-less induction
This method involves feeding the IM through a PWM VSI constructed
drive has been enhanced over the previous technique, and the energy
with a Universal Junction Bridge Block. A proportional-integral
efficiency is satisfactory due to the speed sensor. The equation is
controller is used in the speed-controlling loop to generate the torque
described below (25)
and flux parameter references required to trigger the DTC block. The
The calculation equation for switching times is given by
Direct torque control block calculates and compares the motor flux and
torque to their respective references. At the output, signals for motor

524
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Fig. 7. (a) DTC of IM drive (b) Direct torque control techniques.

Fig. 8. Sensorless IM drive DTC.

Vs = Vsd + jVsq (46) applications. Tom checked the performance of the proposed work for
traction application, the IM drive with 147 kW motor rating is consid-
Where the vectors Vsd and Vsq are calculated from the given sector ered in the simulation circuit but is scaled down to 2.2 kW for hardware
Sin(3π − γ) setup. The techniques are scalar control, six-step VSI inverter, SVM-
TA = TS a (47) based VSI inverter, field-oriented control, Model reference adaptive
Sin(3π)
system-based FOC, DTC, DTC -SVM, and Model reference adaptive
The above-mentioned equations are used to examine a space vector system-based direct torque control-SVM. Two bogies setups were
modulation-based direct torque-regulated induction motor. considered in the closed-loop scheme, it is tested under parameter
variation and low-speed region. From the simulation result, the most
3. Simulation results and discussions significant control methods for an inverter-fed induction machine for
electric traction applications were analyzed by the following results and
The proposed simulation and results of the various control schemes discussions.
for the VSI-fed IM for the traction applications are analyzed using
MATLAB/ Simulink. This research work is mainly based on traction

525
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Fig. 9. Space vector modulation-based Sensorless DTC.

3.1. Six-step voltage source inverter fed parallel connected induction

motors

Hence, the speed control and torque are analyzed for two bogies
under the very high-speed region. Thus, the speed control is analyzed for
five different regions of locomotive traction. The regions are notching,
acceleration-free running, coaching, and breaking. The proposed work is
concentrated on locomotive traction application. A single inverter
feeding two induction motors in parallel must be studied for locomotive
propulsion. As seen in Fig. 10, the six-step voltage source inverter-fed IM Fig. 11. Speed performance of six-step voltage source inverter fed induction
drive achieves a steady state within 5 seconds of simulation. The ripple motors for motor 1 in the low-speed region.
in torque is higher in this conventional method mentioned. The speed
response IM drive under low-speed performance is poor because after proposed work is concentrated on locomotive traction application. For
5 seconds also the speed does not attain a steady state as demonstrated locomotive propulsion, a single inverter powering two parallel induc-
in Figs. 11 and 12. tion motors must be studied for various areas. The proposed induction
drive is analyzed for parameter variation conditions to achieve robust
performance. The parameter variation considered for SVM-based
3.2. Proposed space vector modulation-based voltage power source voltage source inverter-fed induction motors are changes in stator
inverter powered by a parallel-connected induction motors resistance and low-speed region. The motor reaches steady-state very
quickly compared to the conventional six-step inverter mentioned
The performance of the SVM-based VSI IM drive is illustrated in above. In this method, the speed of the motor follows the set speed, so
Fig. 13. The speed and torque are analyzed for two bogies setups. The

Fig. 10. Performance six-step voltage power source inverter fed proposed induction motors in high-speed region a) Motor 1 b) Motor 2.

526
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

resistance variation, but under low speed and low inertia, it’s affected as
illustrated in Fig. 16.

3.4. Proposed sensorless field-oriented control of IM

The performance of sensor-less FOC IM drives with soft computing-

based MRAS techniques is illustrated in Fig. 17. The speed and torque
Fig. 12. Speed performance of six-step voltage power source inverter fed in- are analyzed for two bogies setups. The speed is analyzed for five
duction motors for motor 2 in the low-speed region. different regions of locomotive traction. The regions are notching, ac-
celeration and free running, coaching, and breaking. For locomotive
zero error is observed. traction a single inverter powered two induction motors in parallel
The speed is analyzed for five different regions of locomotive trac- needs to be analyzed for the different regions with variations in pa-
tion. The regions are notching, acceleration-free running, coaching, and rameters. The robust control of sensorless FOC of Induction Motors drive
breaking. Oscillation in speed and ripple in torque is completely reduced is analyzed in conditions such as stator resistance, rotor resistance, low
in this method but the performance is affected (not robust) under low- inertia, and low-speed region. The performance under normal and
speed regions and a 40 % decrease in stator resistance as shown in parameter variation conditions is depicted in Figs. 17 and 18 respec-
Fig. 14. tively. Reference speed is represented in yellow color is in matches the
estimated speed which is denoted in red color as well as the approxi-
mated speed is shown in the color of blue in the rotor speed variation
3.3. Sensor-based FOC of IM in parallel
curve. In the electromagnetic torque curve, the torque ripple tends to be
lesser than the proposed method compared to previous methods are
Fig. 15 shows the performance of the sensor-based FOC IM driving.
demonstrated in Fig. 18. The performance of the drive is not affected in
The speed, torque, and stator current are analyzed for two bogies setups.
this method under stator and rotor resistance variation. Even under low
For locomotive traction, a single inverter powered by two parallel in-
speed and low inertia also performances have been improved compared
duction motors needs to be analyzed for the different regions with
to sensor-based FOC IM drives.
variations in parameters. The robust control of sensor-based FOC of
Induction Motors drive is analyzed in conditions such as stator and rotor
resistance change, very low inertia, and minimum-speed region. Thus, 3.5. Direct torque control of IM drive
the variation in speed is analyzed for five different regions of locomotive
traction. The regions are notching, acceleration-free running, coaching, Figs. 19 and 20 depict the performance behavior of the DTC-based IM
and breaking. As shown in Fig. 15, the motor reaches a steady state very drive, accordingly. The speed and torque are analyzed for two bogies
quickly compared to the conventional six-step VSI method mentioned setups. The speed is analyzed for five different regions of locomotive
above. traction. The regions are notching, acceleration and free running,
In this suggested method, the speed present in the motor follows the coaching, and breaking. For locomotive traction a single inverter pow-
set speed, so zero error. Oscillation in speed and ripple in torque is high ered by two parallel induction motors needs to be analyzed for different
compared to the SVM-based IM drive mentioned above. The perfor- regions with variations in parameters. The reference speed is displayed
mance of the drive is not affected in this method under stator and rotor in blue on the rotor speed curves and it is matched to the rotor speed

Fig. 13. Performance of SVM-based voltage power source inverter fed induction motors in high-speed region a) Bogie 1 b) Bogie 2.

527
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Fig. 14. Performance of space vector modulation based voltage source inverter induction motors a) Stator resistance variation b) Minimum-Speed Zone.

Fig. 15. Performance of sensor-based FOC of induction motors in high-speed region (a) Motor 1 (b) Motor 2.

which is represented in red color. Oscillation in stator current and ripple are notching, acceleration and free running, coaching, and breaking.
in torque is high compared to sensor-less FOC IM drive under parameter The proposed work is concentrated on locomotive traction application.
variation particularly under low-speed region due to hysteresis current For locomotive traction, a single inverter fed a single inverter powered
controller, the performance is affected as shown in Fig. 20. by two parallel induction motors needs to be analyzed. The reference
speed is shown in blue on the rotor speed variation curves and is related
to the rotor speed curve shown in red. Oscillation in stator current and
3.6. Proposed sensorless control direct torque control of IM drive ripple in torque is less compared to conventional DTC IM drive
mentioned above; the under low-speed region also performance is
The speed and torque are analyzed for two bogies setups. The speed improved as shown in Fig. 21. This Sensorless control direct torque
is analyzed for five different regions of locomotive traction. The regions

528
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Fig. 16. Performance of sensor-based FOC of induction motors a) Stator resistance change b) Minimum-speed c) Very Low inertia d) Rotor resistance.

Fig. 17. Performance of sensor less FOC of IM drive.

529
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Fig. 18. Performance of Sensor Less FOC of IM Drive a) Stator Resistance Change b) Minimum Speed.

Fig. 19. Performance of Direct Torque Control of IM Drive.

control is energy efficient and analyzed with the typical DTC. The per- suggested five-phase IM drive under five different regions. The regions
formance of SVM-based DTC of IM drives is illustrated in Figs. 21 and 22 are notching, acceleration and free running, coaching, and breaking.
respectively. The reference speed is displayed in blue, and it is paired with the pre-
dicted speed, which is depicted in green. During acceleration and
3.7. Suggested sensorless control of space vector modulation-based direct coaching periods certain duration alone there is a speed deviation.
torque control Overall the speed error deviation is nil (zero) as depicted in Fig. 25.
The speed performance is analyzed in the resistance to stator change
The proposed DTC-SVM controlling techniques for IM drive for conditions for the suggested five-phase IM drive under five different
traction application are illustrated in Fig. 23. The performance of the regions. The regions are notching, acceleration and free running,
proposed sensorless DTC-SVM-based IM drive with soft computing- coaching, and breaking. The reference speed is displayed in blue, and it
based MRAS techniques is illustrated in Figs. 24–32 respectively. The is paired with the predicted speed, which is depicted in red. During free
speed, torque, and stator current are analyzed for two bogies setups. The running and breaking periods certain duration alone there is a speed
speed is analyzed for five different regions of locomotive traction. deviation. Overall the speed error deviation is nil (zero) as depicted in
The speed performance behavior is analyzed in the very high-speed Fig. 26.
region for the suggested five-phase IM drive under five different re- The speed performance behavior is analyzed in a very minimum
gions. The regions are notching, acceleration and free running, coach- speed region for the suggested five-phase IM drive under five different
ing, and breaking. The proposed work is concentrated on locomotive regions. The regions are notching, acceleration and free running,
traction application. The reference speed is depicted in red, and it is coaching, and breaking. The reference speed is displayed in blue, and it
paired with the calculated speed, which is indicated in yellow. During is matched with the estimated speed, which is depicted in green. During
the coaching period certain duration alone, there is a speed deviation. the free-running period certain duration alone, there is a speed devia-
overall error speed is zero as depicted in Fig. 24. tion. Overall the speed error deviation is nil (zero) as depicted in Fig. 27.
The speed performance is analyzed in medium speed region for the Robust performance is achieved. The speed deviation has been fully

530
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Fig. 20. Performance of Direct Torque Control of IM Drive a) Change in Rotor Resistance Stator Resistance b) Low Inertia c) Low Speed.

decreased in comparison to all other techniques simulated above. suggested five-phase IM drive under five different regions. The regions
The speed performance is analyzed in the below medium speed re- are notching, acceleration and free running, coaching, and breaking.
gion for the suggested five-phase IM drive under five different regions. The reference torque is matched with the estimated torque with reduced
The regions are notching, acceleration and free running, coaching, and torque ripple as depicted in Figs. 29 - 32.
breaking. The reference speed is displayed in blue, and it is matched Oscillation in speed and ripple in torque is higher in the conventional
with the estimated speed, which is depicted in green. During a free- six-step voltage source inverter-fed polyphase IM in parallel, also motor
running period certain duration alone, there is a speed deviation. reaches a steady state after 5 seconds. In the proposed Space vector
Overall the speed error deviation is nil (zero) as depicted in Fig. 28. modulation-based voltage source inverter method, the motor reaches a
Robust performance is achieved. The speed deviation has been fully steady state very quickly compared with the previous method
reduced compared to all other methods simulated above. mentioned, and zero error is observed. Oscillation in speed and ripple in
The torque performance is analyzed in different conditions for the torque is completely reduced but under parameters, variation

531
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Fig. 21. Performance of suggested direct torque control of IM drives (a) Motor 1 (b) Motor 2.

Fig. 22. Direct torque control of induction motors using space vector modulation a) Stator and rotor resistance change b) Minimum-speed region.

performance is sluggish in response. In the third method-Sensor based fulfil the strict performance requirements of contemporary trans-
FOC, the speed and torque performance have been affected compared portation systems by utilizing cutting-edge control techniques like vec-
with the proposed SVM-based IM drive as depicted in Figs. 24 - 32. tor control, DTC, and SVM. This helps to create mobility solutions that
are more sustainable, dependable, and efficient.
4. Hardware implementation The performance obtained in simulation is also tested empirically in
the hardware configuration. Thus, the hardware configuration setup for
The usage of induction motors in traction application such us reli- inverter fed field-oriented controlled sensor less two induction motors
ability, high efficiency and cost effectiveness. The severe operating connected in parallel type with determined parameters using soft
circumstances and performance requirements of these applications call computing-based MRAS is shown in Fig. 33.
for strong, effective, and dependable motor drives. These drives can A Field-Programmable Gate Array (FPGA) is an integrated circuit

532
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Fig. 23. Proposed DTC-SVM control technique block diagram for IM drive.

Fig. 27. Speed Performancee of Proposed Five-Phase IM Drive Inlow-

Speedd Region.

Fig. 24. Speed performance of proposed five-phase IM drive in high-

speed region.

Fig. 28. Speed Performance of proposed five-phase IM drive in below medium

speed region.

Fig. 25. Speed Performance of proposed five-phase IM drive in medium

speed region.

Fig. 29. Torque performance of suggested five-phaseIM drive.

interconnects that enable these blocks to be wired together. Optimizing

the use of these components to achieve desired performance while
minimizing area and power consumption is known as efficient resource
Fig. 26. Speed Performance of Proposed Five-Phase on IM Drive for Change In use. Because FPGAs can process complex control algorithms quickly and
Stator Resistance. precisely, they are very effective at controlling IM drives.
The no-load test is carried out on a single inverter feeding two par-
that can be field-programmable that is, it can be altered by the designer allel motors through soft computing-based MRAS. For a given voltage,
or client after it is manufactured. Like many logic gates that may be different speeds are obtained. Furthermore, the speed variation between
inter-connected in many configurations, FPGAs are made up of an array two motors is reduced under open-loop settings, as shown in Table 1.
of programmable logic blocks and a hierarchy of reconfigurable The voltage value in 150 V and 230 V and the rated current is 0.7 A and

533
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

reference speed of 405, 520, and 605 rpm respectively. The distinction
between the preset and real speed approaches a value of zero therefore
the speed divergence produced between the two distinct motors is lower
than in Table 3. The estimated speed is matched with the reference
speed. The performance metrics of speed error for steady state condition
in closed loop are depicted in Table 4.
The varying speed at one specific voltage is attained by doing the test
on the same setup with various parameters. The voltage value is 150 V
Fig. 30. Torque performancee of proposed five-phaseIM drive for change in and 230 V, the rated current is 0.7 A and 1.3 A kept constant, and then
stator resistance. the speed is estimated for the two motors in the closed loop condition.
The estimated speed for 150 V is 415,530,630 for motor 1 and 411,530
and 626 rpm for motor 2 corresponding to the reference speed of 415,
530, and 635 rpm respectively. However, the speed difference between
the two motors is large in open-loop situations than in closed-loop set-
tings. The estimated speed for 230 V is 415,520,605 for motor 1 and
415,515 and 600 rpm for motor 2 corresponding to the reference speed
of 415, 520, and 605 rpm respectively Table 5 demonstrates that the
inaccuracy is significant. The closed-loop test is used to avoid speed

Table 1
Inverter fed two motors connected in parallel under Steady-state condition
Fig. 31. Torque performance of proposed five-phase IM drive in medium speed.
(open loop).
Applied Rated Rated Set Estimated Estimated
Voltage(V) Current(A) speed(RPM) Speed Speed
(motor1) (Motor 2)
(RPM) (RPM)

150 0.7 405 395 394

0.7 520 500 499
0.7 605 605 605
230 1.3 405 402 402
1.3 520 505 505
1.3 605 602 602

Fig. 32. Torque performance of proposed five-phase IM drive in low speed. Table 2
Speed Error under steady state condition (open loop).

1.3 A kept constant and the speed is estimated for the two motors. The Applied Rated Deviation between set Deviation between set
Voltage(V) Current and rated speed for and rated speed for
performance metrics of speed error for steady state condition are
(A) motor 1 (%) motor 2(%)
depicted in Table 2.
For the closed loop, the same configuration is used. The estimated 150 0.7 2.46 2.71
0.7 3.84 4
speed for 150 V is 404,520,602 for motor 1 and 402,520 and 600 rpm
0.7 0 0
for motor 2 corresponding to the reference speed of 405, 520, and 230 1.3 7.4 7.4
605 rpm respectively. The estimated speed for 230 V is 405,516,605 for 1.3 2.88 2.88
motor 1 and 402,515 and 605 rpm for motor 2 corresponding to the 1.3 4.9 4.9

Fig. 33. Voltage source inverter fed SVM based direct torque-controlled sensor less Induction motors.

534
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Table 3 Table 6
Inverter fed two motors connected in parallel under Steady-state condition Speed Error under parameter variation (open loop).
(closed loop). Applied Rated Deviation between set Deviation between set
Applied Rated Rated Set Estimated Estimated Speed Voltage(V) Current and rated speed for and rated speed for
Voltage(V) Current speed Speed (motor2) (RPM) (A) motor 1 (%) motor 2(%)
(A) (RPM) (Motor 1)
150 0.7 7.22 7.77
(RPM)
0.7 4.9 4.1
150 0.7 405 404 402 0.7 4 2.4
0.7 520 520 520 230 1.3 3.6 4.0
0.7 605 602 600 1.3 2.6 1.8
230 1.3 405 405 405 1.3 1.4 2.1
1.3 520 516 515
1.3 605 605 605

Table 7
Inverter fed two motors connected in parallel under parameter specification
Table 4 variations (closed loop).
Speed Error under steady state condition (closed loop).
Applied Rated Rated Set Estimated Speed E Speed
Applied Rated Deviation between set Deviation between set Voltage(V) Current speed (motor 1) (RPM) (motor 2)
Voltage(V) Current and rated speed for and rated speed for (A) (RPM) (RPM)
(A) motor 1 (%) motor 2(%)
150 0.7 415 415 411
150 0.7 2.4 7.4 0.7 530 530 530
0.7 0 0 0.7 635 630 626
0.7 4.9 8.26 230 1.3 415 415 415
230 1.3 0 0 1.3 520 520 515
1.3 7.6 9.6 1.3 605 605 600
1.3 0 0

differences between two motors. The performance metrics of speed error Table 8
for parameter variation open loop and in closed loop are depicted in Speed Error under parameter variation (closed loop).
Tables 6 and 8, respectively. The speed variance between the two motors Applied Rated Deviation between set Deviation between set
is relatively small under closed-loop analysis, and the system is very Voltage(V) Current and rated speed for and rated speed for
(A) motor 1 (%) motor 2(%)
stable, as shown in Table 7.
The speed performance behavior of the IM drive is analyzed for the 150 0.7 0 9.6
0.7 0 0
suggested method in open loop and closed loop using the present
0.7 7.8 14.2
hardware setup displayed in this paper. Figs. 34 and 35 show the speed 230 1.3 0 0
performance under steady-state conditions and parameter changes 1.3 0 9.6
respectively. Compared to the conventional technique, the suggested 1.3 0 8.2
method minimizes speed error even under parameter fluctuation situa-
tions in a closed-loop system.
evidenced by the comparison of simulation results. This includes
reduced torque ripple as well as speed deviations under steady state and
4.1. Comparative analysis parameter variations like rotor resistance, low inertia, and stator resis-
tance (Tables 9 and 10).
The suggested work has been examined, and by utilizing diverse The vector control, scalar control, as well as direct torquing control
control strategies and methodologies, the performance of two induction technique are analyzed for the drive of the IM as well as the data were
motors that are powered in parallel by an inverter has been improved. analyzed for identifying the advanced controlling technique. Thus, the
Direct torque control (DTC), space vector modulation (SVM) based VSI behavior of speed performance which is present in the IM is decided by
inverter, model reference adaptive system based VSI inverter for traction the setting time of the speed curve. Under steady-state conditions, the
application, field-oriented control, six-step voltage type source inverter speed curve is settled at 1.2 seconds, unstable, 0.01 seconds, and
(VSI), and scalar control are among the control methods. The closed- 0.0001 seconds for scalar control, six-step voltage inverter, direct torque
loop method was tested in low-speed and parameter-varying condi- control, and field-oriented control respectively. Under a change in the
tions, accounting for two bogies setups. The suggested DTC-SVM-based stator and rotor resistance variations, the speed curve is settled at
2-level voltage power source inverter-fed designed polyphase IM in 2.5 seconds, unstable, 0.01 seconds, and 0.0001 seconds for scalar
parallel has improved performance over conventional methods, as control, six-step voltage inverter, field-oriented control, and direct tor-
que control respectively. Under the low-speed region, the speed curve is
Table 5 settled at 8 seconds, 3.5, 2 seconds, and 0.0001 seconds for scalar con-
Inverter fed two motors connected in parallel under parameter specification trol, six-step voltage inverter, direct torque control, as well as field-
variation (open loop). oriented control respectively. Under unbalanced load conditions, the
speed curve is settled at 10 seconds, 3, 7 seconds, and 0.0001 seconds
Applied Rated Rated Set Estimated Speed Estimated
Voltage(V) Current speed (motor 1)(RPM) Speed (motor for scalar control, six-step voltage inverter, direct torque control, as well
(A) (RPM) 2) as field-oriented control respectively. The performance is analyzed for
(RPM) the torque in terms of reduced torque ripple in the IM drive. Under the
150 0.7 415 385 383 steady-state condition, When compared to the traditional technique,
0.7 530 504 508 improved direct torque control reduces torque harmonic ripple by
0.7 625 600 610 100 %. Under a change in a parameter, the torque harmonic ripple is
230 1.3 415 400 398
decreased by 99 % compared to traditional methods. The stability of the
1.3 530 516 520
1.3 615 606 602 system has improved in the proposed techniques, Compared to

535
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Fig. 34. The speed performance of IM drive –steady-state conditions a) open-loop b) closed-loop.

Fig. 35. The speed performance of IM drive – parameter specification variations conditions a) open-loop b) closed-loop.

Table 9
Speed and torque response analysis for change in Rs and Rr.
Different control Speed and torque Performance
techniques
Steady-state condition Under stator and rotor resistance variation

Speed noise Speed error Settling time in Torque Speed noise Speed error Settling time Torque ripple
(%) (%) sec ripple (%) (%) (sec) (%)
(%)

SCT 60 20 1.2 50 72 28 2.5 70

Modified SCT 50 10 1.8 40 50 20 3.1 60
Six step VSI 10 2 Unstable 48 10 20 Unstable 50
SVM based VSI 2 2 0.5 5 4 5 0.7 15
conventional FOC 5 2 0.01 78 5 4 0.01 80
Modified FOC 0 0 0.0001 5 0 1 0.001 5
conventional DTC 3 0 0.0001 35 5 2 0.001 70
Modified DTC 2 0 0.00001 10 4 0 0.00001 15
Modified SVM based 0 0 0.00001 0 0 0 0.00001 1
DTC

conventional DTC, the suggested DTC’s settling time has been improved control, speed, torque, and stator current are analyzed for two bogies
from 0.0001 to 0.00001. Since the torque ripple has been reduced by 99 setup. The speed deviation and torque ripple have been reduced by 99
percent the efficiency of the system have improved. percent compared with the all-other control methods simulated as
Performance of the drive is adequate in the proposed Sensorless FOC depicted in Tables 5 and 6.
method under stator and rotor resistance variation and also under low
speed and low inertia. In the Direct torque control method, performance ➢ Fast dynamic response and enhanced torque control are benefits of
is comparable to FOC IM drive, but under low-speed region, it is not SVM-based DTC. However, there are a number of drawbacks,
adequate due to the hysteresis current controller. In the modified sen- including complexity, expense, computing demands, and suscepti-
sorless Direct torque control, Oscillation in speed and ripple in torque is bility to changes in motor parameters.
less compared with the conventional DTC IM drive mentioned above; ➢ It takes sophisticated design methods, strong hardware, and accurate
furthermore, Performance is also improved in the low-speed area zone. parameter estimation to address these drawbacks and guarantee the
In the modified Space vector modulation-based sensorless Direct torque IM drive system operates as efficiently and reliably as possible.

536
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

Table 10
Speed and torque response analysis in LSR and unbalanced load conditions.
Different control Speed and torque Performance
techniques
Under low-speed region Under unbalanced load

Speed noise Speed error Settling time Torque ripple Speed noise Speed error Settling time Torque ripple
(%) (%) (sec) (%) (%) (%) (sec) (%)

SCT 68 35 8 60 60 25 10 70
Modified SCT 60 20 4.8 50 55 20 4.0 50
Six step VSI 40 10 3.5 58 30 20 3.0 48
SVM based VSI 30 20 4 45 28 35 8 39
conventional FOC 15 10 2 78 18 10 7 78
Modified FOC 2 0 0.001 5 5 1 0.001 5
conventional DTC 10 4 0.1 80 12 5 0.5 80
Modified DTC 6 0 0.0001 15 5 0 0.0001 20
Modified SVM based 0 0 0.0001 2 0 0 0.0001 2
DTC

5. Conclusion interests or personal relationships that could have appeared to influence

the work reported in this paper.
The performance analysis and improvement of inverter powered by
two Induction Motors connected in parallel using different control Acknowledgments
methods such as Scalar control, Six step VSI inverter, SVM-based VSI
inverter, FOC, Model reference adaptive system based FOC, Direct tor- This work was supported by the Researchers Supporting Project
que control-SVM, Model reference adaptive system based Direct torque number (RSP2024R467), King Saud University, Riyadh, Saudi Arabia.
control-SVM for traction application were carried out successfully. The The authors thank the SRM Institute of Science and Technology,
modified Space vector modulation-based sensorless Direct torque con- India, for supporting the study.
trol analyzes speed, torque, and stator current for two bogies setup. In
the Direct torque control method, performance was comparable to FOC References
IM drive, but under low-speed region, it is not adequate due to the
hysteresis current controller. The proposed DTC-SVM-based 2-level [1] K. Koga, R. Ueda, Constitution of v/f control for reducing the steady-state speed
error to zero in IM drive system, IEEE Trans. Ind. Appl. 28 (1992) 463–471,
voltage power source inverter-fed designed polyphase IM in parallel https://doi/10.1109/28.126757.
has enhanced performance compared to conventional methods in the [2] S. Ayman, et al., A five-phase induction machine model using multiple dq planes
presence of reduced torquing ripple and speed deviations under considering the effect of magnetic saturation, IEEE Energy Conv. Cons. Expo. 14
(2014) 287–293, https://doi/10.1109/TIE.2013.2297294.
parameter variations. In the modified sensorless Direct torque control, [3] D. Yadav, et al., Comparison of bioinspired techniques for tracking maximum
Oscillation in speed and ripple in torque is less compared with the power under variable environmental conditions, Int. J. Intel. Syst. 2024 (2024)
conventional DTC IM drive mentioned above; Performance is also 6678384, https://doi.org/10.1155/2024/6678384.
[4] C. Mi, Field-oriented control of IM drives with direct rotor current estimation for
improved in the low-speed area zone. The speed deviation and torque applications in electric and hybrid vehicles, J. Asian Elect. Veh. 5 (2015) 989–992,
ripple have been reduced by 99 percent compared with the other control https://doi.org/10.4130/jaev.5.989.
methods. In a steady-state scenario, enhanced direct torque control [5] B.K. Reddy, et al., Optimal operation of cogeneration power plant integrated with
solar photovoltaics using DLS-WMA and ANN, Int. J. Energy Res. 2024 (2024)
completely eliminates torque harmonic ripple in 100 % comparison to
5562804, https://doi.org/10.1155/2024/5562804.
the conventional method. The settling time for proposed DTC has been [6] M. Habibullah, D.D. Lu, D. Xiao, M.F. Rahman, A simplified finite-state predictive
improved from 0.0001 to 0.00001 compared to conventional DTC. This direct torque control for IM drive, IEEE Trans. Ind. Electron. 63 (2016) 3964–3975,
study can be extended to improve the performance of the DTC-SVM- https://doi.org/10.1109/TIE.2016.2519327.
[7] J.L. Romeral, A. Arias, E. Aldabas, M.G. Jayne, Novel direct torque control (DTC)
based 2-level voltage power source inverter-fed polyphase induction scheme with fuzzy adaptive torque ripple reduction, IEEE Trans. Ind. Elens 50
motor. The future scope of this research can incorporate the optimizing (2004) 487–492, https://doi.org/10.1109/TIE.2003.812352.
control strategies in real-time depending on load demands and operating [8] Y.T. RaoCurrent based restarting method for rotating sensorless IM Drive, in IEEE
Trans. on Energy Conv 38 , 2023, , 2239–2242, 10.1109/TEC.2023.3289065.
conditions by utilizing AI and machine learning approaches and also by [9] J. Ang, Y. Yang, F. Blaabjerg, J. Chen, L. Diao, Z. Liu, Parameter identification of
utilizing advanced thermal modelling and simulation software to fore- inverter-fed induction motors- a review, Energies 11 (2016) 2194, https://doi.org/
cast and control the motor and drive system’s temperature profiles. 10.3390/en11092194.
[10] A. Nurettin, N. İnanç, Sensorless vector control for IM drive at very low and zero
speeds based on an adaptive-gain super-twisting sliding mode observer, IEEE J.
CRediT authorship contribution statement Emerg. Sel. Top. Power Electncs. 11 (2023) 4332–4339, doi.10.1109/
JESTPE.2023.3265352.
[11] I. Nouira, ElBadii, A. Masmoudi, DTC of three-switch three-phase inverter fed im
S. Usha: Writing – review & editing, Writing – original draft, Visu- drives, tenth intern, Conf. Ecol. Veh. Renew. Energy 8 (2015) 222–234, https://
alization, Validation. A. Geetha: Resources, Methodology, Formal doi.org/10.1109/EVER.2015.7112980.
analysis, Conceptualization. P. Geetha: Visualization, Validation, Su- [12] S.M. Mousavi Gazafroodi, A.A. Dashti, Novel MRAS-based estimator for speed-
sensorless IM drive, Iran. J. Elec. Electn. Engg. 10 (2014) 304–313, https://doi.
pervision, Software. Nimay Giri: Writing – review & editing, Writing – org/10.1155/2024/5792804.
original draft, Validation, Methodology. Bidyut Mahato: Resources, [13] M. Comanescu, L. Xu, Sliding mode MRAS speed estimators for sensorless vector
Methodology, Investigation, Conceptualization. T. M. Thamizh Then- control of induction machine, IEEE Trans. Ind. Electron 53 (2006) 146–153,
https://doi.org/10.1109/TIE.2005.862303.
tral: Funding acquisition, Formal analysis, Data curation, Conceptuali-
[14] U. Genc, B. Tekgun, Instantaneous torque error compensation based online torque
zation. Palanisamy R.: Resources, Methodology, Investigation, Formal sharing function for Switched reluctance Machines, Alex. Engg. J. 75 (2023)
analysis. Mohammed Alharbi: Software, Resources, Methodology, 261–270, https://doi.org/10.1109/TIE.2005.862303.
[15] M.Z. Aihsan, A. Jidin, M.M. Azizan, M.I. Fahmi, K. Hasikin, Flexible sector
Funding acquisition, Data curation.
detector-based mismatch supply voltage in direct torque control doubly fed
induction machine: an experimental validation, Alex. Engg. J. 74 (2023) 689–704,
Declaration of Competing Interest https://doi.org/10.1016/j.aej.2023.05.060.

The authors declare that they have no known competing financial

537
S. Usha et al. Alexandria Engineering Journal 108 (2024) 518–538

[16] K. Ohyama, G. Asher, G.M. Sumner, Comparative analysis of experimental industry. Systems, Decision and Control in Energy III, Springer International
performance and stability of sensorless IM drives, IEEE Trans. Ind. Electron 53 Publishing, Cham, 2022, pp. 3–21, https://doi.org/10.1007/978-3-030-87675-3_1.
(2018) 178–186, https://doi.org/10.1109/TIE.2005.862298. [27] S. Adigintla, M.V. Aware, Robust fractional order speed controllers for IM under
[17] U. Sengamalai, S. Chinnamuthu, An experimental fault analysis and speed control parameter variations and low speed operating regions, IEEE Trans. Circuits Syst. II
of an IM using motor solver, J. Electr. Eng. Tech. 12 (2017) 761–768, https://doi. Express Briefs 70 (2023) 1119–1123.
org 10.5370/JEET.2017.12.2.761. [28] N.F.B.A. Rahim, A.W.Y. Khang, A. Hassan, S.J. Elias, J.A.M. Gani, J. Jasmis, and J.
[18] BadsiEl, A New SVPWM Strategy Dedicated to the Control of Delta Inverters, A.J. Alsayaydeh, "Channel Congestion Control in VANET for Safety and Non-Safety
Comput. Math. Electr. Electron. Engin. (COMPEL) 32 (2013) 289–301, https://doi. Communication: A Review," in 2021 6th IEEE International Conference on Recent
org/10.1002/2050-7038.12384. Advances and Innovations in Engineering (ICRAIE), 2021, pp. 1-6, doi: 10.1109/I
[19] S. Usha, C. Subramani, S.K. Padmanaban, Neural network-based model reference CRAIE52900.2021.9704017.
adaptive system for torque ripple reduction in sensorless poly phase IM drive, [29] Satish Kumar Gudey, Mohan Malla, Kiran Jasthi, Srinivasa Rao Gampa, Direct
Energies 12 (2019) 1–25, https://doi.org/10.3390/en12050920. torque control of an IM using fractional-order sliding mode control technique for
[20] A.A. Hossam-Eldin, E. Negm, M.S. Elgamal, K.M. AboRas, Operation of grid- quick response and reduced torque ripple, World Electr. Veh. J. 14 (2023) 137,
connected DFIG using SPWM- and THIPWM-based diode-clamped multilevel https://doi.org/10.3390/wevj14060137.
inverters, IET Gener. Transm. Distrib. 14 (2020) 1412–1419, https://doi.org/ [30] T. Wang, B. Wang, Y. Yu, D. Xu, Discrete sliding-mode-based MRAS for speed-
10.1049/iet-gtd.2019.0248. sensorless IM drives in the high-speed range, IEEE Trans. Power Electron 38 (2023)
[21] Ahmed A. Hossam-Eldin, Negm AbdallahEmtethal; Elgamal, Mohamed S.; AboRas, 5777–5790.
Kareem M.: ’Fault ride-through of grid-connected THIPWM fired DCMLI-based [31] F. Shiravani, P. Alkorta, Jose Antonio Cortajarena, Oscar Barambones, An
DFIG using parallel switched feedback-controlled DVR’, IET Generation 6 ( 14 improved predictive current control for IM drives, Ain Shs. Engg J. 14 (2023)
Transm. amp; Distrib. , 2020, , 945–954, 10.1049/iet-gtd.2019.0215. 102037, https://doi.org/10.1016/j.asej.2022.102037.
[22] S. Potturi, J. Reddy, R. Dash, R. Jena, C.V. Subburaj, C. Dhanamjayulu, Direct [32] I. Fedorchenko, A. Oliinyk, J.A.J. Alsayaydeh, A. Kharchenko, A. Stepanenko, and
Torque Control of IM Using ConvLSTM Based on Gaussian Pillbox Surface, Mat. V. Shkarupylo, “Modified genetic algorithm to determine the location of the
Pros. Engg. 45 (2022) 1–21, https://doi.org/10.1155/2022/4408271. distribution power supply networks in the city,” 2020, doi: 10.5281/
[23] M.K. Metwally, Sensorless speed control of 4-switch three phase inverter fed IM zenodo.5163692.
drives at very low and zero speed, Alex. Engg. J. 52 (2013) 327–336, https://doi. [33] B. Zou, Y. Guo, X. Xiao, B. Yang, B., X. Wang, M. Shi, Y. Tu, Performance
org/10.1016/j.aej.2013.05.003. improvement of matrix converter direct torque control system, Energies 13 (2021)
[24] S. Zhang, S. Li, L. He, J.A. Restrepo, T.G. Habetler, Rotor thermal monitoring 3247, https://doi.org/10.3390/en13123247.
scheme for direct-torque-controlled interior permanent magnet synchronous [34] Z. Hossain, A.K.M., N.B. Hassim, J.A.J. Alsayaydeh, M.K. Hasan, M.R. Islam, A tree-
machines via high-frequency rotating flux or torque injection, Elec. Engg. Sys. Sci. profile shape ultra wide band antenna for chipless RFID tags, Int. J. Adv. Comput.
10 (2021) 1–9, https://doi.org/10.48550/arXiv.2106.02116. Sci. Appl. vol. 12 (4) (2021) 546–550, https://doi.org/10.14569/
[25] S.G. Petkar, V.K. Thippiripati, A novel duty-controlled DTC of a surface PMSM IJACSA.2021.0120469.
drive with reduced torque and flux ripples, IEEE Trans. Ind. Electron. 23 (2021) [35] J.A.J. Alsayaydeh, A. Aziz, A.I.A. Rahman, S.N.S. Salim, M. Zainon, Z.
1–11, https://doi.org/10.1109/TIE.2022.3181405. A. Baharudin, M.I. Abbasi, A.W.Y. Khang, Development of programmable home
[26] V.V. Shkarupylo, I.V. Blinov, A.A. Chemeris, V.V. Dusheba, J.A.J. Alsayaydeh, On security using GSM system for early prevention, ARPN J. Eng. Appl. Sci. vol. 16 (1)
applicability of model checking technique in power systems and electric power (2021) 88–97.

538