Optimized Power Control Strategy for Sensorless Induction Motor

Drive Used in Electric Vehicle Applications

Abhisek Pal*1, Sukanta Das2, and Karuna Kiran3
1
Department of Electrical & Instrumentation Engineering, Thapar Institute of Engineering & Technology, Patiala, India
2
Department of Electrical Engineering, IIT (ISM) Dhanbad, India
3
Department of Electrical Engineering, SVERI'S College of Engineering, Pandharpur, Solapur, India
2024 IEEE 4th International Conference on Sustainable Energy and Future Electric Transportation (SEFET) | 979-8-3503-8399-7/24/$31.00 ©2024 IEEE | DOI: 10.1109/SEFET61574.2024.10717936

*
Email: [email protected]

Abstract— This paper introduces a novel technique for modern sensorless control techniques are matured enough to
minimizing losses in sensorless field-oriented induction motor replace costly speed encoders from the drive systems. This
(IM) drives using a search control (SC) approach based on effectively makes the IM drive adopted in different
improved adaptive quadratic interpolation (AQI). The AQI applications more reliable and compact. Various speed
method is applied within a sensorless IM drive utilizing a rotor sensorless control techniques for IM drives as found in the
flux (RF) model reference adaptive system (MRAS) to literature are described in [3].
demonstrate its effectiveness in reducing overall system On the other hand, the efficiency of IM drives tends to
losses. In this proposed technique, the AQI operates within the decrease significantly when operating under partial load
flux-loop of the speed control system to optimize the flux conditions. Researchers worldwide have made substantial
component of current for efficient drive operation. The efforts to develop effective energy management strategies for
algorithm is straightforward like conventional search controls IM drives to optimize energy utilization and improve
and does not require additional circuitry for practical efficiency. Even a minor improvement in IM efficiency across
implementation. It offers faster convergence without different applications can result in substantial reductions in
compromising the dynamic response of the control loop and is global energy consumption. This highlights the importance
robust against variations in machine parameters. The presented and potential of researching loss minimization techniques
search controller notably enhances the efficiency of the IM (LMT) for IMs. Operating the IM drive at an optimal flux level
drive by minimizing core losses. The present research outcome can notably reduce incurred losses, particularly minimizing
will be highly motivating for e-transportation applications. All machine core losses and enhancing overall efficiency [2]. The
simulations and analyses are conducted using primary industry methods used to determine the optimal flux
MATLAB/Simulink to validate the efficacy of this approach. level for operating IM drives can be broadly classified as: (1)
loss model based control (LMC) [4] – [13] and (2) search
NOMENCLATURE control (SC) [14] - [19]. LMC estimates losses using the
vds ,vqs d-q components of stator voltage (V) machine model and establishes an optimal flux level to reduce
ids , iqs , idr , iqr d-q components of stator and rotor current (A) machine losses. However, the accuracy of these approaches
relies on the correctness of the machine model and the precise
ψ s ,ψ r total stator and rotor flux (Wb)
knowledge of motor parameters. Nonetheless, these methods
ψ ds , ψ qs , ψ dr , ψ qr d-q components of stator and rotor flux (Wb) demonstrate faster convergence times (0.2-2.5 s) as indicated
ψdm , ψ qm d-q components of mutual flux (Wb)
in Table I.
In contrast, SC based techniques address the limitations of
Ls , L m , L r stator, mutual and rotor inductances (H) LMC by identifying optimization points within the input
R s , Rr stator and rotor referred resistance (Ω) power profile. The optimal flux level is determined through
ω e , ω r , ω sl synchronous, rotor and slip speed (rad/s) iterative search methods of control variables while
2
maintaining the required load demand. The Perturb and
σ = 1 − Lm ( Ls Lr ) ,ε leakage factor, error signal
Observe (P&O) technique, widely utilized in industrial drives
Tr = Lr Rr rotor circuit time constant (s) for loss minimization, represents a popular method. In a
^, s, * estimated, stationary, reference quantity conventional approach using fuzzy logic based SC detailed in
d
dt time-derivative operator [14], the optimal flux component of current is set to minimize
input power requirements, resulting in a 50% reduction in
I. INTRODUCTION input power to meet specified load demands according to
simulation results. Another advancement in [15] proposes a
INDUCTION motor (IM) fed by inverter a standard choice neuro-fuzzy SC scheme to generate optimal power for IM
for variable speed AC drive applications especially in electric drive operation. This method dynamically updates
vehicles (EVs), owing to their lower price, higher reliability, membership functions of the controller through back
robustness, and less maintenance requirement [1]. propagation, training the neural network with varying input
The primary techniques utilized to regulate speed in power to achieve optimal performance in both transient and
induction motor (IM) drives are direct torque control (DTC) steady-state operations of interior permanent magnet
and field-oriented control (FOC). FOC is particularly favored synchronous motor drives. In [16], two fuzzy logic based
due to its ability to generate smoother speed and torque controllers are employed to optimize power, leading to
profiles and enhance dynamic response, making it a preferred reported efficiency improvements of 2-8% for a 5 HP machine.
choice for high-performance applications [2]. Moreover, Despite these advantages, SC based methods are often

979-8-3503-8399-7/24/$31.00 ©2024 IEEE

Authorized licensed use limited to: Indian Institute of Technology Patna. Downloaded on November 07,2024 at 10:41:28 UTC from IEEE Xplore. Restrictions apply.
criticized for slow convergence rates and increased torque Lm s
ripples. However, [17] addresses torque ripple reduction by
s
ψ qm =
Lr
( s
ψ qr + Llr iqs . ) (2b)
employing a second-order low-pass filter in the optimal flux
The derivative of the rotor flux components used in the
path (derived from d-axis current) using the golden section s
reference model of RF-MRAS is derived by substituting ψ dm
technique for rotor field-oriented IM drive operations.
s
Similarly, [18] utilizes the golden section method to generate (2a) and ψ qm (2b) into (1a) and (1b) respectively, and then
optimal flux, considering magnetic saturation, hysteresis, and
rearranging.
eddy-current losses for flux estimation in the same drive
d s L s L  d  s
system. A different approach detailed in [19] employs a Ramp ψ dr = r v ds − r  R s + σL s i ds (3a)
method-based pure SC technique with slower convergence dt Lm Lm  dt 
compared to the golden section scheme. This paper also d s L s L  d  s
introduces a hybrid technique combining the Ramp method
ψ qr = r v qs − r  R s + σL s  i qs . (3b)
dt Lm Lm  dt 
and LMC scheme for loss control. A comprehensive review of The rotor flux components in the stationary reference frame
various loss minimization schemes is presented in [20], which utilized within the adaptive model are described as follows
also discusses a hybrid method called ripple correlation control [22]:
strategy integrating LMC and SC techniques to minimize
d s L s s 1 s
losses in IM drives. ψˆ dr = m i ds − ω r ψˆ qr − ψˆ dr (4a)
Extensive review of literature in this area indicates that SC dt Tr Tr
based methods used for minimizing drive losses are not d s L s s 1 s
ψˆ qr = m iqs + ω r ψˆ dr − ψˆ qr . (4b)
affected by variations in machine parameters and are dt Tr Tr
straightforward to implement compared to LMC or hybrid
controllers. However, these techniques are characterized by s* Reference Model
vds ψ drs
relatively slow convergence times (0.5-7 s) according to Table s*
vqs  pψ dr  Lr  v ds  (Rs + σ Ls p)
s s
0  i ds  
s

I. Additionally, the most effective techniques for finding the  s =   s −   s 

s*
i ds  ψ qr  Lm   qs  
p v 0 (Rs + σ Ls p) i qs  
minimum of a function include the golden section method [17] s* ψ qrs
iqs
and adaptive gradient descent method [21], provided the initial Adaptive Model +
interval of uncertainty is known. Otherwise, the adaptive
quadratic interpolation (AQI) method is expected to be more i dss
 1
 pψ̂ drs   − T

− ω̂r   s 
ψ L i 
s
ψ̂ drs
−
 = r   dr  + m  ds 
ψ̂ qrs
efficient. Therefore, the objective of this study is to employ s pψ̂ qrs   1  ψ qrs  Tr  i qss 
iqs   ω̂ −    
 r Tr 
AQI to determine the optimal flux level for induction motor ω̂r
(IM) drive operation under light load torque conditions. The
primary goal, namely selecting the optimal flux using AQI, (a)
achieves rapid convergence. This chosen optimal flux *
minimizes machine core losses, thereby enhancing the overall i
ds
efficiency of the drive system. It is noteworthy that the
proposed algorithm is implemented in a sensorless field- ∆ωr
iqs*
oriented IM drive utilizing a modified rotor flux (RF) based ωr*
model reference adaptive system (MRAS) speed estimator θe
ω̂r ωe
[22] that relies solely on current sensors. This adapted RF-  θe
MRAS speed estimator can function across a wide range, ωsl *
Lr iqs ids *
including zero speed. The proposed loss minimization R r ids iqs
technique, combined with the adopted speed control approach,
demonstrates outstanding performance in enhancing the
efficiency of the entire drive system. Thus, the present research (b)
outcome is highly inspiring for e-transportation applications in Fig. 1. MRAS based speed estimator and AQI algorithm: (a)
terms of a longer run or range. Detailed simulation studies are RF-MRAS structure, (b) schematic diagram of sensorless
conducted using MATLAB/Simulink to validate the vector-controlled IM drive with loss minimization strategy.
effectiveness of the proposed technique.
Fig.1(a) shows the RF-MRAS estimator containing the
II. BRIEF FORMULATION OF RF-MRAS BASED SPEED reference and adaptive models. The voltage and current
ESTIMATOR FOR IM DRIVE components in (3) correspond to the reference values (see
*
The equations representing the stator circuit voltage of an Fig.1(b)). The reference q- and d- current components ( iqs and
IM in a stationary frame are as follows [22]: *
) generated by speed proportional-integral (PI) controller
i ds
s s d s s
v ds = Rs ids + Ls ids + pψ dm (1a) and the AQI based optimum flux selection algorithm for loss
dt minimization respectively are compared against the actual q-
s s d s s and d- current components sensed by two current sensors. The
v qs = Rs iqs + Ls iqs + pψ qm (1b)
dt error signals thus obtained are fed to two PI controllers for the
where, s*
generation of reference voltage commands ( v dss * and vqs ) for
Lm s
s
ψ dm =
Lr
( s
ψ dr + Llr ids ) (2a) inverter to fulfill certain speed set of IM drive. The estimated
speed is compared with the reference speed and the resultant
error signal is fed to a speed PI controller for the generation of

Authorized licensed use limited to: Indian Institute of Technology Patna. Downloaded on November 07,2024 at 10:41:28 UTC from IEEE Xplore. Restrictions apply.
 * p (ids ) depicting loss is given by (see Fig. 2(a))
reference torque or q-component of stator current ( iqs ).The
p (ids ) = P (ids 0 ) + P [i ds 0 , ids1 ](i ds − i ds 0 ) +
unit vector is generated in an indirect way via the calculation (5)
P[i ds 0 ,ids1 , ids 2 ](ids − i ds 0 )(i ds − i ds1 )
of slip speed ( ωsl ). Fig. 1(b) depicts the complete schematic
diagram of a field-oriented IM drive with modified RF-MRAS where, P(ids) is the total drive’s loss, first-order divided
based speed estimator and the proposed LMT. P (i ds 1 ) − P (i ds 0 )
difference, P [i ds 0 , i ds 1 ] = and second-order
i ds 1 − i ds 0
III. LOSS MINIMIZATION TECHNIQUE
P [ids1 ,i ds 2 ] − P [i ds 0 ,i ds1 ]
A loss minimization approach based on SC using adaptive divided difference, P [ids 0 ,ids1 ,ids 2 ] = .
i ds 2 − i ds 0
quadratic interpolation (AQI) for IM drive is integrated into
the outer loop of the control scheme, illustrated in Fig. 1(b), to The turning or lowest point on the p (ids ) curve can be
*
determine the optimal flux component of current ( i ds ) needed obtained by equating its slope to zero as
dp (ids )
for drive operation. The drive loss profile is calculated as the = P [ids 0 ,ids 1 ] + 2 ids P [ids 0 ,ids 1 ,ids 2 ] − (6)
di ds
difference between the power input to the inverter and the shaft P [ids 0 ,ids 1 ,ids 2 ](ids 0 + ids 1 ) = 0
*
power output. Specifically, the optimal i ds minimizes the core Rearranging (6), i ds (≈ i ds _ m ) corresponding to the lowest
loss in this context. The fundamental principle of quadratic
point of the p (ids ) curve is obtained as
interpolation (also known as Powell’s method [23]) involves
fitting successive quadratic interpolation curves to function P [i ds 0 , i ds 1 , i ds 2 ](i ds 0 + i ds 1 ) − P [i ds 0 , i ds 1 ] *
i ds _ m = ≅ i ds . (7)
data, providing a sequence of approximations through an 2 P [i ds 0 , i ds 1 , i ds 2 ]
*
iterative process towards the minimum point. The i ds *
The obtained i ds is minimum, provided the second
corresponding to the minimum loss point determines the derivative of loss curve is non-negative as
required flux, while the torque component of current is
P[i ds 0 ,i ds1 ,i ds 2 ] > 0 (8)
obtained from the speed control loop. The operation of AQI in
this context is detailed in this section. A comprehensive flow-chart for finding the optimum value
i.e. i *ds of the LMT algorithm is given in Fig. 2(b). Its sequence
p (ids ) of operation is described hereunder:
Step1: Initialization of i ds 0 = 0 , i ds 2 = 3 .8 A (at rated flux
ids 2 component of current), ids1 = (ids 0 + ids 2 ) 2 and current
P(ids )
tolerance, ε = 0.05.
*
ids 0 ids1 ids_m i ds Step2: Computation of ids _ m using (7) and measurement of
(a) powers for P (i ds _ m ) and P(ids1 ) .
Start
Initilize ids 0 ,ids1,ids 2 Step3: Verification for current tolerance, ids _ m − ids1 < ε and
tolerance ε
P[ids 0 , ids1 , ids 2 ] > 0 using (8). Finding out i ds _ m .
Compute i ds_m ,
Measure P (ids_m ) and P (ids1 )
Step 4: Checking for the following four cases:
Case I: If i ds _ m < i ds 1 and P (i ds _ m ) < P (i ds 1 ) , then ids0 , ids1 ,
Does
(i − ids1 < ε ) ids2 become ids0 , i ds _ m , ids1 respectively for the next iteration
ds _ m yes
&&
(P [ids 0 , ids 1 , ids 2 ] > 0 ) [see Fig. 3(a)].
fulfills ?
i*ds ≅
(ids_ m + ids1) Case II: If i ds _ m < i ds 1 and P (i ds _ m ) > P (i ds 1 ) , then ids0 ,
no 2
ids1 , ids2 become i ds _ m , ids1 , i ds 2 respectively [see Fig. 3(b)].
ids_ m < ids1 no Stop
yes Case III: If i ds _ m > i ds 1 and P (i ds _ m ) < P (i ds 1 ) , then ids0 ,
ids1 , ids2 become ids1 , i ds _ m , i ds 2 respectively [see Fig. 3(c)].
( )
P ids_m < P(ids1 ) ( )
P ids_m < P(ids1 )
Case IV: If i ds _ m > i ds 1 and P (i ds _ m ) > P (i ds 1 ) then ids0 , ids1
yes no yes no
, ids2 become ids0 , ids1 , i ds _ m respectively for the next iteration
ids 0  ids 0 ids 0  ids _ m ids 0  ids1 ids0  ids0
ids1  ids _ m ids1  ids1 ids1  ids _ m ids1  ids1 [see Fig. 3(d)].
ids 2  ids1 ids 2  ids 2 ids 2  ids 2 ids 2  ids _ m Step 5: Steps 2 to 4 are repeated until the condition specified
in step 3 is satisfied.
Step6: The final optimal value ids = ids _ m + ids1 2 is
*
( )
(b) determined by averaging to minimize the error.
It should be emphasized that the input power is
Fig. 2. Execution of LMT: (a) Approximate optimum point,
monitored and compared with the previous iteration's value,
(b) Flow-chart. following a methodology distinct from that described in [14],
For a given three sets of data points {(idsi , P (idsi )),i = 1,2 ,3} , which involves no additional step size 'H'. In this approach,
the interpolating quadratic polynomial through these points when an upward trend in power consumption is detected, the

Authorized licensed use limited to: Indian Institute of Technology Patna. Downloaded on November 07,2024 at 10:41:28 UTC from IEEE Xplore. Restrictions apply.
flux (or control variable) retains its previous value, achieving smooth speed tracking performance throughout the operation.
the optimal operating flux. Initially, the IM drive operates at The corresponding power profile is depicted in Fig. 5(b).
the rated flux condition, specifically ids2 = 3.8 A. During any Initially, the IM drive operates at rated flux. The average
transient condition arising from sudden changes in speed or energy consumption is measured at 4480 Joules [shown as the
load, the LMT activates automatically once the steady-state 'black' profile in Fig. 5(b)]. However, with the implementation
*
speed error ( Δω r ) drops below 1 rad/s. The optimal flux ( i ds of the LMT for the same speed pattern at 5 seconds, the energy
consumption decreases to 3815 Joules [shown as the 'red'
) adjustment ensures that the drive consistently operates at profile in Fig. 5(b)]. This results in an energy saving of 665
maximum achievable efficiency without impacting power Joules.
requirements at the load side.
The proposed loss minimization algorithm is adaptive in
(
P ids _ m ) P(ids1 ) (
P ids _ m ) P(ids1 )
nature and it works according to the variation in speed of the
drive. The optimization algorithm selects the optimum
operating flux level for drive control. As a result, machine core
loss reduces which in turn brings the overall loss of the drive
ids 0 ids_m ids1 ids 2 ids 0 ids_m ids1 system down without affecting drive operating condition (see
ids 2
Figs. 4-5).
}
}

To consider To eliminate To eliminate To consider

(a) (b)
ω *r ω̂ r
P(ids1 ) (
P ids _ m ) P(ids1 ) (
P ids _ m )
ωr
ids 0 ids1 ids_m ids 2 ids 0 ids1 ids_m ids 2
}

To eliminate To consider To consider To eliminate

(c) (d)
Fig. 3 illustrates various potential conditions at step 4: (a)
i ds _ m < i ds 1 and P (i ds _ m ) < P (i ds 1 ) , (b) i ds _ m < i ds 1 and
ψdr
( ) ( )
P i ds _ m > P (i ds 1 ) , (c) i ds _ m > i ds 1 and P i ds _ m < P (i ds 1 ) , (d) ψqr
( )
i ds _ m > i ds 1 and P i ds _ m > P (i ds 1 ) .

IV. SIMULATION RESULTS: ASSESSMENT OF DRIVE

PERFORMANCE UTILIZING AQI Fig.4. Simulation results at 50 rad/s, 2 Nm load torque: (a)
This section presents simulation results that evaluate the ω *r , ω̂ r , ω r , (b) i ds , (c) ψdr -ψ qr , and (d) loss profile.
effectiveness of the proposed loss minimization algorithm for
operating an IM drive. The validation is conducted by ω r*
designing a simulation model using MATLAB/Simulink. Fig.
1(b) illustrates a representative diagram of the Simulink model ω̂ r
depicting the entire control scheme.
ωr
5.1 Constant Speed Operation
Fig. 4 illustrates a typical simulation outcome under
constant speed conditions of 30 rad/s with a 2 Nm load torque,
where the LMT is engaged at 5 seconds. In Fig. 4(a), it is
evident that the speed tracking performance remains accurate Fig.5. Simulation results for step speed operation: (a) ω *r , ω̂ r ,
even after the LMT is activated. Fig. 4(b) displays a notable ω r , (b) i ds , (c) ψdr -ψ qr , and (d) loss profile.
decrease in the flux component of the stator current, resulting
without AQI
in a reduction of the operating flux level as depicted in Fig. ωr= 100rad/s with proposed AQI
80 500
Efficiency, %

4(c). The flux orientation is effectively maintained post-LMT

ωr = 100rad/s
Loss, W

activation, as shown in Fig. 4(c). The reduction in the flux ωr = 50 rad/s 400
60
level contributes to an 18 W decrease in drive loss, 300
consequently reducing input power consumption by 20% ωr = 10 AQI1
rad/s ωr= 50 rad/s
40 Conv.
Conv2 200
while sustaining the same output power, as illustrated in Fig. without AQI
AQI2
Conv3 ωr = 10rad/s
4(d). Consequently, the overall efficiency of the drive system 20 with proposed AQI AQI 100
0.0 0.5 1.0 0.0 0.2 0.4 0.6 0.8 1.0
improves by 24.5%. Load (pu) Load, pu
(a) (b)
5.2 Step Speed Operation Fig.6. Performance assessment of AQI: (a) drive’s efficiency
Fig. 5 displays the test results for an Indian vehicle drive at variable load torque, and (b) drive’s loss at variable load
cycle. In Fig. 5(a), it is observed that the drive maintains torque.

Authorized licensed use limited to: Indian Institute of Technology Patna. Downloaded on November 07,2024 at 10:41:28 UTC from IEEE Xplore. Restrictions apply.
 V. OVERALL PERFORMANCE ASSESSMENT estimator and AQI based LMT make IM drive more reliable
The efficacy of the proposed algorithm in enhancing the and efficient. The LMT is integrated into the outer loop of the
efficiency of the drive system is further examined across speed control system to determine the optimal level of
various operating points. Fig. 6(a) and (b) demonstrate the operating flux current component at a specific operational
comparative efficiency and corresponding loss of the IM drive state. As a result, the chosen optimal flux level minimizes
system respectively under variable load at 10 rad/s, 50 rad/s losses at that operating state, leading to improved overall drive
and 100 rad/s. The nature of the curves reveals that during efficiency, particularly under low load conditions. Here, IM
constant speed with variable load operation, core loss drive is tested under the vehicle drive cycle and 665 Joules
stabilizes at a constant optimal value while copper loss energy saving is noted for 50 s duration. Analysis further
increases. Conversely, during constant load with variable indicates that this proposed technique achieves rapid
speed operation, copper loss remains constant while core loss convergence (0.1 seconds) compared to other search control
settles at a new optimized value. The proposed algorithm methods, all while maintaining the dynamic response of the
optimizes the flux component of current, leading to a reduction speed control loop. Therefore, this approach could be well-
in flux level and consequently minimizing drive core loss suited for applications in electric transportation powered by
significantly. As a result, a notable enhancement in efficiency induction motors, where achieving high dynamic performance
is observed during light loading conditions compared to full and optimizing efficiency are critical factors.
load conditions. Fig. 6 demonstrates this fact clearly.
ACKNOWLEDGMENT
The efficacy of the proposed algorithm is further
The research was conducted at the Department of Electrical
elaborated through a comparative study with the other
and Instrumentation Engineering, Thapar Institute of
available methods in terms of convergence rate and parameter
Engineering and Technology, Patiala. The author expresses
sensitivity in Table I.
gratitude to the institute for providing Seed grant funding for
TABLE I. Comparative Analysis with Other Available Methods this research work.
Indices
Ref- Mot-
Time of Param-
ere- or Loss
conver- eter REFERENCES
nces Improvement rati- Min.
gence sensiti-
ng Sch.
(s) vity
(HP)
[1] R. Paes, B. Toner, G. Bankay and G. Agashe, "Energy
[4] Ploss reduced by 70% 2.5 1/3
Efficiency improved
optimization for adjustable-speed drive applications:
[5] 0.5 1 Increasing efficiency to cut costs and preserve assets",
by 50 %
Efficiency improved IEEE Ind. Appl. Mag., vol. 27, no. 5, pp. 37-46, Sep/Oct.
[6] 0.5 to 5 1.5
by 12 % 2021.
Unspe-
[8] Ploss reduced by 75% 1.5 High LMC [2] B. K. Bose, Modern power electronics and ac drives,
cified
Unspe- New Delhi: Prentice Hall, 2002.
[9] Ploss reduced by 30% 0.75
cified
[3] R. Kumar, S. Das, P. Syam and A. K. Chattopadhyay,
Unspe
[10] Ploss reduced by 30% 0.2
-cified "Review on model reference adaptive system for
[13] Ploss reduced by 7% 1 2 sensorless vector control of induction motor drives," IET
[14] Ploss reduced by 50% 7 1.5 Electric Power Appl., vol. 9, no. 7, pp. 496-511, Jul.
[15]
Efficiency improved
0.5 0.25
2015.
by 27 %
Efficiency improved Unspe-
[4] M. N. Uddin and S. W. Nam, "New online loss-
[16] 5 minimization-based control of an induction motor
by 2-8 % cified
[17] Pin reduced by 30% 1.8 3 drive," IEEE Trans. Power Electron., vol. 23, no. 2, p.
[18] Pin reduced by 12% 0.5-0.7 3 None SC 926–933, Mar. 2008.
[19] Pin reduced by 21.7% 2 3
Efficiency improved
[5] C. Mademlis, I. Kioskeridis and T. Theodoulidis,
[21] 0.5 0.94 "Optimization of single-phase induction motors—Part I:
by 51.3 %
Prop Efficiency improved Maximum energy efficiency control," IEEE Trans.
-osed by 39.13 % (at 2 Nm 0.1 3 Energy Convers., vol. 20, no. 1, p. 187–195, Mar. 2005.
AQI load)
[19] Pin reduced by 21.6% 1 3 Mediu- Hyb- [6] X.Han, W. Liu, N. Jiao, P. Yao, S. Mao, Z. Li, "Loss
[21] Ploss reduced by 61% 0.5 1.5 m rid minimization control of aircraft wound-rotor
synchronous starter–generator in the starting mode",
All the studies confirm the usefulness of the proposed IEEE Trans. Power Electron., vol.39, no.5, pp.6063-
search control algorithm in improving IM drive’s efficiency 6076, 2024.
with a fast time of convergence.
[7] M. N. Uddin and S. W. Nam, "Development of a
nonlinear and model-based online loss minimization
VI. CONCLUSION control of an IM drive," IEEE trans. Energy Convers.,
In this paper, sensorless vector controlled IM drive in vol. 23, no. 4, pp. 1015-1024, Dec. 2008.
conjunction with a new online search control based efficiency [8] G. Goswami and S. Das, "An online efficient loss model
improvement scheme in the form of adaptive quadratic controller for rotor-field oriented induction motor
interpolation for the loss minimization of IM drive operation
drives," IEEE trans. Energy Convers., vol. 39, no. 1, pp.
is investigated. In this regards, overall system is tested
578-588, Mar. 2024.
rigorously under different operating conditions and it is found
that the scheme works satisfactorily. The modified RF-MRAS

Authorized licensed use limited to: Indian Institute of Technology Patna. Downloaded on November 07,2024 at 10:41:28 UTC from IEEE Xplore. Restrictions apply.
[9] M. Farasat, A. M. Trzynadlowski and M. S. Fadali, [23] A. Pal and S. Das, “Search controller-based online
"Efficiency improved sensorless control scheme for efficiency optimisation strategy for induction motor
electric vehicle induction motors," IET Electr. Syst. drives using modified adaptive quadratic interpolation,”
Transp., vol. 4, no. 4, pp. 122-131, Dec. 2014. IET Power Electron., vol. 13, no. 18, pp. 4282–4290,
[10] J. Yu, W. Pei and C. Zhang, "A loss-minimization port- Dec. 2020.
controlled Hamilton scheme of induction motor for
electric vehicles," IEEE/ASME Trans. Mechatronics,
vol. 20, no. 6, pp. 2645-2653, Dec. 2015.
[11] M. C. D. Piazza, M. Luna and M. Pucci, "Electrical loss
minimization technique for wind generators based on a
comprehensive dynamic modeling of induction
machines," IEEE Trans. Ind. Appl., vol. 53, no. 4, pp.
3696 - 3706, Jul./Aug. 2017.
[12] Y. Shi, R. D. Lorenz and M. A. Valenzuela, "New
developments in loss minimizing control for drives
without compromising torque dynamics," IEEE trans.
Ind. Appl., vol. 52, no. 6, pp. 5327 - 5335, Nov./Dec.
2016.
[13] S. Sridharan and P. T. Krein, "Minimization of system-
level losses in VSI-based induction motor drives: offline
strategies," IEEE Trans. Ind. Appl., vol. 53, no. 2, pp.
1096-1105, Mar./Apr. 2017.
[14] L. Ramesh, S. P. Chowdhury, S. Chowdhury, A. K. Saha
and Y. H. Song, "Efficiency optimization of induction
motor using a fuzzy logic based optimum flux search
controller," in Proc. Int. Conf. Power Electron. Drives
Energy Syst., New Delhi, 2006.
[15] A. H. B. M. Yatim and W. M. Utomo, "Neuro-fuzzy on-
line optimal energy control for variable speed
compressor motor drive system," in Proc. Int. Conf.
Power Electron. Drives Syst., Kuala Lumpur, 2005.
[16] M. N. Uddin and J. Khastoo, "Fuzzy logic-based
efficiency optimization and high dynamic performance
of IPMSM drive system in both transient and steady-
state conditions," IEEE Trans. Ind. Appl., vol. 50, no. 6,
pp. 4251-4259, Nov./Dec. 2014.
[17] C. M. Ta and Y. Hori, "Convergence improvement of
efficiency-optimization control of induction motor
drives," IEEE Tran. Ind. Appl., vol. 37, no. 6, pp. 1746-
1753, Nov./Dec. 2001.
[18] Z. Qu, M. Ranta, M. Hinkkanen and J. Luomi, "Loss-
minimizing flux level control of induction motor
drives," IEEE Trans. Ind. Appl., vol. 48, no. 3, pp. 952-
961, May/Jun. 2012.
[19] C. Chakraborty and Y. Hori, "Fast efficiency
optimization techniques for the indirect vector-
controlled induction motor drives," IEEE Tran. Ind.
Appl., vol. 39, no. 4, pp. 1070-1076, Jul./Aug. 2003.
[20] A. M. Bazzi and P. T. Krein, "Review of methods for
real-time loss minimization in induction machines,"
IEEE Trans. Ind. Electron., vol. 46, no. 6, pp. 2319-
2328, Nov./Dec. 2010.
[21] A. Taheri, A. Rahmati and S. Kaboli, "Efficiency
improvement in DTC of six-phase induction machine by
adaptive gradient descent of flux," IEEE Trans. Power
Electron., vol. 27, no. 3, pp. 1552-1562, Mar. 2012.
[22] A. Pal, S. Das and A. K. Chattopadhyay, "An improved
rotor flux space vector based MRAS for field oriented
control of induction motor drives," IEEE Trans. Power
Electron., vol. 33, no. 6, pp. 5131-5141, Jun. 2018.

Authorized licensed use limited to: Indian Institute of Technology Patna. Downloaded on November 07,2024 at 10:41:28 UTC from IEEE Xplore. Restrictions apply.