See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/380577871

Power Factor Correction Using DCM Technique and Soft Feedback Emphasizing
on Boost Converter's Inductance Calculation

Conference Paper · April 2024

DOI: 10.1109/EPDC62178.2024.10571722

CITATIONS READS

0 311

1 author:

Salar Mosharkesh Barenji

K.N.Toosi University of Technology
5 PUBLICATIONS 3 CITATIONS

SEE PROFILE

All content following this page was uploaded by Salar Mosharkesh Barenji on 15 May 2024.

The user has requested enhancement of the downloaded file.

 28th Electrical Power Distribution Conference, April 23-25, 2024, Zanjan, Iran

Power Factor Correction Using DCM Technique and

Soft Feedback Emphasizing on Boost Converter's
Inductance Calculation
1st Salar Mosharkesh Barenji
K. N. Toosi University of Technology
Tabriz, Iran
[email protected]

Abstract - Boost converters, due to their unique structure, Converter, are widely used. The Bridgeless Rectifier is more
provide the possibility of adjusting the power factor correction efficient at higher power levels, while the Boost Converter is
(PFC). However, improper calculation of hardware parameters such more popular at lower power levels [8]. Figure 2 illustrates the
as inductance and switching frequency, considering the load boost converter structure for power factor correction [9].
connected to the converter's output, can jeopardize the converter's
performance. In other words, the converter's operation may deviate
from the Discontinuous Conduction Mode (DCM). This paper
focuses on the calculation of inductance to achieve optimal
performance, specifically targeting DCM operation for sinusoidal
unidirectional inputs. In other words, a one-sided smoothing
capacitor is not used in the output, emphasizing the PFC aspect.

Keywords: Power factor correction, Boost converter control, Figure 2 - Boost Converter Structure for Power Factor Correction
SMPS, PFC, DCM, Boost Converter Inductance.
Control techniques used in Boost Converter structures
I. INTRODUCTION generally fall under three categories: Continuous Conduction
Nowadays, the use of boost converters has become Mode (CCM), Discontinuous Conduction Mode (DCM), and
widespread due to their unique structure, and they can be found Quasi-Continuous Mode (QCM) [23]. In CCM, which is the
in various equipment. This type of converter is an additive DC continuous mode, the inductor current remains above zero
to DC converter [1]. In general, a boost converter has a structure throughout the switching cycle. The inductor current ripple is
like Figure 1 [2]. typically around 20 to 40 percent of the average input current.
In DCM, the inductor current always returns to zero, and
controlling the duration of its zero state allows for effective
conduction control. Hence, this is referred to as the
discontinuous technique. Contrary to CCM, in DCM, the
inductor is discharged to zero and recharged when the transistor
switch is connected. If a significant increase in load power
occurs in DCM, the converter will operate in CCM at an angle
Figure 1 – Boost Converter Structure of 90 degrees (ϲ=90^°). Thus, the average consumer current will
determine the zeroing of the inductor current ripple. QCM, the
On the other hand, boost converters are employed for active quasi-continuous mode, is a hybrid of CCM and DCM. In this
power factor correction. Generally, for power factor correction, mode, the inductor current reaches zero during some parts of the
either active or passive devices can be used [3]. In the passive cycle due to the transistor switch's on-off action, while in other
method, shunt capacitors are used to neutralize the inductive parts, it remains charged.
effect [4]. However, capacitor banks may cause system
resonance and network harmonics amplification [5]. Harmonic Various techniques can be employed to improve the power
power factors are often created by devices with current rectifiers factor using a boost converter. However, for achieving 100%
in their structure [6]. On the other hand, using capacitor banks power factor correction, the "100% Power Factor Correction
for power factor correction has little effect on harmonic loads Using DCM Control Technique with Software Feedback" can
[7]. In the active method, power factor correction is achieved by be utilized. In this technique, the converter will achieve 100%
converting alternating current to direct current and then power factor correction as long as it operates in DCM and
converting direct current back to alternating current. However, suitable hardware is employed [11]. This article focuses on
controlled rectification must be performed initially to prevent calculating the inductance value used in the boost converter to
the generation of reducing power factor harmonics. In the active achieve DCM performance for non-smoothed DC inputs.
method, two structures, Bridgeless Rectifier and Boost
 II. CALCULATION OF INDUCTANCE
To calculate the inductance, it is advisable to first examine
the converter's performance for smoothed DC inputs and then
proceed with calculations for non-smoothed unidirectional
inputs.
The boost converter operates in two states:
1- When the transistor switch is off
2- When the transistor switch is on Figure 5 - Using Bypass Diode for Initial Charging of Capacitor

In the state where the transistor is on, the structure of the To investigate the performance of the converter, we need to
converter takes the form shown in Figure 3 [12]. derive the equations governing the converter's behavior during
the on and off times of the transistor. Equations 3 to 21 represent
the relationships governing the converter's performance [15].
III. EXAMINING THE TRANSISTOR SWITCH-ON TIME
In the state where the switch is turned on, the inductor is
charged by the power source.
diL
VL = Vi = L (3)
Figure 3 - Boost Converter in Transistor On State dt
diL Vi
In this state, assuming all converter components are ideal, = (4)
dt L
when the switch is turned on, the voltage across the inductor will
be equal to Vin , representing the inductor charging state. In this The rate of change of inductor current is constant, so the
current during the on-time of the transistor switch increases
condition, with the transistor switch on and the diode switch off,
the load is powered through the charge of the capacitor. linearly. Therefore:
∆iL ∆iL Vi
In this state, the inductor charging current can be calculated = = (5)
∆t DT L
using Equation 1. Vi DT
Vi .dt
(∆iL )closed = (6)
L
I = diL = (1)
L

Thus, the energy stored in the inductor can be expressed by IV. INVESTIGATION OF TRANSISTOR SWITCH-OFF TIME
Equation 2.]13[ During the transistor switch-off time, the inductor current is
1 2 directed to the output through the diode. However, since the
W= LI (2) output voltage is constant, the inductor voltage will be obtained
2

In the second state, with the transistor switch turned off, the accordingly.
circuit will take the form shown in Figure 4.]12[ diL
VL = Vi − VO = L (7)
DT
(Vi − VO ).dt
VL = (8)
L . diL

During the discharge of the inductor, or (1 − D), the rate of

change of the inductor current remains constant over time.
Therefore:
Figure 4 - Boost Converter in Transistor Off State ∆iL L ∆i Vi − VO
= (1−D)T = (9)
∆t L
In this state, representing the inductor discharge state, the
(Vi − VO )(1−D)T
induced voltage across the inductor will be equal to Vin − Vo . (∆iL )open = (10)
L
On the other hand, the capacitor charges in this state to supply
the load during the next transistor switch on and diode switch In the steady state, the inductor current will be the same for
off. The presence of the capacitor in this section is essential for both the transistor switch-on and switch-off states, and can be
reducing voltage ripple and storing energy for the transistor expressed as follows:
switch-on and diode switch-off phases.
(∆iL )closed + (∆iL )open = 0 (11)
Usually, a bypass diode or pre-charge diode, as shown in Vi DT (Vi − VO )(1−D)T
Figure 5, is included in the circuit to pre-charge the capacitor in + =0 (12)
L L
the initial moments before the converter starts working. This
pre-charge helps prevent the high initial current flow through the Vi (D + 1 − D) − VO (1 − D) = 0 (13)
inductor and diode in the converter structure [14].
 Based on the obtained equations, the output voltage will be Considering that during the transistor switch-on time, and
given by equation 14: assuming ideal transistor conditions, the input power supply
Vi voltage will be equal to the inductor voltage, we will have:
VO = (14)
1−D I𝑚
𝑉𝑖 = 𝐿 (22)
𝐷𝑇
Additionally, the average inductor voltage must be zero for
periodic operation. By rearranging the variables, we will arrive at Equation 23.
𝑉𝑖 𝐷𝑇
VL = Vi D + (Vi − VO )(1 − D) = 0 (15) I𝑚 = (23)
𝐿
Assuming zero converter losses, we will have:
Furthermore, during the transistor switch-off period, we will
V2
O have:
PO = = VO IO (16)
R
𝑉𝐿 = 𝑉𝑜 − 𝑉𝑖 (24)
Vi Ii = Vi IL (17)
Therefore, we can rewrite equation 24 as equation 25, where
2
V in this equation, FT represents the boost diode's on-time
V2 [ i⁄(1−D)] Vi 2
O
Vi IL = = = (18) duration.
R R R (1−D)2
Vi VO IO I𝑚
IL = = (19) 𝑉𝑜 − 𝑉𝑖 = 𝐿 (25)
R (1−D)2 Vi 𝐹𝑇

The maximum and minimum inductor currents are To obtain F, we rearrange the variables in equation 25.
calculated from the relationships of the average inductor current I𝑚
𝐹= 𝐿 (26)
and changes in inductor current. (𝑉𝑜 −𝑉𝑖 )𝑇

∆iL Vi Vi DT After simplifying the equation 26 for F, we will have:

IL max = IL + = ( + ) (20)
2 R(1−D)2 2L 𝑉𝑖 𝐷
∆iL Vi Vi DT
𝐹= (27)
𝑉𝑜 −𝑉𝑖
IL min = IL − = ( − ) (21)
2 R(1−D)2 2L
On the other hand, for the converter to operate in
Discontinuous Conduction Mode (DCM), it is necessary for the
V. CALCULATION OF INDUCTANCE inductor discharge time, or in other words, the boost diode ON
To achieve DCM performance, it is necessary for the total time, to be shorter than the transistor OFF time.
inductor charging time and discharging time to be less than the
𝐹 < 1−𝐷 (28)
switching signal cycle. Therefore, the inductance of the coil
must be chosen in a way that, as shown in Figure 6 and Therefore, we will have:
considering the parameters of input and output voltage range, 𝑉𝑖 𝐷
maximum output load power, switching frequency, and the < 1−𝐷 (29)
𝑉𝑜 −𝑉𝑖
pulse width of switching, the coil is completely discharged
during the transistor switch-off time. In Figure 6, the blue Equations 22 to 29 are instantaneous relationships, but to
waveform represents the switching signal, and the red waveform calculate the inductance of the coil, it is necessary to examine
illustrates the inductor current. the converter's performance in a steady-state condition. In the
steady state, the average resistance current will be equal to the
average diode current, thus:
I𝐷𝑎𝑣𝑔 = I𝑅𝑎𝑣𝑔 (30)
Figure 7 represents the waveform of the diode current, which
is equal to the inductor discharge current during the transistor
switch-off time.

Figure 6 – Inductor Current Waveform and Switching Signal

For defining the initial variables, we have:

T = Cycle period
Figure 7 – Diode Current Waveform
DT = Transistor switch-on time
FT = Diode switch-on time Considering the triangular shape of the inductor discharge
current waveform, to calculate the average diode current, which
I𝑚 = Maximum inductor current is equal to the inductor current, it is sufficient to multiply the
maximum current by 1⁄2. Additionally, the current calculation
should be performed for one second, so the calculated current value calculation equation, as expressed in Equation 34. It
for one switching cycle is multiplied by the diode conduction should be noted that, due to the presence of a capacitor at the
time and the switching frequency, denoted by the variable 𝑓𝑠𝑤 . converter's output, the variable nature of the input voltage
Moreover, due to the variable nature of the input voltage range amplitude, and considering that the average current calculation
of the converter, the average current value should be considered, is performed over one second, we cannot eliminate I𝑚 from the
resulting in the following equation: averaging process. Therefore, considering these points, the
1
derived equation is presented as Equation 35.
I𝐷𝑎𝑣𝑔 = 𝑓𝑠𝑤 𝑎𝑣𝑔 { I𝑚 𝐹𝑇} (31) 1 L
2
I𝐷𝑎𝑣𝑔 = 𝑎𝑣𝑔{ I𝑚 × I𝑚 } (35)
2 (𝑉𝑜 −𝑉𝑖 )𝑇
Considering the relationship between frequency and period,
we have: By substituting the relationship I𝑚 into Equation 36, we will
1 1 reach the following expression.
I𝐷𝑎𝑣𝑔 = 𝑎𝑣𝑔 { I𝑚 𝐹𝑇} (32)
𝑇 2 1 L 𝑉𝑖 𝐷𝑇 𝑉𝑖 𝐷𝑇
I𝐷𝑎𝑣𝑔 = 𝑎𝑣𝑔 { × } (36)
Considering that the value of T is constant, it should not be 2 (𝑉𝑜 −𝑉𝑖 )𝑇 𝐿 𝐿
used in calculating the average value. After simplifying the equation and excluding the constant
1 1 parameters from the part of calculating the average value, we
I𝐷𝑎𝑣𝑔 = 𝑇 𝑎𝑣𝑔 {I𝑚 𝐹} (33) will have:
𝑇 2

By substituting the equation F, we will have: I𝐷𝑎𝑣𝑔 =

1 L 𝐷𝑇 𝐷𝑇
𝑎𝑣𝑔{ 𝑉𝑖 × 𝑉𝑖 } (37)
2 (𝑉𝑜 −𝑉𝑖 )𝑇 𝐿2
1 I𝑚
I𝐷𝑎𝑣𝑔 = 𝑎𝑣𝑔 { I𝑚 𝐿 } (34)
2 (𝑉𝑜 −𝑉𝑖 )𝑇 Given that the input voltage waveform is a full-wave
In Equation 33 and subsequently in Equation 34, it should be rectified sinusoidal waveform, the average input voltage is
noted that F represents the duty cycle of the diode conduction calculated by multiplying the input voltage by 0.63 [16].
time. Our goal is to determine the critical value of inductance 𝑉𝑖𝑎𝑣𝑔 = 0.63 ∗ 𝑉𝑖 (38)
that maximizes the inductor discharge time, subject to the
condition that (1-D) remains less than or equal to 1, ensuring the Therefore, we will arrive at Equation 39:
converter operates in Discontinuous Conduction Mode (DCM). 1 𝐷2 𝑇 2
Therefore, we need to consider the maximum acceptable value I𝐷𝑎𝑣𝑔 = {(0.63 × 𝑉𝑖 ) × (0.63 × 𝑉𝑖 )} (39)
2 (𝑉𝑜 −𝑉𝑖 )𝑇 𝐿
for F, which is expressed as F=(1-D). For a better understanding,
refer to Figure 8, depicting the voltage waveform during the Equation 40 is a simplified form of Equation 39.
inductor charge and discharge. 1 𝐷2 𝑇 2
I𝐷𝑎𝑣𝑔 = {0.63 × 𝑉𝑖 }2 (40)
2 (𝑉𝑜 −𝑉𝑖 )𝑇 𝐿

Finally, after simplifying Equation 40, we will arrive at

Equation 41.
𝐷2 𝑇 (0.63× 𝑉𝑖 )2
I𝐷𝑎𝑣𝑔 = I𝑅𝑎𝑣𝑔 = (41)
2 (𝑉𝑜−𝑉𝑖 ) 𝐿

Given the Ohm's law for I𝑅𝑎𝑣𝑔 , we have:

𝑉𝑜 𝐷2 𝑇 (0.63× 𝑉𝑖 )2
≈ (42)
𝑅𝐿 2 (𝑉𝑜 −𝑉𝑖 ) 𝐿

Finally, after parameter substitution, we will arrive at

Figure 8 - Inductor Charge and Discharge Voltage Equation 43 to obtain the critical inductance value.
𝑅𝐿 𝐷2 𝑇 (0.63× 𝑉𝑖 )2
In Figure 8, where the waveform is blue and the voltage is 𝐿≈ (43)
2 𝑉𝑜 (𝑉𝑜 −𝑉𝑖 )
positive, the inductor is charging. In the portion where the
waveform is red, and the voltage is negative, the inductor is To achieve DCM operation, the inductance of the coil should
discharging with reversed polarity. According to Figure 8, be smaller than or equal to the calculated value. However,
which illustrates an example of inductor charge and discharge considering that a more significant reduction in inductance can
voltage, the maximum diode conduction time, denoted as F (the lead to an increase in the output voltage, it is preferable to choose
inductor discharge duration), occurs at the input voltage peak. the inductance value slightly less than the calculated inductance.
Since we cannot take the average time and need to consider the 𝑅𝐿 𝐷2 𝑇 (0.63× 𝑉𝑖 )2
peak voltage to calculate the critical inductance value, we 𝐿≤ (44)
2 𝑉𝑜 (𝑉𝑜 −𝑉𝑖 )
calculate the value of F for the input voltage peak using Equation
26, where 𝑉𝑖 is the input voltage peak amplitude. Additionally, Equation 44 is valid when the inductor undergoes charge and
due to the DC nature of the output voltage, the output voltage discharge for each complete switching cycle, representing the
amplitude remains constant. Therefore, (𝑉𝑜 − 𝑉𝑖 ), calculated for calculated inductance as the maximum usable inductance for
the maximum input amplitude, will have a constant value. Due achieving DCM performance. The derived formula's
to this constant value, it should be excluded from the average
performance was investigated through simulation, with nominal 𝑒 2𝐿𝑉𝑜 (𝑉𝑜 −𝑉𝑖 )
𝑅𝑘−1 ≈ (𝐷𝑘−1 )2 𝑇 ×(0.63× 𝑉𝑖 )2
(45)
values as presented in Table I:
A notable point in calculating the value of the load resistance
TABLE I. VALIDATION OF THE DERIVED FORMULA THROUGH
SIMULATION
connected to the output is feedback from the output voltage.
Therefore, Vo utilized is not the nominal voltage; rather, it is
Vin V𝑂 L C F Duty Cycle Load(Ω) feedback from the output voltage.
310 387.5 105uH 520uF 100KHz 10% 1000
Furthermore, by rearranging Equation 44 to Equation 46, the
310 387.5 52uH 520uF 100KHz 10% 500 new pulse width will be provided based on the identified load
310 387.5 21uH 520uF 100KHz 10% 200 value and the voltage SetPoint
2𝐿𝑉𝑆𝑃 (𝑉𝑆𝑃 −𝑉𝑖 )
𝐷𝑘 = √ 𝑒 2 (45)
𝑅 𝑘−1 𝑇 ×(0.63× 𝑉𝑖 )
The input current waveform of the converter for the first
experiment was as shown in Figure 9, and the second and third Therefore, according to Equation 46, to determine the
experiments also exhibited a similar behavior. This pattern appropriate pulse width, we need the desired output voltage,
illustrates the performance of the DCM converter. maximum input voltage, switching signal period, inductance of
the coil, and the load resistance connected to the output.
However, it should be noted that obtaining the value of the load
resistance connected to the output requires the use of feedback
from the output voltage and the pulse width value at the
measurement moment.
Figure 11 illustrates the diagram related to the controller.

Figure 9 - Input Current Waveform for the First Experiment

When the load power consumption exceeded the nominal Fig. 11. Controller Diagram
state, the converter's operation transitioned from DCM to QCM.
Figure 10 illustrates the input current waveform for the first Figure 11 illustrates the controller diagram. As seen in the
experiment, with the difference that the load impedance was diagram, using Equation 45 and feedback from the output
selected as 500 ohms instead of 1000 ohms, leading to the voltage and the pulse width at the time of measurement, the load
elimination of the DCM converter operation. resistance is estimated. Then, using Equation 46, with the
estimated load resistance value and the desired output voltage or
Vsp, a new pulse width is determined to bring the output voltage
amplitude to the desired level.
VI. DESCRIPTION OF THE CONTROL ALGORITHM
In the initial stage, which is performed only once at startup,
the controller generates an initial pulse width based on nominal
values, in accordance with Equation 14, and applies it to the
converter.
Then, in the second phase, following Equation 45, the
controller estimates the resistance value of the load connected to
the output. This is done by using feedback from the output
Figure 10 - Input Current Waveform for a 500-ohm Load voltage and the pulse width at the time of measurement to
estimate the load resistance.
During the operation, the load value may change over time,
and with a constant frequency and pulse width, the output In the third phase, based on the estimated resistance and the
voltage will also change accordingly. Therefore, one of the best command voltage, the controller adjusts the pulse width.
solutions to maintain a constant output voltage alongside Subsequently, the controller, skipping the first phase,
achieving DCM operation is to utilize software feedback[17]. cyclically performs the second and third phases. Thus, with the
For this purpose, firstly, the load value needs to be identified use of the mentioned control algorithm, in addition to precisely
using software feedback from the output voltage. Then, adjusting the output voltage according to the load resistance, due
adjustments to the pulse width can be made. By substituting to the converter's operation in Discontinuous Conduction Mode
Equation 44 with Equation 45, the possibility of identifying the (DCM), power factor correction will also be observed.
load resistance can be achieved.
 However, based on the waveforms obtained from the
simulations, it was noted that the effect of changes in the value
of D becomes apparent after approximately 50 milliseconds.
Therefore, this factor should be considered in hardware design
for better performance. In other words, due to the inherent delay
of the converter, the effect of changes in pulse width becomes
apparent in the output voltage amplitude after about 50
milliseconds. Therefore, the estimation of load resistance and
the application of changes should be performed at intervals of at Fig. 14. Waveform of the Signal Applied to the IGBT Gate
least 50 milliseconds to allow the output voltage to stabilize and
The output voltage of the converter was measured at 389
to ensure more accurate estimation of load resistance.
volts, which is almost equal to the setpoint voltage. The
Figure 12 shows the schematic of the boost converter calculated pulse width and the measured pulse width are also
section, and Figure 13 illustrates the constructed hardware. nearly the same, and any existing difference could be due to
component tolerances and measurement errors.
The practical test results indicate the correct functioning of
the control algorithm in stabilizing the voltage amplitude.
However, the converter's goal is to correct the power factor. In
the subsequent practical test, power factor correction was
examined using a power meter. The test results are shown in
Figure 15, indicating a 100 percent correction of the power
factor. It is noteworthy that the power meter used has an
accuracy of 0.001 for calculating the power factor, clearly
demonstrating the excellent performance of the converter in
correcting the power factor. This correction of the power factor
and stabilization of the output voltage amplitude are achieved
Fig. 12. Schematic of the Constructed Circuit with only one feedback from the output voltage. The practical
test results indicate a 100 percent correction of the power factor.

Fig. 13. Constructed Circuit for Hardware Testing

The nominal values considered for hardware design were Vo

= 387.5V, Vi = 310V (220Vac), Po = 300W, and Fs = 160KHz.
These parameters are loaded into the microcontroller during Fig. 15. Measured Power Factor for an Input Voltage of 220 Volts
programming to enable the controller to utilize them for system
control throughout the control process. Accordingly, based on VIII. CONCLUSION
these nominal values, the minimum load resistance is Considering the attainment of DCM operation in boost
determined to be 500 ohms. Additionally, the nominal pulse converters enables the possibility of correcting power factor to
width is calculated to be 20%, and the inductance value is 100%. Therefore, it is necessary to precisely calculate the value
calculated to be 79uH. of the inductance to maintain DCM performance so that the
converter can maintain its DCM operation for nominal values.
VII. HARDWARE INVESTIGATION RESULTS Consequently, in this paper, the converter performance was
Figure 14 shows the steady-state PWM output of the IGBT initially examined for smoothed DC inputs, followed by
driver. Considering the 30-watt load connected to the converter calculations for unsmoothed DC inputs. The results of the
and the full-wave rectified 220-volt input voltage with a peak of investigations indicated the DCM operation of the converter for
310 volts, the PWM signal in steady state has an approximate nominal converter power. As long as the load power does not
duty cycle of 6 percent. If we calculate the pulse width according exceed the nominal value, the converter's operation remains in
to Equation 5 related to pulse width, we have 𝐷𝑘 = 6.3٪. DCM mode. Hardware tests also confirmed the 100% power
factor correction with DCM operation. The accuracy of the
Figure 14 shows the signal applied to the IGBT gate by the derived relationship for calculating the inductance to achieve
IGBT driver, where the pulse width of the signal applied to the DCM operation was demonstrated.
gate in the steady state of the converter is approximately 6
percent.
 REFERENCES [9] C. W. Clark, "Digital control techniques for power quality improvements
in power factor correction applications," M.Sc. Thesis, University of
[1] A. Emadi, et al., "Integrated power electronic converters and digital British Columbia, 2012.
control," CRC Press, 2017.
[10] E. R. Arenaza and M.-M. Boghiu, "Design and Performance Analysis of
[2] M. H. Rashid, "Power electronics handbook," Butterworth-Heinemann, a High-Power-Density PFC Converter," M.Sc. Thesis, Aalborg
2017. University, 2020.
[3] P. R. Mohanty, A. K. Panda, and D. Das, "An active PFC boost converter [11] S. Mosharkesh Barenji, "Correcting Power Factor by Controlling Boost
topology for power factor correction," 2015 Annual IEEE India Converter Connected to Rectifier," M.Sc. thesis, Dept. Control Eng.,
Conference (INDICON), IEEE, 2015. Faculty of Mechanical Engineering, K. N. Toosi University of
[4] N. Ramzan, et al., "Reactive power compensation through Technology, Tehran, Iran, Spring 2022.
synchronization of networked VFDs," 2016 Clemson University Power [12] Power Integrations, "Active Power Factor Correction – Basics," Power
Systems Conference (PSC), IEEE, 2016. Integrations Inc. AN-53, October 2017.
[5] W. M. Grady and R. J. Gilleskie, "Harmonics and how they relate to [13] S. Maniktala, "Switching Power Supplies A-Z," Elsevier, 2012.
power factor," Proceedings of the EPIR Power Quality Issues &
Opportunities Conference, November 1993. [14] S. Abdel-Rahman, "CCM PFC boost converter design," Infineon, Tech.
Rep. DN 2013–01, 2013.
[6] M. W. Bin Mahmud, A. H. M. Zahirul Alam, and D. A. Rahman,
"Improvement of active power factor correction circuit for switch mode [15] D. W. Hart, "Power electronics," The McGraw-Hill Companies, 2011.
power supply using fly back and boost topology," 2018 7th International [16] ON Semiconductor, "Rectifier Applications Handbook-Reference
Conference on Computer and Communication Engineering (ICCCE), Manual and Design Guide," Denver: ON Semiconductor, Nov 2001.
IEEE, 2018. [17] Salar Mosharkesh Barenji, "100% Power Factor Correction Using DCM
[7] Y. Notohara, et al., "Controlling power factor correction converter for Control Technique with Software Feedback," 9th International
single-phase AC power source without line voltage sensor," The 2010 Conference on Technology and Energy Management (ICTEM),
International Power Electronics Conference-ECCE ASIA-, IEEE, 2010. Behshahr, Mazandaran, Iran, 14-15 February 2024.
[8] L. Liao, T. Wang, and H. Liu, "Control and Modeling of Boost PFC for
the 47~800Hz universal grids," IECON 2020 The 46th Annual
Conference of the IEEE Industrial Electronics Society, 2020.

View publication stats