Journal of Electrical Engineering, Electronics, Control and Computer Science –

JEEECCS, Volume 7, Issue 26, pages 41-52, 2021

Harmonic pollution control of electrical

networks: Comparative study of
controls, by Pulse Width Modulation, Duty
Cycle Modulation and Symmetrical Linear
Duty Cycle Modulation of the three-phase
shunt active Parallel filters
Fabrice Biloa Assolo1, Charles Hubert Kom1, 2, Arnaud Nanfak2, Gildas Martial Ngaleu2
1
Research Laboratory of Computer Science Engineering and Automation
Higher Normal School of Technical Education,
University of Douala, Po Box 2701 Douala, Cameroon
2
Laboratory of Energy, Materials, Modelling and Methods,
Higher National Polytechnic School of Douala
University of Douala, Po Box 2701 Douala, Cameroon
1
[email protected], [email protected], [email protected], [email protected]

Abstract – The increasing use on electrical networks of the current of the source loses its sinusoidal shape and
consumer electronic devices (televisions, computers) and we obtain also a degradation of the power factor.
industrial electronic devices (variable speed drives, high Through therefore, distributors of electrical energy
efficiency lighting) pollutes these networks by harmonic see themselves therefore obliged to impose standards
currents and voltages. The three-phase active shunt
filter is a modern solution for the remediation of three-
and protect themselves against these disturbances.
phase electrical networks. Better pollution control One of the solutions used is passive filtering: it
requires proper sizing of the active shunt filter. Indeed, consists in trapping the currents harmonics in LC
the capacity and adaptability of the filter reside in the circuits, tuned to the ranks of harmonics to be filtered.
quality of the control of the switches of the inverter Rows 5 and 7 are the most commonly filtered.
constituting it. In this article we are interested in However, this solution is moderately effective [1] -
controlling inverters. A comparative study of Pulse [2]. Common use of devices power electronics, made
Width Modulator (PWM) controls, Duty Cycle it possible to design self-adapting harmonic
Modulation inverter (DCM) control and Symmetrical elimination devices called active compensators
Linear Duty Cycle Modulation (SLDCM) control in a
three-phase network. The study shows that the DCM
harmonics, or active filters. These filters the main
inverter gives the best results with regard to the objective of the assets is to compensate for harmonic
reduction of harmonic pollution, and therefore of the currents injected into the network [3] [4] [5].
quality of the network. Our work then turned to the study of the control-
command block of the active filter. There are several
types of control of three-phase shunt filters for this
Keywords: shunt active parallel filter, Harmonic pollution, purpose which have been presented in the literature.
PWM control, DCM control, SLDCM control There are several techniques to control the
compensating current injection such as a Pulse Width
I. INTRODUCTION Modulation (PWM) Duty Cycle Modulation (DCM)
The development of power electronics and the and new duty-cycle modulator structures namely
increase in the powers involved as well as the flexible Symmetrical Linear DCM (SLDCM). PWM
use of semiconductors, an encouraged electrical technique is a modulation technique in which the
engineer to undertake important associations of static frequency can be either constant or variable[6]. In
converters power to electrical machines. These constant frequency PWM method, the frequency is
devices are usually loads that are not linear, which constant and the ON time changes according to the
absorb a non-sinusoidal current and behave like modulating signal. It can be produced by comparing a
harmonic generators. Of moreover, they sometimes reference signal with a carrier signal[6]. In variable
consume reactive energy. Therefore, the waveform of frequency PWM method, the frequency is variable and
ON time or OFF time is constant. In variable
 42 Fabrice Biloa Assolo, Charles Hubert Kom, Arnaud Nanfak, Gildas Martial Ngaleu

frequency PWM method, the frequency is variable and possible extension to other CC-CC converters [14],
ON time or OFF time is constant. There are several even to the important class of DC-AC inverters [15].
different PWM techniques, differing in their methods The article[16] makes the comparative study of IGBT
of implementation. The basics PWM topologies controls between DCM and hysteresis in a three-phase
including single PWM produced by comparing network, it emerges that the DCM control offers the
rectangular reference wave with triangular carrier best performance in terms of harmonic rate reduction.
wave, the multiple PWM and the The new DCM modulator structures are proposed in
Sinusoidal PWM (SPWM) using sinusoidal reference [17]. These are Symmetric Linear DCM (SLDCM)
wave and triangular carrier wave [7]. and General Linear DCM (GLDCM). These two
structures consist of integrator circuits, a hysteresis
The duty cycle modulation (DCM) is a modulation in comparator and advantages of PWM and DCM
which an input signal x is transformed into a train of circuits.
switching wave where duty cycle and period of the In this paper, our objective is to make a comparative
modulated signal vary simultaneously according to study followed by simulation in the software
control signal [8] [7]. The DCM technique has been MATLAB / SIMULINK, of the controls by PWM,
used in several application areas. DCM of the inverter and SLDCM of an SAPF (Shunt
From instrumentation to process control, including Active parallel Filter) for harmonic pollution control
Signal processing. Thus, a scientific article on the application.
application of the principle of modulation of duty The article is organized as follows. In section II, we
cycle in instrumentation has been published [9]. In present the methodology used. In section III, we
this article, a class of new DCM is investigated. It present the simulation results. And we end with a
used the analog signal processing techniques with a conclusion where we highlight the value of this best-
simple operational amplifier circuit. The first papers to performing ordering strategy.
use DCM as an analog-to-digital converter are[10]
[11]. These two papers respectively presented the II. METHOD
feasibility of analog-to-digital conversion by DCM via
virtual simulations and the multi-channel version of A. Principle of parallel active filter
A/D conversion based on the use of duty cycle This filter is connected in parallel on the network. The
modulator cells in parallel, each used as a circuit. principle of the parallel active filter consists in
In [12], buck converters use the DCM control generating currents in opposition of phase to the
technique. In this article, analytical developments, harmonic currents existing on the network and created
numerical analysis, virtual simulations and by the nonlinear loads. In this way, the current
experimentation have demonstrated that DCM control supplied by the energy source remains sinusoidal. The
system provides better characteristics compared to parallel active filter consists of two blocks: the power
standard PWM techniques, so DCM control can be part and the control-command part Figure.1.
considered as a new perspective of control in power
electronics and control engineering [13], with a

Figure 1: Principle of the parallel active filter [18]

 Harmonic pollution control of electrical networks: Comparative study of… 43

Figure 2: Main blocks in the electrical circuit for a shunt active power filter

Concerning Figure 2, we can identify the main

block of our study: the current control because it is
in this block that we will apply in turn the
 p   v v  i 
 q  =  −v
command PWM, DCM inverter and SLDCM to
v  i 
(3)
switch the IGBTs of the inverter. In the second    
current reference block, we will apply the
Instantaneous active and reactive powers can be
instantaneous powers algorithm to identify the
written as the sum of a DC component and an AC
reference currents.
component:
B. Theory of instantaneous powers

 p  p + p 
q  = 
The instantaneous active and reactive power
 (4)
method was developed by Akagi [19] and uses the
Concordia transformation of phase-to-neutral
   q + q 
voltages Vs and line currents Is, in order to
calculate instantaneous active and reactive powers. From equation (4), we can deduce the expressions
The principle of this classical method is now for the components of the charging current along
briefly described. Let respectively be the phase-to- the αβ axes:
neutral voltages of a three-phase network without
zero sequence (connected to a polluting load) and −1
i    V  V    p 
the three load currents, denoted vsa, vsb, vsc and ic1,  i  =  −V  V    −q 
ic2, ic3. The Concordia transformation makes it       (5)
possible to reduce this balanced three-phase system
i   1 V −V   p 
to a two-phase system whose axes are in  = 2 2   
quadrature:  i  V + V V V   −q 

 1 1  V
 sa  By replacing (4) in (5), these currents are expressed
1 − −
Vs  2 2 2   along the αβ axes by:
V  =   Vsb (1)
 s  3 3 3  i   1 V −V   p  1 V  −V    p 
0 2 − 2  Vsc   i  = V 2 + V 2 V V    + V 2 + V 2 V  V    
(6)
       −q    
    −q 

 1 1  I
 ca  Finally, it is easy to obtain the reference currents
 1 − −
 I s  2 2 2   along the abc axes by the inverse transformation
I  =   I cb (2) of Concordia:
 s  3 3 3 
0 2 − 2   I cc 

Instantaneous active power p and instantaneous

reactive power q are defined by:
 44 Fabrice Biloa Assolo, Charles Hubert Kom, Arnaud Nanfak, Gildas Martial Ngaleu

  a change of reference point of the load currents

 1 0  and the network voltages is carried out using
 iref 1    Concordia to switch from three-phase to two-phase
  2 1 3  iref   (7) (αβ). Afterwards the reactive and active powers are
 iref 2  = −
3 2

2  iref  
 calculated. At this stage, to obtain the AC
 iref 3   
  components, two low-pass filters are used
− 1 −
3 respectively for the active and reactive power. The
 2 2  calculation of the reference currents in the αβ
benchmark is carried out thanks to the alternating
Figure 3 gives the implementation algorithm of the components of the powers of the voltage and of the
instantaneous power method to obtain the regulated voltage of the inverter. Finally to obtain
reference currents. The two-phase current method the final result we apply the inverse transform of
works in the space of αβ currents. It therefore Concordia
requires fewer calculations while being more :
precise and more robust than other methods.

Figure 3: Generation of reference currents by the instantaneous power method

C. Control by PWM of the active filter invert

In [20] the different PWM techniques are
presented: the PWM with symmetric regular sampling,
the PWM with asymmetric regular sampling and the
PWM with natural sampling. The technical diagram of
the PWM is given in Figure 4 each group of transistor-
diode assembled in parallel forms a bi-controllable
switch. The previous technical diagram first uses a
regulator which determines the reference voltage of
the inverter (modulator) from the difference between
the measured current and its reference. The latter is
then compared with a triangular signal called a high
frequency carrier in this work it is a 200 Hz setting the
switching frequency. The output of the comparator
provides the command order of the switches.Figure 5
gives the Simulink model of the PWM control where
the reference currents come from the instantaneous
power unit PQ.

Figure 4: PWM control principle

 Harmonic pollution control of electrical networks: Comparative study of… 45

Figure 5: Command by PWM in MATLAB/SIMULINK

voltage Vs (t) is related to the associated switching

control signal, the switching being ensured by an
D. Control by DCM of the inverter of the active
operational amplifier in its non-linear operation. The
three-phase shunt filter
saturation voltages of the amplifiers are +Vsat and -
Vsat while +VCC and -VCC are their supply voltages
The duty cycle modulation technique (DCM) The mathematical model DCM in our shunt filter is
translates new perspectives in inverter control
 V if u +  u −
techniques. DCM consists in transforming a signal Vs (t ) =  sat
−Vsat if u  u
Vref (reference intensity) into a modulated signal Vs + −

(t).± Vcc, produces an all-or-nothing modulated (8)

signal, thus protecting the transmitted signal from   V +  V if V (t ) = V
 u+ = 
sat 1 ref s sat
noise. Non-inverting DCM is used in this case
because it varies according to the modulation signal.  − Vsat +  1 ref if Vs (t ) = −Vsat
V
Its operating principle is described in figure 6

With
R1
=
R1 + R2
(9)
R2
1 =
R1 + R2
The supply voltage of the capacitor is governed by a
differential equation:
duc (t )
RC + uc (t ) = Vs (t ) (10)
dt

A solution of this equation is:

uc (t ) = (ui − u f )e − t / RC + u f (11)
Figure 6: Non-inverter duty-cycle modulator [17]
The capacitor charge and discharge times are included
The principle of duty cycle modulation (DCM) is respectively 0 <t <T1 and T1 <t <T2 It is established
based on encoding in the duty cycle Rm=Ton (t)/Tm that:
(t) of a square wave signal of the reference analog   V + ( − 1)Vsat 
T1 = − ln  1 ref
  V − (1 −  )V 
signal Vref (t). A DCM circuit therefore behaves like (12)
a voltage controlled oscillator. It consists therefore in  1 ref sat 

  V − ( − 1) Vsat 
a modulator involving a high relaxation frequency of
the oscillator and a low frequency of the control
T2 = − ln  1 ref
  V + (1 +  ) V 
(13)
signal to be modulated. Thus, the resulting modulated
 1 ref sat 
 46 Fabrice Biloa Assolo, Charles Hubert Kom, Arnaud Nanfak, Gildas Martial Ngaleu

 ()
Pm =
 V − ( − 1) Vsat  V + ( − 1)Vsat Vsat (1 +  ) log(
1+ 
T =  ln 1 ref +  ln 1 ref 1−
)
1Vref + (1 +  )Vsat 1Vref − (1 −  )Vsat
 ( V ) 2 − ( ( − 1) V )2 
 T = RCln   
1 ref sat
(14) 1
 (1Vref ) − ( (1 −  ) Vsat )2   fm0 =
1+ 
2
   2 ln( )
 = RC  1−
 1
 f min = ()
The duty cycle of the non-inverting modulator is:
  ( V ) − ((1 +  )Vsat ) 2 
2

T1 (Vref )  ln  1 max 2 2 
  (1Vmax ) − (( − 1)Vsat ) 
Rm = 
Tm (Vref )
(15)
 0  1
we deduct: 

1Vref − ( − 1)Vsat With the basic frequency of the modulator fm0

ln( ) previously selected, the minimum frequency fm
1Vref + (1 + )Vsat
Rm (Vref ,  , Vsat ) = (16) previously established α=0.003081723734398
(1Vref ) 2 − (( +)Vsat )2 τ =RC with fmlim is the limit modulator frequency. Our
ln( )
(1Vref ) 2 − ((1 − )Vsat ) 2 work on optimization is based on the formulation
given in [16] and for the chosen fm0 equal to 200
KHz, the frequency fmlim is obtained by calculating
fmin for 𝛼 very small. We obtain the following
In order to improve the efficiency of the electronic optimum parameters of DCM:
parts of the non-inverting DCM modulator, we are 𝛼=0.003081723734398 or
optimizing our modulator. The aim is to find the α=R1/(R1+R2) and 1-α=R2/(R1+R2)
optimal characteristic parameters α, τ and E of the
modulator providing an excellent level of operation. After calculating the value τ=0.000486670905896s,
For this purpose, the effects of non-linearity and R1=330Ω, R2=10KΩ, C=47nF, R=1KΩ
variation of the characteristic parameters were
studied. The optimal quality of the modulator Figure.7 represents the Simulink diagram of the DCM
obtained is observed through its characteristic modulator. The latter's input is Iref and its output is
quantities: Rm(Vref,,α,Vsat) and f m = 1 cmd. Its parameters were determined in the previous
T (Vref ,  ,Vsat , ) paragraph.
with fm being the frequency of the modulator and
Rm(Vref,,α,Vsat) the duty cycle.
Optimizing the MRC modulator therefore means
maximizing Pm(α,Vsat) or minimizing - Pm(α,Vsat)
Which is the slope. Our optimization criterion is
therefore Pm (α,Vsat) or

Figure 7: Command by DCM in MATLAB/SIMULIN

 Harmonic pollution control of electrical networks: Comparative study of… 47

E. Symmetrical Linear Duty-Cycle Modulation

Figure. 8 gives the operating principle of the SLDCM. Now let's calculate ton and toff
The hysteresis comparator obtained the signal and the + −
integrator modulates by integrating the difference During ton, VS=+Vsat and v0 goes from v0 to v0
between the modulated signal and the reference RC

v0−
signal, a triangular signal is a return to the input of the ton = − v0+
dv0 ()
comparator.
Vsat − Vef
R1 Vsat
ton = 2 RC ()
R2 Vsat − Vref

− +
During ton, VS=-Vsat and v0 goes from v0 to v0
R1 Vsat
t0 FF = 2 RC ()
R2 Vsat + Vref

So, = ton + tOFF ()

Figure 8: Symmetrical Linear Duty-Cycle Modulation [17] R1 Vsat2

T = 4 RC ()
R2 Vsat2 − Vref2
The mesh law applied to the voltages on the first non-
inverting modulator gives the equations below:  Which represents the duty cycle is equal to:

vo R2 + Vs R1 = 0 () ton 1  Vref 

= = 1 +  ()
R T 2  Vsat 
vo = − 1 Vs ()
R2
The duty cycle of SLDCM is a linear equation (30)
consisting only of the passive components. When
vo represents both the integrator output voltage and
Vref = 0v the respective values of its minimum
the input to the control signal switching block. The
periods and are duty cycles are equal to:
current in the capacitor C is defined by:
R1
T0 = 4 RC
R1 R2
v0+ = + Vsat ()
R2 1
=
R1 () 2
v0− = − Vsat
R2
III. SIMULATION AND RESULTS
The current in the capacitor C is defined by:
Vs − Vref dvc
A. Simulation parameter
ic = =C () Simulations in MATLAB/SIMULINK were carried
R dt out under the specifications found in Table 1 below.
Where, vc is the voltage across the capacitor. The parameters of the simulations are grouped
The output voltage v0 of the integrator is given by: together in table 1. Regarding the DCM parameters,
we use the data from[16] whose optimization
parameters were determined in section II with a
v0 = Vref − vc frequency of 200 hz. on the other hand the parameters
() of the SLDCM use the data of [17]which were
 vc = Vref − v0 provided experimentally in this article. The frequency
value of SLDCM is 200HZ. Finally, the frequency of
So, the PWM carrier is 200Hz compared to a frequency
RC of 50Hz to produce the IGBT control slots.
dt = (dVref − dv0 ) () The simulated electrical network is three-phase
Vs − Vref three-wire and the MATLAB/Simulink software is
used. The pollutant load is a conventional three-phase
If dVref = 0 then Vref is a constant so rectifier connected in series in an electrical network

RC
dt = − dv0 ()
Vs − Vref
 48 Fabrice Biloa Assolo, Charles Hubert Kom, Arnaud Nanfak, Gildas Martial Ngaleu

The general structure of the studied system is given in

TABLE 1: SIMULATION PARAMETERS Figure. 9, the system is constituted:
Vs=220V,f=50Hz,
Source Rs= 0.25x10-3Ω • A three-phase power source,
Lc=0.023 mH, • A non-linear charge,
Load Rc=0,82mΩ, =0,78mΩ, • An active three-phase shunt filter connected
Lc=2,6mH in parallel to the mains,
Lf=3mH, C=8mF • A filter control unit.
Filter
R1=330Ω, R2=10KΩ,
DCM C=47nF, R=1KΩ
R1=10KΩ, R2=20KΩ,
SLDCM C=10nF, R=5.000KΩ

Figure 9: The general structure of the system studied

Figure 10: Waveform of the load and source Is(t) before filter
 Harmonic pollution control of electrical networks: Comparative study of… 49

Figure 11: Order of source current harmonics and TDH before shunt active filter

Figure 12: Waveform of the load

Figure. 10 shows the course of the current Is(t) in waveform, i.e. a THD of 28.197%. Figure. 12 is a
the network before filtering for phase (a), the spectral waveform of the load current main actors of said
analysis of this current is presented in Figure. 11 deformations because it does not consume the
source harmonics spectrum contains the harmonics of sinusoidal energy produced by the source it then
rank 5,7,13,17 and 19. And we observe the presence injects into the circuit of harmonics.
of several ranks of harmonics disturbing the current
.

Figure 13: Waveform of the current source Is(t) after filter: control by PWM

Figure 14: Order of source current harmonics and TDH after shunt active filter: control by PWM

Figure. 13 and 14 are the characteristics obtained After filtering the current Is(t) by the active shunt
from the source current after filtering using the PWM filter controlled by PWM control, we obtain the signal
control, thus obtaining respectively the waveform in Figure. 13. We observe a significant improvement
which has become sinusoidal again, proof that there in the waveform which has become sinusoidal and
has been a reduction in harmonics and the harmonic whose spectrum given in Figure 14 allowing us to
rate which has passed from 28.19% THD to 1.72%, affirm that there is a clear decrease in harmonics, i.e.
the result of which complies with the IEEE-519 a THD of 1.72%.
standard which must meet the condition that the THD
must be less than 5%.
 50 Fabrice Biloa Assolo, Charles Hubert Kom, Arnaud Nanfak, Gildas Martial Ngaleu

Figure 15: Waveform of the current source Is(t) after filter: control by DCM

Figure 16: Order of source current harmonics and TDH after shunt active filter: control by DCM

By applying the control by Modulation with DCM to the curve and whose current spectrum Is(t) in Figure
our active shunt filter we obtain the signal Is(t) in 16 shows a greater reduction in harmonics, i.e. THD =
Figure 15 which shows a significant improvement in 0.79%.

Figure 17: Waveform of the current source Is(t) after filter: control by SLDCM

Figure 18: Order of source current harmonics and TDH after shunt active filter: control by SLDCM

By applying the control by Modulation with the level of the curve and whose current spectrum Is
SLDCM to our active shunt filter we obtain the signal (t) in Figure 18 shows a reduction in harmonics of less
Is(t) in Figure 17 which shows a slight degradation at than 5%, that is to say THD = 3.59%.
TABLE 2: INTERPRETATION OF PERFORMANCE FOR THE TWO ORDERS
PWM Control DCM Control SLDCM Control
THD before filtering 28,19% 28,19% 28,19%
THD after filtering 1.72% 0.79% 3.59%.
Main voltage 220V 220V 220V
Controllability of the
good good Bad
switching frequency
Waveform of output current Is good Good Fairly good
 Harmonic pollution control of electrical networks: Comparative study of… 51

Table 2 gives an interpretation of the results the literature for harmonic reduction. It is clearly
obtained for Hysteresis and Cyclic Ratio Modulation established that all these methods give a THD ˂ 5%
(DCM) control. THD of the current at phase (a) Is(t) thus complying with the IEEE-519 standard, but the
after filtering calculated on the first 20 harmonic ranks control strategy using DCM with the shunt active
is 1.72% for PWM control. 0.79% for DCM control, filter topology proposed in this work is more flexible
and 3.59% for SLDCM which corresponds to the due to its simplicity of implementation and allows
IEEE 519 standard that limits the THD ˂ 5%, but it good controllability of the switching frequency of the
can be seen that control by DCM has better inverter switches that make up the active filter.
performance in terms of harmonic elimination, it is
easy to use, ensures operation at a fixed frequency.
NOMENCLATURE
CONCLUSION
PWM Pulse Wave Modulation
In this article, we have presented the method of
DCM Duty Cycle Modulation
instantaneous powers for the creation of disturbance
General linear Duty
currents; a mathematical model has been developed. GLDCM
Cycle Modulation
We used PWM control DCM and SLDCM commands
Linearized Non-
for switching IGBTs. Their principle is based on the
LNIDCM Inverter Duty Cycle
modulation of a reference voltage. This voltage,
Modulation
called the modulating signal at the input of the
Non-Inverter Duty
modulator, will be transposed into a modulated signal NIDCM
Cycle Modulation
in the form of square waves representing the control
Total Harmonic
signals of the IGBTs. These two systems were THD
Distorsion
implemented and observed under the same conditions
Reference or control
in an active shunt filter. Regarding the THD rate in Vref
voltage
the electrical circuit, the DCM gives the best
Vsat Saturation voltage
performance because it is 0.79% against 1.72% for
Vsa,Vsb,Vsc Output voltage
the PWM and 3.59% for SLDCM. but it can be seen
Integrator output
that the DCM control has better performance in terms Vo
voltage
of harmonic elimination and is positioned in the
Vc Capacitor voltage
literature as a promising method due to its flexibility
and ease of implementation in industrial applications.
It will also be interesting to carry out an in-depth REFERENCES
study between the intelligent controls techniques [1] J. W. Dixon, J. J. Garcia, et L. Moran, « Control system for
neural networks with DCM applied to control the three-phase active power filter which simultaneously
inverter in an active shunt filter. compensates power factor and unbalanced loads », IEEE
transactions on Industrial Electronics, vol. 42, no 6, p.
TABLE 3: COMPARISON OF METHODS IN THE 636‑641, 1995.
LITERATURE [2] A. Sahbani, M. L.-B. Braiek, M. A. Dami, et M. Jemli,
PWM « Commande d’un filtre actif triphasé shunt par logique
Intersective
Hysteresis
DCM floue », 2005.
Control
Control
Control [3] H. Chaoui et P. Sicard, « Adaptive fuzzy logic control of
[16] permanent magnet synchronous machines with nonlinear
[21]
friction », IEEE Transactions on Industrial Electronics, vol.
Waveform of 59, no 2, p. 1123‑1133, 2011.
good Fairly
output current good [4] T. Yorozu, M. Hirano, K. Oka, et Y. Tagawa, « Electron
good
spectroscopy studies on magneto-optical media and plastic
TDH before substrate interface », IEEE translation journal on magnetics
filtering 20,17% 26,57% 28,19 in Japan, vol. 2, no 8, p. 740‑741, 1987.
[5] A. F. H. Nohra, M. Fadel, et H. Y. Kanaan, « A novel
TDH after instantaneous power based control method for a four-wire
filtering 2.33% 3,12% 0.79% SAPF operating with highly perturbed mains voltages », in
2016 IEEE International Conference on Industrial
Technology (ICIT), 2016, p. 1236‑1241.
Signal quality No noise Brute No noise [6] M. Moranchel, F. Huerta, I. Sanz, E. Bueno, et F. J.
Rodríguez, « A comparison of modulation techniques for
Controllability modular multilevel converters », Energies, vol. 9, no 12, p.
of the 1091, 2016.
Yes No Yes [7] S. K. Peddapelli, Pulse Width Modulation. De Gruyter
switching
frequency Oldenbourg, 2016.
[8] L. N. Nnem, B. M. Lonla, G. B. Sonfack, et J. Mbihi,
« Review of a multipurpose duty-cycle modulation
technology in electrical and electronics engineering »,
The performance of the proposed active shunt filter Journal of Electrical Engineering, Electronics, Control and
control strategy was compared with the control Computer Science, vol. 4, no 2, p. 9‑18, 2018.
strategy adopted with some other control methods [9] J. Mbihi, B. Ndjali, et M. Mbouenda, « Modelling and
simulation of a class of duty-cycle modulators for industrial
proposed in the literature. Table 3 shows the instrumentation », 2005.
comparison of various control techniques proposed in
 52 Fabrice Biloa Assolo, Charles Hubert Kom, Arnaud Nanfak, Gildas Martial Ngaleu

[10] J. Mbihi, F. Beng, martin kom, et L. Nneme, « A Novel [20] L. F. Rafanotsimiva, G. Besançon, D. Georges, E. J. R.
Analog-to-digital conversion Technique using nonlinear Sambatra, et J. M. Razafimahenina, « Modélisation
duty-cycle modulation », International Journal of Electronics multimodèle et commande par Compensation Parallèle
and Computer Science Engineering, vol. 7, janv. 2013. Distribuée d’un système SMIB », MADA-ENELSA, vol. 1, p.
[11] J. Mbihi et L. N. Nneme, « A multi-channel analog-to-Digital 18‑26, 2013.
conversion technique using parallel duty-cycle modulation », [21] A. Morsli et A. Tlemçani, « Dépollution des Réseaux
International Journal of Electronics and Computer Science Electriques Basse Tension utilisant un Filtre Actif Parallèle à
Engineering, vol. 1, no 3, p. 826‑833, 2012. deux niveaux contrôlé par l’Algorithme P-Q », vol. 53, no 2,
[12] J. Mbihi et L. N. Nneme, « A novel control scheme for buck p. 7, 2012.
power converters using duty-cycle modulation »,
International Journal of power electronics, vol. 5, no 3‑4, p.
185‑199, 2013.
[13] B. L. Moffo, J. Mbihi, et L. N. Nneme, « A low cost and high
quality duty-cycle modulation scheme and applications »,
International Journal of Electrical, Computer, Energetic,
Electronic and communication Engineering, vol. 8, no 3, p.
82‑88, 2014.
[14] L. N. Nneme et J. Mbihi, « Modeling and simulation of a new
duty-cycle modulation scheme for signal transmission
systems », American Journal of Electrical and Electronic
Engineering, vol. 2, no 3, p. 82‑87, 2014.
[15] J. Mbihi, « Dynamic modelling and virtual simulation of
digital duty-cycle modulation control drivers », International
Journal of Electrical, Computer, Energetic, Electronic and
Communication Engineering, vol. 11, no 4, p. 472‑477, 2017.
[16] P. NNa et E. Ndjakomo, « Harmonic Pollution Control of the
Electrical Network by Three-Phase Shunt Active Filter:
Comparative Study of Controls, by Hysteresis and by Duty
Cycle Modulation », World Academy of Science, Engineering
and Technology International Journal of Energy and Power
Engineering, 2020.
[17] G. Ngaleu, C. Kom, T. Yeremou, S. Eke, et A. Nanfak,
« Design of New Duty-Cycle Modulator Structures for
Industrials Applications, an Alternative to Pulse-Width
Modulation », European Journal of Electrical Engineering,
vol. 23, p. 103‑111, juin 2021, doi: 10.18280/ejee.230203.
[18] A. T. Boum, G. B. Djidjio Keubeng, et L. Bitjoka, « Sliding
mode control of a three-phase parallel active filter based on a
two-level voltage converter », Systems Science & Control
Engineering, vol. 5, no 1, p. 535‑543, 2017.
[19] L. Benchaita et S. Saadate, « A comparison of voltage source
and current source shunt active filter by simulation and
experimentation », IEEE transactions on power systems, vol.
14, no 2, p. 642‑647, 1999.