International Journal of Power Electronics and Drive Systems (IJPEDS)

Vol. 14, No. 3, September 2023, pp. 1748~1758

ISSN: 2088-8694, DOI: 10.11591/ijpeds.v14.i3.pp1748-1758  1748

Comparative study between PI and SMC controllers for DFIG

using fuzzy wind power estimator

Atef Jemmali, Khaled Elleuch, Hafedh Abid, Ahmed Toumi

Laboratory of Sciences and Techniques of Automatic Control and Computer Engineering (Lab-STA),
Department of Electrical Engineering, National School of Engineering of Sfax, University of Sfax, Sfax, Tunisia

Article Info ABSTRACT

Article history: In this paper, we are interested in the control of the rotor current of the
doubly feed induction generator (DFIG), that is used in wind turbines. In this
Received Jan 31, 2023 context, we present a comparative study between two types of controllers.
Revised Mar 14, 2023 The first is based on a classic PI proportional-integral control, while the
Accepted Mar 29, 2023 second is based on the SMC sliding control. The command chain uses a
fuzzy estimator of the wind power, which makes it possible to determine the
reference speed of DFIG. The system includes essentially a turbine, two
Keywords: converters, one on the rotor side and the other on the grid side, and a filter.
The grid-side converter is used to maintain a constant DC bus voltage and to
DFIG control rotor reactive power. The rotor side converter is used to control the
Doubly-fed speed of the DFIG to guarantee maximum power extracted from the turbine.
PI controllers In addition, this converter allows controlling the reactive power through the
SMC control of the rotor current according to the axis "d" of the frame linked to
Wind turbine the rotating field. The simulation results show the effectiveness of sliding
mode control (SMC) for the considered system.
This is an open access article under the CC BY-SA license.

Corresponding Author:
Atef Jemmali
Laboratory of Sciences and Techniques of Automatic Control and Computer Engineering (Lab-STA)
Department of Electrical Engineering, National School of Engineering of Sfax, University of Sfax
Sfax, Tunisia
Email: [email protected]

1. INTRODUCTION
Wind power is one of the main sources of energy used by humanity. Indeed, it is known as a clean
renewable energy and it is free. The popular development in the world and the growing needs for electrical
energy in the industrial and domestic fields has led to think of other sources of energy such as renewable
energy which has several advantages [1], [2]. After the recurrent global oil crises, the exploitation and
installation of wind energy sources are increasingly recommended and encouraged. Indeed, there are two
main configurations of wind turbines. The first configuration, which does not use static energy converters,
operates at a fixed speed. While the second configuration, which uses static converters installed between the
generator and the network, operates at variable speed. This last configuration allows wind energy to be
converted at different wind speeds.
In wind systems, several types of generators are used such as the permanent magnet synchronous
generator (PMSG), the asynchronous machine and the double-feed induction machine [3]. In the literature of
wind energy system (WES), several large-power wind turbines installed in the world use the doubly-feed
induction-generator (DFIG). This latest, presents an advantage in the production levels of electrical energy
for variable speeds, which can reach 30% of synchronism speed, hence a significant range of wind speed is
exploited [4]. The control strategy of DFIG consists of keeping the DC bus voltage constant and controlling

Journal homepage: http://ijpeds.iaescore.com

Int J Pow Elec & Dri Syst ISSN:2088-8694  1749

the rotor active power through the control of the rotor current according to the quadratic axis of the rotating
frame of reference. The DC bus is located between two converters. The last converter is followed by an
(L, r) type filter [5], [6].
The control of DFIG is based on the orientation of the flux of the stator in the two-phase reference
(d, q) according to the axis d which leads to the decoupling of the quantities between the two axes d and q of
the rotor [7]–[9]. In the literature of wind energy conversion system (WECS) using DFIG, several control
techniques are proposed. One of the best-known methods are the classic proportional integral (PI) controller
[10], [11]. A linear quadratic regulator (LQR) controller is used to control the DFIG based WES as shown in
[12]. This controller giving stability but it can lose this due to variations of parameters or operating point. So,
to obtain a robust control, a sub-optimal H∞ regulator was used. The work in [13] describes a nonlinear
model predictive control (NMPC) that is presented to a WECS based on DFIG. The strategy consists to use
torque-current control loop engender the rotor reference voltage and speed control loop that yields the
reference of the torque. The works in [14]–[16] explain how sliding mode control (SMC) is used to change
the inner current loop of the conventional vector control. The gains of control are chosen using positive semi-
definite barrier function to evade chattering effect. Fuzzy logic control constitutes an approach used by
several researchers. Indeed, this approach can give robust regulators to the change of parameters machine or
grid, the fuzzy regulators can be developed without need of the mathematical model system, it is enough to
know the dynamic behavior [17]. Despite the effectiveness of these controllers, they remain complex
especially at the implementation. SMC approach combined with fuzzy logic control is applied to control a
variable speed of WECS using DFIG [18]. The adaptive neuro-fuzzy inference system uses maximum power
point tracking controller for DFIG based WECS [19]. This approach resulted in less chatter compared to the
conventional PI controller for a wide wind range.
The main contribution consists to use a new sliding mode control, which has the same structure as
the proportional-integral controller. Also, the paper includes a comparative study between the integral
proportional controllers with that of the proposed sliding mode. A fuzzy supervisor of the T-S type is used to
predict the optimal speed of the turbine, which allows extracting the maximum energy from the wind.
This paper is organized as follows: The first part of the next section describes WECS based on
DFIG. Then, the second and third parts respectively present mathematical models of the turbine and the
DFIG. The control strategy is described in section 3. Indeed, the first part is reserved to the presentation of
the classical PI control strategy applied to the DFIG. The second part is concerned to the elaboration of
control law basis on SMC. While the third part describes the new T-S fuzzy supervisor. This latest allows
predicting, at any time and wind speed, the optimal speed for the turbine to extract the maximum energy from
wind. Discussions and simulation results are shown in section 4. Then, the paper is closed by a conclusion.

2. MODEL OFWIND ENERGY CONVERSION SYSTEM

2.1. System description
The system contains essentially a turbine, two converters and a filter. The first converter controls the
operating point of DFIG, while the second AC/DC converter is used to keep constant the DC bus voltage and
to control both the active and reactive rotor powers injected to the electrical grid. In this structure, the wind
turbine rotates with variable speed. The rotor side converter is used to control the speed of the DFIG to
guarantee maximum power extracted from the turbine. The structure of connections with the grid is displayed
in the Figure 1 [20], where RSC and GSC stand for rotor side converter and grid side converter respectively.

grid

L r
AC/DC DC/AC
DFIG (RSC) (GSC)
wind

Figure 1. General architecture of wind energy plant based on DFIG

2.2. Model of wind turbine

It is well known that a Wind Turbine transforms wind energy into mechanical energy on the shaft of
the DFIG using a speed reducer. The power at the turbine shaft can be written according to the following
formula [21], [22]:

Comparative study between PI and SMC controllers for DFIG using fuzzy wind power … (Atef Jemmali)
1750  ISSN: 2088-8694

1
𝑃𝑎 = 𝜋𝜌𝑣 3 𝑅2 𝐶𝑝 (𝜆, 𝛽) (1)
2

with:
𝑅𝛺𝑡
𝜆= (2)
𝑣

Where 𝐶𝑝 , 𝛽, 𝜆, 𝛺𝑡 , 𝑅, 𝜌 and 𝑣 define respectively power coefficient, pitch angle, tip-speed-ratio, turbine
speed, radius of the turbine, air density and wind speed. Each turbine is defined by its own power coefficient
which is limited by a maximum power called the Betz limit equal to 0.59. The following Figure 2. Shows the
evolution of 𝐶𝑝 as a function of 𝜆.

2.3. DFIG model

The most standard structures of WECS use the DFIG, as shown in Figure2. In this case, the turbine
is linked to DFIG through a gearbox [23]. By transforming the inertia and the viscous friction of the turbine
to the DFIG, the global mechanical dynamics of the rotor is written as follows:
𝑑𝛺
𝐽 + 𝑓𝛺 = 𝛤𝑎 − 𝛤𝑔 (3)
𝑑𝑡

where, J: total inertia of the wind turbine, 𝛤𝑎 : electromagnetic torque of the DFIG, f: total coefficient of
viscous friction of the turbine and 𝛤𝑔 : turbine mechanical torque on the shaft of the machine. The electrical
dynamic model of the DFIG is described in the d-q rotating synchronous frame. The equations voltage of
stator and rotor are given respectively by (4) and (5):

𝑉𝑑𝑠 = 𝑅𝑠 𝐼𝑑𝑠 + 𝜓̇𝑑𝑠 − 𝜔𝑠 𝜓𝑞𝑠

{ (4)
𝑉𝑞𝑠 = 𝑅𝑠 𝐼𝑞𝑠 + 𝜓̇𝑞𝑠 + 𝜔𝑠 𝜓𝑑𝑠

𝑉𝑑𝑟 = 𝑅𝑟 𝐼𝑑𝑟 + 𝜓̇𝑑𝑟 − (𝜔𝑠 − 𝜔𝑟 )𝜓𝑞𝑠

{ (5)
𝑉𝑞𝑟 = 𝑅𝑟 𝐼𝑞𝑟 + 𝜓̇𝑞𝑟 + (𝜔𝑠 − 𝜔𝑟 )𝜓𝑑𝑠

The 𝜔𝑠 and 𝜔𝑟 are respectively the pulsations of the stator and rotor currents. 𝐼𝑑𝑠 ,𝐼𝑞𝑠 , 𝐼𝑑𝑟 and 𝐼𝑞𝑟 represent
respectively the stator and rotor currents. 𝑅𝑠 and 𝑅𝑟 represent respectively the stator and rotor windings
resistances. The expressions of the stator and rotor fluxes are written as (6) and (7).

𝜓𝑑𝑠 = 𝐿𝑠 𝐼𝑑𝑠 + 𝑀𝑠𝑟 𝐼𝑑𝑟

{𝜓 = 𝐿 𝐼 + 𝑀 𝐼 (6)
𝑞𝑠 𝑠 𝑞𝑠 𝑠𝑟 𝑞𝑟
𝜓𝑑𝑟 = 𝐿𝑟 𝐼𝑑𝑟 + 𝑀𝑠𝑟 𝐼𝑑𝑠
{𝜓 = 𝐿 𝐼 + 𝑀 𝐼 (7)
𝑞𝑟 𝑟 𝑞𝑟 𝑠𝑟 𝑞𝑠

𝐿𝑠 , 𝐿𝑟 and 𝑀𝑠𝑟 represent respectively the inductances of the stator, the rotor and the mutual. The expression
of the electromagnetic torque is written as (8).
𝑀𝑠𝑟
𝛤𝑎 = 𝑛𝑝 (𝐼𝑑𝑟 𝜓𝑞𝑠 − 𝐼𝑞𝑟 𝜓𝑑𝑠 ) (8)
𝐿𝑠

The 𝑛𝑝 denote the pole pairs number of DFIG. The stator field is oriented along the d axis of frame. The grid
is considered stable and perfect and the stator resistance of DFIG is neglected. Therefore, we can write the
(9) and (10).

𝜓𝑑𝑠 = 𝜓𝑠 = 𝐿𝑠 𝐼𝑑𝑠 + 𝑀𝑠𝑟 𝐼𝑑𝑟

{𝜓 = 0 = 𝐿 𝐼 + 𝑀 𝐼 (9)
𝑞𝑠 𝑠 𝑞𝑠 𝑠𝑟 𝑞𝑟

𝑉𝑑𝑠 = 0
{𝑉 = 𝜔 𝜓 (10)
𝑞𝑠 𝑠 𝑑𝑠

The electromagnetic torque equation becomes (11).

Int J Pow Elec & Dri Syst, Vol. 14, No. 3, September 2023: 1748-1758
Int J Pow Elec & Dri Syst ISSN:2088-8694  1751
𝑀𝑠𝑟
𝛤𝑎 = −𝑛𝑝 𝐼𝑞𝑟 𝜓𝑠 (11)
𝐿𝑠

The equations of both powers become (12).

𝑀𝑠𝑟
𝑃𝑠 = − 𝑉𝐼
𝐿𝑠 𝑠 𝑞𝑟
{ 𝑀𝑠𝑟 𝜓𝑠 (12)
𝑄𝑠 = − 𝑉𝑠 𝐼𝑑𝑟 + 𝑉𝑠
𝐿 𝑠 𝐿𝑠

Based on (5) and (9) the rotor state equations are expressed as (13),
1
̇ =
𝐼𝑑𝑟 (−𝑅𝑟 𝐼𝑑𝑟 + 𝑉𝑑𝑟 + 𝑠𝜔𝑠 𝐿𝑟 𝜎𝐼𝑞𝑟 )
𝐿𝑟 𝜎
{ 1 𝑀𝑠𝑟 𝑉𝑠
(13)
̇ =
𝐼𝑞𝑟 (−𝑅𝑟 𝐼𝑞𝑟 + 𝑉𝑞𝑟 − 𝑠𝜔𝑠 𝐿𝑟 𝜎𝐼𝑑𝑟 − 𝑠 )
𝐿𝑟 𝜎 𝐿𝑠

2
𝑀𝑠𝑟
where 𝜎 = 1 − and s denote respectively the dispersion coefficient and the slip speed. The block diagram
𝐿𝑟 𝐿 𝑠
of a simplified DFIG model [24] can be determined by using (11), (12) and (13) and can be described by
Figure 3.

Figure 2. Variation of the power coefficient Cp

Vdr I dr +
+ 1 M srVs Qs
−
Rr + pLr Ls
+ +
ss Lr
Vs2
s Ls
ss Lr
−
Vqr np
+ 1
−
M srVs
Rr + pLr I qr Ls Ps s Ta
−
M V
s sr s
Ls

Figure 3. Simplified DFIG model block diagram

Comparative study between PI and SMC controllers for DFIG using fuzzy wind power … (Atef Jemmali)
1752  ISSN: 2088-8694

3. DFIG CONTROL
The DFIG control chain has two nested loops, one for the current and another for the speed. The
internal current loop admits a faster dynamic than that of the speed. Indeed, the quadratic current reference
value is deduced from the external velocity loop [25], [26]. The optimal reference speed is deduced from the
MPPT characteristic, which describes the evolution of power as a function of turbine speed for different
values of wind speed [27].
In this section, we propose two types of current controllers. In the first subsection, we present the
classical proportional integral controller. Whereas, in the second subsection a new sliding mode control,
which has the same architecture as the proportional-integral controller is designed. In the third and fourth
subsections are respectively reserved to the speed controller and the presentation of T-S fuzzy supervisor.

3.1. Current PI controller

The PI controller is used for linear system. In our case, the DFIG is a nonlinear system. However, it
is necessary to decouple the variables of the direct and quadrature axes. To appear the transfer functions for
both axes, the vector control is achieved by eliminating the coupling terms by compensating the term
𝑀 𝑉
(−𝑠𝜔𝑠 𝐿𝑟 𝜎𝐼𝑞𝑟 )) on the d-axis and the term (𝑠𝜔𝑠 𝐿𝑟 𝜎𝐼𝑑𝑟 + 𝑠 𝑠𝑟 𝑠 ) on the q-axis. This allows for each of the
𝐿𝑠
axes to be controlled independently of the other with its own PI controller. The procedure for controlling the
direct and quadratic rotor currents of the DFIG by the PI controller is similar. Indeed, after decoupling the
magnitudes of the two direct and quadratic axes, the current regulation block diagrams are identical. We are
content to represent a single loop whose reference and measured quantities are Idq to say that it is valid for
the both axes (d and q) as shown in Figure 4. The quadratic reference current can be computed basis on the
electromagnetic torque (11). Whereas, reference current along the direct axis is maintained constant to cancel
the reactive power according to (12).

idq − mes
idq − ref 1
Ri ( p)
+ − Lr p + Rr

Figure 4. The current loop

The current regulator equation:

𝐾𝑖𝑖
𝑅𝑖 (𝑝) = 𝐾𝑖𝑝 + (14)
𝑝

The current transfer function:

𝐾
𝑝+ 𝑖𝑖
𝑖𝑑𝑞−𝑚𝑒𝑠 (𝑝) 𝐾𝑖𝑝 𝐾𝑖𝑝
= (15)
𝑖𝑑𝑞−𝑟𝑒𝑓 (𝑝) 𝐿𝑟 𝜎 𝑝2 +𝑅𝑟+𝐾𝑖𝑝 𝑝+ 𝐾𝑖𝑖
𝐿𝑟 𝜎 𝐿𝑟 𝜎

The regulation parameters are given by:

𝐾𝑖𝑝 = 2𝜉𝜔 𝑖𝐿𝑟 𝜎 − 𝑅𝑟 (16)

𝐾𝑖𝑖 = 𝐿𝑟 𝜎𝜔𝑖2 (17)

The DFIG control block diagram using PI type controller can be illustrated by Figure 5.

3.2. Current sliding mode controller (SMC)

The SMC approach is considered as a robust control. However, to guarantee the convergence of
currents errors towards zero for all reference values, we propose in this subsection the development of a
controller basis on the theory of sliding modes. The control law includes two components, as shown in the
following [28]:

𝑢 = 𝑢𝑒𝑞 + 𝑢𝑠

Int J Pow Elec & Dri Syst, Vol. 14, No. 3, September 2023: 1748-1758
Int J Pow Elec & Dri Syst ISSN:2088-8694  1753

grid
Vds
Vqs PLL
s
L r

DFIG RSC GSC

wind
I rabc
 Vdc
 mes  s − − r MLI
2 
I rd s − −r Loop Vdc
abc abc 2
s −r
control
+ dq Grid Qr
np
 −
I rq dq
+ +
side
 mes s ss Lr s
M Vs
+ Ls Wind
ss Lr
−
I qr − ref  ref speed
I dr − ref + + PI + T-S Fuzzy
PI + PI
− − − supervisor
I rd I rq  mes

Figure 5. Block diagram of DFIG control using PI regulator

The first one is known as equivalent control term and the second is the switching control term. In
the case of DFIG, two surfaces are defined, one for the current along the direct axis and another for the
current along the quadrature axis as shown by (18):
𝑡
𝑠1 = 𝐼𝑑𝑟−𝑟𝑒𝑓 − 𝐼𝑑𝑟 + 𝑐1 ∫0 (𝐼𝑑𝑟−𝑟𝑒𝑓 − 𝐼𝑑𝑟 )𝑑𝑡
{ 𝑡 (18)
𝑠2 = 𝐼𝑞𝑟−𝑟𝑒𝑓 − 𝐼𝑞𝑟 + 𝑐2 ∫0 (𝐼𝑞𝑟−𝑟𝑒𝑓 − 𝐼𝑞𝑟 )𝑑𝑡

with 𝑐1 and 𝑐2 are two positive constants. The derivative of the sliding surface gives:
̇ + 𝑐1 (𝐼𝑑𝑟−𝑟𝑒𝑓 − 𝐼𝑑𝑟 )
𝑠̇1 = −𝐼𝑑𝑟
{ (19)
̇ + 𝑐2 (𝐼𝑞𝑟−𝑟𝑒𝑓 − 𝐼𝑞𝑟 )
𝑠̇2 = −𝐼𝑞𝑟

using (13) which gives:

𝑅𝑟 𝑉𝑑𝑟
𝑠̇1 = 𝐼 − − 𝑠𝜔𝑠 𝐼𝑞𝑟 + 𝑐1 (𝐼𝑑𝑟−𝑟𝑒𝑓 − 𝐼𝑑𝑟 )
𝐿𝑟 𝜎 𝑑𝑟 𝐿𝑟 𝜎
{ 𝑅𝑟 𝑉𝑞𝑟 𝑀𝑠𝑟𝑉𝑠 (20)
𝑠̇2 = 𝐼 − +𝑠 + 𝑠𝜔𝑠 𝐼𝑑𝑟 + 𝑐2 (𝐼𝑞𝑟−𝑟𝑒𝑓 − 𝐼𝑞𝑟 )
𝐿 𝜎 𝑞𝑟
𝑟 𝐿𝑟 𝜎 𝐿𝑠 𝐿𝑟 𝜎

the equivalent control laws along the direct and quadrature axes 𝑉𝑑𝑟−𝑒𝑞 and 𝑉𝑞𝑟−𝑒𝑞 are calculated from
𝑠̇1 = 0 and 𝑠̇2 = 0 which allows writing:

𝑉𝑑𝑟−𝑒𝑞 = (𝑅𝑟 − 𝑐1 𝐿𝑟 𝜎)𝐼𝑑𝑟 − 𝑠𝜔𝑠 𝐿𝑟 𝜎𝐼𝑞𝑟 + 𝑐1 𝐿𝑟 𝜎𝐼𝑑𝑟−𝑟𝑒𝑓

{ 𝑀𝑠𝑟 (21)
𝑉𝑞𝑟−𝑒𝑞 = (𝑅𝑟 − 𝑐2 𝐿𝑟 𝜎)𝐼𝑞𝑟 + 𝑠𝜔𝑠 𝐿𝑟 𝜎𝐼𝑑𝑟 + 𝑠 𝑉𝑠 + 𝑐2 𝐿𝑟 𝜎𝐼𝑞𝑟−𝑟𝑒𝑓
𝐿𝑠

in the sliding mode approach the reaching condition, 𝑠𝑖 𝑠̇𝑖 ≺ 0 , must be verified. So, 𝑠1 𝑠̇1 ≺ 0 and 𝑠2 𝑠̇2 ≺ 0
must be verified.

𝑅𝑟 (𝑉𝑑𝑟−𝑒𝑞 +𝑉𝑑𝑟−𝑠 )
𝑠1 𝑠̇1 = 𝑠1 ( 𝐼 − − 𝑠𝜔𝑠 𝐼𝑞𝑟 + 𝑐1 (𝐼𝑑𝑟−𝑟𝑒𝑓 − 𝐼𝑑𝑟 ))
𝐿𝑟 𝜎 𝑑𝑟 𝐿𝑟 𝜎
{ (22)
𝑅𝑟 (𝑉𝑞𝑟−𝑒𝑞 +𝑉𝑞𝑟−𝑠 ) 𝑀𝑠𝑟 𝑉𝑠
𝑠2 𝑠̇2 = 𝑠2 ( 𝐼𝑞𝑟 − +𝑠 + 𝑠𝜔𝑠 𝐼𝑑𝑟 + 𝑐2 (𝐼𝑞𝑟−𝑟𝑒𝑓 − 𝐼𝑞𝑟 ))
𝐿𝑟 𝜎 𝐿𝑟 𝜎 𝐿𝑠 𝐿𝑟 𝜎

Basis on (21) the (22) becomes as follow:

(−𝑉𝑑𝑟−𝑠 )
𝑠1 𝑠̇1 = 𝑠1
𝐿𝑟 𝜎
{ (−𝑉𝑞𝑟−𝑠 )
(23)
𝑠2 𝑠̇2 = 𝑠2
𝐿𝑟 𝜎

if we choose 𝑉𝑑𝑟−𝑠 = 𝑘1 𝑠𝑖𝑔𝑛(𝑠1 ) and 𝑉𝑞𝑟−𝑠 = 𝑘2 𝑠𝑖𝑔𝑛(𝑠2 ) with 𝑘1 ≻ 0 and 𝑘2 ≻ 0,the reaching conditions
are verified as shown (24).
Comparative study between PI and SMC controllers for DFIG using fuzzy wind power … (Atef Jemmali)
1754  ISSN: 2088-8694

𝑠1 𝑘1
𝑠1 𝑠̇1 = − 𝑠𝑖𝑔𝑛(𝑠1 ) = −𝐾1 |𝑠1 | ≺ 0
𝐿𝑟 𝜎
{ 𝑠2 𝑘2 (24)
𝑠2 𝑠̇2 = −𝑠2 𝑠𝑖𝑔𝑛(𝑠2 ) = −𝐾2 |𝑠2 | ≺ 0
𝐿 𝜎
𝑟

The expressions of the global commands along the direct and quadrature axes are given by:

𝑉𝑑𝑟 = 𝑉𝑑𝑟−𝑒𝑞 + 𝑉𝑑𝑟−𝑠 = 𝑅𝑟 𝐼𝑑𝑟 − 𝑠𝜔𝑠 𝐿𝑟 𝜎𝐼𝑞𝑟 + 𝑐1 𝐿𝑟 𝜎(𝐼𝑑𝑟−𝑟𝑒𝑓 − 𝐼𝑑𝑟 ) + 𝑘1 𝑠𝑖𝑔𝑛(𝑠1 )

{ 𝑀𝑠𝑟 𝑉𝑠 (25)
𝑉𝑞𝑟 = 𝑉𝑞𝑟−𝑒𝑞 + 𝑉𝑞𝑟−𝑠 = 𝑅𝑟 𝐼𝑞𝑟 + 𝑠𝜔𝑠 𝐿𝑟 𝜎𝐼𝑑𝑟 + 𝑠 + 𝑐2 𝐿𝑟 𝜎(𝐼𝑞𝑟−𝑟𝑒𝑓 − 𝐼𝑞𝑟 ) + 𝑘2 𝑠𝑖𝑔𝑛(𝑠2 )
𝐿𝑠

the speed regulator provides the quadratic reference current of the rotor while the reference according to the
𝜓
direct axis of rotor is equal to 𝑠 to cancel the reactive power [29]. The block diagram of DFIG control using
𝑀𝑠𝑟
SMC can be illustrated as follow Figure 6.

grid
Vds
Vqs PLL
s
L r

DFIG RSC GSC

wind
I rabc

 mes s − −r MLI
Vdc
2
I rd abc abc
 Vdc
dq dq s − −r Loop
I rq 2
control
 mes r Vrd Vrq
+ s − r Grid
np
 − I rq SMC I rd
side
Qr
s
I qr − ref
I dr − ref PI

Wind speed T-S fuzzy 

ref
supervisor + −
 mes

Figure 6. Block diagram of DFIG control using SMC

The dynamics of the current loop is very fast so that it can be neglected in front of that of the speed
loop, which makes it possible to trace the speed regulation chain in Figure 7. The speed regulator equation:
𝑘𝑖𝛺
𝑅𝛺 (𝑝) = 𝑘𝑝𝛺 + (26)
𝑝

The speed transfer function:

𝑘
𝑝+𝑘 𝑖𝛺
𝛺𝑚𝑒𝑠 (𝑝) 𝑘𝑝𝛺 𝑝𝛺
= (27)
𝛺𝑟𝑒𝑓 (𝑝) 𝐽 𝑝2 +(𝑓+𝑘𝑝𝛺 )𝑝+𝑘𝑖𝛺
𝐽 𝐽

Regulation parameters:

𝑘𝛺𝑝 = 2𝜉𝜔𝛺 𝐽 − 𝑓 (28)

𝑘𝛺𝑖 = 𝐽𝜔𝛺2 (29)

3.4. T-S fuzzy supervisor

Based on knowledge of extreme wind speeds such as (V min, Vmax) that the turbine can support, as
well as the corresponding optimal angular speeds (Ω opt1, Ωopt2) for each of the corresponding wind speeds as

Int J Pow Elec & Dri Syst, Vol. 14, No. 3, September 2023: 1748-1758
Int J Pow Elec & Dri Syst ISSN:2088-8694  1755

well as the optimal powers provided. However, it is possible to compute, exploiting a T-S fuzzy type
supervisor, the optimal velocity Ωopt and the electrical converted power Popt for each wind speed situated in
the range (Vmin, Vmax). The membership functions of the T-S fuzzy supervisor are shown by Figure 8.

𝛺𝑟𝑒𝑓 = 𝛺𝑜𝑝𝑡 = 𝑀1 𝛺𝑜𝑝𝑡1 + 𝑀2 𝛺𝑜𝑝𝑡2 (30)

𝑃𝑟𝑒𝑓 = 𝑃𝑜𝑝𝑡 = 𝑁1 𝑃𝑜𝑝𝑡1 + 𝑁2 𝑃𝑜𝑝𝑡2 (31)

3 3
𝑉 3 −𝑉𝑚𝑖𝑛 3
𝑉𝑚𝑎𝑥
Where, 𝑁1 = 33
and 𝑁2 = 33
describe the member ship functions for converted power. P opt1 and Popt2
𝑚𝑖𝑛
𝑉𝑚𝑎𝑥 𝑚𝑖𝑛
𝑉𝑚𝑎𝑥
define the optimum powers for the extremes wind speed.

 ref  mes
R ( p) 1
+ − Jp + f

Figure 7. The speed loops Figure 8. Membership functions for wind

4. SIMULATION RESULTS AND DISCUSSION

To check the performance of the two control techniques: PI control and SMC control, we used
MATLAB/Simulink software. The parameters of the DFIG and the turbine are indicated by the Table 1. The
wind speed variation is trapezoidal between 6 m/s and 13 m/s to 9 m/s as shown in Figure 9. It gives the
DFIG speed variation between 80 rad/s and 170 rad/s to 116 rad/s. The Figures 10 to 15 show respectively
speed variation, quadratic and direct rotor currents evolution, three phases rotor currents, stator power
variation and the electromagnetic torque variation.
In Figure 10 the speed follows its reference determined by the T-S fuzzy supervisor. According to
the Figure 11, Figure 12 and Figure 13, it is noted that the PI control method generates oscillations at the
level of the rotor currents, which is not the case for the SMC control method. To avoid the phenomenon of
chattering for the control technique by SMC the switching function has been changed by the saturation
function and with an adequate choice of its constant ki. The effect of rotor current oscillations resulting to the
PI controller affects both powers injected to the network as shown by Figure 14 and the electromagnetic
torque Figure 15.

Table 1. Parameters wind turbine

Value Parameters Value Parameters
2 Mw DFIG power 2.5 mH Mutual inductance
0.0026 Ω Resistance of stator 2 Number of pole pairs
0.0029 Ω Resistance of rotor 75 Gear reduction ratio
2.587 mH Inductance of stator 40 Radius of turbine
2.587 mH Inductance of rotor

Figure 9. Wind speed variation Figure 10. Speed variation

Comparative study between PI and SMC controllers for DFIG using fuzzy wind power … (Atef Jemmali)
1756  ISSN: 2088-8694

Figure 11. Quadratic rotor current evolution Figure 12. Direct rotor current evolution

Figure 13. Three phase rotor current Figure 14. Stator active power variation

Figure 15.Electromagnetic torque variation

5. CONCLUSION
This paper describes two types of controllers namely the popular classical controller PI and the
SMC for a WECS using the DFIG. Although the PI-based control strategy looks simply, it has severe
disadvantages that make it ineffective for various applications. In addition, a T-S type fuzzy supervisor is
developed to predict the optimal turbine speed and power that it can convert from the wind. The presented
simulation results prove that SMC could be an attractive solution for a WECS exploiting the DFIG. However,
the chattering phenomenon keep an important problem with this approach. The results demonstrate that the

Int J Pow Elec & Dri Syst, Vol. 14, No. 3, September 2023: 1748-1758
Int J Pow Elec & Dri Syst ISSN:2088-8694  1757

SMC strategy is more interesting and efficient in terms of maximizing power conversion, compared to the
traditional PI technique.

REFERENCES
[1] H. Abdelli, A. Mezouar, M. Bendjebbar, and K. Belgacem, “Synthesis of smc algorithms applied to wind generator,”
International Journal of Power Electronics and Drive Systems, vol. 12, no. 1, pp. 404–412, 2021, doi:
10.11591/ijpeds.v12.i1.pp404-412.
[2] G. Abad, J. López, M. A. Rodríguez, L. Marroyo, and G. Iwanski, “Doubly Fed Induction Machine: Modeling and Control for
Wind Energy Generation,” in Doubly Fed Induction Machine: Modeling and Control for Wind Energy Generation, 2011. doi:
10.1002/9781118104965.
[3] A. Tounsi and H. Abid, “Generator and grid side converter control for wind energy conversion system,” International Journal of
Power Electronics and Drive Systems, vol. 12, no. 3, pp. 1832–1844, 2021, doi: 10.11591/ijpeds.v12.i3.pp1832-1844.
[4] K. Ouari, T. Rekioua, and M. Ouhrouche, “Nonlinear model predictive controller of a variable speed wind turbine driven doubly
fed induction generator,” Journal of Electrical Systems, vol. 9, no. 2, pp. 243–255, 2013.
[5] R. Benadli, B. Khiari, and A. Sellami, “Three-phase grid-connected photovoltaic system with maximum power point tracking
technique based on voltage-oriented control and using sliding mode controller,” 2015 6th International Renewable Energy
Congress, IREC 2015, 2015, doi: 10.1109/IREC.2015.7110963.
[6] K. Noussi, A. Abouloifa, H. Katir, I. Lachkar, and F. Giri, “Nonlinear control of grid-connected wind energy conversion system
without mechanical variables measurements,” International Journal of Power Electronics and Drive Systems, vol. 12, no. 2, pp.
1139–1149, 2021, doi: 10.11591/ijpeds.v12.i2.pp1139-1149.
[7] X. Lin, K. Xiahou, Y. Liu, and Q. H. Wu, “Design and Hardware-in-the-Loop Experiment of Multiloop Adaptive Control for
DFIG-WT,” IEEE Transactions on Industrial Electronics, vol. 65, no. 9, pp. 7049–7059, 2018, doi: 10.1109/TIE.2018.2798566.
[8] R. Mahroug, M. Matallah, and S. Abudura, “Modeling of wind turbine based on dual DFIG generators,” International Journal of
Power Electronics and Drive Systems, vol. 13, no. 2, pp. 1170–1185, 2022, doi: 10.11591/ijpeds.v13.i2.pp1170-1185.
[9] A. Boyette, “Control-command of a double-fed induction generator with storage system for wind power generation,” Université
Henri Poincaré-Nancy I, 2006.
[10] H. Abu-Rub, M. Malinowski, and K. Al-Haddad, “Power Electronics for Renewable Energy Systems, Transportation and
Industrial Applications,” Power Electronics for Renewable Energy Systems, Transportation and Industrial Applications, vol.
9781118634035, pp. 1–795, 2014, doi: 10.1002/9781118755525.
[11] A. S. L. V. Tummala, “A robust composite wide area control of a DFIG wind energy system for damping inter-area oscillations,”
Protection and Control of Modern Power Systems, vol. 5, no. 1, 2020, doi: 10.1186/s41601-020-00170-y.
[12] K. M. Haneesh and T. Raghunathan, “Robust Control of DFIG Based Wind Energy System Using an H∞ Controller,” Journal of
Electrical Engineering and Technology, vol. 16, no. 3, pp. 1693–1707, 2021, doi: 10.1007/s42835-021-00699-4.
[13] L. L. Rodrigues, J. Sebastian Solis-Chaves, O. A. C. Vilcanqui, and A. J. S. Filho, “Predictive Incremental Vector Control for
DFIG with Weighted-Dynamic Objective Constraint-Handling Method-PSO Weighting Matrices Design,” IEEE Access, vol. 8,
pp. 114112–114122, 2020, doi: 10.1109/ACCESS.2020.3003285.
[14] A. S. L. V. Tummala, “Modified vector controlled DFIG wind energy system based on barrier function adaptive sliding mode
control,” Protection and Control of Modern Power Systems, vol. 4, no. 1, 2019, doi: 10.1186/s41601-019-0119-3.
[15] M. I. Martinez, G. Tapia, A. Susperregui, and H. Camblong, “Sliding-mode control for DFIG rotor- and grid-side converters
under unbalanced and harmonically distorted grid voltage,” IEEE Transactions on Energy Conversion, vol. 27, no. 2, pp. 328–
339, 2012, doi: 10.1109/TEC.2011.2181996.
[16] S. Boubzizi, H. Abid, A. El hajjaji, and M. Chaabane, “Comparative study of three types of controllers for DFIG in wind energy
conversion system,” Protection and Control of Modern Power Systems, vol. 3, no. 1, 2018, doi: 10.1186/s41601-018-0096-y.
[17] H. Karimi-Davijani, A. Sheikholeslami, H. Livani, and M. Karimi-Davijani, “Fuzzy Logic Control of Doubly Fed Induction
Generator Wind Turbine 1 1,” World Applied Sciences Journal, vol. 6, no. 4, pp. 499–508, 2009, doi: 10.5281/zenodo.4308924.
[18] D. Kairous and R. Wamkeue, “DFIG-based fuzzy sliding-mode control of WECS with a flywheel energy storage,” Electric Power
Systems Research, vol. 93, pp. 16–23, 2012, doi: 10.1016/j.epsr.2012.07.002.
[19] A. A. Chhipa et al., “Adaptive neuro‐fuzzy inference system‐based maximum power tracking controller for variable speed wecs,”
Energies, vol. 14, no. 19, 2021, doi: 10.3390/en14196275.
[20] Y. Majdoub, A. Abbou, and M. Akherraz, “High-performance MPPT control of DFIG with optimized flux reference in presence
of nonlinear magnetic characteristic,” International Journal of Power Electronics and Drive Systems, vol. 13, no. 2, pp. 1195–
1208, 2022, doi: 10.11591/ijpeds.v13.i2.pp1195-1208.
[21] M. W. K. Mbukani and N. Gule, “PLL-Based Sliding Mode Observer Estimators for Sensorless Control of Rotor-Tied DFIG
Systems,” IEEE Transactions on Industry Applications, vol. 55, no. 6, pp. 5960–5970, 2019.
[22] B. Beltran, M. E. H. Benbouzid, and T. Ahmed-Ali, “Second-order sliding mode control of a doubly fed induction generator
driven wind turbine,” IEEE Transactions on Energy Conversion, vol. 27, no. 2, pp. 261–269, 2012, doi:
10.1109/TEC.2011.2181515.
[23] S. Mensou, A. Essadki, T. Nasser, and B. Bououlid Idrissi, “A direct power control of a DFIG based-WECS during symmetrical
voltage dips,” Protection and Control of Modern Power Systems, vol. 5, no. 1, 2020, doi: 10.1186/s41601-019-0148-y.
[24] M. N. Musarrat and A. Fekih, “A Fault-Tolerant Control Framework for DFIG-Based Wind Energy Conversion Systems in a
Hybrid Wind/PV Microgrid,” IEEE Journal of Emerging and Selected Topics in Power Electronics, vol. 9, no. 6, pp. 7237–7252,
Dec. 2021, doi: 10.1109/JESTPE.2020.3034604.
[25] S. Ghosh, Y. J. Isbeih, R. Bhattarai, M. S. El Moursi, E. F. El-Saadany, and S. Kamalasadan, “A Dynamic Coordination Control
Architecture for Reactive Power Capability Enhancement of the DFIG-Based Wind Power Generation,” IEEE Transactions on
Power Systems, vol. 35, no. 4, pp. 3051–3064, 2020, doi: 10.1109/TPWRS.2020.2968483.
[26] S. M. Tripathi, A. N. Tiwari, and D. Singh, “Optimum design of proportional-integral controllers in grid-integrated PMSG-based
wind energy conversion system,” International Transactions on Electrical Energy Systems, vol. 26, no. 5, pp. 1006–1031, 2016,
doi: 10.1002/etep.2120.
[27] H. M. Seoudy, M. A. Saadeldin, and W. A. Mohamed, “Design and implementation of optimal controller for DFIG-WT using
autonomous groups particle swarm optimization,” International Journal of Power Electronics and Drive Systems, vol. 13, no. 3,
pp. 1813–1821, 2022, doi: 10.11591/ijpeds.v13.i3.pp1813-1821.

Comparative study between PI and SMC controllers for DFIG using fuzzy wind power … (Atef Jemmali)
1758  ISSN: 2088-8694

[28] H. Abid, “Takagi-Sugeno Fuzzy Controller and Sliding Mode Controller for a Nonholonomic Mobile Robot,” Mathematical
Problems in Engineering, vol. 2021, 2021, doi: 10.1155/2021/7703165.
[29] S. Sharma, J. P. Mishra, and S. Datta, “Sliding mode power control of a DFIG based variable speed wind energy conversion
system,” 12th IEEE International Conference Electronics, Energy, Environment, Communication, Computer, Control: (E3-C3),
INDICON 2015, 2016, doi: 10.1109/INDICON.2015.7443335.

BIOGRAPHIES OF AUTHORS

Atef Jemmmali obtained the Master degree in automatic, from higher school of
sciences and techniques of Tunisia 2009. He is currently member of Laboratory of Sciences
and Techniques of Automatic Control and Computer Engineering (Lab-STA) in National
School of Engineering of Sfax, University of Sfax, Tunisia. He is currently PhD student in
National School of Engineering of Sfax. His research interests include development of control
approaches for hybrid renewable energy systems. He can be contacted at email:
[email protected].

Khaled Elleuch obtained the M.S degree on January 1996 in electrical

engineering from the ESSTT of Tunis, Tunisia. He is member of Laboratory of Sciences and
Techniques of Automatic and Computer Engineering (Lab-STA) in National Engineering
School of Sfax, University of Sfax. He obtained the Ph.D. degree from National Engineering
School of Sfax ENIS, Tunisia on June 2010. Currently, he is an associate professor in the
Higher Institute of Biotechnology of Sfax ISBS, Tunisia. His researches focus on identification
and control of nonlinear systems, photovoltaic energy and wind energy. He can be contacted at
email: [email protected].

Hafedh Abid had an electrical engineering diploma from the National School of
Engineering of Sfax, Tunisia in 1989, then a diploma in Electrical and Electronic from the
High School of Technical Sciences of Tunis in 1995 and in 1996 the Aggregation in Electric
Genius. From 1996 to 2006, he was a ‘Technologies’ Teacher of the Electric Department in
High Institute of Technologies of Sousse. He is working now as a Professor at National School
of Engineering of Sfax, University of Sfax, Tunisia. His current research interests are fuzzy
systems and photovoltaic systems. He is the author and co-author of numerous national and
international publications. He can be contacted at email: [email protected].

Ahmed Toumi is an electrical engineer from the Sfax Engineering National

School, he has the DEA in instrumentation and Measurement from University of Bordeaux,
France in 1981 and the Thesis degree from the University of Tunis in 1985. He joined the
National Engineering School as an Associate Professor, since 1981. In 2000, he had the
University Habilitation (HDR) from the Engineering School of Sfax, Tunisia. He was working
as a director of the Electrical Engineering Department in ENIS. Study and research areas touch
modeling, stability of electric machines and electrical networks. He can be contacted at email:
[email protected].

Int J Pow Elec & Dri Syst, Vol. 14, No. 3, September 2023: 1748-1758