| Journal of Collective Sciences and Sustainability
Received: 24 September 2026; Revised: 08 November 2026; Accepted: 13 November 2026; Published Online: 19 November 2026.
J. Collect. Sci. Sustain., 2026, 2(4), 26418 | Volume 2 Issue 4 (December 2026) | DOI: https://doi.org/10.64189/css.26418
© The Author(s) 2026
This article is licensed under Creative Commons Attribution NonCommercial 4.0 International (CC-BY-NC 4.0).
Design of a PI-Controlled PWM Three-Phase Rectifier
for DC Voltage Regulation in Low-Speed PMSG Wind
Energy Systems
Ika Noer Syamsiana,
1,*
Raditya Pratitoa,
1
Rahma Nur Amalia,
1,
Asfari Hariz Santoso
1,
and Arwin Datumaya
Wahyudi Sumari
2,
1
Electrical Power Systems Study Program, Department of Electrical Engineering, State Polytechnic of Malang, Malang, 65141, East
Java, Indonesia
2
Office of Special High-Rank Officer, Indonesian Air Force Headquarters, Jakarta Timur, DKI Jakarta, 13870, Indonesia
*Email: ikanoersyamsiana@polinema.ac.id (Ika Noer Syamsiana)
Abstract
Wind energy in Malang, Indonesia, generally has low wind speeds, causing voltage fluctuations in the outputs
of permanent magnet synchronous generators (PMSGs). This study aims to design and evaluate a three-phase
rectifier based on pulse width modulation (PWM) with a proportional-integral (PI) controller to improve DC
voltage stability in low-speed wind energy conversion systems. Evaluation was carried out through
MATLAB/Simulink and Proteus simulations with variations in the input voltage, duty cycle, and resistive (R)
and resistive–inductive (R–L) load conditions. The simulation results revealed that under open-loop
conditions, the maximum DC voltage was close to the theoretical value of 64.8 V. At the same time, the duty
cycle variation affected the minimum voltage and produced a voltage ripple of approximately 9–10%. The PI
controller reduces the voltage ripple to approximately 1–3% and improves the stability of the DC output
voltage at input voltages of 36 V and 48 V. The results show that the PI controller can improve the regulatory
performance of the PWM rectifier voltage under varying input voltage and load conditions. This study
contributes to the development of a PMSG-based wind energy conversion system for areas with low wind
potential. Although this study is based on simulation results, it provides a basis for further experimental
validation to evaluate the performance of the proposed system under real operating conditions.
Keywords: DC Voltage regulation; Low-speed wind energy; Permanent Magnet Synchronous Generators (PMSGs);
PI Controller; PWM rectifier; Three-phase rectifier.
1. Introduction
The global and national energy crisis has accelerated the transition to sustainable renewable energy sources.
Dependence on fossil fuels not only raises concerns about resource scarcity but also significantly exacerbates
greenhouse gas emissions and climate change.
[1]
As a vast archipelagic country with abundant natural
resources, Indonesia has experienced increasing energy consumption alongside a growing demand for
renewable energy. Indonesia possesses substantial renewable energy reserves, with an estimated total
capacity of 419 GW. These reserves include 75 GW from hydropower, 23.7 GW from geothermal energy, 32.6
GW from bioenergy, 207.8 GW from solar energy, 60.6 GW from wind energy, and 19.3 GW from microhydro
sources .
[2]
Wind energy in Indonesia is recognized as a viable renewable resource because of its considerable potential,
estimated at approximately 60.6 GW. However, the installed capacity remains limited to approximately 157
MW, with a target of 255 MW by 2025.
[2]
More recent data from the Ministry of Energy and Mineral Resources
of the Republic of Indonesia indicate an even greater national potential of approximately 154.6 GW, while the
installed capacity by 2024 had only reached approximately 152.3 MW, concentrated mainly in high-wind
regions such as South Sulawesi (the 75 MW Sidrap and 72 MW Jeneponto wind farms); the government
targets an additional 5 GW of wind capacity by 2030 under the 2025–2035 National Electricity Supply
Business Plan (ESBP), while low-wind regions such as East Java have not yet been prioritized for
development.
[3]
A more recent comprehensive review corroborates these figures and highlights that the
country's wind development remains geographically concentrated, with most remaining potential still
unexploited outside a handful of high-wind provinces.
[4]
This limitation is largely due to relatively low average
wind speeds, which restrict large-scale power generation, as well as insufficient extensive studies and
confidence in selecting optimal sites, particularly in remote and offshore areas.
[5]
Innovations such as low-
wind-speed turbines are being researched to increase wind energy utilization in regions with low wind
speeds in Indonesia, potentially making wind power economically competitive with fossil fuels.
[6]
The
minimum wind speed required to generate power is 3 m/s. Wind performance varies across three locations:
Sampang, Surabaya, and Nganjuk. The annual average wind speeds in Sampang, Surabaya, and Nganjuk are
2.34 m/s, 3.03 m/s, and 1.97 m/s, respectively. Wind in these regions can generate 1,284 MW, 1,008 MW, and
1,199 MW per year, assuming a constant temperature coefficient of approximately 6%. The capacity factors of
the wind power plants in Sampang, Nganjuk, and Surabaya are 30.02%, 30.01%, and 30.00%, respectively.
These activities could reduce greenhouse gas emissions by 0.904, 0.709, and 0.846 tons of CO2 per year in the
surrounding atmosphere.
[7]
These wind speed data indicate that Indonesia generally has low wind speeds.
In East Java Province, particularly in Malang city, the wind energy potential is characterized by relatively low
and variable average wind speeds of approximately 3–6 m/s. This condition presents challenges for Wind
Power Plants (WPGP), especially during the conversion of mechanical energy into electrical energy. Currently,
permanent magnet synchronous generators (PMSGs) are widely used in power generation because of their
high efficiency. However, the flux density of permanent magnets decreases with increasing temperature,
complicating voltage management; hybrid-excitation PMSG designs have been proposed to counteract this
temperature sensitivity while keeping the generator’s output voltage within safe limits across varying wind
speeds.
[8]
Wind energy conversion systems (WECSs) based on a PMSG connected to the grid have been
proposed and implemented. The possibility of achieving shorter pole pitches allows machines to operate at
low speeds, eliminating the need for gearboxes or enabling the use of single-stage gears for more compact
designs.
[9]
The relationship between the fluctuations in the wind speed and the electrical output of a PMSG can be
explained through basic wind-energy physics. The kinetic energy contained in a moving mass of air is
proportional to the square of the wind speed:

(1)
where m is the mass of air and v is the wind speed; consequently, the power extracted from the wind is
proportional to the cube of the wind speed,


(2)
where ρ is the air density and A is the swept area of the turbine blades.
[10]
Because the wind speed in low-wind
regions such as Malang fluctuates continuously within a narrow range, even small variations in v produce
relatively large variations in the mechanical power delivered to the PMSG shaft. This fluctuating mechanical
input directly translates into a fluctuating electrical frequency and voltage at the PMSG terminals since the
generator’s output voltage and frequency are proportional to its rotational speed. As a result, the DC voltage
produced after rectification also fluctuates so that the voltage and frequency characteristics of the PMSG
output are strongly influenced by fluctuations in the wind speed, making rectifiers and control systems crucial
for efficiently maintaining power quality.
Three-Phase Rectifiers are crucial in Wind Power Plant (WPGP) systems for converting three-phase AC
voltage from Permanent Magnet Synchronous Generators (PMSG) into stable DC voltage. The use of
conventional rectifiers without control systems often leads to significant issues, including high current
harmonics, DC voltage ripple, and reduced power conversion efficiency. These conditions degrade power
quality, cause power losses, and negatively impact the components and overall performance of wind power
plant (WPGP) systems.
[11]
Indrianto and Hariyadi’s research demonstrated that three-phase rectifiers
operating with resistive–inductive loads generate significant current harmonics, which adversely affect the
power quality of electrical systems.
[12]
Therefore, the integration of pulse width modulation (PWM) along with
control systems is essential in power conversion systems.
1.1. Related work
Previous studies on three-phase rectifier control for PMSG-based and general power conversion systems can
be grouped according to their control approach. A first group of studies examined passive or diode-based
rectifiers connected to PMSG wind turbines; Rahimi modeled a grid-connected PMSG wind turbine assisted by
a diode rectifier and boost converter and reported that although simple and low-cost, such uncontrolled
rectifiers provide no means of regulating the DC-link voltage and are strongly affected by variations in
generator speed and load, whereas Tchoumtcha et al. proposed a synergetic control scheme for a stand-alone
PMSG that simultaneously regulates current, maximum power extraction, and DC-link voltage under
fluctuating load and uncertain turbine parameters.
[13,14]
A second group focused on active or PWM-based
rectifiers as an alternative to passive rectification; Alili et al. demonstrated that a Vienna-rectifier-based PWM
converter connected to a PMSG wind turbine can regulate DC-link voltage more effectively than a passive
rectifier, while Rajaei et al. further showed that Vienna-rectifier topologies reduce switching and conduction
losses for low-power PM-based wind energy conversion systems; in the same vein, Annisa et al. simplified
firing-angle generation for a three-phase controlled rectifier through a Simulink-based cosine-comparator
gate-control logic, Acikgoz et al. combined a type-2 fuzzy neural network with artificial-bee-colony
optimization to regulate the DC-link voltage of a three-phase PWM rectifier under unbalanced supply and
regenerative-mode conditions, and Li et al. proposed a model-predictive direct power control method that
reduces the computational burden of PWM rectifier control without sacrificing regulation accuracy.
[15-19]
A
third group of studies investigated proportional-integral (PI) control applied at the DC-link or rectifier output;
Bouafia et al. combined fuzzy logic and PI control for direct power control of a three-phase PWM rectifier to
regulate the output voltage; Sanchis and Peñarrocha-Alós proposed an experimental step-response-based
tuning method for PI controllers; and Zhou et al. reported that a well-tuned PI controller can maintain a near-
sinusoidal input current with low distortion while preserving system power-factor reliability.
[20-22]
Furthermore, Nurwati et al. reinforced that discrete PI control in PWM rectifiers can significantly enhance
system performance.
[23]
Beyond rectifier-specific applications, similar trial-and-error and optimization-based
PI tuning approaches have proven effective for other power-electronic converters that supply wind and
renewable-energy loads; Chincholkar et al. introduced a normalized-error PI controller to overcome
saturation limitations in a DC-DC buck converter; Uc et al. validated PI-gain optimization using local-and-
global search algorithms on a DC-DC boost converter; and Mrabet et al. compared trial-and-error Kp/Ki
optimization against metaheuristic tuning for a wind-energy MPPT controller, reinforcing that trial-and-error
PI tuning remains a practical and sufficiently accurate baseline for this class of system.
[24-26]
A fourth group
examined rectifier control performance under variations in speed, input voltage, and load; Shwetha and
Guruswamy applied a whale-optimization-algorithm-tuned PI controller to a three-phase PWM rectifier and
achieved a rise time of 0.0126 s, a settling time of 0.06 s, and a DC voltage ripple below 3%, whereas Alzahrani
compared controlled and uncontrolled rectifiers for low-power wind turbines and reported that fully
controlled rectifiers provide a wider regulation range under varying firing angles at the cost of additional
control-circuit complexity.
[27,28]
Finally, a fifth group of studies validated their proposed control strategies
across multiple platforms, combining MATLAB/Simulink simulation with either circuit-level software or
physical hardware testing to confirm consistency between numerical and experimental results.
[22,27,28]
However, most of these studies were conducted under highly stable source conditions and do not accurately
reflect the characteristics of wind power generation at low and variable wind speeds.
1.2. Research gap
Previous research has developed various rectifier topologies and control methods for wind energy conversion
systems based on permanent magnet synchronous generators (PMSGs).
[29-31]
However, studies on the effects of
the duty cycle-varying PWM on the DC output voltage and voltage ripple under open-loop conditions,
especially across input voltages representing changes in the wind speed, still need further investigation. In
addition, the influence of resistive (R) and resistive–inductive (R–L) load characteristics, as well as the
effectiveness of proportional–integral (PI) controllers in improving DC voltage stability and reducing ripples
compared with uncontrolled conditions, requires quantitative evaluation of the same rectifier configuration.
Studies integrating system-level simulation in MATLAB/Simulink and circuit-level simulation in Proteus
should also be developed to provide a more comprehensive performance analysis. Therefore, this study
evaluates PWM-based three-phase rectifiers with PI controllers across various input voltages, duty cycles, and
R and R–L loads by comparing open-loop and closed-loop performance. This study's main contribution is a
quantitative evaluation of DC voltage regulation and voltage-ripple reduction across two simulation
platforms, providing a basis for developing and experimentally validating low-speed PMSG-based wind
energy conversion systems.
1.3. Proposed solution
To address these gaps, in this study, a PI-controlled PWM three-phase rectifier intended for a low-speed PMSG
wind energy system is designed. The control variable is the DC output voltage of the rectifier, which is
regulated by adjusting the PWM duty cycle applied to the six MOSFET switches on the basis of the error
between the reference voltage and the measured DC output. The system is first analyzed under open-loop
conditions with manually set duty cycles of 30%, 50%, 74%, and 95% to characterize the rectifier's inherent
behavior and subsequently evaluated under closed-loop PI control at three-phase input voltages of 48 V and
36 V, which are used as representative operating points for higher- and lower-wind-speed conditions,
respectively (see Section 2.2 regarding the fixed-voltage-source representation of the PMSG input), rather
than being derived from a calibrated wind-speed-to-voltage PMSG model. Both resistive (R) and resistive-
inductive (R-L) loads are tested to evaluate the influence of load characteristics on system performance.
System-level control design and PI tuning are performed in MATLAB/Simulink, while circuit-level
performance, including the output voltage waveform, current, power, and ripple, is evaluated using Proteus.
1.4. Research objective
This study aims to design and evaluate a PI-controlled PWM three-phase rectifier for regulating the DC output
voltage of a low-speed PMSG wind energy system under variable input-voltage, duty-cycle, and load
conditions.
1.5. Main contribution
The main contributions of this study are as follows: (1) an application-specific PI-controlled PWM three-phase
rectifier design for low-speed PMSG operation; (2) a comparative assessment of open-loop and closed-loop
responses under different input voltages, duty cycles, and R/R-L loads; and (3) a cross-platform evaluation
using MATLAB/Simulink and Proteus, demonstrating a reduction in the DC voltage ripple from approximately
9–10% to 1–3%.
2. Methods
Testing and data collection in this study were carried out by varying the three-phase source voltage (V
in
) and
load conditions Resistive (R) and Resistive-Inductive (RL) and then observing the output results in the form of
the DC voltage, current, power, and voltage ripple for each test condition. The following are the testing
parameters used shown in Table 1.
Table 1: Testing Parameters
NO
Input Voltage Variation
(V
in
)
Duty Cycle Variation
(%)
Resistive Load Variations
(R)
Resistive-Inductive Load Variations
(Z)
1.
48 V
30%
50%
74%
95%
450 Ω
650 Ω
1000 Ω
1936 Ω (1548 Ω + 3,7 H)
645 Ω (516 Ω + 1,23 H)
323 Ω (258 Ω + 0,62 H)
2.
36 V
30%
50%
74%
95%
450 Ω
650 Ω
1000 Ω
1936 Ω (1548 Ω + 3,7 H)
645 Ω (516 Ω + 1,23 H)
323 Ω (258 Ω + 0,62 H)
In the data analysis stage, data need to be collected to assess the performance of the three-phase rectifier is
shown in Table 2. All measurements listed in Table 2 (Multimeter, Oscilloscope) were performed using
Proteus's built-in virtual instruments during circuit-level simulation, not physical hardware devices; no
physical multimeter or oscilloscope was used to produce the results reported in this study.
Table 2: Analysis Parameters
No
Measuring tool
Test
1.
Multimeter
Set the potentiometer to measure DC voltage. Connect the multimeter probe to the
rectifier output terminals. Check the result on the display.
2.
Oscilloscope
Connect the oscilloscope probe to the rectifier output terminal. Change the
rectifier's PWM duty cycle and observe the voltage waveform on the display.
3.
Oscilloscope
Connect the oscilloscope probe to the rectifier output terminal. Change the
rectifier's PWM duty cycle and watch the voltage graph on the display to detect any
ripple.
4.
Multimeter
Adjust the potentiometer to measure the current. Connect the multimeter probe in
series with the positive DC terminal and monitor the current flowing to the load.
5.
Calculation
The power on the load is obtained from Ohm's law calculations, which is the result
of multiplying the measured DC voltage by the load current.
All measurements listed in Table 2 (Multimeter, Oscilloscope) were performed using Proteus's built-in virtual
instruments during circuit-level simulation, not physical hardware devices; no physical multimeter or
oscilloscope was used to produce the results reported in this study.
2.1. Research framework
The stages of this research follow the flow shown in Fig. 1. This study begins with a literature review on PI-
controlled PWM three-phase rectifiers for low-speed PMSG wind energy systems, followed by the conceptual
design of the control system. The control system is then implemented and tested; if the test result is not yet
satisfactory, the design is revised and reimplemented, and the cycle is repeated until the control system
performs as intended. Once the control system design is successful, data collection is carried out for all test
conditions (duty cycle, input voltage, and load variations), followed by data analysis and, finally, the
formulation of conclusions.
2.2. System configuration
The block diagram of the overall low-speed WPGP system considered in this study is shown in Fig. 2. Wind
energy is captured by the wind turbine and converted into three-phase AC voltage by the PMSG. This AC
voltage is converted into DC voltage by the three-phase rectifier, which is the focus of this study; the resulting
DC voltage is subsequently stepped down and regulated by a buck converter before being used to charge a 24
V
DC
battery and, through an inverter, supply the AC load; the buck-converter-to-battery charging stage follows
the general charge/discharge control strategies reviewed for renewable-energy storage systems in Banguero
et al.
[32]
The configuration of the three-phase rectifier subsystem is shown in Fig. 3: the PMSG (represented by
a three-phase AC source of 48 V or 36 V) supplies the three-phase rectifier built from six MOSFET switches;
the rectifier's DC output is measured by a voltage sensor of the resistive-divider type described in Jittakort
and Aurasopon’s study for accurate low-cost embedded voltage/current sensing, and this measurement is fed
back to an Arduino UNO running the PI control algorithm, which adjusts the PWM duty cycle applied to the
MOSFET switches so that the output voltage is regulated toward the desired setpoint.
[33]
Fig 1: Research flowchart.
Fig 2: Overall Low-Speed WPGS system block diagram.
Fig 3: Three-Phase rectifier subsystem block diagram closed-loop system.
The closed-loop control structure is shown in Fig. 4. A 48 V (or 36 V) reference/setpoint voltage is compared
with the measured DC output voltage at a summing junction to produce an error signal. This error is
processed by the PI controller running on the Arduino UNO, whose output determines the PWM duty cycle
applied to the three-phase rectifier (the plant). The output voltage of the rectifier is continuously measured by
the voltage sensor and returned as the feedback signal, allowing the controller to keep the output voltage
close to the setpoint despite variations in load or input voltage.
Fig 4: Closed-Loop PI control block diagram.
2.3. Three-Phase PWM rectifier design
The three-phase rectifier is built using six MOSFET switches arranged in a three-phase full-bridge
configuration, switched according to a PWM pattern generated by the Arduino UNO and applied through three
IR2110 gate-driver ICs; the Arduino UNO program controlling this PWM pattern runs with a fixed sampling
time of 10 ms, as implemented in the Proteus circuit simulation shown in Fig. 5. A DC-link capacitor is
connected at the rectifier output to smooth the rectified voltage before it reaches the load; on the basis of the
ripple-tolerance design calculation, the required capacitance was 0.0062 F (6200 µF), and since a 6200 µF
capacitor was not commercially available, the closest standard value, 6800 µF, was selected for
implementation. The PWM carrier (triangular-wave) signal that drives the six MOSFET switches is generated
at a fixed switching frequency of 5 kHz. Under ideal, unmodulated six-pulse rectification, the theoretical DC
output voltage is given by Eq. 6 in Section 3 (V
do
= (3√2/π) × V
AC
). The rectifier output is tested using resistive
(R) loads of 450 Ω, 650 Ω, and 1000 Ω and resistive-inductive (R-L) loads of 1548 Ω + 3.7 H, 516 Ω + 1.23 H,
and 258 Ω + 0.62 H, as summarized in Table 1.
Fig 5: Proteus simulation circuit of the three-phase PWM rectifier.
2.4. Open-Loop PWM operation
Under open-loop operation, the PWM duty cycle is set manually (30%, 50%, 74%, and 95%) without any
voltage feedback so that the inherent behavior of the rectifier can be characterized independently of the
controller. Each duty cycle is tested at a three-phase input voltage of 48 V with every R and R-L load listed in
Table 1, and the maximum voltage (V
max
), minimum voltage (V
min
), current, power, and voltage ripple are
recorded for each condition. All the results reported in this study were obtained entirely through simulation:
system-level control design and PI-gain tuning were carried out in MATLAB/Simulink, and circuit-level
verification (output voltage waveform, current, power, and ripple) was carried out in Proteus. No physical
hardware prototype was used to produce the results reported in this manuscript.
2.5. Design and tuning of the PI controller
The PI controller output is computed from the P equation and the I equation, where e(t) is the error between
the reference voltage (Sp) and the measured output voltage (Pv):
  (3)


(4)
   (5)
The gains in Kp and Ki were obtained through the trial-and-error tuning method described in Section 2
(increasing Kp to speed up the response and then increasing Ki to remove steady-state error). The
MATLAB/Simulink model of the closed-loop PI-controlled rectifier is shown in Fig. 6. In the
MATLAB/Simulink model, the PI controller output (modulation index) is bounded by a saturation block with
an upper limit of +1 and a lower limit of -1 before being used to generate the PWM switching signals. In the
corresponding embedded implementation, the Arduino UNO program (executed within the Proteus circuit
simulation) maps this modulation index to its native 8-bit PWM range (0–255), corresponding linearly to a 0–
100% duty cycle. A rate limit block in the Simulink model restricts the rate of change of the modulation index
to ±0.5 per second to prevent the controller from overreacting (oscillating) to ripple in the feedback signal,
thereby improving the overall stability of the control system. The final gains in Kp and Ki obtained for every
tested condition (input voltage, load type, and load value), together with the resulting closed-loop V
max
, V
min
,
current, power, and ripple, are listed in Table 3. In the MATLAB/Simulink model, no fixed sample time was
manually configured; the simulation instead ran using Simulink's automatic (variable-step) solver setting.
The Kp and Ki gains were tuned separately for each combination of input voltage and load (Table 3), not as a
single fixed gain set. In each case, Kp first increased until the output voltage responded quickly, and Ki
subsequently increased to remove the steady-state error, whereas the output-voltage response remained
stable under the load being tuned (Section 3.1). The tuning was therefore empirical: no transfer-function-
based design, formal stability analysis, or optimization algorithm was applied, and the gains in Table 3 (Kp =
0.27–1.61; Ki = 0.14–2.88) are not claimed to be optimal, and their spread between conditions reflects this
per-condition tuning.
Table 3: Closed-Loop PI controller performance - full detail (R and R-L, 48 V and 36 V)
V
in
(V)
Load
Kp
Ki
V
max
(V)
V
min
(V)
Current (A)
Power (W)
Ripple (%)
48
450 Ω (R)
1.05
0.5
66.5
60.1
0.14
8.862
2.9
48
650 Ω (R)
0.27
0.68
66.3
60.3
0.1
6.33
2.7
48
1000 Ω (R)
0.94
2.05
66
59.9
0.06
3.777
2.8
48
1548 Ω + 3.7 H (R-L)
1.61
2.43
66.1
60.8
0.04
2.538
2.4
48
516 Ω + 1.23 H (R-L)
1.56
2.88
66.2
60.5
0.12
7.602
2.6
48
258 Ω + 0.62 H (R-L)
1.45
0.69
66.2
60.0
0.25
15.775
2.8
36
450 Ω (R)
0.5
0.5
49.4
46.8
0.1
4.81
1.6
36
650 Ω (R)
1.07
1.86
49.3
46.6
0.07
3.3565
1.6
36
1000 Ω (R)
1.12
1.23
48.6
46.9
0.04
1.91
1.0
36
1548 Ω + 3.7 H (R-L)
0.63
0.14
49.2
46.2
0.03
1.431
1.8
36
516 Ω + 1.23 H (R-L)
1.13
0.9
49.4
46.6
0.09
4.32
1.7
36
258 Ω + 0.62 H (R-L)
1.25
1.66
49.4
46.5
0.19
9.1105
1.7
Fig 6: MATLAB/Simulink model of the Closed-Loop PI-Controlled PWM rectifier.
3. Results and discussion
3.1. Open-Loop PWM rectifier performance
Testing was carried out in two stages, namely, the tuning stage and the data collection stage. In the initial
stage, MATLAB/Simulink was used to tune the proportional (Kp) and integral (Ki) parameters of the PI
controller through the trial and error method so that the values of Kp and Ki were obtained to ensure a stable
output voltage response under various load conditions. After the optimal Kp and Ki values were obtained, the
final testing and data collection stage was conducted using Proteus, which was employed to evaluate the
performance of the three-phase rectifier circuit comprehensively through circuit-level simulation, including
the output voltage waveform, current, power, and ripple across different duty cycles, input voltages (V
in
), and
load types. In this study, waveform testing was performed using an oscilloscope with two measurement
channels, namely, Channel A and Channel B, each functioning to observe signals at different measurement
points in the circuit. The use of two channels aimed to allow simultaneous observation of the rectifier output
voltage and the PWM switching signal within a single display. The following is an explanation of each channel
used.
Duty cycle testing using a resistive load
This test aims to determine the effect of varying PWM duty cycles on the rectifier’s output voltage under open-
loop conditions (without a PI controller) using a purely resistive load (R). The duty cycle variations tested
were 30%, 50%, 74%, and 95%, with a three-phase input voltage (V
in
) of 48 V.
Duty cycle variation = 30%
Table 4 lists the measured V
max
, V
min
, current, power, and ripple of the three resistive loads at a 30% duty
cycle, and Figs. 7–9 show the corresponding output waveforms for the 450 Ω, 650 Ω, and 1000 Ω loads,
respectively.
Table 4: Resistive load test data - Duty = 30%
Duty cycle
width (%)
Load
Test data
V
max
(V)
V
min
(V)
Current(A)
Power(W)
Ripple (%)
Linear load (resistor/r (Ω))
30%
450
66.1
62.4
0.15
9.915
1.7%
650
66.3
61.5
0.1
6.63
2.2%
1000
66.1
62.9
0.06
3.966
1.4%
Duty cycle variation = 50%
Table 5 lists the measured V
max
, V
min
, current, power, and ripple of the three resistive loads at a 50% duty
cycle, and Figs. 10–12 show the corresponding output waveforms for the 450 Ω, 650 Ω, and 1000 Ω loads,
respectively.
Table 5: Resistive load test data - Duty = 50%
Duty cycle
width (%)
Load
Test data
V
max
(V)
V
min
(V)
Current(A)
Power(W)
Ripple (%)
Linear load (resistor/r (Ω))
50%
450
66.2
49.8
0.15
9.93
8.2%
650
66.1
48.5
0.1
6.61
8.9%
1000
65.8
48.7
0.06
3.948
8.6%
Duty cycle variation = 74%
Table 6 lists the measured V
max
, V
min
, current, power, and ripple of the three resistive loads at a 74% duty
cycle, and Figs. 13–15 show the corresponding output waveforms for the 450 Ω, 650 Ω, and 1000 Ω loads,
respectively.
Table 6: Resistive load test data - Duty = 74%
Duty cycle
width (%)
Load
Test data
V
max
(V)
V
min
(V)
Current(A)
Power(W)
Ripple (%)
Linear load (resistor/r (Ω))
74%
450
65.6
47.5
0.14
9.184
9.2%
650
65.8
47.9
0.1
6.58
9.1%
1000
65.3
46.7
0.06
3.918
9.6%
Duty cycle variation = 95%
Table 7 lists the measured V
max
, V
min
, current, power, and ripple of the three resistive loads at a 95% duty
cycle, and Figs. 16–18 show the corresponding output waveforms for the 450 Ω, respectively.
Table 7: Resistive load test data - Duty = 95%
Duty cycle
width (%)
Load
Test data
V
max
(V)
V
min
(V)
Current(A)
Power(W)
Ripple (%)
Linear load (resistor/r (Ω))
95%
450
64.7
47
0.15
9.705
9.1%
650
65.4
47.2
0.1
6.54
9.3%
1000
65.6
48
0.06
3.936
8.9%
Fig 7: Resistive load waveform 450Ω - Duty = 30%
Fig 8: Resistive load waveform 650Ω - Duty = 30%
Fig 9: Resistive load waveform 1000Ω - Duty = 30%
Fig 10: Resistive load waveform 450Ω - Duty = 50%
Fig 11: Resistive load waveform 650Ω - Duty = 50%
Fig 12: Resistive load waveform 1000Ω - Duty = 50%
Fig 13: Resistive load waveform 450Ω - Duty = 74%
Fig 14: Resistive load waveform 650Ω - Duty = 74%
Fig 15: Resistive load waveform 1000Ω - Duty = 74%
Fig 16: Resistive load waveform 450Ω - Duty = 95%
Fig 17: Resistive load waveform 650Ω - Duty = 95%
Fig 18: Resistive load waveform 1000Ω - Duty = 95%
Duty Cycle testing using a Resistive-Inductive load
This test aims to determine the effect of varying PWM duty cycles on the output voltage of the rectifier under
open-loop conditions (without a PI controller) using a resistive-inductive load (R-L). The duty cycle variations
tested were 30%, 50%, 74%, and 95%, with a three-phase input voltage (V
in
) of 48 V for each of the three R-L
load values in Table 1. The results are listed in Table 8.
Table 8: Resistive–Inductive load test data (All Duty Cycles)
Duty (%)
Load
V
max
(V)
V
min
(V)
Current (A)
Power (W)
Ripple (%)
30
1548 Ω + 3.7 H
66.2
61.8
0.04
2.648
2
30
516 Ω + 1.23 H
66.2
61.4
0.13
8.606
2.2
30
258 Ω + 0.62 H
66.4
61.6
0.25
16.6
2.2
50
1548 Ω + 3.7 H
65.6
47.8
0.04
2.624
9.1
50
516 Ω + 1.23 H
66
49.1
0.13
8.58
8.5
50
258 Ω + 0.62 H
66
49.3
0.25
16.5
8.4
74
1548 Ω + 3.7 H
65.5
48.8
0.04
2.62
8.4
74
516 Ω + 1.23 H
65.4
47.9
0.12
7.848
8.9
74
258 Ω + 0.62 H
65
47.5
0.25
16.25
9
95
1548 Ω + 3.7 H
65.3
48
0.04
2.612
8.8
95
516 Ω + 1.23 H
65
47.2
0.12
7.8
9.2
95
258 Ω + 0.62 H
64.8
46
0.25
16.2
9.8
The R-L loads show the same ripple trend across duty cycles as the resistive loads in Tables 4–7; the two load
types are compared in Section 3.4.
3.2. Analysis of Open-Loop Duty-Cycle effects
Analysis was conducted to evaluate the effect of varying the PWM duty cycle width on the output voltage (V
DC
)
under open-loop conditions. The data used were average values obtained from testing with resistive and
resistive-inductive loads at an input voltage of 48 V, and these data were compared with the theoretical
DC voltage of a 6-pulse rectifier. The theoretical voltage value (V
do
) can be formulated as follows.



   (6)
With:

= Theoretical DC voltage of a 6-pulse rectifier (V)
 = Peak-to-average voltage conversion factor (constant 3√2/π)

= AC Voltage (V)
Table 9 compares the measured V
max
, V
min
, and average V
DC
with the theoretical V
do
for each duty cycle, where
the average V
DC
is the mean over the six loads of (V
max
+ V
min
)/2.
Table 9: Duty cycle variation to output voltage
Duty(%)
V
max
(V)
V
min
(V)
V
DC
Average (V)
V
do
Theory (V)
Error V
max
to V
do
(%)
30
66,1 – 66,4
61,4 – 62,9
64,1
64,8
+2,0% s.d. +2,5%
50
65,6 – 66,2
47,8 – 49,8
57,4
64,8
+1,2% s.d. +2,2%
74
65,0 – 65,8
46,7 – 48,8
56,6
64,8
+0,3% s.d. +1,5%
95
64,7 – 65,6
46,0 – 48,0
56,2
64,8
-0,2% s.d. +1,2%
The data shown in Table 9 indicate that V
max
remains relatively constant across all duty cycle variations (64.7–
66.4 V) and is close to the theoretical value of V
do
= 64.8 V, with an error margin of less than 2.5%. This finding
demonstrates that V
max
is determined by the peak waveform of the 6-pulse rectifier rather than by the duty
cycle width. Conversely, V
min
decreases significantly as the duty cycle increases, from 61.4–62.9 V at a 30%
duty cycle to approximately 46–49 V at duty cycles between 50% and 95%. This reduction in V
min
substantially decreases the average V
DC
, from 64.1 V at a 30% duty cycle to approximately 56–57 V at duty
cycles of 50% or greater. This phenomenon is evident in Fig. 19, where the output voltage across all load
conditions both resistive and resistive-inductive converges at each duty cycle point. This highlights that
changes in the output voltage (V
out
) are more strongly influenced by variations in the duty cycle than by the
type or magnitude of the load used.
In Fig. 19, the output voltage of all six loads sharply decreases, by approximately 7–8 V, from a 30% to a 50%
duty cycle and then remains nearly constant up to 95%. At a 30% duty cycle, the six loads yield V
DC
values of
63.8–64.5 V, very close to the DC output voltage of the rectifier, V
do
= 64.8 V, whereas at duty cycles of 50–
95%, the V
DC
of all the loads stabilizes at approximately 55.4–58 V. Analysis was carried out to evaluate the
effect of varying the PWM duty cycle width on the ripple in the output voltage (V
DC
) under open-loop
conditions. The ideal ripple factor of a 6-pulse rectifier without a filter is approximately 4.2%, indicating that
the measured ripple does not originate from the natural rectification process but rather from the interaction
between PWM switching and the capacitor filter. The ripple value can be formulated as follows.
Fig 19: Variation of duty cycle on output voltage
At a 30% duty cycle, a 450 Ω resistive load is used.



   (7)



 (8)








  (9)
With:

= Peak-to-peak ripple value (V)

= Peak Voltage on the Graph (V)

= Lowest Voltage on the Graph (V)

= RMS value of the ripple (V)
The ripple is evaluated from the V
max
and V
min
read on the oscilloscope waveform (Figs. 7–18) because the
capacitor-filtered output consists of repeated charge–discharge arcs, so its AC component can be
characterized by the peak-to-peak value. The RMS value of this ripple is approximated by V
pp
/(2√3), which is
exact for a triangular waveform, and is normalized by the mean value (V
max
+ V
min
)/2 so that the result is a
ripple factor comparable with the 4.2% of the ideal six-pulse rectifier.
[34]
This approximation was checked
against that of the ideal case: for an unfiltered six-pulse output, V
max
= V
p
and V
min
= V
p
cos 30°, so Eqs. 8 and 9
give 4.15%, within approximately 1.2% (relative) of the exact value of 4.20% obtained from RF = √(FF² −
1).
[34]
Figs. 8 and 9 give 4.15%, within approximately 1.2% (relative) of the exact value of 4.20% obtained from
RF = √(FF² − 1).
[34]
For the same arc-shaped waveform, for which the waveforms in Figs. 13–18 resemble, the
true RMS value of the AC component is 0.299 V
pp
, approximately 3.6% above the triangular approximation
(0.289 V
pp
), whereas the sinusoidal approximation V
pp
/(2√2) (0.354 V
pp
) would overestimate the ripple by
approximately 22% (for example, 9.8% would become approximately 12%). Because the same method is
applied to every open-loop and closed-loop condition, the ripple values in Tables 3–8 and 10 are mutually
consistent and comparable; they remain estimates from the V
max
and V
min
readings since the recorded
waveforms were not integrated numerically. The calculation of ripple values for each test variation was
carried out according to the method applied in this study. The results of the calculations for each testing
condition are presented in the following table. The results are summarized in Table 10.
Table 10: Duty cycle variation to Ripple
Duty (%)
Ripple P-P(V)
Ripple RMS(V)
Ripple (%)
Ripple Factor 6-Pulse 6 Theory
Gap to Ceiling (pp)
30
3,2 – 4,8
0,92 – 1,39
1,4% – 2,2%
4,2%
-2,8 pp s.d. -2,0 pp
50
16,4 – 17,8
4,73 – 5,14
8,2% – 9,1%
4,2%
+4,0 pp s.d. +4,9 pp
74
16,7 – 18,6
4,82 – 5,37
8,4% – 9,6%
4,2%
+4,2 pp s.d. +5,4 pp
95
17,3 – 18,8
4,99 – 5,43
8,8% – 9,8%
4,2%
+4,6 pp s.d. +5,6 pp
The data in Table 10 indicate that the ripple increases between the 30% and 50% duty cycles (rising from
1.4%–2.2% to 8.2%–9.1%) and then tends to remain stable, although it slightly increases from the 50% to
95% duty cycles. Moreover, each ripple value at duty cycles between 50% and 95% significantly exceeds the
theoretical ripple factor for a 6-pulse rectifier (4.2%), with differences reaching 4% to 5.6%.
The same trend is shown graphically in Fig. 20: the ripple of all six loads increases from 30% to 50% of the
duty cycle and then remains nearly constant, and the overlapping lines indicate that the ripple is governed by
the duty cycle rather than by the load type or magnitude. Because the ripple at duty cycles of 50–95% is well
above 4.2% of the unfiltered six-pulse rectifier, it is attributed to the interaction between PWM switching and
the RC filter, as noted above, rather than to the rectification process itself.
Fig 20: Duty cycle variation against Ripple.
3.3. Effect of Input-Voltage Variation and Closed-Loop PI-Controlled performance
Analysis was conducted to evaluate the effects of changes in three-phase input voltages (48 V and 36 V) on the
output voltage (V
DC
) under closed-loop conditions using a proportional-integral (PI) controller with Kp and Ki
values obtained through trial-and-error tuning. The data used were average values obtained from testing
with resistive and resistive-inductive loads, and these were compared with the theoretical DC voltage of a
6-pulse rectifier. The theoretical voltage value (V
do
) can be formulated as follows.
At an input voltage condition of 48 V



   (10)
At an input voltage condition of 36 V



   (11)
With:

= Theoretical DC voltage of a 6-pulse rectifier (V)
 = Peak-to-average voltage conversion factor (constant 3√2/π)

= AC Voltage (V)
Table 3 presents the complete closed-loop test results for every combination of input voltage (48 V, 36 V),
load type (R, R-L), and load value, including the Kp and Ki gains used for each condition. The ripple values in
Table 3 are given with one decimal and are recomputed from the listed V
max
and V
min
with Eqs. 7–9 for
consistency with the open-loop tables; a comparison with the open-loop ripple is discussed in Section 3.5.
Table 11 shows that the V
max
values from the Proteus simulation closely approach the theoretical V
do
under
both input voltage (V
in
) conditions, with an error margin of only +2.19% at V
in
= 48 V and +1.28% at V
in
= 36
V. These findings are consistent with those of the duty cycle analysis in Section 3.2, which demonstrated that
V
max
is influenced primarily by the peak of the rectified waveform rather than by the control parameters
applied. In other words, regardless of the efforts of the proportional-integral (PI) controller to adjust the duty
cycle, the output voltage remains “maintained” around its theoretical limit. The average value in Table 11 is
the mean of the V
max
values of the six loads in Table 3; thus, the error expresses the deviation of V
max
from V
do
.
Table 11: Input voltage variations compared to theory
V
AC
(V)
Theory V
do
(V)
V
max
Range(V)
Average V
max
(V)
Error (%)
48
64,80
66,0 – 66,5
66,22
+2,19%
36
48,60
48,6 – 49,4
49,22
+1,28%
Table 12 shows that the ratio of V
min
to the input voltage remains nearly constant at 1.26 for V
in
= 48 V and
1.29 for V
in
= 36 V. Hence, even when the input voltage falls from 48 V to 36 V, V
min
remains proportionally
close to the peak of the rectified waveform, which indicates that the PI controller operates consistently at both
input-voltage operating points (representing higher and lower wind speeds) and avoids the large decrease in
V
min
observed under open-loop operation at higher duty cycles (Section 3.2).
Table 12: Interpretation and V
min
Ratio to Input Voltage
V
AC
(V)
Average V
min
(V)
V
min
Ratio (V
AC
)
Interpretation
48
60,27
1,26
Ripple is pressed near the natural peak of the rectifier.
36
46,60
1,29
Consistent – The PI control still keeps V
min
near the rectifier's natural peak
even when V
AC
drops.
The lines for all resistive and resistive–inductive loads nearly overlap at both input voltages: the output
voltage (V
DC
) decreases from approximately 63 V (62.95–63.45 V) at V
in
= 48 V to approximately 48 V (47.7–
48.1 V) at V
in
= 36 V. The closed-loop response therefore follows the input voltage and is not noticeably
affected by the load type or magnitude (Fig. 21).
Fig 21.: Input voltage variation to output voltage
3.4. Effect of load characteristics (Resistive vs. Resistive-Inductive Loads)
A comparison between resistive (R) and resistive-inductive (R-L) loads under open-loop duty-cycle testing at
V
in
= 48 V (Tables 4–8) reveals that the load type has only a minor effect on the ripple. At a 30% duty cycle, the
R-L loads produced a ripple of 2.0–2.2%, close to the 1.4–2.2% obtained with R loads; at duty cycles of 50–
95%, the R-L ripple (8.4–9.8%) and the R-load ripple (8.2–9.6%) were in the same range. The load currents
differed (0.04–0.25 A for R-L versus 0.06–0.15 A for R) because the load resistances differed (258–1548 Ω
versus 450–1000 Ω), but this did not result in a different ripple or V
min
trend. Hence, as concluded in Section
3.2, the open-loop ripple is governed primarily by the PWM duty cycle rather than by the type or magnitude of
the connected load.
3.5. Open-Loop and Closed-Loop comparison
Table 13 summarizes the open-loop and closed-loop results side by side. Under open-loop operation, the
voltage ripple ranged from approximately 1.4% at a 30% duty cycle to nearly 9.8% at duty cycles of 50–95%
for both R and R-L loads. After the PI controller was applied under closed-loop operation, the ripple was
reduced to approximately 2–3% at V
in
= 48 V and approximately 1–2% at V
in
= 36 V for both R and R-L loads
(see Table 3 for the full per-condition breakdown), i.e., an overall range of approximately 1–3% across the
tested input voltages and load types. This demonstrates that compared with open-loop operation, closed-loop
PI control provides a substantial and consistent improvement in the DC output quality, regardless of the load
characteristics. This comparison is limited to steady-state voltage ripple, which is consistent with the scope of
the underlying thesis, which evaluates only steady-state conditions; transient-response metrics such as rise
time, settling time, and overshoot were not computed and are not included in this study.
Table 13: Open-Loop and Closed-Loop comparison
Operating Condition
Control Mode
Ripple (%)
V
in
= 48 V, R/R-L, duty 30%
Open-loop
1.4-2.2
V
in
= 48 V, R/R-L, duty 50-95%
Open-loop
8.2-9.8
V
in
= 48 V, R/R-L
Closed-loop (PI)
2-3
V
in
= 36 V, R/R-L
Closed-loop (PI)
1-2
4. Conclusion
The test results revealed that under open-loop conditions, the maximum output voltage (V
max
) of the three-
phase rectifier tends to remain stable in the range of 64.7 V to 66.4 V, which is close to the theoretical value of
V
do
= 64.8 V, with an error margin below 2.5%. This finding indicates that V
max
is influenced primarily by the
inherent characteristics of the 6-pulse rectifier waveform rather than by changes in the duty cycle. On the
other hand, the minimum voltage (V
min
) and ripple are strongly affected by the duty cycle, as the ripple
increases rapidly from approximately 2% at a 30% duty cycle to nearly 10% at duty cycles between 50% and
95%, which is far above the theoretical limit of 4.2%. After the proportional-integral (PI) control system was
applied under closed-loop conditions, the ripple was reduced to only 1%–3% across various input voltage
conditions, indicating that, in the simulations, this control system maintained the output voltage with low
ripple across the tested conditions. These findings confirm that the objectives of this study have been
achieved. A PWM controller based on proportional-integral (PI) is designed and modeled, consisting of six
MOSFETs, an Arduino UNO, and three IR2110 driver ICs, and is tested through MATLAB/Simulink and Proteus
simulations only. The proportional-integral (PI) control system was also successfully implemented through a
voltage feedback scheme that automatically adjusts the duty cycle, and its performance was tested using
Proteus simulations covering voltage, current, ripple, and power. This study demonstrates that the quality
assessment of the PWM-based rectifier output cannot be judged solely by the peak voltage but must consider
the effect of the duty cycle on the minimum voltage and ripple. Fundamentally, these simulation results
indicate that a proportional-integral (PI) control system tuned using the trial and error method can serve as a
simple yet sufficient approach to maintaining DC voltage stability in low-speed wind power generation
systems (WPGS). This study's novelty lies in three main aspects. First, a quantitative analysis of the effect of
the PWM duty cycle on the minimum voltage and voltage ripple at the same rectifier under open-loop
conditions is performed. The simulation results revealed that the maximum voltage was relatively close to the
theoretical value of 64.8 V. In contrast, the voltage ripple increased from 1.4–2.2% at a 30% duty cycle to 8.2–
9.8% at a 50–95% duty cycle, exceeding the theoretical value of 4.2% for an ideal six-pulse rectifier. Second,
the rectifier performance under open-loop and closed-loop conditions was evaluated using PI controllers at
two input voltages, namely, 48 V and 36 V, as well as six resistive (R) and resistive–inductive (R–L) load
variations in one structured test scenario. Third, system-level simulation integration using MATLAB/Simulink
and circuit-level simulation using Proteus were used to evaluate the characteristics. The effectiveness of the
PI feedback was evaluated by comparing the open-loop and closed-loop performance at 36 V and 48 V input
voltages with R and R–L load variations. An evaluation in MATLAB/Simulink and Proteus revealed that the PI
controller reduced the voltage ripple from approximately 9–10% to 1–3% under the simulated conditions
tested. Thus, this study's main contribution is a structured quantitative analysis of the effects of the duty cycle,
input voltage variation, load characteristics, and PI control under the same rectifier configuration, providing a
basis for further development and experimental validation of rectifiers in low-speed PMSG applications.
Acknowledgment
Not applicable.
CRediT Author Contribution Statement
Ika Noer Syamsiana: Conceptualization, Methodology, Writing – Review & editing. Raditya Pratito:
Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Writing – Original
draft. Rahma Nur Amalia: Investigation, Resources, Software, Supervision, Validation. Asfari Hariz Santoso:
Investigation, Resources, Software, Supervision, Validation. Arwin Datumaya Wahyudi Sumari: Data
Curation, Project Administration, Resources, Software. All authors have read and approved the final version of
the manuscript for publication and agree to be accountable for all aspects of the work, ensuring that questions
related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Funding Declaration
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-
profit sectors.
Conflict of Interest
There is no conflict of interest.
Artificial Intelligence (AI) Use Disclosure
The authors declare that artificial intelligence (AI)-assisted tools were used only for language refinement,
grammar improvement, and manuscript structuring purposes during the preparation of this work. All
technical content, experimental implementation, results, and interpretations were independently developed
and verified by the authors.
Supporting Information
Not applicable.
References
[1]
S. Karlilar Pata, M. Balcilar, Decarbonizing energy: Evaluating fossil fuel displacement by renewables in
OECD countries, Environmental Science and Pollution Research, 2024, 31, 31304-31313, doi:
10.1007/s11356-024-33324-8.
[2]
N. A. Pambudi, R. A. Firdaus, R. Rizkiana, D. K. Ulfa, M. S. Salsabila, S. Suharno, Renewable energy in
Indonesia: current status, potential, and future development, Sustainability, 2023, 15, 2342, doi:
10.3390/su15032342.
[3]
Hingga 2030, Pemerintah Bidik Tambahan Kapasitas Terpasang PLTB 5 GW, 2024,
https://www.esdm.go.id/id/media-center/arsip-berita/hingga-2030-pemerintah-bidik-tambahan-
kapasitas-terpasang-pltb-5-gw-, Accessed 15 August 2026.
[4]
N. A. Pambudi, D. K. Ulfa, I. R. Nanda, I. M. Gandidi, A. Wiyono, M. K. Biddinika, B. Rudiyanto, L. H. Saw,
The future of wind power plants in Indonesia: Potential, challenges, and policies, Sustainability, 2025,
17, 1312, doi: 10.3390/su17031312.
[5]
J. Langer, M. Zaaijer, J. Quist, K. Blok, Introducing site selection flexibility to technical and economic
onshore wind potential assessments: New method with application to Indonesia, Renewable Energy,
2023, 202, 320–335, doi: 10.1016/j.renene.2022.11.084.
[6]
J. Langer, S. Simanjuntak, S. Pfenninger, A. J. Laguna, G. Lavidas, H. Polinder, J. Quist, H. P. Rahayu, K.
Blok, How offshore wind could become economically attractive in low-resource regions like Indonesia,
Iscience, 2022, 25, 104945, doi: 10.1016/j.isci.2022.104945.
[7]
A. Y. Megawati, F. F. Haryani, Sukarmin, Sarwanto, D. Wahyuningsih, Pujayanto, Supurwoko, An
overview of wind energy to optimize initial potential in Java, E3S Web of Conferences, EDP Sciences,
475, 12, doi: 10.1051/e3sconf/202447503001.
[8]
N. A. Elsonbaty, M. A. Enany, M. M. Elymany, Hybrid permanent magnet synchronous generator as an
efficient wind energy transducer for modern wind turbines, Wind Engineering, 2024, 48, 1203-1223,
doi: 10.1177/0309524X241256957.
[9]
A. Jain, S. Shankar, V. J. Vanitha, Power generation using permanent magnet synchronous generator
(PMSG) based variable speed wind energy conversion system (WECS): An overview, Journal of Green
Engineering, 2018, 4, 477-504, doi: 10.13052/jge1904-4720.742.
[10]
F. Aryanto, M. Mara, M. Nuarsa, Pengaruh kecepatan angin dan variasi jumlah sudu terhadap unjuk
kerja turbin angin poros horizontal, Dinamika Teknik Mesin, 2013, 3.
[11]
T. Thentral, R. Rathakrishnan, V. Anbalagan, K. Dhandapani, U. Sengamalai, P. Ramasamy, Mitigation of
current harmonics in multi-drive system, International Journal of Power Electronics and Drive Systems,
2022, 13, 113, doi: 10.11591/ijpeds.v13.i1.pp113-121.
[12]
R. Indrianto, S. Hariyadi, PENGUKURAN HARMONIK PADA PENYEARAH TIGA FASA DENGAN
MENGGUNAKAN MATLAB SIMULINK, Jurnal Penelitian: Politeknik Penerbangan Surabaya, 2020, 5,
316-323, doi: 10.46491/jp.v5i4.547.
[13]
M. Rahimi, Modeling, control and stability analysis of grid connected PMSG based wind turbine
assisted with diode rectifier and boost converter, International Journal of Electrical Power & Energy
Systems, 2017, 93, 84-96, doi: 10.1016/j.ijepes.2017.05.019.
[14]
D. B. Tchoumtcha, C. T. Dagang, G. Kenne, Synergetic control for stand-alone permanent magnet
synchronous generator driven by variable wind turbine, International Journal of Dynamics and Control,
2024, 12, 2888-2902, doi: 10.1007/s40435-024-01384-w.
[15]
A. Alili, M. B. Camara, B. Dakyo, Vienna Rectifier-Based Control of a PMSG Wind Turbine Generator,
Processes, 2022, 10, 413, doi: 10.3390/pr10020413.
[16]
A. Rajaei, M. Mohamadian, A. Yazdian, Efficiency analysis of vienna rectifier-based grid interface
systems for low-power pm wecs, Iranian Journal of Science and Technology, Transactions of Electrical
Engineering, 2019, 43, 577-586, doi: 10.1007/s40998-019-00185-8.
[17]
S. Annisa, M. F. Immanuel, S. Pratama, M. Ashari, D. P. Saragi, A Novel Simulink-Based Implementation
of Cosine Comparator Logic for Gate Control in Three-Phase Controlled Rectifiers, Techno. Com, 2025,
24, 1086, doi: 10.62411/tc.v24i3.13369.
[18]
H. Acikgoz, C. Yildiz, R. Coteli, B. Dandil, DC-link voltage control of three-phase PWM rectifier by using
artificial bee colony based type-2 fuzzy neural network, Microprocessors and Microsystems, 2020, 78,
103250, doi: 10.1016/j.micpro.2020.103250.
[19]
M. Li, X. Wu, S. Huang, G. Liang, Model predictive direct power control using optimal section selection
for PWM rectifier with reduced calculation burden, International Journal of Electrical Power & Energy
Systems, 2020, 116, 105552, doi: 10.1016/j.ijepes.2019.105552.
[20]
A. Bouafia, F. Krim, J. P. Gaubert, Design and implementation of high performance direct power control
of three-phase PWM rectifier, via fuzzy and PI controller for output voltage regulation, Energy
conversion and management, 2009, 50, 6-13, doi: 10.1016/j.enconman.2008.09.011.
[21]
R. Sanchis, I. Peñarrocha-Alós, A new method for experimental tuning of PI controllers based on the
step response, ISA transactions, 2022, 128, 329-342, doi: 10.1016/j.isatra.2021.09.008.
[22]
Z. Zhou, J. Song, Y. Yu, Q. Xu, X. Zhou, Research on high-quality control technology for three-phase PWM
rectifier, Electronics, 2023, 12, 2417, doi: 10.3390/electronics12112417.
[23]
T. Nurwati, G. Gradianto, B. Siswoyo, B. Mosaddegh, I. N. Satiawan, Discrete Implementation of a PI
Controller for Three-Phase PWM Rectifier, Jurnal EECCIS, 2024, 18, 58-62, doi:
10.21776/jeeccis.v18i2.1666.
[24]
S. Chincholkar, M. Tariq, M. Poshtan, M. Sharaf, Normalized error-based PI controller and its
application to the DC–DC buck converter, Mathematics, 2024, 12, 240, doi: 10.3390/math12020240.
[25]
L. D. M. Uc, V. M. R. Rivera, D. A. U. Sosa, B. Saldivar, Experimental Validation of PI Controller
Optimization Using LPO and PSO for a DC–DC Boost Converter., Sci, 2026, 8, 250, doi:
10.3390/sci8090250.
[26]
N. Mrabet, C. Benzazah, C. Mohssine, E. A. Ahmed, K. Driss, R. Badr, L. Ilyas, Enhancing wind energy
conversion efficiency: a novel MPPT approach using P&O with ADRC controllers versus PI controllers
with Kp and Ki optimization via genetic algorithm and ant colony optimization, Cleaner Energy
Systems, 2024, 9, 100159, doi: 10.1016/j.cles.2024.100159.
[27]
G. Shwetha, K. P. Guruswamy, An improved WOA of PI control for three phase PWM rectifier, Indonesian
Journal of Electrical Engineering and Computer Science, 2025, 38, 39-49, doi:
10.11591/ijeecs.v38.i1.pp39-49.
[28]
A. Alzahrani, Experimental Investigation of Controlled and Uncontrolled Rectifiers for Low-Power
Wind Turbines, Applied Sciences, 2023, 13, 4120, doi: 10.3390/app13074120.
[29]
T. R. S. de Freitas, P. J. M. Menegáz, D. S. L. Simonetti, Rectifier topologies for permanent magnet
synchronous generator on wind energy conversion systems: A review, Renewable and Sustainable
Energy Reviews, 2016, 54, 1334-1344, doi: 10.1016/j.rser.2015.10.112.
[30]
V. Yaramasu, A. Dekka, M. J. Durán, S. Kouro, B. Wu, PMSG‐based wind energy conversion systems:
survey on power converters and controls, IET Electric Power Applications, 2017, 11, 956-968, doi:
10.1049/iet-epa.2016.0799.
[31]
F. Blaabjerg, M. Chen, L. Huang, Power electronics in wind generation systems, Nature Reviews
Electrical Engineering, 2024, 1, 234-250, doi: 10.1038/s44287-024-00032-x.
[32]
E. Banguero, A. Correcher, Á. Pérez-Navarro, F. Morant, A. Aristizabal, A review on battery charging and
discharging control strategies: Application to renewable energy systems, Energies, 2018, 11, 1021, doi:
10.3390/en11041021.
[33]
J. Jittakort, A. Aurasopon, A digitally controlled adaptive current interface for accurate measurement of
resistive sensors in embedded sensing systems, Journal of Sensor and Actuator Networks, 2025, 14, 82,
doi: 10.3390/jsan14040082.
[34]
M. Rashid, Power Electronics: Devices, Circuits, and Applications. Pearson, 4
th
ed., 2014,
https://books.google.co.id/books?id=KRD9mAEACAAJ.
Publisher Note: The views, statements, and data in all publications solely belong to the authors and
contributors. GR Scholastic is not responsible for any injury resulting from the ideas, methods, or products
mentioned. GR Scholastic remains neutral regarding jurisdictional claims in published maps and institutional
affiliations.
Open Access
This article is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License,
which permits the non-commercial use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as appropriate credit to the original author(s) and the source is given by providing a link to the
Creative Commons License and changes need to be indicated if there are any. The images or other third-party
material in this article are included in the article's Creative Commons License, unless indicated otherwise in a
credit line to the material. If material is not included in the article's Creative Commons License and your
intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this License, visit:
https://creativecommons.org/licenses/by-nc/4.0/
© The Author(s) 2026