Conference lnLernaLlonale des Lnergles 8enouvelables (ClL8'13)

Sousse, 1unlsle - 2013

1

Control strategies of the mechanical speed for the wind turbine

N. Cherfia *, D. Kerdoun ** and A. Boumassata **.


LGEC Research Laboratory, Department of Electrical Engineering,
Constantine 1 University, 25000 Constantine, Algeria
Phone: +213780102155, e-mail: msn822009@live.fr, kerdjallel@yahoo.fr, boumassata-r10@hotmail.fr




Abstract- A winds turbine role is to convert the kinetic energy of
wind into energy electric. Its components are designed to
maximize the conversion. Different control structures have been
developed according to the characteristic power speed. In our
case, we propose the control strategy of the wind turbine to
maximize power with control of the speed. This control strategy
consists in determining the speed of the turbine which provides
the maximum generated power.

Keywords - Wind, turbine, DFIG, control, PI

I. INTRODUCTION

During the last years, there was a strong penetration of the
renewable resources of energy in the network of power supply.
The wind power production played and will continue to play a
very important role in this domain for years to come.
Wind turbines have base of the doubly fed induction
machine (DFIG) undoubtedly appeared as one of the high
technologies for the manufacturers of wind turbines,
Demonstrating that it is about an actual cost, Effective and a
reliable solution.
The first part is dedicated to the description and modelling
of wind turbines based on physical equations responsive
operation.
The second part, we present a mathematical model of the
(DFIG) will simulate the model in generator mode.
The third is devoted to the study of the technique of indirect
control power and compare the simulation results by PI
Controller with Anticipation and PI Controller with Phase
Advance.


II. MODEL OF THE TURBINE
The model is based on the characteristics of steady state
power of the turbine [1].

P
m
=
P
m
P
mt
P
mt
= C
p
. P
mt
=
1
2
C
p
(z)pnR
2
I
1
3
(1)
With z =
U
1
R
v
1
(2)
0
1
: Rotation speed before multiplier.
R :rotor radius 35.25 m
: air density, 1.225 kg.m
-3

C
p
= (z, [) = C
1
[
C
1
x
i
-C
3
[ -C
4
exp[
C
S
x
i
+C
6
z (3)
with :
1

=
1
+0.O8
-
0.035

3
+1
et C
1
= u.S176 ; C
2
= 116 ;
C
3
= u.4 ;C
4
= S ;C
5
= 21 ; C
6
= u.uu68 [1].



Fig.1 The power factor for different angles of stalls

Characteristics of C
p
in terms of for different values of the
pitch angle are shown in figure 1. The maximum value of
C
p
(C
pmux
= u.4SSS ) is reached of =2 and =10.01. This
particular value of is defined as the nominal value z
nom
[1]
,[2].

III. DYNAMIC MODEL OF THE DOUBLY FED
INDUCTION MACHINE
A commonly used model for the doubly fed induction
generator (DFIG) is the Park model. The electrical equations
of the DFIG in the Park reference frame are given as follows
[2], [3]:

_
u
xd
= R
x
i
sd
+
dq
sd
dt
-
s

sq
u
xq
= R
x
i
sq
+
dq
sq
dt
+
s

sd

(4)

_
u
rd
= R
r
i
d
+
dq
rd
dt
-

q
u
rq
= R
x
i
q
+
dq
rq
dt
+

d

(5)

The stator and rotor flux are given as:
_

xd
= I
x
i
sd
+I
m
i
d

xq
= I
x
i
sq
+I
m
i
q

(6)

_

d
= I

i
d
+I
m
i
sd

rq
= I

i
q
+I
m
i
sq

(7)

In these equations, R
s
, R
r
, L
s
and L
r
are respectively the
resistances and the inductances of the stator and the rotor
windings,L
m
is the mutual inductance.

0 5 10 15 20 25
-0.1
0
0.1
0.2
0.3
0.4
X: 10.01
Y: 0.4353
report of the tip speed lambda
P
o
w
e
r

c
o
e
f
f
i
c
i
e
n
t

C
p
B=2
B=10
B=15
B=5
B=20
Conference lnLernaLlonale des Lnergles 8enouvelables (ClL8'13)
Sousse, 1unlsle - 2013
2

v
sd
, v
sq
, v
rd
, v
rq
, i
sd
, i
sq
, i
rd
, i
rq
,
sd
,
sq
,
rd
,
rq
are the
d and q components of the stator and rotor voltages, currents
and flux, whereas m
r
is the rotor speed in electrical degree.

The electromagnetic torque is expressed as:

C
cm
= p(
sd
. i
sq
-
sq
. i
sd
) (8)

Stator and rotor variables are both referred to the stator
reference Park frame. With the following orientation, the d
component of the stator flux is equal to the total flux whereas
the q component of the stator flux is null figure. 2.




Fig.2 Determination of the electrical angles in Park reference frame

sd
=
s
,
sq
= u (9)

By replacing (9) in (6) and (8), the electromagnetic torque
can be given as follows:

C
cm
= -p
L
m
L
s
i
q

sd
(10)

Assuming that the resistance of the stator winding R
s
is
neglected, and referring to the chosen reference frame, the
voltage equations and the flux equations of the stator winding
can be simplified in steady state as follows:

_
:
sd
= u
:
sq
= :
s
=
s

s

(11)
_

xd
= I
x
i
sd
+I
m
i
d
u = I
x
i
sq
+I
m
i
q

(12)

From (12), the equations linking the stator currents to the
rotor currents are deduced below:

_
i
sd
=
q
s
L
s
-
L
m
L
s
i
d
i
sq
= -
L
m
L
s
i
q

(13)

The active and reactive powers at the stator side are defined
as:





_
P
x
= :
xd
i
sd
+:
sq
i
sq

x
= :
xq
i
sd
+:
sd
i
q

(14)

Taking into consideration the chosen reference frame, the
above power equations can be written as follows:

_
P
x
= :
s
i
sq

x
= :
x
i
sd

(15)
Replacing the stator currents by their expressions given in (15),
the equations below are obtained:

_
P
s
= -:
s
L
m
L
s
i
q

s
=

s
q
s
L
s
-

s
L
m
L
s
i
d

(16)

The block diagram of the DFIG model in Park reference frame
is depicted in figure 3, assuming a constant stator voltage (vs).




Fig.3 Block diagram of the DFIG model

IV. REGULATION WITH BUCKLE OF POWER
to improve the control system the DFIG, we will introduce
an additional loop control of active and reactive power in the
block diagram of the control loop without power so that each
axis controller contains two PI control, one to control the
power and the other rotor current (figure 4) [4]










u
rd

u
rq
Q
x

P
x

1
R
r
+p(L
r
-
L
m
2
L
x
)

-
I
m
I
s
I
s

-
I
m
I
s
I
s

m
r
(L
r
-
L
m
2
L
x
)
m
r
(L
r
-
L
m
2
L
x
)
F
x
2
L
x
m
x

m
r
L
m
L
x
q
x

1
R
r
+p(L
r
-
L
m
2
L
x
)

|
rq

|
rd

0
d,q frame
Stator axis
0
m

0
x


0
r

O

xa

O

ra

O

r

O

d

O

q

F

x

q
x

O

x

Conference lnLernaLlonale des Lnergles 8enouvelables (ClL8'13)
Sousse, 1unlsle - 2013
3








Fig.4 Schema block indirect regulation with loop power

V. SYNTHESIS OF THE REGULATOR PI FOR
THE CONTROL OF THE POWER
It is a simple and easy to implement controller. Figure 5
shows a closed loop system corrects by a PI controller. In our
case, the transfer function is in the form h
p
+
h
t
p
as shown in
Figure 5

Fig.5 diagram of the control system by PI

The transfer function of the open loop (FTBO) with
regulators is written as follows [2 ,6 ,7]:

F
B
=
p+

.
1

+p
(17)

Applying the method of compensation pole transfer
function 1, we obtain the following equality:

=
R
r
L
r

(18)

After compensation, the following are obtained FBO:
F
B
=

L
r
p
(19)

This gives us the transfer function following closed loop:
F
B
=
1
1+p
(20)
With: =
L
r

and = [1 -
L
m

L
s
L
r

: is the response time of the system that we fix the order of
10ms.In this case, the gains of the PI controllers are expressed
in terms of machine parameters and response time as follows:
_

p
=
L
r

=
R
r

(21)
We used the method of compensation poles for its speed; it
is clear that n 'not the only valid method for the synthesis of PI
controller.

VI. MODEL OF THE MULTIPLIER
MECHANICAL PART

The mechanical party of the turbine includes three
directional blades pitch and of length R. They are fixed to a
drive shaft in a rotation speed

, a multiplier Connected of
gain G. This multiplier causes the electric generator. We can
model all the three blades as one and the same mechanical
system characterized by the sum of all the mechanical
characteristics. Due to the aerodynamic blade design, we
believe that the coefficient of friction with respect to the air is
very small and can be neglected. Also, the speed of the turbine
being very low, Friction losses will be negligible compared to
the friction losses on the side of the generator. On the basis of
these hypotheses, is then obtained a mechanical model
consisting of two masses as shown in figure 6, the validity,
relative to the complete model of the turbine, has already been
verified [1], [5].
1
R
r
+pL
r

p
+

|
p

r




DFIG

PI

I
m

s
I
s
I
sq

PI

-
I
s
I
m
I
s

-
I
s
I
m
I
s

PI
PI
I
sq
2

s
I
s

P
s

s

i
d

i
q
I
qc

I
dc

i
qc

i
dc

P
sc

sc
Conference lnLernaLlonale des Lnergles 8enouvelables (ClL8'13)
Sousse, 1unlsle - 2013
4



Fig.6 Mechanical model of the wind turbine
With:

: Moment of inertia of the turbine to the equivalent

inertia of the three blades of the turbine.

: the moment of inertia of the DFIG.

: The coefficient due to viscous friction of DFIG.

: mechanical torque on the shaft of the DFIG.

: the rotational speed of the DFIG.

: the electromagnetic torque of the DFIG.

By considering that the multiplier is ideal, That is to say,
the mechanical losses are negligible, it is then modelled by the
following two equations:

C
m
=
C
t

(22)
0
m
=G.0
t
(23)

From Figure 6, we can write the fundamental equation of
dynamics of the mechanical system of the mechanical shaft
DFIG by:

[

+
m

dU
m
dt
+. 0
m
= C
m
-C
cm
(24)

The total inertia J :

= [

+
m
(25)

VII. CONTROL STRATEGIES OF THE TURBINE
The control strategy is to adjust the torque appearing on the
tree turbine speed so as to fix a reference. To achieve this, we
will use a speed control [2] ,[3].
According to equation (24) and (25):

dU
m
dt
=
1

. (C
m
-. 0
m
-C
cm
) (26)
The electromagnetic torque is:

C
cmc
= PI. (0
c
-0
mcc
) (27)
0
c
= 0. 0
tubncc
(28)



The reference speed of the turbine corresponds to the
optimum value corresponding to the speed ratioI
pax
(fixed
on the blade angle to 2).

0
tubncc
=
x
C
mcx
.v
R
(29)

The couple thus determined by the controller is used as a
reference torque of the turbine model as can be seen in figure 7.
Fig.7 Block diagram of the controller with mechanical speed


VIII. CORRECTION OF THE SPEED OF THE
TURBINE

Different technologies can be considered markers for
control of the speed; in our case we describe two types of
regulations:

A. PI Controller with Anticipation
We consider a correction proportional integral (PI):

C
cmc
= [b
1
+
b
0
s
. (0
c
-0
mcc
) (30)

b
1
: proportional gain and b

: integral gain, correction

parameters are determined. It is necessary to increase the
parameter b

to reduce the action of

g
wind torque. The
natural frequency and damping coefficient are given by:

n
= _
b
0

and =
++b
1
b
0
.
o
n
2
(31)

Can impose a time response and damping factor given was:

b
0
=
n
2
. and b
1
=
2.b
0
.{
o
n
- - (32)

F
u
1
6

1
6

1
]x +

C
m

C
m

R
urh|n
F
u

C
p
.
p
2
. x. F
u
3
.
1

m

PI G
2
up
. F
u
R

C

r


Control device with control of the speed
mechanical
Turbine
Multiplier
Tree

m


G
0
t

0
m

C
cm

C
t

f
C
m

Conference lnLernaLlonale des Lnergles 8enouvelables (ClL8'13)
Sousse, 1unlsle - 2013
3


B. PI Controller with Phase Advance
We consider a correction proportional integral (PI):

C
cmc
=
u
1
.s+u
0
.s+1
. (0
c
-0
mcc
) (33)

a
1
,a

etx are the parameters to determine the corrector

and s is the Laplace variable. It is necessary to increase the
parameter a

to reduce the action of

g
wind torque. The
natural frequency and damping coefficient are given by:

n
= _
u
0
+
.
and =
++u
1
u
0
+
.
o
n
2
(34)

IX. SIMULATION RESULTS:

The simulation is done by imposing active and reactive
power reference (P
rI
, Q
rI
), While the machine is entailed in
variable speed
P
c
= p. P
mopt
,
c
= u,
with:
: Performance DFIG;
P
mopt
: The optimal mechanical power
the simulation results are given by the following figures.

Fig.8 Mechanical speed with anticipation control

Fig.9 Electrical active power produced with anticipation control


Fig.10 Electrical reactive power produced with anticipation control




Fig.11 Mechanical speed with phase advance control


Fig.12 Electrical active power produced with phase advance control
0 1 2 3 4
0
1000
2000
3000
4000
5000
time [s]

M
e
c
h
a
n
i
c
a
l

s
p
e
e
d

(
t
r
/
m
i
n
)
anticipation
0 1 2 3 4
-16
-14
-12
-10
-8
-6
-4
-2
0
2
x 10
5
time [s]

a
c
t
i
v
e

p
o
w
e
r

[
W
]
anticipation


Pmes
Pref
0 1 2 3 4
-1
-0.5
0
0.5
1
x 10
5
time [s]

r
e
a
c
t
i
v
e

p
o
w
e
r

[
V
A
R
]
anticipation


Qmes
Qref
0 1 2 3 4
0
500
1000
1500
2000
2500
time [s]

M
e
c
h
a
n
i
c
a
l

s
p
e
e
d

(
t
r
/
m
i
n
)
phase advance
0 1 2 3 4
-10
-8
-6
-4
-2
0
2
x 10
5
time [s]

a
c
t
i
v
e

p
o
w
e
r

[
W
]
phase advance


Pmes
Pref
Conference lnLernaLlonale des Lnergles 8enouvelables (ClL8'13)
Sousse, 1unlsle - 2013
6


Fig.13 Electrical reactive power produced with phase
advance control

X. CONCLUSION

In our work, we have established the model of the
machine using its power equations in the dq axis system
related to synchronization. We have also developed the
method of vector control power of the machine to know the
order.
The two methods of maximizing power were almost granny
gives results (speed mechanical, power active and reactive)
the only difference and the transient time in the case of PI
with big anticipation and compared PI phase advance.

APPENDIX























NOMINAL POWER =1.5(MW)
STATOR PER PHASE RESISTANCE =0.012 ()
ROTOR PER PHASE RESISTANCE=0.021 ()
STATOR LEAKAGE INDUCTANCE= 2.0372.10
-004
(H)
ROTOR LEAKAGE INDUCTANCE= 1.7507.10
-004
(H)
MAGNETIZING INDUCTANCE= 0.0135 (H)
NUMBER OF POLES PAIRS=2
MOMENT OF INERTIA= 1000 (KG.M
2
)
FRICTION COEFFICIENT =0.0024


RFERENCES
[1] Z. Lubosny, Wind Turbine Operation in Electric Power Systems, Berlin,
Germany: Springer, 2003.
[2] S.El Aimani, Modlisation De Diffrentes Technologies Doliennes
Intgres Dans Un Rseau De Moyenne Tension.
[3].F.Poitiers, Etude Et Commande De Gnratrices Asynchrones Pour
Lutilisation De Lnergie Eolienne,2003.
[4] A. Mehdary, tude dune chaine de conversion dnergie olienne
base dune aroturbine ,6
mes
Journes des doctorants, Laboratoire des
sciences de linformation et des systmes LSIS, universit de St Jrme,
Marseille,2009
[5] J. Usaola, P. Ledesma, J. M. Rodriguez, J. L. Fernadez, D. Beato, R.
Iturbe,J. R. Wihelmi, Transient stability studies in grids with great wind
power penetration. Modeling issues and operation requirements, Proceedings
of the IEEE PES Transmission and Distribution Conference and Exposition,
September 7-12, 2003, Dallas (USA).
[6].Moriarty P.J. Butterfield S.B. Wind Turbine Modeling Overview for
Control Engineers, In Proc .American Control Conf., St.louis, MO, 2009.
[7] Multon B. Obin G. Gergaud O. Ben Ahmed H.

Le Gnie Electrique dans
le vent : tat de lart et recherches dans le domaine de la gnration olienne,
JCGE03, Saint-Nazaire, 2003.














0 1 2 3 4
-1
0
1
2
3
4
x 10
4
time [s]

r
e
a
c
t
i
v
e

p
o
w
e
r

[
V
A
R
]
phase advance


Qmes
Qref