Journal of Operation and Automation in Power Engineering

Vol. 11, No. 4, Dec. 2023, Pages: 258-267

http://joape.uma.ac.ir

Torque Ripple Reduction of the Position Sensor-less Switched

Reluctance Motors Applied in the Electrical Vehicles
A. Dejamkhooy∗, A. Ahmadpour

Department of Electrical Engineering, University of Mohaghegh Ardabili, Ardabil, Iran

Abstract— The Switched Reluctance Motors (SRMs) not only are low cost for industry applications, but also they could work in various
conditions with high reliability and efficiency. However, usage of these motors in high speeds applications under discrete mode causes
decreasing the efficiency. In this paper, a new optimized control method based on the various Torque Sharing Functions (TSFs) and
optimization algorithms is proposed for Minimum Torque Ripple Point Tracking (MTRPT) of a 4-phase SRM with 6/8 poles. In this method,
turn-on and commutation angles are controlled based on the lookup table. The proposed method could adjust the rapid variations of the
current in the starting mode of SRM. To show the robustness of the proposed approach, a real case study is considered, the control method
is applied in an Electric Vehicle (EV) mechanism, and its performance is assessed in various motion states such as acceleration, breakage,
and steady-state. Also, the position sensor for the studied EV is neglected, which could reduce the extra costs. There are two various
scenarios considered for solving the problem. First, the turn-off and turn-on angles are controlled, and the commutation angle is fixed.
The results show the robustness of the proposed method with about 90 % diminishing the torque ripple, compared to when all mentioned
angles are fixed. In the second step, based on a lookup table, instead of using complex analytical methods, the turn-on angle is controlled.
Therefore, a variable turn-on angle proportional to the applied speed is applied to the commutation control system of SRM. Besides, a
lookup table is created to restrain the reduction of the turn-off angle. The simulation results are compared to other previous methods, and
the worth of the proposed method is shown.

Keywords—Switched Reluctance Motors, Minimum Torque Ripple Point Tracking, Torque Sharing Functions, Electric Vehicles.

1. I NTRODUCTION a car is accelerating and needs extra power, it needs continuous

torque without fluctuating. Also, when an obstacle is reached,
1.1. Motivation the force should be reduced. So, a quick reaction is necessary
The usage of Electrical Vehicles (EVs) has various benefits, to reduce the torque. In the cases of conventional controllers, the
which finally, efficiently decrease the emissions of greenhouse fluctuations of the accelerator and brake pedals have caused user
gases. Therefore, interest in various traction systems is enhanced dissatisfaction. Therefore, it needs an optimal control to mitigate
for EVs and Hybrid EVs (HEVs). In recent years, several electrical the mentioned fluctuations.
motors have been introduced, such as induction motors, brushless
DC motors, and Switched Reluctance Motors (SRMs) [1, 2]. The 1.2. Literature Review
SRMs, due to their simpler structures, lower cost, and higher
resilience under various conditions, are primarily used in EVs [3]. The Torque Sharing Function (TSF) for current profiling is
An SRM includes poles for rotor and stator, separately. Also, SRMs the most common method among the Minimum Torque Ripple
have no Permanent Magnet (PM), and windings are employed in Point Tracking (MTRPT) techniques [4–6]. The TSF also includes
the stator. This causes motor works with high speeds as well as various types as linear, cubic, exponential, and so on, which are
high-temperature conditions. The variations of reluctance help to studied in plenty of papers. In [7], a novel nonlinear type of TSF
produce torque in SRMs. Owing to salient poles on both sides of is proposed to increase the efficiency and obtain MTRPT. The
the motor, the electrified winding attracts the nearest rotor pole authors of [4] proposed both online and offline TSF that work
and the phase inductance variates. It should be noted that the with a variation of operation point and flux linkage, respectively.
rotational motion of SRM is enabled due to a control drive of its In the same way as [4], a novel offline TSF is applied to SRM
phase by a power converter. that could be controlled by the profile of flux linkage [8]. A new
However, the special structure of SRM, i.e. doubly salient, and its TSF for SRM in [9] is proposed to reach a forecastable control. A
excitation produced by a single source, lead to non-linear output novel forecastable multi-objective optimization model is proposed
characteristics. Therefore, due to this drawback, the ripple of torque in [10] that encourages the TSF control.
and acoustic noise are obtained with high values. Braking and The mentioned studies don’t consider the turn-off/on angles in their
accelerating in cars is a continuous mechanical movement. When proposed methods. The control of SRMs, due to their nonlinear
features, is a challenge. To cope with this problem, the model of
SRMs is often linearized, and the controller is designed. In [11],
the Dahlin controller family is proposed for adaptive control in
Received: 30 Nov. 2021 SRM. A comprehensive comparison of various linearized models
Revised: 24 Jun. 2021 is carried out in [12]. Besides, the authors proposed a new fixed PI
Accepted: 11 Sep. 2022
∗ Corresponding author: controller that is based on SRM parameters. The authors of [13]
E-mail: [email protected] (A. Dejamkhooy) presented a new small-signal model of optimal control of speed
DOI: 10.22098/JOAPE.2023.9908.1694 and current in SRM drive. In [14], the PI controller is composed
Research Paper of TSF that could control the hysteresis current.
@2023 University of Mohaghegh Ardabili. All rights reserved The analytic modeling is one of the main challenges to reaching

258
Journal of Operation and Automation in Power Engineering, Vol. 11, No. 4, Dec. 2023 259

a robust control system of SRMs. In order to cope with this

issue, this paper presents a novel method that could consider the
TSF at the same time that the SRM controller is designed, and
it minimizes the back-EMF effects. In the next step, a robust,
fast, and low-cost controller is needed for SRM control. In [15],
a predictive current controller is proposed for the coupled SRM
based on the net flux method. The results of [15] show the
enhancement of the torque regulation. An analysis of dynamic
current control methods for SRM drives is done in [16]. In this
study, the authors concluded that the PI controllers have better
results than others due to lower sensitivity to the errors of models.
A current deadbeat controller with active thermal management is
used in [17] for optimal MTRPT of SRM. However, the applied
method couldn’t be a worthwhile technique for Direct Torque
Control (DTC) in SRMs due to obtaining the reference value Fig. 1. The structure of the studied 4-phase SRM with 8 stator poles and 6
in a single cycle causes the high effort of the control process. rotor poles
A fact in [18] is validated by experimental results that various Phase A Phase B Phase C Phase D

methods could be effective in some cases and act weakly in

others. In this recent study, the authors presented the performance
and sensitivity of discrete-time PI controller, deadbeat control,
and Dahlin controller in both dynamic and steady-state mode
for PM–Synchronous Motor (PMSM). In [19], a novel design of
SRM is proposed with the aim of MTRPT. Also, the optimization
results are evaluated with experimental outputs of a prototype of
the proposed structure. The application of SRM in EV is studied
in [20], and a new control strategy is proposed to minimize the
torque ripple based on TSFs. A new TSF is presented in [21] for
torque ripple reduction of SRMs, which could establish the new Fig. 2. The structure of 4-phase asymmetric bridges–power converter used
current reference generation strategy. The propulsion applications in SRM drive
with SRMs are the issue of [22] that effort to reduce their torque
ripple and acoustic noise in EVs and electric bikes. In a recent 2. M ODELING OF SRM D RIVE
work [23], the authors tried to propose an indirect torque control noindent With neglecting the hysteresis effects and eddy
strategy for SRMs by introducing a novel TSF. currents, the relation of voltage per phase could be considered as
In [24], a novel control based on field-oriented control without follow:
sensor is proposed to detect the current fault of an induction motor.
Moreovere, authors of [25] used the presented method in [24], and dλs (φs , is )
VS = Rs is + (1)
combined it with Direct Torque Control (DTC) to control a three dt
phase induction motor. In [26], a vector control method is applied where, VS is supplied voltage, Rs is the resistance of s-th phase,
on the three phase induction motor under faults conditions. In is indicates phase current of s-th phase, t is time, and λs shows
[27], the Direct Instantaneous Torque Control (DITC) is poroposed the phase flux linkage of s-th phase that is a function of position
for generating/braking mode of SRM. The results show the higher (φs ) and current, and could be expressed as follows:
efficiency and lower torque ripple in comparison with DTC. In λs (φs , is ) = Ls (φs , is ) is
[28], a sliding mode DITC cruise control is applied in the SRM where, Ls is the phase inductance of s-th phase. By replacing
drive with aim of MTRPT and speed tracking. In [29], a novel (2) in (1), the new equation of voltage could be as (2), which
hybrid control method based on Model Predictive Control (MPC) includes three terms of dropped voltage by resistance, dropped
and TSF has been presrnted to decrease the torque ripple of SRM. voltage by inductance, and back Electro–Magnetic Force (EMF).
The Maximum Torque per Ampere (MTPA) control strategy as a d dis
novel method is studied in [30] and applied on a three phase SRM VS =Rs is + {Ls (φs , is ) is } = Rs is + Ls (φs , is )
to decrease the torque ripple. dt dt
(2)
d
+ ωis {Ls (φs , is )}
dφ
1.3. Paper Contributions The electromagnetic torque, Te , could be calculated as follows:
In this paper, the authors proposed the following contributions:
dω
• Preparing a novel control strategy of SRM based on TSFs to Te (φs , is ) = TL + J + Bω (3)
dt
reduce the torque ripple.
• The turn-on angle and commutation angle are controlled
based on the lookup table. Un-Aligned Position Aligned Position
• The control method is applied in an EV mechanism, and its STATOR
performance is studied.
• Also, the proposed method could adjust the rapid variations ROTOR
LK
of the current in the starting mode of SRM.
π/12 Region

1.4. Paper Organization Actual

The paper organization is as follows. In Section 2, the dynamic

model of SRM is studied. In Section 3, the proposed method is
presented. Section 4 presents the results of the simulation, and Idealized Motor
a discussion of the results is debated. Finally, the conclusion is Position
presented in Section 5. π/6 Generator Torque π/6 Motoring Torque
Fig. 3. The inductance of 4/8 poles, four phases SRM
A. Dejamkhooy and A. Ahmadpour : Torque Ripple Reduction of the Position Sensor-less Switched Reluctance... 260

ω
_ Tr Tr-ph
ωr + PI Controller TSF +
_

ϕ T Hysteresis Controller Power Converter

i
Lookup Table
T–i –ϕ ϕ

Fig. 4. The inductance of 4/8 poles, four phases SRM

Tph* per pole, there are two switches. When phase A should be turned
Phase A, Phase B, Phase C, Phase D on, the gate command is given to the switches T1 and T2 , and
ϕ
Tph* the supply voltage is placed on both ends of the stator phase
Phase A, Phase B, Phase C, Phase D windings (Vph = Vdc ). Therefore, the current starts to increase,
Trefϕ
Tph* which is a function of stator windings inductance and the voltage
Phase A, Phase B, Phase C, Phase D ϕ of the power source. In this case, there will be a little power loss.
ϕ When phase A has to be turned off, the switches T1 and T2 will
Tph* ϕon {ϕon + ϕov}Tref ϕoff
Tph* Phase A, Phase B, Phase C, Phase D be turned off (Vph = –Vdc ). However, due to the inductor current
ϕ (a) Phase A, Phase B, Phase C, Phase D ϕ
Tref ϕ cannot being cut off suddenly, a huge voltage is generated at both
ϕon {ϕon + ϕov} ϕoff
Tph* ends that tend to maintain the current. To solve this problem, two
Trefϕ Phase A, Phase B, Phase C, Phase D ϕ diodes are installed in the circuit to remove the energy stored in
ϕon {ϕon + ϕov} Tref ϕoff
Tph*
the phase without damaging the switches. When the switches are
ϕ turned off, these diodes form a path for the inductor current, which
ϕon {ϕon + ϕov}Trefϕ ϕoff Phase A, Phase B, Phase C, Phase D ϕ
ϕon {ϕon + ϕov} ϕoff flows from negative to positive poles. In this way, the supply
Tph*
Tph*
ϕ Phase A, Phase B, Phase C, Phase D voltage is placed inversely on both ends of the considered phase
Trefϕ Phase A, Phase B, Phase C, Phase D ϕ and the power supply is charged. Obviously, if the supply voltage
ϕon {ϕon + ϕov} ϕoff
Tph* is higher, the both of increasing and decreasing rates will also be
Trefϕ (b) Phase A, Phase B, Phase C, Phase D ϕ faster. The current equation could be formed as follows:
ϕon {ϕon + ϕov} Tref ϕoff
Tph*  
ϕ ϕ
dis 1 dLs (φs , is )
ϕϕon {ϕonon++ϕϕ
{ϕ } } Tref
ovov
ϕ ϕoff
ϕoff
Phase A, Phase B, Phase C, Phase D = VS − Rs is − is ω (6)
*
on
dt Ls (φs , is ) dφ
TTph* ph

ϕϕ PhaseA,
Phase A, Phase
PhaseB,
B, Phase
PhaseC,
C,Phase
PhaseDD
ϕ According to Fig. 2, the power converters used in this paper are
ϕon {ϕon + ϕov} Tref ϕoff
Tph* asymmetric bridges with two switches and two diodes in each
ϕ Phase A, Phase B, Phase C, Phase D ϕ phase. The main advantage of this converter is its control flexibility,
ϕon {ϕon + ϕov} TTrefref ϕoff
Tph*
which allows each phase to be controlled separately, especially
ϕ (c) Phase A, Phase B, Phase C, Phase D ϕ ϕ
in high-speed applications (like EVs) when commutation occurs
ϕϕonon {ϕ + ϕ ov}} Tref
on + ϕov
{ϕon ϕϕoffoff between the adjacent phases. Besides, the main disadvantage may
Tph*
be the need for high-power components, which increases the cost
ϕ Phase A, Phase B, Phase C, Phase D ϕ
ϕon {ϕon + ϕov}
Tref
ϕoff and size.

ϕ
ϕon {ϕon + ϕov} Tref ϕoff 3. C ONTROL M ETHOD BASED ON TSF
ϕ The output torque in the SRMs is obtained from the sum of
ϕon {ϕon + ϕov} ϕoff the output torques in each phase, and each phase can produce
the desired torque in a certain range. So, to distribute the torque
(d) generation task between all phases, it is necessary to identify the
Fig. 5. The curve of TSFs, (a) LTSF, (b) STSF, (c) ETSF, (d) CTSF positions of each phase that can produce the highest torque at
these positions. For example, consider a 4/6 poles, three phases
where, TL is load, inertia moment is introduced by J, ω is SRM. In this SRM, the rotor pole-pitch is π/2. Therefore, each
rotational speed, and B indicates the friction coefficient. Besides, stator pole produces clockwise torque in half of the period (π/4)
the electromagnetic torque could be calculated as follows: and counterclockwise torque in the other half. On the other hand,
because the motor is three-phase, each phase must produce torque
i2s d
Te (φs , is ) = {Ls (φs , is )} (4) at π/6 in the same direction as the principal component, for
2 dφs completion π/2. Now, assuming a 4/8 poles, four phases SRM,
Finally, the description of the relation between the electro–magnetic the rotation period of the rotor is equal to π/3, and each pole in
and the mechanical domains of the SRM could be as follows: each period can produce π/6 of torque in the same direction as the
i2s d dω principal component. Due to the number of phases being 4, each
{Ls (φs , is )} − TL = J + Bω (5) phase should produce the desired torque at π/12. Fig. 3 shows
2 dφs dt the inductance curves and areas for generating clockwise and
The behavior of an SRM could be described by (2) and (5). counterclockwise torques for the last mentioned SRM [31]. In this
In these equations, ω and is are the state variables, and their figure, let’s consider the π/6 region that could generate torque. In
multiplication shows the nonlinear behavior of SRM. Therefore, a this region, an π/12 sub-region should be selected, which includes
development controller should linearize this behavior. The structure the best performance in producing torque. If the selected region is
of the studied SRM is shown in Fig. 1. closed to the non-aligned position, due to the low value of dL/dφ,
at the beginning of each phase excitation, a large current needs to
Fig. 2 shows the structure of the used power converter as a be injected to produce torque. Therefore, a high value of power in
supply source of the SRM. As seen in this structure, per phase the commutation region will be lost. Also, if this region is closed
Journal of Operation and Automation in Power Engineering, Vol. 11, No. 4, Dec. 2023 261

ω
_ Tr Tr-ph
ωr + PI Controller TSF +
_

ϕov ϕ T Hysteresis Controller Power Converter

ω Turn-ON angle ϕon
Controller i
Lookup Table
T–i –ϕ ϕ
ω Commutation angle
Controller

(a)

Start

Input data: ωr,

objective torques, and
motor characteristics

Create a look-up table of torque

PI controller for generated
with angle and current
the reference torque

Choosing optimal value for

Ki and Kp with IPPOA

Control the angles

TSF

Hysteresis Controller

Power convertor

Torque ripple is lower

than max. allowable
ripple ?

End

(b)
Fig. 6. The proposed control method for SRM: a) block diagram, b) flowchart

Table 1. The characteristic of the studied SRM

Parameter Value Parameter Value

Number of phases (m) 3 Inertia (J) 0.05
Number of stator poles (Ns ) 8 Maximum current 450
Number of rotor poles (Nr ) 6 Maximum flux 0.486
Stator resistance per phase (Rs ) 0.2 Source Voltage (VS ) 300
Non–aligned inductance 0.0067 Friction coefficient (B) 0.02
Saturated aligned inductance 0.0015 Aligned inductance 0.0236
A. Dejamkhooy and A. Ahmadpour : Torque Ripple Reduction of the Position Sensor-less Switched Reluctance... 262

Fig. 11. The applied pulses to switches of phase A

Table 2. The summarized results for case 1

Fig. 7. The magnetic characteristic of the studied SRM
TSF type Tav Tmax Tmin Torque Ripple (%)
(N.m) (N.m) (N.m)
LTSF 17.53 19.28 15.87 19.42
STST 17.53 19.77 15.58 23.89
ETSF 17.53 20.77 15.51 30.06
CTSF 17.53 19.61 15.58 22.96

to the aligned position, the value of dL/dφ is low again. However,

due to a large amount of inductance near the aligned position and
the lack of rapid current discharge, it causes a delay in current
reduction at commutation and produces negative torque. For this
purpose, by omitting the saturation effect, the value dL/d
Fig. 8. The relations between torque–position–current of the studied SRM φ can be calculated between the aligned and non–aligned positions.
Then, the π/12 sub-region with the highest dL/dφ will be selected
as the desired area. In the case of considering the saturation effect,
since the inductance is a function of both rotor position and current,
proportional currents should be proposed for different torques.
Then, the areas of π/12 with the highest level will be selected.
Due to the applied current to one phase cannot be changed with a
high slope, it is not possible that the produced torque to be zero by
one phase at a special moment and bring the output torque of the
other phase to the desired value in a short moment. Therefore, for
the moments of commutation, the amount of torque produced by
each phase as well as its changing process should be determined.
Therefore, the TSF has been introduced that the output torque
of each phase changes based on specific mathematical functions.
This method is a control process to produce a continuous uniform
(a) torque to reduce the torque ripple. Also, it uses some special
sharing functions for generating the reference torques.
Fig. 4 illustrates a block diagram of the TSF control method for
SRM with four phases. This method includes some main steps.
First, a reference speed (ω r ) is determined and compared with the
motor speed. Second, the resulting error is given to a PI control
block to produce the reference signal of the torque. Third, this
signal is decomposed into TSFs proportional to each of the phases,
while the sum of the reference torques of all phases at any given
moment is equal to the total reference torque. Fourth, the reference
output torque is compared with the actual torque signals of each
(b)
Fig. 9. The speeds and torque obtained in case 1, (a) The reference,
obtained and error speeds, (b) The obtained torque

Fig. 10. The rotor position variations in case 1

Fig. 12. The current variation of the DC source
Journal of Operation and Automation in Power Engineering, Vol. 11, No. 4, Dec. 2023 263

linear function based on the rotor position. The equations of LTSF

are as follows:
 φ−φon
 Tref φov φon ≤ φ < φon + φov
∗
Tph = T ref φon + φov ≤ φ < φof f − φov (7)
φ f −φ
Tref ofφov φof f − φov ≤ φ < φof f


where, φon , φof f , and φov are the turn-on, turn-off, and
overlapping angles, respectively. Besides, Tref is the total desired
torque. The mentioned parameters are determined in the LTFS
curve for four phases in Fig. 5(a).

(a)
3.2. The Sinusoidal TSF
In the Sinusoidal TSF (STSF), the output torque of each phase
in the commutation interval is a sine function according to the
position of the rotor, which is defined as with (8). Also, the curve
of STSF for four phases is plotted in Fig. 5(b).
  
2 π φ−φon
 Tref sin 2 φov

 φon ≤ φ < φon + φov
∗
Tph = Tref   φon + φov ≤ φ < φof f − φov
 Tref sin2 π φof f −φ

φof f − φov ≤ φ < φof f

2 φov
(8)
3.3. The Exponential TSF
The defined function of the Exponential TSF (ETSF) is as
(b) follows:
Fig. 13. The flux and current variation of all phases, (a) Flux, (b) Current h  2
i
1 − exp − (φ−φ on )

T φon ≤ φ < φon + φov
 ref φov


phase, and their errors are applied to the torque hysteresis control ∗
Tph = Tref   φon + φov ≤ φ < φof f − φov
2
block. Fifth, the hysteresis block applies switching pulses to the  Tref exp − (φof f −φov −φ)

φof f − φov ≤ φ < φof f

φov
electronic power converter for starting the motor [32]. Note that
to generate the reference torques of each phase, the rotor position (9)
signal is also necessary, which is obtained from the position sensor The curve of ETSF for four phases is shown in Fig. 5(c).
data or estimating methods. In this paper, the second method is
used.
3.4. The Cubic TSF
3.1. The Linear TSF The Cubic TSF (CTSF) is defined base on a cubic algebraic
The Linear TSF (LTSF) indicates that the torque produced by function. As seen in (10), the desired function is non-linear. The
each phase in the commutation interval could be determined as a curve of CTSF for four phases is shown in Fig. 5(d).

   2  3 
T 3 φ−φ on
− 2 φ−φon
φon ≤ φ < φon + φov
 ref

φov φov


∗
Tph = Tref  φon + φov ≤ φ < φof f − φov (10)
 
φ−φof f +φov 2 φ−φof f +φov 3

   
 Tref 1−3 + 2 φof f − φov ≤ φ < φof f

φov φov

Table 3. The summarized results for case 2

and φov are interdependent, the number of TSF parameters can
TSF Tav Tmax Tmin Torque Ripple be reduced to only two parameters. The relationship between TSF
type (N.m) (N.m) (N.m) (%) parameters is as follows:
LTSF 17.52 18.93 16.08 16.22 φof f = φon + φov + ∆φ (11)
STST 17.52 19.52 15.94 20.42
ETSF 17.52 20.16 15.11 28.94 where, ∆φ is the stroke angle that is the displacement angle
CTSF 17.52 19.46 16.07 19.31 between consecutive phases, and is defined as follows:
2π
∆φ = (12)
mNr
3.5. The Optimization Criteria of TSFs where, m is the number of phases, and Nr is the number of rotor
poles. Besides, a new parameter as the central angle is defined
Although the TSF could reduce the torque ripple of SRM, the that two consecutive phases have the same torque (equal to half of
copper losses are dependent on the TSF type and its parameters. ∗
Tph ) at this point. This angle, based on Fig. 4, could be defined
Therefore, selecting a robust TSF is essential. The selected TSF as follow:
type also affects the speed range with acceptable torque ripple as
well as the phase current peak. Because the values of φon , φof f , φC = φon + φov /2 = φof f − ∆φ − φov /2 (13)
A. Dejamkhooy and A. Ahmadpour : Torque Ripple Reduction of the Position Sensor-less Switched Reluctance... 264

Fig. 14. The results of torque for phase A in case 1, (a) LTSF, (b) STSF, (c) ETSF, (d) CTSF

The torque ripple could be calculated as follow:

Tmax − Tmin
Tripple (%) = × 100 (15)
Tavg
The proposed control method based on TSF is evaluated in two
different modes: 1) the introduced TSFs in the previous section
are used without controlling turn-on, turn-off, and commutation
angles; 2) the same TSFs are employed and turn-on and
commutation angles are controlled. The results of two cases are
compared at high speeds.

Fig. 15. The variation of angles in case 2

4.1. The Results of Case 1
In the cases of minimum turn-on angle and maximum turn-off In this case, based on Fig. 4, the torque is determined from the
angle, the maximum overlapping angle could be obtained as feedback of the PI controller. Also, the load toque is 15 N.m. The
φM
ov
AX
= τ /4, which τ is the pole pitch. curve of reference and obtain speeds, and the error between these
speeds are shown in Fig. 13(a). Moreover, the obtained torque
3.6. The Controlling of Angles from this error is presented in Fig. 13(b).
The variation of the rotor position is shown in Fig. 10. As seen
In some works, the turn-on angle is pulled back until the in this figure, the period of rotor position reduces with increasing
torque error reaches the desired value. The aim of this paper is the rotor speed and is fixed in steady-state. The applied pulses
based on the reduction of turn-on angle and combination with the to switches of phase A are based on Fig. 11. Also, the current
commutation angle control method in order to reduce the torque variations of the DC source are presented in Fig. 12. The results
ripple at high speeds. The variations of turn-on angle and speed of voltages and currents of all phases are shown in Fig. 13.
are stored in a lookup table. Then, the turn-on angle, in proportion During the simulation, to limit the peak current in the phase
to the speed, tends to the aligned angle. In this method, the values turned-on moment, the turn-on angle for phase A is adjusted to
of turn-on angle are optimized by optimization algorithms. 5.37◦ according to the positive and negative torque generation
If speed increases, the turn-on and turn-off angles decrease, intervals. Also, the optimal turned-off angle is slightly less than
which leads to an increment of the current peak at the beginning the phase-aligned angle (59◦ ) to prevent the production of negative
of the phase conduction. So, to prevent a large increase in current torque, which leads to a high torque ripple. The commutation
and prevent torque drop, the commutation range will be raised by angle of all phases is also set at a constant value of 5.6◦ . In this
increasing the commutation angle. Therefore, the optimal values case, all three angles, i.e. turn-on, turn-off, and commutation, have
of the commutation angle are always within the expected range a constant value for each phase at different speeds. The reference
and are increased in proportion to the increment of speed that is and actual torques for phase A with various TSFs are shown in
stored in a lookup table. The final sketch of the proposed method Fig. 14. The summarized results from these figures are listed in
for controlling the SRM used in EVs is shown in Fig. 6. Table 2. According to Table 2, it can be seen that the torque
ripple, in this case, is slightly higher. Therefore, using this idea
4. T HE S IMULATION R ESULTS that by increasing the speed, bringing the turn-on angle closer
The characteristic of the studied SRM is tabulated in Table 1. to the non-aligned angle, and combining it with increasing the
Also, the magnetic characteristic of this SRM is shown in Fig. 7, commutation angle, in the next step, the torque ripple could be
in which some parameters like aligned, non-aligned, and saturated improved.
aligned inductances, as well as the maximum allowable motor
current, could be extracted. The torque–position–current curve is
4.2. The Results of Case 2
another important feature of each SRM, which is shown in Fig. 8.
As seen in this figure, in φ = 0, the stator and rotor poles are The simulation method, in this case, is based on the block
in the aligned position, and in φ = π/6 are in the non-aligned diagram in Fig. 6 with the same reference speed and a load torque
position. The flux of each phase could be calculated as follow: of case 1. The turn-on angle variations, which are estimated at
Z different speeds using the lookup table, are shown in Fig. 15. As
λs = [VS − Rs Is ] dt (14) seen in this figure, the turn-on angle tends to be the non-aligned
angle as the velocity increases to decrease the torque ripple at
Journal of Operation and Automation in Power Engineering, Vol. 11, No. 4, Dec. 2023 265

Fig. 16. TThe results of torque for phase A in case 2, (a) LTSF, (b) STSF, (c) ETSF, (d) CTSF

high speeds. Besides, the variations of the commutation angle are

shown in Fig. 1. It should be noted that in this case, the turn-off
angle is set to a constant value (44◦ for phase A). The turn-off
angle of the other phases is set with a favorable shift relative to
phase A. The results of phase A for various TSFs are plotted in
Fig. 16, and the summarized results obtained from Fig. 16 are
shown in Table 3.
Based on Table 3, it can result that the torque ripple has improved
compared to case 1, due to the favorable changes in turn-on and
commutation angles in proportion to the speed increase. It causes
avoiding two states: entering the current into the negative zone of
the torque and generating the peak current in the starting the phase
conduction. (a)
As shown in Fig. 16, the input of the power converter is the
torque signals, to generate the requested voltage. Therefore, the
consumption power of the switches is related to the torque values.
Table 3 shows the torque ripples with various methods that indicate
the power loss of switches. So, by applying the proposed method,
the power loss will be reduced.

4.3. Applying the proposed method in a case study: electrical

vehicle
As mentioned before, an EV uses the negative and positive
torques, continuously. Therefore, in this section, these movements
are simulated and the proposed method is applied. The PI controller (b)
has two main coefficients as Kp and Ki , which are optimized
using the Improved Prey Predator Optimization Algorithm (IPPOA) Fig. 17. The requested features of an SRM employed in EV, (a) speed, (b)
[33] under each movement, which leads to better tracking of the torque
reference velocity curve. In this case, the variations of reference
and actual speeds are presented in Fig. 17(a), and the needed
torque is plotted in Fig. 17(b). The simulation results for turn-on
and commutation angles are shown in Fig. 18. The reference torque
waveform and the actual torque for phase A for various TSFs
are shown in Fig. 19. The obtained values for torque ripple show
that it is again improved compared to case 2, due to the optimal
selection of PI coefficients. This is the most common control of
an SRM based on TSFs with a speed controller that turn-on and
commutation angles are applied based on a lookup table. As the
turn-on angle reduces to the non-aligned angle, the copper losses
are also reduced more. Therefore, the drive efficiency will be
enhanced. In this paper, the turn–on angle is reduced to 36◦ , while
in other methods it reached 33◦ .

5. C ONCLUSIONS
In this paper, a novel method is proposed and evaluated for the
aim of MTRPT of an SRM applied in EV, based on various TSFs. Fig. 18. The variation of angles of an SRM employed in EV, (a) turn-on
The simulations are performed in various cases, and the results angle, (b) commutation angleF
show better performance of the controller:
A. Dejamkhooy and A. Ahmadpour : Torque Ripple Reduction of the Position Sensor-less Switched Reluctance... 266

•In the first case, all angles are assumed as fixed, and the
average torque ripple is obtained at about 24%. In this case,
the LTSF was better than other TSFs.
• In the second case, the speed was higher than in case
1, and the angle database is considered with a lookup
table to decrease the computational volume. To avoid large
fluctuations in current, at the beginning and end of the
conduction range, the phase shift values are optimally
selected. In this case, the turn-on and commutation angles
are controlled and the mean value of the torque ripple was
about 20%, which is declined in comparison with case 1.
• Finally, the proposed control method is applied in an EV,
which needs to continuously brake and accelerate. First, it
accelerated, and then braked to assess the performance of the
method and its response speed. The results are acceptable
and the angles variated as fast as in case 2.
• The coefficients of the PID are optimized with IPPOA that
could response very faster than other methods.
It should be noted that if this control method is performed for
SRM with more poles, much less torque ripple will be obtained.

R EFERENCES
[1] I. Boldea, ,L.N. Tutelea, ,L. Parsa, and D. Dorrell, "2014.
Automotive electric propulsion systems with reduced or
no permanent magnets: An overview," IEEE Trans. Indust.
Electron., 61(10), pp.5696-5711.
[2] E. Bostanci, ,M. Moallem, A. Parsapour, and B. Fahimi,
"Opportunities and challenges of switched reluctance motor
drives for electric propulsion: A comparative study," IEEE
Trans. Transp. Electrif. ,3(1), pp.58-75, 2017.
[3] M.V. de Paula, and T.A. dos Santos Barros, "A New Flux
Linkage Estimation with Drift Cancellation Technique for
Switched Reluctance Machines," Electronics, 9(3), p.405,
2020.
[4] H. Li, ,B. Bilgin, and A. Emadi, "An improved torque sharing
function for torque ripple reduction in switched reluctance
machines, IEEE Trans. Power Electron., 34(2), pp.1635-1644,
2018.
[5] M. Hamouda, ,A. Abdel Menaem, ,H. Rezk, ,M.N.
Ibrahim, and L. Számel, "Comparative evaluation for
an improved direct instantaneous torque control strategy
of switched reluctance motor drives for electric
vehicles," Mathematics, 9(4), p.302, 2021.
[6] G Mariappan, . and K. Lakshmanan, "An enhanced control
method for torque ripple minimization of switched reluctance
motor using hybrid technique," J. Intell. Fuzzy Syst. ,
(Preprint), pp.1-24, 2022.
[7] A.K. Rana, and A.R. Teja, "A mathematical torque ripple
minimization technique based on a nonlinear modulating
factor for switched reluctance motor drives," IEEE Trans.
Indust. Electron., 69(2), pp.1356-1366, 2021.
[8] H. Li, ,B. Bilgin, and A. Emadi, "An improved torque
sharing function for torque ripple reduction in switched
reluctance machines," IEEE Trans. Power Electron., 34(2),
pp.1635-1644, 2018.
[9] H. Cai, H. Wang, M. Li, S. Shen, Feng, Y. and J. Zheng,
"Torque ripple reduction for switched reluctance motor with
optimized PWM control strategy," Energies, 11(11), p.3215,
2018.
[10] S. Song, R. Hei, R. Ma, and W. Liu, . Model Predictive
Control of Switched Reluctance Starter/Generator With
Torque Sharing and Compensation," IEEE Trans. Transp.
Fig. 19. The results of torque for phase A of the SRM employed in EV, Electrif., 6(4), pp.1519-1527, 2020.
(a) LTSF, (b) STSF, (c) ETSF, (d) CTSF
[11] M.V. de Paula, dos Santos T.A. Barros, H.S. Moreira, Catata,
E.H., Villalva, M.G. and E. Ruppert Filho, "A dahlin cruise
control design method for switched reluctance motors with
minimum torque ripple point tracking applied in electric
vehicles," IEEE Trans. Transp. Electrif., 2020
Journal of Operation and Automation in Power Engineering, Vol. 11, No. 4, Dec. 2023 267

[12] S.S. Ahmad, and G. Narayanan, "Linearized modeling of Technol. Mag. 69(7), pp.7172-7186, 2020.
switched reluctance motor for closed-loop current control. [23] C. Li, C. Zhang, J. Liu, and D. Bian, "A High-Performance
IEEE Trans. Indust. Appli., 52(4), pp.3146-3158, 2016. Indirect Torque Control Strategy for Switched Reluctance
[13] Q. Ma, ,E.R. Ayman, and J.S. Lai, "September. Small-signal Motor Drives. Math. Probl. Eng., 2021.
modeling and speed controller design for switched reluctance [24] A. Gholipour, M. Ghanbari, E. Alibeiki, and M. Jannati,
motor drives," IEEE Energy Conversion Cong. Exposi. "Sensorless FOC Strategy for Current Sensor Faults in
(ECCE) (pp. 4497-4504). IEEE, 2018. Three-Phase Induction Motor Drives," J. Oper. Autom. Power
[14] W. Peng, ,J. Pelletier, ,Y. Mollet, and J. Gyselinck, Eng., 2021.
September. Torque Sharing Function and Firing Angle [25] H. Dahmardeh, M. Ghanbari, and S.M. Rakhtala, "A novel
Control of Switched Reluctance Machines-Hysteresis Current combined DTC method and SFOC system for three-phase
Control Versus PWM," Int. Conf. Elec. Machines (ICEM) induction machine drives with PWM switching method. J.
(pp. 1717-1723). IEEE, 2018. Oper. Automa. Power Eng., 11(1), pp. 1-10, 2023.
[15] S. Mehta, I. Husain, and P. Pramod, September. Predictive [26] M. Nikpayam, M. Ghanbari, A. Esmaeli, and M. Jannati,
current control of mutually coupled switched reluctance "Vector Control Methods for Star-Connected Three-Phase
motors using net flux method," IEEE Energy Conversion Induction Motor Drives Under the Open-Phase Failure. J.
Cong. Expos. (ECCE) (pp. 4918-4922). IEEE, 2019. Oper. Autom. Power Eng., 10(2), pp.155-164, 2022.
[16] S. Mehta, ,P. Pramod, and I. Husain, June. Analysis of [27] M.V. de Paula, E.O.H. Catata, E. Ruppert Filho, dos
dynamic current control techniques for switched reluctance Santos Barros, T.A. and Williamson, S., November. Direct
motor drives for high performance applications. In 2019 IEEE Instantaneous Torque Control of Switched Reluctance
Transportation Electrification Conference and Expo (ITEC) Machines for Low Torque Ripple Regenerative Braking,"
(pp. 1-7). IEEE, 2019. Brazilian Power Electro. Conf. (COBEP) (pp. 1-5). IEEE,
[17] X. Zhang, ,Q. Yang, ,M. Ma, ,Z. Lin, and S. Yang, 2021.
. A switched reluctance motor torque ripple reduction [28] M.V. De Paula, and T.A. dos Santos Barros, "A sliding mode
strategy with deadbeat current control and active thermal DITC cruise control for SRM with steepest descent minimum
management," IEEE Veh. Technol. Mag. 69(1), pp.317-327, torque ripple point tracking," IEEE Trans. Indust. Electron.,
2019. 69(1), pp.151-159, 2021.
[18] S. Walz, R. Lazar, Buticchi, G. and M. Liserre, "Dahlin-based [29] P. Ren„ J. Zhu, Jing, Z., Z. Guo, and A. Xu, "Minimization
fast and robust current control of a PMSM in case of low of torque ripple in switched reluctance motor based on
carrier ratio," IEEE Access, 7, pp.102199-102208, 2019. MPC and TSF," IEEE Trans. Elect. Electron. Eng., 16(11),
[19] S. Mehta, Kabir, M.A., Pramod, P. and Husain, I. Segmented pp.1535-1543, 2021.
Rotor Mutually Coupled Switched Reluctance Machine for [30] Z. Yu, C. Gan, K. Ni, R. Qu, and W. Kong, "Dual Three-
Low Torque Ripple Applications," IEEE Trans. Indus. Appli., Phase Flux-Modulated Switched Reluctance Motor Drive
2021. With Maximum Torque per Ampere Strategy," IEEE Trans.
[20] M. Hamouda, A.A. Menaem, H. Rezk, M.N. Ibrahim, and Indust. Appli., 57(2), pp.5806-5817, 2021.
Számel, L., "An improved indirect instantaneous torque [31] V.P. Vujičić, "Minimization of torque ripple and copper losses
control strategy of switched reluctance motor drives for light in switched reluctance drive," IEEE trans. power electron.,
electric vehicles," Energy Reports, 6, pp.709-71, 2020. 27(1), pp.388-399, 2011.
[21] Z. Xia, B. Bilgin, S. Nalakath, and A. Emadi, "A New [32] J. Guo, X. Ma, and A. Ahmadpour, "Electrical–mechanical
Torque Sharing Function Method For Switched Reluctance evaluation of the multi–cascaded induction motors under
Machines With Lower Current Tracking Error," IEEE Trans. different conditions," Energy, 229, p.120664, 2021.
Indust. Electron., 2020. [33] A. Dejamkhooy, and A. Ahmadpour, "Optimal UC and
[22] B. Bilgin, B. Howey, A.D. Callegaro, J. Liang, M. Kordic, economic dispatching with various small energy resources in
J. Taylor, and A. Emadi, "Making the case for switched the micro-grid using IPPOA and IMILP," Energy Reports, 7,
reluctance motors for propulsion applications," IEEE Veh. pp.7572-7590, 2021.