European Journal of Electrical Engineering

Vol. 21, No. 5, October, 2019, pp. 457-464

Journal homepage: http://iieta.org/journals/ejee

Direct Power Control of a Wind Turbine Based on Doubly Fed Induction Generator
Ibrahim Yaichi1*, Abdelhafid Semmah1, Patrice Wira2
1
Electrical Engineering Dept., University of Djillali Liabes, Sidi Bel Abbes 022000, Algeria
2
IRIMAS Laboratory, University of Haute Alsace, Mulhouse 68093, France

Corresponding Author Email: [email protected]

https://doi.org/10.18280/ejee.210508 ABSTRACT

Received: 26 April 2019 This paper proposes a direct power control (DPC) strategy for the doubly-fed induction
Accepted: 19 August 2019 generator (DFIG) operating under variable wind speeds. Under this strategy, the active and
reactive powers of the DFIG are directly controlled by the AC/DC converter, whose switching
Keywords: states were selected from a switch table. Besides, a two-level hysteresis corrector was selected
to control the active and reactive powers, ensuring the dynamic performance of the DFIG. The
Pulse Width Modulation (PWM), Doubly
effectiveness of the proposed DPC strategy was compared through MATLAB simulation with
Fed Induction Generator (DFIG), Field
the field oriented control (FOC), a classical proportional-integral (PI) control strategy for wind
Oriented Control (FOC), Direct Power
turbines. The results show that the DPC strategy adjusted the instantaneous active and reactive
Control (DPC)
powers in the grid perfectly with respect to their references, and realized the absorption of
sinusoidal currents with a unity power factor. The proposed DPC strategy has a great
application potential in wind power generation.

1. INTRODUCTION Neghouchi in 1998, for the direct control of the active and
reactive instantaneous powers of the three-phase PWM
Faced with the problems posed by fossil fuels and nuclear rectifier without network voltage sensors [2].
fission, the first and best possible answer would be to save The principle of direct control was proposed by Noguchi et
energy and use it sparingly, avoiding wasting it. But man al. [1] and it was developed later for several applications. The
cannot do without her. This is why it must obligatorily develop goal was to eliminate the modulation block and internal loops
the already existing means of substitution and look for new by replacing them with a switch board whose inputs are the
ones. These means of substitution of which one speaks, it is of errors between the reference values and the measurements [2].
course the "renewable energy". Among these, wind energy This type of command considers the converter associated
clearly appears prominently, not as a replacement for with the machine as a set where the control vector is
conventional sources, but as extra energy. Indeed, the potential constituted by the switching states. Its main advantages are the
energy of moving air masses represents a considerable deposit speed of the dynamic torque response and the low dependence
worldwide. This energy offers two great advantages, since it is on machine parameters [3, 4].
totally clean and renewable. The proposed DPC technique reduces current and power
In the field of high power drives, there is a new and original ripple. The results of the proposed controller systems
solution, using an alternative machine operating in a particular simulation (FOC) and (DPC) are compared for different step
mode. This is the Doubly Fed Induction Generator (DFIG), it changes in the active/reactive power. The common goal of
is a three-phase asynchronous machine with a wound rotor that these controls was to ensure the sampling of sinusoidal
can be powered by two voltage sources, one to the stator and currents while guaranteeing a unit power factor with
the other to the rotor. decoupled control of the active and reactive powers.
The vector control (more specifically the one with The effectiveness of these techniques is verified by
orientation of the stator flux) will allow us to realize a control simulation under Matlab/ Simulink.
independent of the active and reactive power of the DFIG, by This DPC control has the advantages over vector control
using classical regulators of Proportional-Integral (PI) type. (FOC) to solve the problem of fluctuations in power and
The purpose of the vector control is to control the current generated by the DFIG on the power grid.
asynchronous machine as an independently excited direct
current machine where there is a natural decoupling between The power converter Grid side

the magnitude controlling the flux (the excitation current) and

that related to the torque (the armature current) [1]. Wind
AC/DC DC/AC

To circumvent the sensitivity problems to parametric

variations, we have developed control techniques in which the DFIG B us
DC
Power
Grid
active and reactive powers are estimated from the only
electrical quantities accessible to the rotor. Among these
The power converter
methods, direct power control (DPC) appeared to be Ωmec

competitive with the vector control technique. The first

configuration of this type of control was proposed by T. Figure 1. DFIG based wind turbine scheme

457
2. MODELING OF THE WIND TURBINE SYSTEM

The diagram of the DFIG wind turbine connected to the grid,

including the various mechanical and electrical quantities used
to model the electromechanical conversion chain, is shown in
Figure 1 [5].

2.1 Model of the wind

The characteristics of the wind will determine the amount

of energy that can actually be extracted from the wind farm [6].
The wind speed will be modeled in deterministic form by a
sum of several harmonics:

𝑉𝑣 = 8 + 0.2 𝑠𝑖𝑛(0.1047. 𝑡) + 2 𝑠𝑖𝑛(0.2665. 𝑡) +

0.2 𝑠𝑖𝑛(3.6645. 𝑡) (1)
Figure 2. Coefficient of power of DFIG
Wind energy comes from the kinetic energy of the wind.
Indeed, if we consider a mass of air, m, which moves with The multiplier adapts the speed (slow) of the turbine to the
velocity v, the kinetic energy of this mass is given by: speed of the generator (Figure 3). This multiplier is modeled
mathematically by the following equations [12]:
1
𝐸𝑐 = 𝑚𝑣 2 (2)
2

If, during the unit of time, this energy could be completely

recovered by means of a propeller that sweeps a surface 𝑆,
located perpendicularly to the direction of the wind speed, the
instantaneous power supplied would be, then:
1
𝑃𝑣 = 𝜌𝑆𝑣 3 (3)
2
Figure 3. Diagram of the wind turbine
The Betz limit characterizes the ability of the wind turbine
to capture wind energy. The corresponding power is given by 𝐶𝑎𝑒𝑟
𝐶𝑔 = (8)
[7]: 𝐺

Ω𝒎𝒆𝒄
1
𝑃𝑡 = 𝐶𝑝 (𝛽, 𝜆) 𝜌𝜋𝑅2 𝑣 3 (4) Ω𝑡𝑢𝑟𝑏𝑖𝑛𝑒 = (9)
𝑮
2

The maximum power that can be collected by a wind turbine With: 𝐺: Gain of the speed multiplier.
is calculated by the Betz limit: The mass of the wind turbine is transferred to the turbine
shaft in the form of an inertia Jturbine and comprises the mass
1 of the blades and the rotor mass of the turbine.
𝑃𝑡𝑚𝑎𝑥 = 𝐶𝑝𝑚𝑎𝑥 (𝛽, 𝜆) 𝜌𝜋𝑅2 𝑣 3 = 0.593𝑃𝑣 (5)
2
𝐽𝑡𝑢𝑟𝑏𝑖𝑛𝑒
𝐽= + 𝐽𝑔 (10)
Thus, the notion of aerodynamic efficiency 𝜂 of the wind 𝐺2
turbine can be defined by the ratio:
The fundamental equation of the dynamics makes it
𝐶𝑝 𝐶𝑝 possible to determine the evolution of the mechanical speed
𝜂= = (6)
𝐶𝑝𝑚𝑎𝑥 0.539 from the total mechanical torque (𝐶𝑚𝑒𝑐 ) applied to the rotor:

𝑑Ω𝑚𝑒𝑐
2.2 Power coefficient 𝐽 = 𝐶𝑚𝑒𝑐 (11)
𝑑𝑡

The power coefficient 𝐶𝑝 depends on the number of blades J: is the total inertia that appears on the rotor of the generator.
of the rotor and their geometrical and aerodynamic shapes This mechanical torque takes into account, the
(length and profile of the sections). These are designed electromagnetic torque Cem produced by the generator, the
according to the characteristics of the site, the desired nominal viscous friction couple 𝐶𝑓 , and the torque 𝐶𝑔 .
power, the type of regulation (in pitch or stall) and the type of
The mechanical torque applied to the rotor:
operation (fixed or variable speed) [8, 9].
Numerical approximations have been developed in the 𝑑Ω𝑚𝑒𝑐
literature to model the coefficient 𝐶𝑝 and different expressions 𝐽 = 𝐶𝑚𝑒𝑐 = 𝐶𝑔 − 𝐶𝑒𝑚 − 𝐶𝑓 (12)
𝑑𝑡
have been proposed [10, 11].
The resistant torque due to friction is modeled by a
 ( + 0.1) coefficient of viscous friction f:
C p (  ,  ) = (0.5 − 0.067(  − 2)) sin[
18.5 − 0.3(  + 2) (7)
𝐶𝑓 = 𝑓 Ω𝑚𝑒𝑐 (13)
−0.00184( − 3)(  − 2)]

458
 𝑑𝜑𝑑𝑠
0.5 𝑉𝑑𝑠 = 𝑅𝑠 𝐼𝑑𝑠 + − 𝜔𝑠 𝜑𝑞𝑠
𝑑𝑡
𝑑𝜑𝑞𝑠
0.45 𝑉𝑞𝑠 = 𝑅𝑠 𝐼𝑞𝑠 + − 𝜔𝑠 𝜑𝑑𝑠
𝑑𝑡
𝑑𝜑𝑑𝑟 (20)
0.4 𝑉𝑑𝑟 = 𝑅𝑟 𝐼𝑑𝑟 + − (𝜔𝑠 − 𝜔𝑟 )𝜑𝑞𝑟
𝑑𝑡
𝑑𝜑𝑞𝑟
𝑉𝑞𝑟 = 𝑅𝑟 𝐼𝑞𝑟 + − (𝜔𝑠 − 𝜔𝑟 )𝜑𝑑𝑟
Cp (N.m)

0.35
𝑑𝑡
{ ,
0.3

0.25 Expression of active and reactive power:

0.2 𝑃𝑠 = 𝑉𝑑𝑠 𝐼𝑑𝑠 + 𝑉𝑞𝑠 𝐼𝑞𝑠

{ (21)
𝑄𝑠 = 𝑉𝑞𝑠 𝐼𝑑𝑠 − 𝑉𝑑𝑠 𝐼𝑞𝑠
0 0.2 0.4 0.6 0.8 1
Time (s)
𝑃𝑟 = 𝑉𝑑𝑟 𝐼𝑑𝑟 + 𝑉𝑞𝑟 𝐼𝑞𝑟
{ (22)
Figure 4. Power coefficient of DFIG 𝑄𝑟 = 𝑉𝑞𝑟 𝐼𝑑𝑟 − 𝑉𝑑𝑟 𝐼𝑞𝑟

3. MODELLING OF THE DOUBLY FED INDUCTION 5. FIELD ORIENTED CONTROL STRATEGY (FOC)
GENERATOR
The purpose of the vector control is to control the
The DFIG has three stator equations and three rotor asynchronous machine as an independently excited DC
equations: machine where there is a natural decoupling between the
magnitude controlling the flux (the excitation current) and that
𝑑𝜑𝑟𝑎𝑏𝑐 related to the torque (the armature current). This decoupling
[𝑉𝑠𝑎𝑏𝑐 ] = −[𝑅𝑠 ][𝐼𝑠𝑎𝑏𝑐 ] +
{ 𝑑𝑡
(14) provides a very fast torque response, a large speed control
𝑑𝜑𝑠𝑎𝑏𝑐
[𝑉𝑟𝑎𝑏𝑐 ] = [𝑅𝑟 ][𝐼𝑟𝑎𝑏𝑐 ] + range and high efficiency for a large steady state load range
𝑑𝑡
[13].
[𝜑 ] = −[𝐿𝑠𝑠 ][𝐼𝑠𝑎𝑏𝑐 ] + [𝑀𝑠𝑟 ][𝐼𝑠𝑎𝑏𝑐 ] Examination of the expression of the torque of the machine
{ 𝑠𝑎𝑏𝑐 (15) shows that it results from a difference between two quadrature
[𝜑𝑟𝑎𝑏𝑐 ] = −[𝑀𝑠𝑟 ][𝐼𝑟𝑎𝑏𝑐 ] + [𝐿𝑟𝑟 ][𝐼𝑟𝑎𝑏𝑐 ]
components of the stator current and rotor flux which has a
complex coupling between the machine magnitudes. The
The four-inductor matrices are: working reference for the control is that related to the rotating
field so that the axis (𝑑) coincides with the direction desired
𝐿𝑠 𝑀𝑠 𝑀𝑠 𝐿𝑟 𝑀𝑟 𝑀𝑟
flow, which can be rotor, stator, or air gap. Thus, it is possible
[𝐿𝑠𝑠 ] = [𝑀𝑠 𝐿𝑠 𝑀𝑠 ] ;[𝐿𝑟𝑟 ] = [𝑀𝑟 𝐿𝑟 𝑀𝑟 ] (16)
to orient the different flows of the machine [14, 16].
𝑀𝑠 𝑀𝑠 𝐿𝑠 𝑀𝑟 𝑀𝑟 𝐿𝑟
The stator flux with the conditions:
𝐿𝑠 and 𝐿𝑟 are respectively the stator and rotor inductors. 𝜑 = 𝜑𝑑𝑠 = 𝜑𝑠
𝑀𝑠 and 𝑀𝑟 are respectively the mutual stator and rotor { 𝜑 =0 (23)
𝑞𝑠
inductances.
The rotor flow with the conditions:
[𝑀𝑠𝑟 ] = [𝑀𝑟𝑠 ]𝑇 =
4𝜋
𝑐𝑜𝑠 𝜃 𝑐𝑜𝑠(𝜃 −
3
) 𝑐𝑜𝑠( 𝜃 − 2𝜋/3) 𝜑 = 𝜑𝑑𝑟 = 𝜑𝑟
4𝜋
{ 𝜑 =0 (24)
𝑀 𝑐𝑜𝑠(𝜃 − 2𝜋/3) 𝑐𝑜𝑠 𝜃 𝑐𝑜𝑠(𝜃 − ) (17) 𝑞𝑟
3
4𝜋
[ 𝑐𝑜𝑠(𝜃 − 3
) 𝑐𝑜𝑠( 𝜃 − 2𝜋/3) 𝑐𝑜𝑠 𝜃 ] For a DFIG, the relation of the electromagnetic torque is
given by the following equation:
The dynamic equation is:
3 𝐿𝑚
𝐶𝑒𝑚 = 𝑝(𝜑𝑞𝑠 𝑖𝑑𝑟 − 𝜑𝑑𝑠 𝑖𝑞𝑟 ) (25)
𝑑Ω 2 𝐿𝑠
𝐽 = 𝐶𝑒𝑚 − 𝐶𝑟 − 𝑓Ω (18)
𝑑𝑡
Starting from Eq. (25), a decoupling can be performed in
such a way that the torque will be controlled solely by the
4. MATHEMATICAL EQUATIONS OF THE DFIG IN current iqr and thus the flow by the current 𝑖𝑑𝑟 . The final
THE PARK REFERENCE (d-q) relationship of the couple is:

Magnetic equations: 3 𝐿𝑚
𝐶𝑒𝑚 = − 𝑝(𝜑𝑠 𝑖𝑞𝑟 ) (26)
2 𝐿𝑠
𝜑𝑑𝑠 = 𝐿𝑠 𝐼𝑑𝑠 + 𝑀 𝐼𝑑𝑟
𝜑𝑞𝑠 = 𝐿𝑠 𝐼𝑞𝑠 + 𝑀 𝐼𝑞𝑟 𝑉𝑑𝑠 = 0
(19) {𝑉 = 𝑉 = 𝜔 𝜑 (27)
𝜑𝑑𝑟 = 𝐿𝑟 𝐼𝑑𝑟 + 𝑀 𝐼𝑑𝑠 𝑞𝑠 𝑠 𝑠 𝑠

{ 𝜑𝑞𝑟 = 𝐿𝑟 𝐼𝑞𝑟 + 𝑀 𝐼𝑞𝑠

𝜑𝑠 = 𝐿𝑠 𝐼𝑑𝑠 + 𝑀 𝐼𝑑𝑟
{0=𝐿 𝐼 +𝑀𝐼 (28)
Stator and rotor voltages: 𝑠 𝑞𝑠 𝑞𝑟

459
 We get the following expressions for the active and reactive direct torque control applied to electrical machines. The goal
powers: is to directly control active and reactive power in a PWM
rectifier. The errors between the reference values of the
𝐿𝑚
𝑃𝑠 = −𝑉 𝐼 instantaneous active and reactive powers and their
𝐿 𝑠 𝑞𝑟
{ (29) measurements are introduced in two hysteresis comparators
𝜑 𝐿
𝑄𝑠 = 𝑉𝑠 𝑠 − 𝑉𝑠 𝑚 𝐼𝑑𝑟 which determine the switching state of the semiconductors,
𝐿𝑠 𝐿𝑠
with the help of a switchboard and the value of the sector
𝑉𝑠 where the voltage of the generator [17].
By drawing: 𝜑𝑠 = from equation (27), the expression of
𝜔𝑠
the reactive power becomes: 6.1 General principle of DPC

𝑉𝑠2 𝑉𝑠 𝐿𝑚 The overall structure of the DPC, using a predefined

𝑄𝑠 = − 𝐼𝑑𝑟 (30)
𝐿𝑠 𝜔𝑠 𝐿𝑠 switching table, applied to the three-phase machine-side
converter is shown in Figure 5. It is similar to that of the direct
The expression of the powers (29) can thus be simplified in torque control (DTC). Instead of torque and rotor flux, it is the
the following way: active and reactive stator power that is the controlled quantities.
𝑀
𝑃 = −𝑉𝑠 𝑖𝑞𝑟 The principle of the DPC consists in selecting a sequence of
𝐿𝑠
{ 𝑀 𝑉2
(31) switching commands (Sa, Sb, Sc) of the semiconductors
𝑄 = −𝑉𝑠 𝑖𝑑𝑟 + 𝑠 constituting, from a switching table. The selection is made on
𝐿𝑠 𝐿𝑠 𝜔𝑠
the basis of the errors (𝜀𝑃𝑠 and 𝜀 Q s ) between the references of
the active and reactive powers ( Ps∗ and Q∗s ) and the real
6. DIRECT POWER CONTROL OF DFIG values(𝑃𝑠 and Q s ), provided by two comparators to hysteresis
of digitized outputs HP and HQ respectively, as well as on the
Direct Power Control (DPC) is based on the concept of sector (zone) in which the vector of the rotor flux is [18, 19].

Transformer

Filter

DC Grid
Hp Switching table
εPs
Ps-ref
+ Si DFIG
- Hq
εQs
Qs-ref
+
-
Z(i)
Ir
Ps Qs α/β
Estimator: Ps, Qs ir-αβ
Vs
Sector
Vs-αβ a,b,c

Figure 5. DPC configuration of the DFIG

In order to achieve a switching table providing simultaneous hysteresis comparators, the measured values of the powers
control of the active and reactive powers, in all sectors, it is being estimated from the following relationships [21, 22]:
essential to study the variations caused by the application of
3 𝐿𝑚
each of the control vectors on the latter, and this at the same 𝑃𝑠 = − 𝑉𝜑
2 𝜎𝐿𝑠 𝐿𝑟 𝑠 𝑟𝛽
time. During a complete period of rotor voltage. The control { (32)
3 𝑉𝑠 𝑉𝑠 𝐿𝑚
vectors selected in this switching table must ensure the 𝑄𝑠 = ( 𝜑 − 𝜑 )
2 𝜎𝐿𝑠 𝑠 𝜎𝐿𝑠 𝐿𝑟 𝑟𝛼
restriction of the reference tracking error of the two active and
reactive powers, simultaneously. From where,
6.2 Estimation of active and reactive power 𝐿𝑚
𝜑𝑟𝛼 = 𝜎𝐿𝑟 𝑖𝑟𝛼 + 𝜑𝑠
𝐿𝑠
Instead of measuring the powers on the network, capturing 𝜑𝑟𝛽 = 𝜎𝐿𝑟 𝑖𝑟𝛽
the rotor currents, and estimate 𝑃𝑠 and 𝑄𝑠 . This approach gives ⃗⃗⃗⃗𝑠 |
|𝑉 (33)
early control of the powers in the stator windings. That is, the |𝜑
⃗⃗⃗⃗𝑠 | =
𝜔𝑠
one established by neglecting the resistance of the stator phase. 𝐿2𝑚
We can find the relations of Ps and Q s according to the two { 𝜎=1−
𝐿𝑠 𝐿𝑟
components of the rotor flux in the reference frame (𝛼𝑟 − 𝛽𝑟 )
[19, 20, 23]. If, by introducing the angle δ between the stator and rotor
The active and reactive powers are controlled by two flux vector, 𝑃𝑠 and 𝑄𝑠 become:

460
 3 𝐿𝑚
𝑃𝑠 = − 𝜔 |𝜑 ||𝜑𝑟 | 𝑠𝑖𝑛 𝛿 Table 1. Optimal vector selection table [21]
2 𝜎𝐿𝑠 𝐿𝑟 𝑠 𝑠
{ 3 𝜔𝑠 𝐿 (34)
𝑄𝑠 = |𝜑𝑠 |( 𝑚 |𝜑𝑟 | 𝑐𝑜𝑠 𝛿 − |𝜑𝑠 |) Rp 1 -1
2 𝜎𝐿𝑠 𝐿𝑟
Rq 1 0 -1 1 0 -1
1 V5 V7 V3 V6 V0 V2
The derivative of the two equations in (34) gives: Rotor 2 V6 V0 V4 V1 V7 V3
flux 3 V1 V7 V5 V2 V0 V4
𝑑𝑃𝑠 3 𝐿𝑚 𝑑|𝜑 | 𝑠𝑖𝑛 𝛿
=− 𝜔 |𝜑 | 𝑟 sector 4 V2 V0 V6 V3 V7 V5
𝑑𝑡 2 𝜎𝐿𝑠 𝐿𝑟 𝑠 𝑠 𝑑𝑡
{ 𝑑𝑄 3 𝐿𝑚 𝜔𝑠 𝑑(|𝜑𝑟 | 𝑐𝑜𝑠 𝛿)
(35) 5 V3 V7 V1 V4 V0 V6
𝑠
= |𝜑𝑠 | 6 V4 V0 V2 V5 V7 V1
𝑑𝑡 2 𝜎𝐿 𝐿 𝑠 𝑟 𝑑𝑡

As can be seen in (23), these last two expressions show that Nevertheless, the fundamental reality on which the DPC is
the active and reactive stator powers can be controlled by the based is that the movement of the rotor flux in the machine
modification of the relative angle δ between the stator and follows a continuous progression in time and that it will seem
rotor flux vectors and their amplitudes. to cross each sector one by one if it is sampled sufficiently.
The study in Table 2 indicates that if the rotor flux was for
6.3 Choice of hysteresis comparators example in sector 2 and the vector 𝑉3 had just been applied,
the variation of the reactive power measured at the stator must
In order to obtain very good dynamic performances, the inevitably be negative since the vector 𝑉3 decreases the
choice of a two-level hysteresis corrector seems to be the reactive power to the stator. If this had not been the case, we
simplest and best solution for controlling the active and would have to admit that our estimate of the sector is no longer
reactive power. These comparators (Figure 6) must make it fair and that the flow would rather be in sector 1 or 5. Table 3
possible to control the exchange of the active and reactive presents the signs of the variations of the instantaneous active
power between the DFIG and the power grid in both directions and reactive powers for each input voltage vector of the
and with the two modes of hypo and hyper-synchronous rectifier according to the sector where the voltage of the
operation of the DFIG. generator is located. By choosing the appropriate output vector,
it is possible to select the signs of variation of the active and
reactive powers independently.
Hp
Hq
Table 2. DFIG parameters
+1 +1
ԑPs Parameter name Symbol Value Unit
-ΔPs 0 +ΔQs ԑQs
-ΔQs Rated power Pn 1.5 MW
+ΔPs 0 Rated current In 1900 A
-1 Rated DC-Link voltage UDC 1200 V
-1
Stator rated voltage Vs 398/690 V
Stator rated frequency f 50 Hz
Rotor inductance Lr 0.0136 H
Figure 6. Hysteresis comparators: active power, reactive Stator inductance Ls 0.0137 H
power Mutual inductance M 0.0135 H
Rotor resistance Rr 0.021 Ω
Similar to DTC, the DPC for DFIG is based on the selection Stator resistance Rs 0.012 Ω
of a rotor voltage vector such that errors between measured Number of pair of poles p 2 -
and reference quantities are reduced and maintained between
the limits of the hysteresis bands. Table 3. Wind turbine parameters

6.4 Switching table Parameter name Symbol Value Unit

The power coefficient Cp 𝑚𝑎𝑥 0.48 -
Rotor radius R 35.25 m
To select the optimum rotor voltage vector, it is necessary Speed multiplier gain G 90 -
to know the relative position of the rotor flux in the six sextants. The density of the air 𝜌 1.225 kg/m3
A three-phase inverter with two voltage levels can produce Moment of total inertia J 1000 Kg.m2
eight different combinations; these eight combinations Viscous friction fr 0.0024 N.m.s-1
generate eight voltage vectors that can be applied to the rotor coefficient
terminals of the DFIG.
The division of the complex plane into six angular zones
𝑍𝑖 (i = 1.....6) can be determined by the following relation: 7. SIMULATION RESULTS
𝜋 𝜋 𝜋 𝜋 The stator current generated by the DFIG which has a
− + (𝑖 − 1) ≤ 𝑍𝑖 < + (𝑖 − 1) (36)
6 3 6 3
sinusoidal shape (Figure 7 and Figure 8), which means a good
quality of energy supply to the power grid.
It follows that Table 2 of the optimal vectors is derived in
From the results of the simulation obtained in Figure.9 and
the same way by giving priority to the control of the active
Figure.10, we can say that the control DPC response time is
power on the reactive power. The signals of HP and HQ thus
very fast compared to the FOC control with high performance
the rotor flux vector position δ, represent the inputs of this (tracking instructions, very fast response time, no overshoot,
truth table, whereas the switching states𝑆𝑎 ,𝑆𝑏 ,𝑆𝑐 represent its minimal static error).
output.

461
 4000
According to Figure 11, the power factor obtained by the 2000
0
3000
two control strategies is variable according to the values of the -2000
0.325
0.33
0.335
0.34
0.345

Statoric current (A)

2000
active and reactive powers. To keep a unit power factor on the
1000
network side, the stator reactive power set point will be kept
0
zero (Qs-ref = 0MVAR), this is the objective of the next test. -1000
Figure 12 illustrates the reactive power injected into the -2000
array by the assembly. We set ourselves a zero instruction to -3000
have a unity power factor on the network side. -4000
0 0.5 1 1.5
Figure 13 shows that the unit power factor (PF) more stable Time (s)

for DPC versus FOC.

Figure 7. Stator currents (DPC)

4000 2000

0
3000
-2000
0.12
0.14
0.16
0.18

2000
Rotoric current (A)

1000

-1000

-2000

-3000

-4000
0 0.5 1 1.5
T ime (s)

Figure 8. Rotor currents (DPC)

6
x 10
2

6
x 10
1.5
1.5

1 0.5 Ps-ref
FOC
0
DPC
Active power [W]

-0.5 5
0.5 x 10
-1
15
-1.5
10
0 -2
0 0.01 0.02 0.03 0.04 0.05 5

0
-0.5 -5

0.8 0.81 0.82 0.83 0.84

-1

-1.5
0 0.5 1 1.5
Time [s]

Figure 9. Active power (FOC, DPC)

6
x 10
2
Qs-ref
FOC
1.5 DPC

1
6
x 10
Reactive power [VAR]

1
0.5
0
5
x 10
0 -1
15

0.98 1 1.02
-0.5
10

-1 5

0
-1.5
0.01 0.02 0.03

-2
0 0.5 1 1.5
Time [s]

Figure 10. Reactive power (FOC, DPC)

462
 1
FOC
REFERENCES
0.9 DPC

0.8 [1] Noguchi, T., Tomiki, H., Kondo, S., Takahashi, I. (1996).
0.7 Direct power control of PWM converter without power
source voltage sensors. In IAS ’96. Conference Record
Power Factor

0.6
0.8
of the 1996 IEEE Industry Applications Conference
0.5
0.6 Thirty-First IAS Annual Meeting, 2: 941-946.
0.4
0.4 https://doi.org/10.1109/IAS.1996.560196
0.3
0.2 [2] Çadirci, I., Ermiş, M. (1992). Double-output induction
0.2
0.82 0.84 0.86 0.88 0.9
generator operating at subsynchronous and
0.1 supersynchronous speeds: Steady-state performance
0 optimisation and wind-energy recovery. IEE Proceedings
0 0.2 0.4 0.6 0.8 1
Time (s) B (Electric Power Applications), 139(5): 429-442.
https://doi.org/10.1049/ip-b.1992.0053
Figure 11. Power factor of DFIG (FOC, DPC) [3] Shehata, E.G. (2015). Sliding mode direct power control
of RSC for DFIGs driven by variable speed wind turbines.
16
x 10
5
Alexandria Engineering Journal, 54(4): 1067-1075.
Qs-ref
FOC
https://doi.org/10.1016/J.AEJ.2015.06.006
14
DPC [4] Bossanyi, E., Burton, T., Sharpe, D., Jenkins, N. (2000).
12
Wind Energy Handbook. New York: Wiley.
10
http://library.uniteddiversity.coop/Energy/Wind/Wind_
Reactive power (Var)

8
Energy_Handbook.pdf
6
[5] Kahla, S., Soufi, Y., Sedraoui, M., Bechouat, M. (2015).
4
On-Off control based particle swarm optimization for
2
maximum power point tracking of wind turbine equipped
0
by DFIG connected to the grid with energy storage.
-2 International Journal of Hydrogen Energy, 40(39):
-4
0 0.5 1 1.5 13749-13758.
Time (s)
https://doi.org/10.1016/J.IJHYDENE.2015.05.007
[6] Amrane, F., Chaiba, A. (2015). A hybrid intelligent
Figure 12. Reactive power (FOC, DPC) control based on DPC for grid-connected DFIG with a
fixed switching frequency using MPPT strategy. In: 4th
1.2 International Conference on Electrical Engineering
FOC
DPC (ICEE), Boumerdes, Algeria, pp. 1-4.
1 https://doi.org/10.1109/INTEE.2015.7416678
[7] Şeker, M., Zergeroğlu, E., Tatlicioğlu, E. (2016). Non-
1

0.8 0.8 linear control of variable-speed wind turbines with

Power Factor

0.6 permanent magnet synchronous generators: A robust

0.6 backstepping approach. International Journal of Systems
0.4
Science, 47(2): 420-432.
0.2
0.4 https://doi.org/10.1080/00207721.2013.834087
0.8 0.805 0.81
[8] Çadirci, I., Ermiş, M. (1992). Double-output induction
0.2 generator operating at subsynchronous and
supersynchronous speeds: steady-state performance
0
0 0.5 1 1.5
optimisation and wind-energy recovery. IEE Proceedings
Time (s) B (Electric Power Applications), 139(5): 429-442.
https://doi.org/10.1049/ip-b.1992.0053
Figure 13. Power Factor of DFIG (FOC, DPC) [9] Aboulem, S., Boufounas, E.M., Boumhidi, I. (2017)
Optimal tracking and robust intelligent based PI power
controller of the wind turbine systems. Intelligent
8. CONCLUSION Systems and Computer Vision, Fez, Morocco.
https://doi.org/10.1109/ISACV.2017.8054981
This paper has been dedicated to the modelling, simulation [10] Barambones, O. (2019). Robust wind speed estimation
and analysis of a wind turbine operating at variable speed. The and control of variable speed wind turbines. Asian
DPC technique is applied for the tow side (grid and rotor side). Journal of Control, 21(2): 856-867.
The proposed scheme has the advantages of simple algorithm, https://doi.org/10.1002/asjc.1779
good dynamic performances, good quality of energy supplied [11] Poitiers, F., Bouaouiche, T., Machmoum, M. (2009).
to the electricity network. Advantages of the DPC control Advanced control of a doubly-fed induction generator for
structure of very high dynamic response, no use of nested wind energy conversion. Electr Power Syst Res, 79(7):
loops, coordinate or modulator transformations, or decoupling 1085-1096. https://doi.org/10.1016/J.EPSR.2009.01.007
of current components. Based on the obtained results, it can be [12] Boyette, A. (2006). Contrôle-commande d’un générateur
concluded that the robust control method can be a very asynchrone à double alimentation avec système de
attractive solution for devices using DFIG, such as wind stockage pour la production éolienne. UniversitéHenri
energy conversion systems. Poincaré- Nancy 1.

463
[13] Sawada, M., Itamiya, K. (2011). A design scheme of https://doi.org/10.1109/ISIE.2007.4374755
model reference adaptive control system with using a [22] Bouyekni, A., Taleb, R., Boudjema, Z., Kahal, H. (2018).
smooth parameter projection adaptive law. In SICE A second-order continuous sliding mode based on DPC
Annual Conference 2011, pp. 1704-1709. for wind-turbine-driven DFIG. Elektroteh Vestn, 85: 29-
https://ieeexplore.ieee.org/abstract/document/6060241 36.
[14] Rudraraju, V.R.R., Nagamani, C., Ilango, GS. (2015). A [23] Shehata, E.G. (2015). Sliding mode direct power control
control scheme for improving the efficiency of DFIG at of RSC for DFIGs driven by variable speed wind turbines.
low wind speeds with fractional rated converters. Int J Alexandria Eng J., 54: 1067-75.
Electr Power Energy Syst, 70: 61-9. http://doi.org/10.1016/j.aej.2015.06.006
https://doi.org/10.1016/J.IJEPES.2015.01.032
[15] Ko, H.S., Yoon, G.G., Kyung, N.H., Hong, W.P. (2008).
Modeling and control of DFIG-based variable-speed NOMENCLATURE
wind-turbine. Electr Power Syst Res, 78(11): 1841-1849.
https://doi.org/10.1016/J.EPSR.2008.02.018 AC Alternating Current
[16] Mouilah, K., Abid, M., Naceri, A., Allam, M. (2014). DC Direct Current
Fuzzy Control of a Doubly Fed Induction Generator for PF Power Factor
Wind Turbines’. J. Electr. Eng. JEE, 14(4): 352-357. 𝑪𝒑 Power coefficient
[17] Abosh, A.H., Zhu, Z.Q., Ren, Y. (2017). Reduction of 𝑹 Blade radius (m)
torque and flux ripples in space vector modulation-based 𝑹𝒔 , 𝑹𝒓 Stator and rotor resistances (Ω)
direct torque control of asymmetric permanent magnet 𝑳𝒔 , 𝑳𝒓 Self inductance of stator and rotor (H)
synchronous machine. IEEE Transactions on Power 𝑴 Mutual magnetizing inductance
Electronics, 32(4): 2976-2986. 𝝋𝒔 , 𝝋𝒓 Stator and rotor flux (Wb)
https://doi.org/10.1109/TPEL.2016.2581026 𝑪𝒆𝒎 Electromagnetic torque (Nm)
[18] Antoniewicz, P., Kazmierkowski, M.P. (2008). Virtual- 𝒗 Wind speed (m/s)
flux-based predictive direct power control of AC/DC 𝑱 Inertia moment of the moving element (kgm2)
converters with online inductance estimation. IEEE 𝝀 Ratio of speed
Trans. Ind. Electron., 55(12): 4381-4390. 𝝆 Air density
https://doi.org/10.1109/TIE.2008.2007519 𝜷 Pitch angle
[19] Bektache, A., Boukhezzar, B. (2018). Nonlinear 𝒇𝒓 Viscous friction coefficient
predictive control of a DFIG-based wind turbine for 𝒑 Number of pair poles
power capture optimization. Int J Electr Power Energy 𝑮 Mechanical speed multiplier
Syst, 101: 92-102. 𝝎𝒔 Synchronously rotating angular speed (rad/s)
https://doi.org/10.1016/J.IJEPES.2018.03.012
𝝎𝒓 Electrical angular rotor speed (rad/s)
[20] Datta, R., Ranganathan, V.T. (2001). Direct power
𝑽𝒔 , 𝑽𝒓 Stator and rotor voltage (V)
control of grid-connected wound rotor induction machine
𝑰𝒅𝒓 , 𝑰𝒒𝒓 Direct and quadrature component of the rotor currents
without rotor position sensors. IEEE Trans. Power
Electron., 16(3): 390-399. (A)
https://doi.org/10.1109/63.923772 𝑰𝒅𝒔 , 𝑰𝒒𝒔 Direct and quadrature component of the stator
[21] Abad, G., Rodriguez, M.A., Poza, J. (2007). Predictive currents (A)
direct power control of the doubly fed induction machine 𝒈 Slip
with reduced power ripple at low constant switching Ω Mechanical speed (rad/s)
frequency. In 2007 IEEE International Symposium on 𝑷 Active power (W)
Industrial Electronics, pp. 1119-1124. 𝑸 Reactive power (Var)

464