IEEE Global Energy Conference (GEC) 2022, October 26-29, 2022, Batman, Turkey 282

Advanced Control Techniques for Doubly-Fed

Induction Generators Based Wind Energy
Conversion Systems
Hamid Chojaa Aziz Derouich Seif Eddine Chehaidia
Laboratory of Technologies and Laboratory of Technologies and Industrial Mechanics Laboratory
Industrial Services Industrial Services Badji Mokhtar Annaba University
Higher School of Technology, USMBA Higher School of Technology, USMBA Box 12, Annaba 23000, Algeria
Fez, Morocco Fez, Morocco [0000-0003-3475-6503]
[0000-0001-9408-4687] [0000-0003-4598-5002]
Mohamed I. Mosaad
Othmane Zamzoum Yanbu Industrial College (YIC)
Mohammed Taoussi Laboratory of Technologies and Alnahdah, Yanbu Al Sinaiyah, Yanbu
Laboratory of Technologies and Industrial Services 46452, Saudi Arabia
Industrial Services Higher School of Technology, USMBA [0000-0002-6531-5455]
Higher School of Technology, USMBA Fez, Morocco
Fez, Morocco [0000-0001-9339-3963]
[0000-0002-8425-0677]
Mourad Yessef
Habib Benbouhenni Laboratory of Engineering, Modeling
Research Unit on Embedded Systems, and Systems Analysis
Research Center on Scientific and Faculty of sciences Dhar El Mahraz,
Technical Information CERIST, USMBA
Algiers, Algeria Fez, Morocco
[0000-0001-8253-4863] [0000-0001-6344-7174]

Abstract— Due to their advantages especially the capability DFIG, it is recommended to use control strategies that allow
of operation at wide wind speed ranges, doubly-fed induction the control of the generated power by the DFIG, namely: the
generators (DFIGs) are widely used in wind energy conversion active power that will be controlled at the reference power
2022 Global Energy Conference (GEC) | 978-1-6654-9751-0/22/$31.00 ©2022 IEEE | DOI: 10.1109/GEC55014.2022.9987088

systems (WECSs). This paper introduces a performance
enhancement of WECS-DFIG by extracting the maximum
power from WECS-DFIG in addition to active and reactive
powers control. The maximum power point is extracted by
capturing the optimal speed by an intelligent controller. While
Integral Sliding Mode Controller (ISMC) and the Field
Oriented Control (FOC_PI) are presented for active and
reactive powers control. The proposed controllers were applied
to control a 1.5 grid-connected DFIG. All simulations were done
using Matlab Software.
Keywords— ANNC, Doubly Fed Induction Generator,
Integral Sliding Mode Controller (ISMC), Wind Energy
Conversion System (WECS).
I. INTRODUCTION
Due to the huge demand for electrical power and the
environmental concerns of global warming, many renewable
sources of energy are integrated into the electrical grids
recently [1]. Among these sources PV, wind, fuel cells, and
biomass[2-5]. Wind energy is one of the dominant source of
generated by the turbine to ensure better efficiency, and the
reactive power that will be kept at zero in order to keep a
UPF. Fig. 1 presents the schematic diagram of this system.
Currently, AI based controllers (Human expertise or data
driven approaches) and nonlinear control techniques are
widely used to improve the DFIG performances [3-6, 8-10,
12]. In [10], an intelligent Fuzzy Gain Scheduling of PI
controller is proposed, the obtained results shown the
effectiveness of the fuzzy tuning of P and I gains, however
the tuning of fuzzy rules and membership function is too
much time taken and requires high human expertise. In [12]
a comparative PI Neural Network controller is proposed and
compared to conventional Sliding Mode and Backstepping
controller.
Transformer Grid
Wind Turbine RSC GSC
Gearbox
V DFIG
Lf Rf

renewable energy as 24% of the renewable sources is from β

t g Filter

wind [6, 7]. This extensive use of wind energy is due to its
various advantages, although the nonlinearities and wind PMW PMW
speed dependence [8]. This wind speed dependence requires
a generator that has the capability of operation under a ABC ABC
dq dq
variable speed range; Double-Fed Induction Generator P
Vdq Vdq

(DFIG) [9]. Thus, it is essential to have a controller that can g Ps*

Power factor
control
track the maximum power by maximum power point tracking Cem*
ISMC
Vdc_mes
technique (MPPT) whatever the wind speed [1, 10]. Using V
MPPT-ANN Vdc* +
_
PI Qref=0
DFIG is an interesting solution because of its rigid structure,
long life, and the ability to control active as well as reactive Fig. 1. Diagram of the system.
powers with variable speed [6, 11]. In order to have a high
quality of electrical power generated by a WECS based on

978-1-6654-9751-0/22/$31.00 ©2022 IEEE

Authorized licensed use limited to: Consortium - Algeria (CERIST). Downloaded on December 30,2022 at 01:12:36 UTC from IEEE Xplore. Restrictions apply.
IEEE Global Energy Conference (GEC) 2022, October 26-29, 2022, Batman, Turkey 283

The envisaged technic was the training of network based on A. MPPT Technique
error measurement. Despite its effectiveness the major The MPPT method seeks to improve the performance of
drawback of their aforementioned paper is the use of 30 WECS by extracting the maximum wind energy. The speed
neurons is the hidden layer, which increases the overfitting controller aims to control the speed by regulating the
phenomenon. Regarding variable structure controllers, the electrical power of the generator according to the λopt by
nonlinear sliding mode control (SMC) approach is one of the forcing the torque of the electromagnetic generator to be
most widely used in DFIG control due to its robustness and equal to its desired reference value, as. 𝐶𝑒𝑚−𝑟𝑒𝑓 = 𝐶𝑒𝑚 .
disturbance rejection. As one can see in [11] the use of Turbine Gearbox High Speed Shaft
Integral Sliding Mode Control (ISMC) improves the steady
λ RΩ
state system response’s, In [7] Integral Terminal Sliding β t t 1 g

V G
Mode control (ITSMC) is proposed. The main disadvantage
of ITSMC is the important number of tunable parameters, Cp
which complicates the optimization of controller. However,
the chattering phenomenon is the main drawback of this Pt Ct
1
Tem
1
+_ g
Ωt G Js + Fv
controller. In a recent previous work [8], a comparative study
Tem_ref
between serval nonlinear control strategies was presented. V
Conducted simulations show that Adaptive Backstepping
Controller performs well compared to SMC and PI. In [13], λoptV g_ref

G +_ Controller
Super-Twisting Sliding Mode Control based Artificial Neural R Control speed
(STSMC-ANN) is proposed. The synergitisation of ANN g

with variable structure controller increase the machine

performance, even in presence of disturbance and
Fig. 2. MPPT method.
uncertainties, however, the controller is highly complex
which requires a lot of calculation and increases the cost of
The electromagnetic reference torque 𝐶𝑒𝑚−𝑟𝑒𝑓 allows a
implementation in field wind turbines.
rotational speed to be equal to its reference value 𝛺ref that is
In order to achieve the control objective of wind turbine
obtained at the speed controller output (Fig.2).
working under low wind speed conditions, an intelligent
strategy based on the neural network is proposed in this work. B. ANN controller
Where ANN controller is used at the turbine level, in order to ANN is considered as a very performant technique for
improve the precision of the reference electromagnetic torque modeling complex nonlinear systems [14]. Since WECS is a
communicated to the DFIG. In the machine side ISMC complex nonlinear system, presenting high nonlinearities, the
controller is used. As a summary, this paper presents some use of ANN could overcome the nature of the system by
improvements in the performance of WECS-DFIG through mapping collected inputs and outputs data sets due to its high
three aspects: computational capability. Fig.3 shows the internal structure
of the proposed MPPT controller. It is composed of the input
 MPPT to track the maximum power by controlling the layers with two neurons that represent both mechanical and
electromagnetic torque of the DFIG. wind speeds, and the output layers with one neuron that
 Regulating the active and reactive power by ISMC in represents the reference electromagnetic torque. After
order to overcome the problems uncounted when network optimization, 20 neurons in hidden layer is used
using FOC-PI controller. based on obtained error between measured and reference
electromagnetic torque.
This article is structured as follows: Firstly, the turbine is
modeled and a MPPT with speed control based an ANN
controller is presented. Then, section 3 introduces the
dynamic modelling of the DFIG. Subsequently, section 4
examines the proposed ISMC. Section 5 is reserved for a
comparative analysis of the ISMC with the FOC-PI to control
the DFIG. Finally, section 6 summarizes the conclusion.
II. WECS MODEL
The expression of the aerodynamic power is given by:
1
𝑃𝑎𝑒𝑟 = 𝐶𝑝 𝑃𝑉 = 𝜌𝜋𝑅2 𝑉 3 𝐶𝑝 (𝜆, 𝛽) (1) Fig. 3. Structure of the ANN.
2

In this research paper, the Cp (λ,β) variations are modeled C. DFIG model
by the following equations: The dynamic equation of DFIG can be expressed as follows
[2-6]:
116 −21
Cp(𝜆, 𝛽) = 0 . 5 ( − 0.4𝛽 − 5) 𝑒𝑥𝑝 ( ) + 068 𝜆
𝜆𝑖 𝜆𝑖
1 1 0.035
𝜆𝑖
=
𝜆+0.08𝛽
−
𝛽 3 +1
(2)
𝛺𝑡 𝑅
{ 𝜆=
𝑉

Authorized licensed use limited to: Consortium - Algeria (CERIST). Downloaded on December 30,2022 at 01:12:36 UTC from IEEE Xplore. Restrictions apply.
IEEE Global Energy Conference (GEC) 2022, October 26-29, 2022, Batman, Turkey 284

𝑑 𝑇1
𝑉𝑠𝑑 = 𝑅𝑠 𝑖𝑠𝑑 + 𝜙𝑠𝑑 − 𝜔𝑠 𝜙𝑠𝑞 𝑉 = 𝑆𝑑𝑞 𝑆𝑑𝑞 (12)
𝑑𝑡 2
𝑑
𝑉𝑠𝑞 = 𝑅𝑠 𝑖𝑠𝑞 + 𝜙𝑠𝑞 + 𝜔𝑠 𝜙𝑠𝑑
𝑑𝑡
(3) The derivative of (12) yields:
𝑑
𝑉𝑟𝑑 = 𝑅𝑟 𝑖𝑟𝑑 + 𝜙𝑟𝑑 − (𝜔𝑠 − 𝜔𝑟 )𝜙𝑟𝑞 𝑆𝑑
𝑑𝑡
𝑑 𝑉̇ = ( 𝑖̇̇ − 𝑓1 − 𝛼𝑑 𝑠𝑖𝑔𝑛(𝑒𝑑 (𝑡)))
{ 𝑉𝑟𝑞 = 𝑅𝑟 𝑖𝑟𝑞 + 𝑑𝑡
𝜙𝑟𝑞 + (𝜔𝑠 − 𝜔𝑟 )𝜙𝑟𝑑 𝜎𝐿𝑟 𝑟𝑑−𝑟𝑒𝑓
𝑆𝑞
+ ( 𝑖̇̇ − 𝑓2 − 𝛼𝑞 𝑠𝑖𝑔𝑛 (𝑒𝑞 (𝑡)))
𝜎𝐿𝑟 𝑟𝑞−𝑟𝑒𝑓
3 𝑆𝑑 𝑆𝑑 𝑆𝑞 (13)
⃗𝑠 × 𝐼𝑠∗ } = 3 (𝑉𝑠𝑑 𝑖𝑠𝑑 + 𝑉𝑠𝑞 𝑖𝑠𝑞 )
𝑃𝑠 = 𝑅𝑒{𝑉 = 𝛾𝑑 − 𝛼𝑑 𝑠𝑖𝑔𝑛(𝑒𝑑 (𝑡)) + 𝛾
{ 2
3
2
3 (4) 𝜎𝐿𝑟 𝜎𝐿𝑟 𝜎𝐿𝑟 𝑞
⃗𝑠 × 𝐼𝑠∗ } = (𝑉𝑠𝑞 𝑖𝑠𝑑 − 𝑉𝑠𝑑 𝑖𝑠𝑞 )
𝑄𝑠 = 𝐼𝑚{𝑉 𝑆𝑞
2 2
− 𝛼 𝑠𝑖𝑔𝑛 (𝑒𝑞 (𝑡))
𝜎𝐿𝑟 𝑞
III. APPLICATION OF THE ISMC
IV. RESULTS
The SMC strategy is a nonlinear control method, known
by its robustness even in presence of dynamic or parametric To examine the proposed control performance, the
uncertainties, thanks to its variable structure, where system validation test is realized in Matlab/Simulink environment on
rich the sliding surface and keep it whatever the operating 1.5 MW WECS-DFIG. The comparative study based on
point is. However, the drawback of SMC that limits the serval criterion is summarized in Table 1 The numerical
application of this control is the chattering problem [7, 9, 14]. parameters of the system are given in Tables 2 and 3,
To confront this problem, an ISMC strategy is developed in presented in the Appendix section.
this paper [15, 16]. A. Maximum power point tracking
𝑋̇ = 𝑓(𝑥, 𝑡) + 𝑔(𝑥, 𝑡)𝑢 (5) The proposed MPPT control technique will be investigated
under wind speed variation shown in Fig. 4. As the wind
where: speed varies, the proposed controller tried to keep the value
𝑅 𝑀𝑉𝑠𝑞 𝑅𝑟 𝑀𝜓𝑠𝑑 of 𝐶𝑝 (𝜆, 𝛽) at a constant or semi-constant value of 0.48,
− (𝜎𝐿𝑟 )2 (𝜎𝐿𝑟 𝑖𝑟𝑑 + )+ + which represents the maximum desired value when the TSR
𝑟 𝜔𝑠 𝐿𝑆 𝐿𝑆 (𝜎𝐿𝑟 )2
𝜔𝑟 𝑀 reached its optimum desired value of 8.2, Fig. 5 and 6. The
(𝜎𝐿𝑟 𝑖𝑟𝑞 + 𝑉 )
𝜎𝐿𝑟 𝜔𝑠 𝐿𝑠 𝑠𝑑 proposed ANN-MPPT succeed in tracing the reference rotor
𝑓(𝑥, 𝑡) = (6)
𝑅 𝑀𝑉
− (𝜎𝐿𝑟 )2 (𝜎𝐿𝑟 𝑖𝑟𝑞 + 𝑠𝑑) speed perfectly as depicted in Fig. 7.
𝑟 𝜔 𝐿 𝑠 𝑠
𝑅𝑟 𝑀𝜓𝑠𝑞 𝜔𝑟 𝑀
[ + 𝐿𝑠(𝜎𝐿𝑟 )2 + 𝜎𝐿𝑟 (𝜎𝐿𝑟 𝑖𝑟𝑑 + 𝜔𝑠 𝐿𝑠 𝑉𝑠𝑞 ) ]

1
0
𝜎𝐿𝑟
𝑔(𝑥, 𝑡) = [ 1 ] (7)
0
𝜎𝐿𝑟

with 𝑋 = [𝑖𝑟𝑑 𝑖𝑟𝑞 ]. The switching variables and their

derivatives are defined by:
Fig. 4. Wind speed profile (m/s).
𝑆𝑑 = 𝑒𝑑 (𝑡) + 𝛼𝑑 𝑒𝐼 (𝑡)
{ (8)
𝑆𝑞 = 𝑒𝑞 (𝑡) + 𝛼𝑞 𝑒𝐼 (𝑡)

𝑆𝑑̇ = 𝑒𝑑̇ (𝑡) + 𝛼𝑑 𝑒̇ 𝐼 (𝑡) = 0

{ ̇ (9)
𝑆𝑞 = 𝑒𝑞̇ (𝑡) + 𝛼𝑞 𝑒̇ 𝐼 (𝑡) = 0

Where: 𝑒𝑑𝑞 (𝑡) = 𝑖𝑟𝑑𝑞 − 𝑖𝑟𝑑𝑞−𝑟𝑒𝑓 , 𝛼𝑑𝑞 are constants

coefficients, and 𝑒𝐼̇ (𝑡) = 𝑠𝑖𝑔𝑛 (𝑒𝑑𝑞 (𝑡)).

𝑖̇̇ 𝑟𝑑 − 𝑖̇̇ 𝑟𝑑−𝑟𝑒𝑓 + 𝛼𝑑 𝑒̇ 𝐼 (𝑡) = 0 Fig. 5. Aerodynamic power coefficient Cp ( 𝜆, 𝛽).

{ (10)
𝑖̇̇ 𝑟𝑞 − 𝑖̇̇ 𝑟𝑞−𝑟𝑒𝑓 + 𝛼𝑞 𝑒̇ 𝐼 (𝑡) = 0

Substituting (5) into (10) yields:

𝑈𝑟𝑑 𝑖̇̇ 𝑟𝑑−𝑟𝑒𝑓 − 𝑓1 − 𝛼𝑑 𝑠𝑖𝑔𝑛(𝑒𝑑 (𝑡))

[𝑈 ] = 𝑔−1 [ ] (11)
𝑟𝑞 𝑖̇̇ 𝑟𝑞−𝑟𝑒𝑓 − 𝑓2 − 𝛼𝑞 𝑠𝑖𝑔𝑛 (𝑒𝑞 (𝑡))

We define the Lyapunov function below:

Fig. 6. Tip Speed Ratio TSR (rad/s).

Authorized licensed use limited to: Consortium - Algeria (CERIST). Downloaded on December 30,2022 at 01:12:36 UTC from IEEE Xplore. Restrictions apply.
IEEE Global Energy Conference (GEC) 2022, October 26-29, 2022, Batman, Turkey 285

(b)
Fig. 7. Rotor Speed (rad/s). Fig. 9. (a): Stator active Ps, (b): Reactive power Qs.

B. Active and reactive powers control

Fig.8 and 9 show also that the currents irq and ird, the active
power and reactive powers are tracking their reference
perfectly for both control strategies. However, the ISMC
strategy has the best performance, namely a minimum
overshoot, fewer oscillations, and fast dynamic response.
Fig. 10 shows that the stator currents under ISMC are the best
sinusoidal form compared to the FOC_PI currents. From Fig. (a)
11, we can see that the total harmonic distortion (THD) of the
stator current is under FOC_PI is 0.91% while under ISMC
is to 0.38%. Note that the THD is reduced when ISMC is
used. As in Fig. 12, the DC link voltage achieves its reference
value perfectly with minimum fluctuations for the ISMC
strategy compared to FOC-PI strategy.

(b)
Fig. 10. Stator current isabc: (a) FOC_PI, (b) ISMC.

(a)

(a)

(b)
Fig. 8. Generator current, (a): ird and (b): irq.

(b)
Fig. 11. THD for the current isa: (a) ISMC and (b) FOC_PI.

Table 1 represents the comparison study between the FOC

used PI controller and ISMC strategy in terms of set point
(a) tracking precision, the response time, the static error, and the
phase current isa THD. This study reveals good improvements
by ISMC. These improvements incorporate the minimization
of the harmonics in the stator current signals and response
time optimization.

Authorized licensed use limited to: Consortium - Algeria (CERIST). Downloaded on December 30,2022 at 01:12:36 UTC from IEEE Xplore. Restrictions apply.
IEEE Global Energy Conference (GEC) 2022, October 26-29, 2022, Batman, Turkey 286

Stator Resistance;Rs 1.2.10-2 (Ω)

Rotor Resistance; Rr 2.1.10-2 (Ω)
Stator Inductance; Ls 1.37.10-2 H)
Rotor Inductance; Lr 1.36.10-2 (H)
Mutual Inductance, M 0.0135 (H)
Number of Pair of Poles, p 2
DC Bus and RL Filter
Fig. 12. Dc bus voltage. DC-Link Capacitor; C 8×10-3 (F)
Filter Inductance; Lf 0.005 (H)
TABLE 1. COMPARATIVE STUDY BETWEEN THE FOC-PI AND THE ISMC.
Performance FOC-PI ISMC Filter Resistance; Rf 0.012 (Ω)
Response Time (ms) 401.8 295
Static Error (%) 4.44 1.61
REFERENCES
Set-Point Tracking Good Very good
Precision Medium High [1] T. Burton, N. Jenkins, D. Sharpe, and E. Bossanyi, Wind
Current 𝒊𝒔𝒂 THD (%) 0.91 0.38 energy handbook: John Wiley & Sons, 2011.
[2] S. Liaquat, M. S. Fakhar, S. A. R. Kashif, A. Rasool, O.
Saleem, and S. Padmanaban, "Performance analysis of
CONCLUSION APSO and firefly algorithm for short term optimal
scheduling of multi-generation hybrid energy system,"
This research work focuses on improving the WTs IEEE Access, vol. 8, pp. 177549-177569, 2020.
dynamic performance of the WECS based on a DFIG. The [3] G. S. Chawda, A. G. Shaik, M. Shaik, S. Padmanaban, J.
comparison study between the FOC-PI controller and the B. Holm-Nielsen, O. P. Mahela, et al., "Comprehensive
ISMC had been carried out under the same scenario review on detection and classification of power quality
conditions. The mathematical model equations of the two disturbances in utility grid with renewable energy
controllers and the simulation results have been given in penetration," IEEE Access, vol. 8, pp. 146807-146830,
detail. 2020.
The MPPT method with the control of speed based on the [4] W. S. Hassanein, M. M. Ahmed, M. O. abed el-Raouf, M.
intelligent control ANNC is also developed in this work to G. Ashmawy, and M. I. Mosaad, "Performance
ensure that the power coefficient Cp can be set at its desired improvement of off-grid hybrid renewable energy system
maximum value. Therefore, the maximum energy can be using dynamic voltage restorer," Alexandria Engineering
captured. Journal, vol. 59, pp. 1567-1581, 2020.
The results show that the use of the ISMC control strategy [5] H. S. Ramadan, S. Padmanaban, and M. I. Mosaad,
has improved the DFIG dynamic performances. In addition, "Metaheuristic-based Near-Optimal Fractional Order PI
this result shows the superiority of ISMC compared to the Controller for On-grid Fuel Cell Dynamic Performance
Enhancement," Electric Power Systems Research, vol.
FOC-PI in terms of dynamic response, the reference tracking,
208, p. 107897, 2022.
precision, THD of the stator currents and the static error. The
[6] W. Strielkowski, L. Civín, E. Tarkhanova, M.
main advantage of this strategy is its good insensitivity to Tvaronavičienė, and Y. Petrenko, "Renewable energy in
parametric uncertainties and disturbances. the sustainable development of electrical power sector: A
For future work, the implementation of the studied control review," Energies, vol. 14, p. 8240, 2021.
strategy will be realized on the dSPACE DS1104 board. [7] S. E. Chehaidia, H. Kherfane, H. Cherif, B. Boukhezzar,
L. Kadi, H. Chojaa, et al., "Robust Nonlinear Terminal
APPENDIX Integral Sliding Mode Torque Control for Wind Turbines
Considering Uncertainties," IFAC-PapersOnLine, vol.
TABLE 2. THE TURBINE PARAMETERS 55, pp. 228-233, 2022.
Parameters Value [8] H. Chojaa, A. Derouich, M. Taoussi, S. E. Chehaidia, O.
Nominal Power Zamzoum, M. I. Mosaad, et al., "Nonlinear Control
1.5 (MW)
Strategies for Enhancing the Performance of DFIG-Based
Blade Radius; R 35.25 (m) WECS under a Real Wind Profile," Energies, vol. 15, p.
Gearbox Gain; G 90 6650, 2022.
[9] M. I. Mosaad, N. A. Sabiha, A. Abu-Siada, and I. B. Taha,
Moment of Inertia; 𝐽𝑡 103 (Kg.m2) "Application of Superconductors to Suppress
Friction Coefficient; 𝑓𝑣 2.4.10-2 (N.m.s-1) Ferroresonance Overvoltage in DFIG-WECS," IEEE
Transactions on Energy Conversion, vol. 37, pp. 766-777,
2021.
[10] S. E. CHEHAIDIA, A. ABDEREZZAK, H.
TABLE 3: THE DFIG PARAMETRES
KHERFANE, N. GUERSI, H. CHERIF, and B.
BOUKHEZZAR, "Fuzzy gain scheduling of PI torque
Parameters Value controller to capture the maximum power from variable
Rated Power; Pn 1.5 (MW) speed wind turbines," in 2020 IEEE 2nd International
Conference on Electronics, Control, Optimization and
Stator Rated Voltage;Vs 698 (V) Computer Science (ICECOCS), 2020, pp. 1-6.
Stator Rated Frequency; f 50 (Hz)

Authorized licensed use limited to: Consortium - Algeria (CERIST). Downloaded on December 30,2022 at 01:12:36 UTC from IEEE Xplore. Restrictions apply.
IEEE Global Energy Conference (GEC) 2022, October 26-29, 2022, Batman, Turkey 287

[11] H. Chojaa, A. Derouich, S. E. Chehaidia, O. Zamzoum,

M. Taoussi, and H. Elouatouat, "Integral sliding mode
control for DFIG based WECS with MPPT based on
artificial neural network under a real wind profile,"
Energy Reports, vol. 7, pp. 4809-4824, 2021.
[12] A. Nouriani and H. Moradi, "Variable speed wind turbine
power control: A comparison between multiple MPPT
based methods," International Journal of Dynamics and
Control, vol. 10, pp. 654-667, 2022.
[13] I. Sami, S. Ullah, S. U. Amin, A. Al-Durra, N. Ullah, and
J.-S. Ro, "Convergence enhancement of super twisting
sliding mode control using artificial neural network for
DFIG based wind energy conversion systems," IEEE
Access, 2022.
[14] R. Abdelli, D. Rekioua, T. Rekioua, and A. Tounzi,
"Improved direct torque control of an induction generator
used in a wind conversion system connected to the grid,"
ISA transactions, vol. 52, pp. 525-538, 2013.
[15] C. Hamid, A. Derouich, M. Taoussi, O. Zamzoum, and A.
Hanafi, "An improved performance variable speed wind
turbine driving a doubly fed induction generator using
sliding mode strategy," in 2020 IEEE 2nd International
Conference on Electronics, Control, Optimization and
Computer Science (ICECOCS), 2020, pp. 1-8.
[16] H. Chojaa, A. Derouich, M. Taoussi, O. Zamzoum, and
M. Yessef, "Optimization of DFIG wind turbine power
quality through adaptive fuzzy control," in International
Conference on Digital Technologies and Applications,
2021, pp. 1235-1244.

Authorized licensed use limited to: Consortium - Algeria (CERIST). Downloaded on December 30,2022 at 01:12:36 UTC from IEEE Xplore. Restrictions apply.