DUAL ACTIVE BRIDGE CONVERTER

FOURTH DRAFT

Power Electronics Converter

Name: SAYAN CHATTERJEE

Roll Number: 246102019

Course Instructor: Prof.CHANDAN KUMAR

October 30, 2024

1
 Dual Active Bridge Converter

1 Introduction
The dual active bridge, a bidirectional, controlled, dc-dc converter with high power capabilities, is made
up of dc-link capacitors, a high frequency transformer, eight semiconductor devices, and an energy transfer
inductor. A more straightforward description of the converter would be a standard full-bridge with a
controlled rectifier. This converter’s symmetry—its primary and secondary bridges are identical—allows
it to manage power flow in both directions, which is why it’s used for a variety of high-power applications.
[1].
The topology is shown in Fig. 1, where V1 and V2 are the dc-link voltages, L1 is the leakage inductance
of the transformer plus any necessary external energy transfer inductance , and S1−4 and Q1−4 are
the controllable semiconductor switches. Depending on the needs for power transfer, there are several
configurations and constructions for the Dual Active Bridge system. A 2011 study titled ”High-frequency
design considerations of dual active bridge 1200 V SiC MOSFET DC-DC converter” explored how SiC
and high frequency could be used to improve the dual active bridge’s power density even more. [2].
Various control strategies [3] are formulated for the control and power transfer of DAB converter.They
are broadly classified as follows. Single-Phase Shift (SPS) involves phase-shifting only one bridge, offering
limited control. Double-Phase Shift (DPS) provides a wider control range by phase-shifting both bridges.
Triple-Phase Shift (TPS) introduces a third phase shift, further enhancing control flexibility. Equal-Phase
Shift (EPS) maintains equal phase shifts between the bridges, simplifying control but limiting the output
voltage range. The choice of control strategy depends on factors such as desired output voltage range,
power level, efficiency, and control complexity.

1.1 origin
In ”A three-phase soft-switched high power density DC/DC converter for high power applications” [4]
from 1988, the Dual Active Bridge circuit was first employed with a thyristor. (DAB) operation. It
was mainly designed for three phase power transfer through DAB. This paper was further modified to
journal to include indepth control analysis in 1991 [5]. One conference paper by the title ”Performance
characterization of a high power dual active bridge DC/DC converter” shows use of IGBT for a 50 KW
setup [6].

2 Power Flow Analysis

Power flow analysis in power electronic converters is the study and calculation of how electrical power
is transported, processed, and distributed among various components of a power converter system. It
entails studying the steady-state operating conditions of the system to ensure efficient and dependable
functioning.

2.1 Simple AC circuit power flow analysis

. Two totem-poled switching devices, or two switches operating in complementary fashion to convert a
DC to AC signal, make up each full-bridge. These switches are powered by complementary square-wave
pulses. The switching frequency of the converter (Fs ) is the frequency at which these complementary
devices switch on. High frequency switching will be used in this application to take advantage of SiC’s
excellent physical qualities and minimize the size of passive components. The magnetizing inductance
of an isolation transformer is minimal at high frequencies, and its leakage inductance is the sole way
to model the transformer. Fig. 2 reflects an equivalent system, which will be used to derive the power
equation for the converter and Fig. 5 shows the equivalent voltage and current flow through leakage
inductance in ideal case.[7] Power transfer between the two bridges in a dual-active bridge is analogous
to the power flow between two voltage buses in a power system. Consider two voltage sources connected
by a line reactance shown in Fig. 2.

VI = per phase sending end voltage

VO = per phase receiving end voltage

2
 S1
S3
D1 D3
S1 S3

D1
Switch L
D3 +
Switch
T
1 Switch Switch VB
C
_
C
+
VB
_ n:1 S4
S2 S4
S2
D2
D4 D2 D4

Switch Switch Switch Switch

Figure 1: DUAL ACTIVE BRIDGE TOPOLOGY.[1]

V<0° jɷL V1=VL

Figure 2: Power Transfer Between Voltage Bus

VO lags VI by an angle ϕ
X = reactance of per transmission line.
On the per phase basis power on sending end,

SI = PI + jQI = VI I ∗
(1)
SO = PO + jQO = VO I ∗

where SO is total output power, PI ,PO and QI ,QO are active and reactive input and output power and
I is the current flowing from input to output which can be written as
VI − VO
I=
jX
∗ ∗ (2)
V I − VO
I∗ =
−jX

From equation (1) and (2), we get

VI (VI ∗ − VO ∗ )
SI = .
−jX
(3)
VO (VI ∗ − VO ∗ )
SO =
−jX

3
Now
VI = |VI | ̸ 00 so,,VI = VI∗ = |VI |
(4)
VO = |VO | ̸ − ϕ = |VO | e−jϕ .
Using above values equation (3) becomes

|VI | V) j1  2

SO = PO + jQO = sin ϕ + |VO | − |VI | |VO | cos ϕ (5)
X X
The active power transfer between the two buses can be written as depending on power angle or load
angle ϕ.
|VI | |VO |
PO = sin ϕ (6)
X
This clearly tells that power transfer happens due to phase shift B.

2.2 Single Phase Shift Control Strategy

.
The Single Phase Shift (SPS) Control Strategy is widely used in Dual Active Bridge (DAB) converters
to control power flow between two DC buses. The DAB converter topology is widely used in power
conversion applications, including bidirectional power flow in battery energy storage systems, electric
cars, and renewable energy systems.

2.3 Double Phase Shift Control Strategy

The Dual Phase Shift (DPS) Control Strategy in a Dual Active Bridge (DAB) converter is an advanced
control method that provides more flexibility and improved performance compared to the simpler Sin-
gle Phase Shift (SPS) control strategy. It addresses some of the limitations of SPS, such as increased
circulating currents and reduced efficiency at light loads.

2.4 Extended Phase Shift Control Strategy

The Extended Phase Shift (EPS) Control Strategy for Dual Active Bridge (DAB) converters provides
greater flexibility than Single Phase Shift (SPS) and Dual Phase Shift (DPS) strategies. EPS enhances
the performance of DAB converters, especially under light loads, by lowering circulating currents and
increasing efficiency across various operating circumstances.

2.5 Triple Phase Shift Control Strategy

The Triple Phase Shift (TPS) Control Strategy for a Dual Active Bridge (DAB) Converter is a more
flexible and efficient control method than Single Phase Shift (SPS) and Dual Phase Shift (DPS) control
strategies. It is intended to increase the converter’s efficiency by lowering circulating currents, improving
zero-voltage switching (ZVS) circumstances, and increasing overall power transfer management, partic-
ularly at low loads.By individually managing these three phase shifts, the TPS technique provides for
tighter control over power transmission and avoids inefficiencies such as circulating currents, which arise
when power is not optimally delivered [8].

2.6 DAB power flow analysis

.
Phase-shifting the pulses of one bridge in relation to the other allows one to direct the flow of power
in the dual active bridge. Phase shift modulation (PSM) is a type of control that routes power between
the two dc busses so that power from the leading bridge reaches the lagging bridge.A voltage differential
across the leaking inductance is produced by the applied square waves in Fig. 5(c), which serves as an
illustration of this idea. Power delivery from the primary bridge to the secondary bridge is instantiated
by shifting the secondary bridge pulses by +ϕ, taking into account the control pulses for switches S1,4
of the primary bridge and S5,8 of the secondary bridge. Likewise, electricity is sent to the major bridge
when the secondary bridge is moved by -ϕ to become the leading bridge.

4
 500

1000

1000

-1000

50

-100

2000

-2000

Figure 3: Single Phase Shift Modulation Scheme waveform.

The symmetry of the current waveform IL through the leakage inductance allows for the following
power flow analysis to be developed using a half switching period. The inductor current waveform can
be expressed as:

diL (t) vi (t) − vo (t)

= (7)
dt Lk
Each half cycle can be divided into two intervals: Interval 1 occurs between (0 < θ< ϕ) and interval
2 is defined as (ϕ < θ < π). Considering the depiction of the current waveform in Fig. 5(c), solving for
(1) gives the following expressions during the two time durations.

vo I1 + I2
vi + = Lk , for 0 < t < dT (8)
n dT
vo I1 − I2
vi − = Lk , for dT < t < T (9)
n (1 − d)T
With n being the turns ratio of the transformer, T being the duration of a half-cycle of the period, I1 and
I2 being the inductor current during switching instances, and d being the phase shift duty percentage of
the two bridges, which will be referred to as the duty cycle of the converter. The inductor current, input
current and output current of DAB over half cycle is shown in Fig. 4. From Fig. 4(a), the following

5
 I1

I2

-I2

-I1

t1 t2

I1

I2

dT
T

n.Io
I1
I2

dT

Figure 4: (a) leakage inductance current with some important parameters. (b) input current (ii ). (c)
output current on the primary side of the converter (nio ), both used to calculate the proposed averaged
model.[7]

relation can be derived between I1 , t1 , I2 , t2 and dT .

I1 I2
= (10)
t1 t2
t1 + t2 = dT (11)
the set of (2) and (3) can be solved, obtaining

T  vo vo 
I1 = 2 d + vi − (12)
2Lk n n

6
 T  vo 
I2 = 2vi d − vi + (13)
2Lk n
!
2 vno d + vi − vno
t1 = T (14)
2 vi + vno

!
2vi d − vi + vno
t2 = T . (15)
2 vi + vno

The output average current injected into the output cell, comprising the load and the output capacitor,
is calculated using Fig. 4(c) as
 
1 1 1 1
lo = I1 t1 − I2 t2 +(1−d) T I2 +(1−d) T (I1 −I2 ) (16)
nT 2 2 2

Using (6)–(9), the expression of the output average current (10) can be simplified as

(1 − d) dT vi
lo = . (17)
nLk
Using a similar analysis to the one described previously to obtain the average output current, the
average input current can be obtained as
 
1 1 1 1
ll = I2 t2 − I1 t1 + (1 − d) T (I1 + I2 ) . (18)
T 2 2 2

Once again, using (6)–(9), the expression of the input average current can be simplified as

(1 − d) dT vo
ll = (19)
nLk
From this the average output power can be derived:

(1 − |d|)dT vi vo
P = vo Io = (20)
nL

3 Generalized Model of DAB Converters

A grid connected inverter can be considered as a current source as the grid connected inverter volt-
age depends on the grid and only current is supplied by the inverter. For the veneralized model the
high frequency transformer magnetizing current is considered to be very small, the voltage drop across
transistor-diode pairs is negligible, and transistor switching transients are negligible. Using Phase Shift
Modulation, the voltage at the transformer input side, vin , can only have two states: 1) +Vin when
transistors S1 and S4 are ON; and 2) Vin when transistors S2 and S3 are ON. Therefore,

vin (τ ) = s1 (τ )Vin (τ ) (21)

where the switching function at input bridge, s1 (τ ), is


T T
s1 (τ ) = 1, 0 ≤ τ < − 1, ≤ τ < T (22)
2 2

where T = 1/fs . Throughout the analysis, τ is used to denote the time within a period, referenced to
the rising edge of s1 . Similarly, at output bridge

vo (τ ) = s2 (τ )Vout (τ ) (23)

and the switching function at output bridge, s2 (τ ), is


dT T dT dT T dT
s2 (τ ) = 1, ≤τ < + − 1, 0 ≤ τ < or + ≤ τ < T. (24)
2 2 2 2 2 2

7
If we select transformer current, il , and output capacitor voltage, vo , as state variables, then the state
equations of a DAB converter can be derived as
dil (τ ) s1 (τ ) s2 (τ )
= Vin (τ ) − Vout (τ ) (25)
dτ L L
and
dvo (τ ) 1 s2 (τ ) Io
=− vo (τ ) + il (τ ) − (26)
dτ RCo Co Co
Equations (20) and (21) are both time-varying and nonlinear. To derive a linear time-invariant model,
the next step is to apply averaging. The basic idea behind averaging is to represent a state variable, x(),
during the interval t − T ≤ τ ≤ t, using its Fourier series. In equation (21), Io is the equivalent output
current. R is the equivalent parallel resistance at the output due to parasitics and C0 is the output dc
link capacitance.
∞
X
x(τ ) = ⟨x⟩k (t)ejkωs τ (27)
k=−∞

where ωs = 2πfs , and complex number <x>k (t) is the kth coefficient in the Fourier series.
1 t 1 t 1 t
Z Z Z
−jkωs τ
⟨x⟩k (t) = x(τ )e dτ. = x(τ ) cos kωs τ dτ − j x(τ ) sin kωs τ dτ (28)
T t−T T t−T T t−T
This formulation is equivalent of <x>k (t) , where τ = t − T + s and s ∈ [0, T ).
(a) leakage inductance current with some important parameters. (b) input current (ii ). (c) output
current on the primary side of the converter (nio ), both used to calculate the proposed averaged model
Conventional state-space average can be applied when the ripples in state variables are small, which, in
fact, only considers dc terms (k=0) in (21) and (22). However, the transformer current in a DAB converter
is purely ac, meaning ripple is large and the dc term is zero. The generalized averaging technique uses
more terms in the Fourier series to represent more details in the model. In the case of DAB converters,
it is natural to include more terms (both k=0 and k=±1) in the Fourier series to represent the averages
of state variables.
the derivative of the kth coefficient for variable x is
 
d d
⟨x⟩k (t) = x (t) − jkωs ⟨x⟩k (t) (29)
dt dt k
 
d
where dt x represents the average of the differential of a state variable in (20) and (21). The kth
k
coefficient of the product of two variables x and y is
∞
X
⟨xy⟩k = ⟨x⟩k−i ⟨y⟩i . (30)
i=−∞

Consider that the 1st coefficient and the 1st coefficient in Fourier series are complex conjugates, the
product of zeroth coefficient terms is
⟨xy⟩0 = ⟨x⟩0 ⟨y⟩0 + 2 (⟨x1R ⟩⟨y1R ⟩ + ⟨x1I ⟩⟨y1I ⟩) (31)
and the 1st coefficient terms are
⟨xy⟩1R = ⟨x⟩0 ⟨y⟩1R + ⟨x⟩1R ⟨y⟩0 (32)
⟨xy⟩1I = ⟨x⟩0 ⟨y⟩1I + ⟨x⟩1I ⟨y⟩0 (33)
where the subscripts “R” and “ I” mean the real and the imaginary parts of a complex number, respec-
tively. Applying (24) and (26)– (28) to the state-space equations in (20) and (21), the zeroth and 1st
coefficients of state variables Il and vo are
d⟨il ⟩0 1 2
= +
dt L⟨s1 ⟩0 ⟨vi ⟩0 L⟨s1 ⟩1R ⟨vi ⟩1R
2
+ (34)
L⟨s1 ⟩1I ⟨vi ⟩1I − L1 ⟨s2 ⟩0 ⟨vo ⟩0
2 2
+ ⟨s2 ⟩1R ⟨vo ⟩1R + ⟨s2 ⟩1I ⟨vo ⟩1I
L L

8
 d⟨il ⟩1R
= ωs ⟨il ⟩1I
dt
1
+ (⟨s1 ⟩0 ⟨vi ⟩1R + ⟨s1 ⟩1R ⟨vi ⟩0 ) (35)
L
1
− (⟨s2 ⟩0 ⟨vo ⟩1R + ⟨s2 ⟩1R ⟨vo ⟩0 )
L
d⟨il ⟩1I
= −ωs ⟨il ⟩1R
dt
1
+ (⟨s1 ⟩0 ⟨vi ⟩1I + ⟨s1 ⟩1I ⟨vi ⟩0 ) (36)
L
1
− (⟨s2 ⟩0 ⟨vo ⟩1I + ⟨s2 ⟩1I ⟨vo ⟩0 )
L
and
d⟨vo ⟩0 1 1
= − ⟨iN ⟩0 − ⟨vo ⟩0
dt Co RCo
1 2
+ ⟨s2 ⟩0 ⟨il ⟩0 + ⟨s2 ⟩1R ⟨il ⟩1R (37)
Co Co
2
+ ⟨s2 ⟩1I ⟨il ⟩1I
Co
d⟨vo ⟩1R 1 1
= − ⟨iN ⟩1R − ⟨vo ⟩1R + ωs ⟨vo ⟩1I
dt Co RCo
(38)
1 1
+ ⟨s2 ⟩0 ⟨il ⟩1R + ⟨s2 ⟩1R ⟨il ⟩0
Co Co
d⟨vo ⟩1I 1 1
= − ⟨iN ⟩1I − ⟨vo ⟩1I − ωs ⟨vo ⟩1R
dt Co RCo
(39)
1 1
+ ⟨s2 ⟩0 ⟨il ⟩1I + ⟨s2 ⟩1I ⟨il ⟩0 .
Co Co
It is assumed that the dynamics of the input voltage and the load is much slower that those of the DAB
converter. Therefore, <vi >0 = Vi , <vi >1R = <vi >1I = 0, <iN >0 = iN , and <iN >1R = <iN >1I = 0.
Next, it is necessary to find the coefficients of switching functions s1 (t) and s2 (t). Because of the fixed
50% duty ratio, the zeroth coefficients of s1 (t) and s2 (t) are both zero as the dc component of perfectly
symmetrical square wave is zero.
⟨s1 ⟩0 = ⟨s2 ⟩0 = 0. (40)
The 1st coefficients of s1 (t) and s2 (t) are

⟨s1 ⟩1R = 0
2 (41)
⟨s1 ⟩1I = −
π
and
2 sin dπ
⟨s2 ⟩1R = −
π (42)
2 cos dπ
⟨s2 ⟩1I = − .
π
Finally, substitute (35)–(37) into (29)–(34), and organize differential equations into matrix represen-
tation in (38) , –¿ where X=[(vo )0 (il )1R (il )1I (il )0 (vo )1R (vo )1I ]T . In (38), (il )0 , (vo )1R , and (vo )1I are
all zero if their initial values are zero. The dynamics of il0 , (vo )1R , and vo1I are decoupled from the rest

9
of system.
 
1 4 sin (dπ) 4 cos (dπ)
 − RC −
πCo
−
πCo
0 0 0 
o
1
 
 2 sin (dπ)
 
0 ωs 0 0 0

 0 −
Co 
 
 πL 
 0 0  
 2 cos  
(dπ) 
d
 −ωs 0 0 0 0   2 
Vi
πL
 
X=

4 sin (dπ) 4 cos (dπ)  −
X +  0  .
dt  πL I

0 0 0 0  N
   0 0 
πL πL
  

 2 sin (dπ) 1   0 0 
 0 0 0 − − ωs 
πCo RCo 0 0
 
 
 2 cos (dπ) 1 
0 0 0 − −ωs −
πCo RCo
(43)
Therefore, the state-space representation in (38) can be equivalently reduced to
(a) leakage inductance current with some important parameters. (b) input current (ii ). (c) output
current on the primary side of the converter (nio ), both used to calculate the proposed averaged model

  
− RC 1
− 4 sin(dπ) − 4 cos(dπ) − C1o
    
vo0 πCo πCo vo0 0  
d  o
vi
it1R  = 2 sin(dπ) ×  it1R  +  0 0 (44)
 
πL 0 ωs 
iN
dt

2
it1I 2 cos(dπ)
−ωs 0 it1I − πL 0
πL

The state equation in (39) indicates that the dynamics of DAB dc–dc converters can be represented using
the zeroth coefficient of output voltage and the 1st coefficients of transformer current as state variables.
If more terms in the Fourier series were used, then the model would be more accurate. However, the
resulting model would be too complex to provide insightful information for controller design.

4 Steady State Analysis of Dab with SPS

In single phase shift modulation, the steady state operation of DAB can be segregated into four intervals
based on the operation of switches [15]. Thus, the equation of voltage across and current through the
inductor will also vary in the four intervals.
The voltage across the legs of the primary and secondary side bridges will be

Vp = Vd ; Vs = Vo (1)

Then, the secondary voltage referred to the primary side will be:
VO
VS′ = − = −VO′ (2)
n
Now, the voltage across the inductor and thus the current through the inductor will be given by:

VLK = VP − VS′ = Vd + VO′ (3)

The switches S1 and S4 of the primary bridge will continue to conduct, and gate pulses are given to
switches Q1 and Q4 of the secondary bridge in the second interval. The voltages VP,VS and VLK for
this interval will be given by the following equations:

VP = Vd ; VS′ = VO′ (4)

VLK = Vd − VO′ (5)

In the third interval, switches S1 and S4 of the primary bridge are turned off while S2 and S3 are switched
on and Q1 and Q4 on the secondary bridge will continue to conduct. The voltages VP,VS, and VLK for
this interval are:

VP = −Vd ; VS′ = VO′ (6)

′
VLK = −(Vd + VO ) (7)

10
During the fourth interval, switches S2 and S3 are conducting on the primary bridge, and switches Q2
and Q3 start conducting now on the secondary side. The equation for the voltages are:

VP = −Vd ; VS′ = −VO′ (8)

VLK = −Vd + VO′ (9)

Assuming VP greater than VS, the current through the inductor will rise or fall depending on the voltage
across the inductor. Corresponding voltages VP,VS,VLK and inductor current (ILK) are shown in Fig.
3. Further, the input and output current can be written in terms of the inductor current for the four
intervals are given below.
diLK
L = VLK = VP − VS′ (10)
dt
Assuming the change in inductor current to be linear, in the first interval:
I1 + I2 Vd + VO′
= (11)
DT L
Or
Vd + VO′
I1 + I2 = DT (12)
L
Input and output currents are:
ii = iLK ; io = −iLK (13)
where, T=Ts2,ii= input current, io= output current. In the second interval:
I1 − I2 Vd − VO′
= (14)
(1 − D)T L
Or
Vd − VO′
I1 − I2 = (1 − D)T (15)
L
ii = iLK = io (16)

The waveform for single phase shift is shown in :

Figure 5: (a) Primary voltage(b) secondary voltage(c) inductor current for forward power flow.[9]

Similarly the waveforms can be seen in for reverse mode :

11
 Figure 6: (a) Primary voltage(b) secondary voltage(c) inductor current for reverse power flow.[9]

4.1 Extended Phase shift(eps)

For overcoming the deficiencies of SPS control strategy, EPS control is used. In EPS control the switch
pairs in one bridge is provided inner phase shift while another switch pairs are switched same as in
SPS.The waveforms for EPS technique in forward and reverse conduction mode are shown in below :

Figure 7: (a) Primary voltage(b) secondary voltage(c) inductor current for forward power flow.[9]

12
 Figure 8: (a) Primary voltage(b) secondary voltage(c) inductor current for for reverse power flow.[9]

References
[1] B. Zhao, Q. Song, W. Liu, and Y. Sun, “Overview of dual-active-bridge isolated bidirectional dc–dc
converter for high-frequency-link power-conversion system,” IEEE Transactions on Power Electronics,
vol. 29, no. 8, pp. 4091–4106, 2014.
[2] A. Kadavelugu, S. Baek, S. Dutta, S. Bhattacharya, M. Das, A. Agarwal, and J. Scofield, “High-
frequency design considerations of dual active bridge 1200 v sic mosfet dc-dc converter,” in 2011
Twenty-Sixth Annual IEEE Applied Power Electronics Conference and Exposition (APEC), 2011, pp.
314–320.
[3] Y. A. Harrye, K. Ahmed, G. Adam, and A. Aboushady, “Comprehensive steady state analysis of
bidirectional dual active bridge dc/dc converter using triple phase shift control,” in 2014 IEEE 23rd
International Symposium on Industrial Electronics (ISIE), 2014, pp. 437–442.
[4] R. De Doncker, D. Divan, and M. Kheraluwala, “A three-phase soft-switched high power density dc/dc
converter for high power applications,” in Conference Record of the 1988 IEEE Industry Applications
Society Annual Meeting, 1988, pp. 796–805 vol.1.
[5] ——, “A three-phase soft-switched high-power-density dc/dc converter for high-power applications,”
IEEE Transactions on Industry Applications, vol. 27, no. 1, pp. 63–73, 1991.

13
[6] M. Kheraluwala, R. Gasgoigne, D. Divan, and E. Bauman, “Performance characterization of a high
power dual active bridge dc/dc converter,” in Conference Record of the 1990 IEEE Industry Applica-
tions Society Annual Meeting, 1990, pp. 1267–1273 vol.2.

[7] A. Rodrı́guez, A. Vázquez, D. G. Lamar, M. M. Hernando, and J. Sebastián, “Different purpose

design strategies and techniques to improve the performance of a dual active bridge with phase-shift
control,” IEEE Transactions on Power Electronics, vol. 30, no. 2, pp. 790–804, 2015.
[8] F. Corti, V. Bertolini, A. Reatti, E. Cardelli, and M. Giallongo, “Comparison of control strategies for
dual active bridge converter,” in 2022 IEEE 21st Mediterranean Electrotechnical Conference (MELE-
CON), 2022, pp. 902–907.
[9] A. Kumar, A. H. Bhat, and P. Agarwal, “Comparative analysis of dual active bridge isolated dc to
dc converter with single phase shift and extended phase shift control techniques,” in 2017 6th Inter-
national Conference on Computer Applications In Electrical Engineering-Recent Advances (CERA),
2017, pp. 397–402.

14