1

A New Method of Maximum Power Point Tracking for DFIG Based

Wind Turbine

Shabani, A. Deihimi
Department of electrical engineering
Bu Ali Sina university, Hamedan, I. R. Iran
a.shabani@basu.ac.ir


Keywords: wind turbine, DFIG, mppt, rotational speed



Abstract
In this paper, different operational regions of
doubly fed induction generator (DFIG) based
wind turbine (WT), from viewpoints of rotor
speed, generated power, tip speed ratio ()
and the angle of blades of the WT's rotor, is
studied and classified. Then a new fast and
explicit method of maximum power point
tracking (mppt) will be proposed. The method
is based on the difference between optimum
and current rotational speed of the shaft of
WT. The proposed method is compared with
another method to unfold the superior one.
This comparison will be done based on the
speed of operation and quality of generated
power and the results shows the priority of the
proposed method.


1. INTRODUCTION
The soaring use of fossil fuels and their
depletion over the last two decades combined
with a growing concern about pollution of
environment have led to a boost for renewable
energy generation. This accelerated drive has
led to a tremendous progress in the field of
renewable energy systems during last decade.
Now a day wind energy conversion systems
(WECS) becomes an essential part of modern
electric networks. In the grid of today
penetration of wind turbines (WTs) is
growing rapidly in size and number. This
increase in the size of wind turbines leads to
further reduction in the cost of wind power.
And on the other hand, the decrement of cost
of wind power redound to intensification in
the use of WECSs.
Along different types of WTs, the DFIG
based wind turbine has caught most of
interest. This is because of: 1) in comparison
with constant speed WTs, DFIG based WT
operates in much wider range of wind
velocity and it's production has a better power
quality. 2) in comparison with synchronous
generator based WT (PMSG or WRSG),
DFIG based WT has less manufacturing cost
[1].
In DFIG based WT by using back-to-back
PWM inverters between the grid and the rotor
circuit (see fig. 1), and employing vector
control techniques, the active and reactive
powers handled by the machine can be
controlled independently [2]. Since the stator
is directly connected to the grid, the stator
flux is constant over the entire operating
region. Therefore, the torque can be
10-E-REN-1691

A New Method of Maximum Power Point Tracking for DFIG Based Wind Turbine

25
th
International Power System Conference
2

maintained at its rated value even above the
synchronous speed [3].
However this type of WT is a variable speed
one, but it's range of operational speeds is
restricted by the rate of power of it's rotor side
inverter. In usual, the rate of power of
converter is between 0.1 through 0.4 of
generator rated power, and hence the
maximum slip (S
max
) of generator would be
0.1 through 0.4, too [4].
Same as rotational speed, the rated power of
the DFIG based WT, is dictated by rated
power of generator and of course the rated
power of it's inverter [5].
Based on the maximum value of speed and
rated generating power, different controlling
regions may be exposed and in each region,
an especial controlling law




Fig. 1 a simple structure of DFIG based wind turbine

governs. These controlling algorithms are
described by maximum power point tracking
(mppt) controller [6].
In recent years, many papers has been
published about different methods of mppt
algorithm, some of them used sophisticated
controlling method like sliding mode [6]-[7],
and some used adaptive control [8] or Rotor
Position Phase Lock Loop (PLL) [9], [10].
But drawback of these method is their
complicated methods used by them. And
meanwhile none of them describes the
controlling algorithm of all operational
regions.
In this paper, controlling method of whole
operational regions are described and a new
simple and explicit mppt algorithm is
proposed.



2. AERODYNAMIC POWER
Encountering of the wind with the blades of
the WT's rotor leads to creation of mechanical
power in the rotor. The value of this power
depends on the velocity of the wind,
rotational speed of WT's rotor, the angle of
rotor's blade and the structure of blades. The
produced power may be expressed as:

P
M
=
1
2
. p
A
. n. R
2
. C
p
([, z)I
wnd
3

(1)

Where P
M
is the mechanical power extracted
from the wind in N.m/s; is the air density
(1.225kg/m3); V is the wind speed (m/s) and
R is the wind turbine rotor radius (m). C
p
is
the Efficiency coefficient which is formulated
by use of the structural data of the WT's rotor,
blades angle () and tip speed ratio () which
may be described as:

z =
R.
wt
:
wnd

(2)

Where
wt
stands for rotational speed of WT's
rotor.
C
p
can be described as:

C
popt
(z
opt
, [) = u.22(
116
z

-u.4[
-S)cxp(
-12.S
z

)
(3)
Where

1
z

=
1
z
opt
+u.u8[
-
u.uSS
[
3
+1

(4)

opt
is a value of which make the maximum
value of C
p
(see fig. 2).
Fig. 3 illustrates mechanical power curves for
each wind speed versus
wt
with use of
equations (1) through (4). And effect of on
the extracted power is shown in fig. 4 [1].

Back to back
converter
DFIG
Gear
box
WT's rotor


Fig. 3 harves
o

Fig. 4 diffe
speed of 12m


OPERATI
As can be
generator i
back to
converter,
controlled.
flow throu
A New
Fig. 2 C
sted mechanic
of the wind fro
erent mechanic
m/s which are
de
IONAL RE
seen in fig
is connected
back conv
the generat
. The maxi
ugh the con
Method of M
25
t

Cp versus .


cal power vers
om 4m/s to 22

cal power for
results of fluc
egree)
EGIONS
g. 1, the ro
d to the util
verter. By
ted power o
imum powe
nverter (from
Maximum Pow
th
Internation


sus wt , spee
2m/s.

constant wind
ctuations in
otor circuit
lity grid via
y using th
of WT may b
er which c
m the grid

er Point Track
nal Power Sys
3
eds
d
(in
of
a a
his
be
an
to
th
us
po
th
ro
be


W
ge
T
a

P
t

W
gr
ge
ro




C
P

king for DFIG

stem Confere
he rotor or
sually betw
ower of gen
he maximu
otational sp
etween the r

utcd
(1 -
Where
rate
enerator.
he rated pow
function of
t
= (1 -S)
Where P
t
is t
rid, P
s
is the
enerator and
otor circuit
Fig. 5 DF
C
p
V
cut in
V

rmax
P
t

r
P
max
C
pmax
G Based Wind
ence
r from the
ween 0.1 t
nerator and
um slip o
peed of g
range of [6]
-|S
mux
|)

ut
d
is the s
wer of WT
f stator or ro
)P
s
=
1 -S
s
the delivere
e power del
d P
r
is the p
[5].
FIG based WT
V
r max
Co
V
Var.
r

Opt.
=0
I
d Turbine
rotor to th
through 0.4
d this ratio
of generato
generator s
] :

tcd
(1 +|S
m
ynchronous
may be rep
otor rated po
S
|P

|
ed power of
livered just
power deliv
T controlling r
V
rated
Co
Co

ons.
r

Var.
=0
II

he grid) is
4 of rated
determines
or. So the
should laid
mux
|)
(5)
s speed of
presented as
ower:
(6)
f WT to the
by stator of
vered by the
regions
ons.
r
ons. P
t
>0
III
V
w
V
w
V
w
V
cut out
s
d
s
e
d
)
f
s
)
e
f
e

w
w
w

A New Method of Maximum Power Point Tracking for DFIG Based Wind Turbine

25
th
International Power System Conference
4

It is evident that the limits of maximum
allowable generating power and maximum
and minimum allowable rotational speed of
generator should not be exceeded. Based on
these limitations, different operational regions
may be introduced.
Fig. 5 shows different operational regions
which may be considered in controlling of a
DFIG based WT.
As can be seen in fig. 5, there are three major
controlling regions, which are discussed here:

A. Region I
This region is between V
cut in
and V
r max
. When
the wind velocity is less than V
cut in
, the WT
doesn't produce any electrical power and the
incoming mechanical power is wasted in
rotational components of WT. At speed of cut
in, which depends on the type of WT is
between 2.5 m/s through 4 m/s, producing the
electrical power which is commanded by
mppt unit, would be begun. The transferring
ratio of gear box should be chosen in a way
that when the C
p
has its maximum value, the

r cutin
corresponds to V
cutin
be as:

K
gcubox
.
cut n
= (1 -|S
mux
|)
g utcd

(7)

And the upper limit of this region would be as
upper limit in (5). In whole this region, the
should have it's optimum value, which leads
to maximum of C
p
and hence, for each wind
speed, the maximum extractable power would
yields (see fig. 2 and 3).
And the angle of blades would be kept zero,
because the maximum mechanical power has
been not reached yet.

B. Region II
This region laid between two wind speed of
V
r max
and V
rated
. in this region, the speed of
the generator's rotor
Reached to it's maximum value and should
not be exceeded (should be kept constant at
it's maximum value). So the controlling
algorithm should operate in a way that
increase of wind speed doesn't lead to
increase in rotational speed of WT's rotor.
This means that in this region, the is not at
it's optimum value any more and would be
reduced (see fig. 2).
In this region the value of is still zero.

C. Region III
This condition is the last situation that a WT
would operate in. the lower wind speed limit
of this region is V
rated
and the upper limit is
V
cut out
. if the wind speed exceed this limit the
operation of WT would be stopped and the
braking system would engaged because
mechanical damage is expected.
In this region, the rotational speed of WT is
still kept constant at it's maximum value.
and C
p
are not on their optimum value and
further more, the angle of blades has
increased from zero to kept the mechanical
power at it's rated value (see fig. 4).
The increment command of angle of blades is
issued by mppt algorithm.


3. PROPOSED MPPT ALGORITHM
Basis of all these controlling regions shown in
fig. 5, is on contrasting of mechanical and
electrical torque which with neglecting the
damping factor and spring constant, it can be
formulated as:
I
m
-I
c
= [
J
Jt

(8)

Whit knowing that:

P = I. (9)

It is evident that with convenient adjustment
of recommended power (P
set
), changes of
can be controlled.
Now controlling algorithm of different
regions may be expanded.

A. Region I
As mentioned before, in this region, it is need
to follow the optimum and hence according
to (2), the should be changed linearly with
change of wind speed. So with regards to (8)
and (9), if the should be increased, the
value of P
set
must be set much lower than


mechanica
decreased,
much high
equation w

P
sct
= P
M

Where
correspond

Fig. 6 disp

Fig. 6 illus
At first, as
the mechan
of WT is
wind speed
moment of
instant, fl
neglectable
would be P
moment, th
is K(
3
-
from P
M
. t
will led
process, P
M
P
set
tracks
approachin
) reduces to
At the nex
speed has
pervious st
unchanged
to P
21
. The
this case P
of mechan
by kinetic
A New
al power (P
M
the value
her than P
M

would do it a
-K(
opt
-

opt
is
ding to
opt
.
placement of o
mppt
strates the w
sume that th
nical power

2
. Now th
d from V
2
t
f WT's rota
luctuation
e and hen
P
23
. Accord
he discrepan
-
2
), whi
this differen
the
r
to
P
M
moves f
s the trajec
ng P
set
to P
3
o zero.
xt step, its
a reductio
tate, at the f
d and P
M
ha
en P
set
woul
P
set
is more
nical input p
energy sto
Method of M
25
t
M
) and if the
e of P
set
sh
. the propos
as well:
-

)
the rotati

operation poin
t method
way equatio
he wind spe
r has settled
here is an in
to V
3
. Due
ting compo
of rotation
nce the o
ding to (10
ncy betwee
ich would b
nce between
increase.
from P
23
to
ctory of B
3
the stateme
assumed t
on from V
2
first momen
as a step d
ld be P
21
+K
than P
M
, an
power will
ored in rota
Maximum Pow
th
Internation
e should b
hould be s
sed followin
(1
ional spe
nt in proposed
n (10) work
eed is V
2
, an
d at P
2
and
ncrease in t
to enormo
onents, at fir
nal speed
output pow
0), in the fir
en P
M
and P
be subtract
n P
M
and P
s
During th
oward P
3
, an
to P
3
. Wi
ent K(
3

that the win
2
to V
1
. Lik
nt,
r
rema
down from P
K(
2

1
).
nd deficien
compensat
ating mass

er Point Track
nal Power Sys
5
be
set
ng
0)
ed
d
ks.
nd

he
ous
rst
is
wer
rst
P
set

ed
set
,
his
nd
ith

r

nd
ke
ain
P
2

In
cy
ed
of
W
in
W
co
in

B
In
co
to

C
T
ta
th
th
pr
in
ha
de
it
Si
pr
on
to
is
of
[1

4.
A
in
W
be
in
re
ch
in
to
th
th
m
se
fi
su
A
of
sm
king for DFIG
stem Confere
WT's rotating
n
r
and it w
With proper
ontrol the r
n produced p
B. Region II
n this regi
onstant, so t
o P
M
, which

r
should be
C. Region II
he controll
aken in this
he mechanic
he pitch con
rocess. As
ncreasing t
arvesting
ecreased an
's upper lim
ince the p
rocess, has
n the struct
oday, the m
s about 10 d
f pitch con
12].
. SIMULAT
All simulatio
n MATLAB
WT which h
ench is av
nformation
efer to [5] a
hosen as 5e
n [3], is com
o unfold su
he changes
he input, th
m/s, in secon
econd 150,
g. 7). The
upposed to b
As can be se
f proposed
moother tha
G Based Wind
ence
g parts, and
will last unti
r choosing
rate and ac
power.
II
ion, it is n
the value o
h means tha
zero.
III
ling algorit
region too,
cal power h
ntrol takes a
can be s
the angle
mechanic
nd hence in
mit.
pitch contr
it's own dy
ture of win
maximum ch
degrees per
ntroller can
TION RES
ons here ha
B/simulink.
has been c
vailable in
about mode
and [13]. the
e6. the othe
mpared wit
uperior one
of power t
he wind spe
nd 70 it red
it increases
e rated wi
be 12 m/s.
een in fig. 8
d algorithm
an the othe
d Turbine
d this lead t
il acced
of K, its p
curacy of f
need to ke
f P
set
shoul
at the rate o
thm of reg
, and in add
has reached
a part in the
seen in fig
of blades
cal powe
nput power
rol is a
ynamic whi
nd turbine.
hanges in bl
r second [1]
be found i
SULTS
as been acc
. The para
considered
appendix.
eling of WT
e value of K
er algorithm
th proposed
. This algo
to control t
eed assume
duces to 9
s to 10 m/s
ind speed
8, the produ
m (the bo
er one and
o reduction
de to zero.
possible to
fluctuations
ept the
r
d be equals
f change of
gion II, is
dition, since
to it's limit,
controlling
g. 4, with
s (), the
er would
remains in
mechanical
ch depends
In WTs of
lade's angle
]. Structure
in [11] and
complished
ameters of
as the test
for more
T, one may
K in (10) is
m suggested
d algorithm
orithm uses
the WT. as
ed to be 10
m/s and in
s again (see
of WT is
uced power
lt one) is
it's power
n
o
s
r

s
f
s
e
,
g
h
e
d
n
l
s
f
e
e
d
d
f
t
e
y
s
d
m
s
s
0
n
e
s
r
s
r

A New Method of Maximum Power Point Tracking for DFIG Based Wind Turbine

25
th
International Power System Conference
6

quality is better, too. Another drawback of
suggested mppt algorithm in [3].


Fig. 7 profile of wind speed, applied for comparison


Fig. 8 comparison of produced power by proposed
mppt algorithm (the bolt one) and the mppt algorithm
suggested in [3].


Fig. 9 amplitude of .


Fig. 10 rotational speed of WT's rotor.

Now that the superiority of proposed
algorithm has been unfolded, the other
outputs of this algorithm are shown. In fig. 9,
the amplitude of , is shown which according
to fig. 2 should be about 7.2. fig. 10, shows
the rotational speed of WT's rotor. Fig.11,
illustrates the electrical torque of generator.
Fig. 12, presents the mechanical input power
(bolt one) which delivered by the WT's shaft
and electrical produced power delivered to the
utility grid. In these simulations, demanded
reactive power is supposed to be zero. As can
be seen in fig. 12, the generated electrical
power has an overshoot at time 70s, and an


Fig. 11 generated electrical torque of WT.


Fig. 12 mechanical (bolt line) and electrical (thin line)
power.


Fig. 13 terminal out coming d and q currents.

undershoot at time 150 s. these are because of
the kinetic energy stored in rotating mass
which unleashes during reduction in
rotational speed, and stores during increment

A New Method of Maximum Power Point Tracking for DFIG Based Wind Turbine

25
th
International Power System Conference
7

. this affects terminal currents shown in fig.
13, which exposes terminal out coming d and
q currents.


5. CONCLUSION
Different operating and controlling regions of
DFIG based wind turbine from viewpoints of
rotor speed, generated power, tip speed ratio
() and the angle of blades of the wind
turbine's rotor, is studied and classified. Based
on these regions, a new simple method of
maximum power point tracking algorithm is
presented which is based on of rotor. This
algorithm applied to model of a typical DFIG
based WT in MATLAB/simulink, and results
proves the efficiency of new method.


APENDIX
Wind turbine and generator data:
Drive train:
J
gen
=121.5, J
wt
=6.41e6,
K
se
=1.4035e4, D
e
=500e3,
K
gear
=168.3

Generator:

P=4, n
s
=20, n
r
=80,
f=50Hz, V
an
=690V
S
b
=2.5MW, R
s
=4.29e-3,
L
s
=0.257e-3, R
r
=7.68e-2,
L
r
=4.858e-3, L
m
=8.0428e-3

Wind turbine:

Slip(S
max
)=0.25,
cut-in wind speed: 4 m/s,
cut-out wind speed: 22 m/s,
rated wind speed:12m/s,
turbine rotor radius: 41.5m

REFERENCE
[1] T. Ackermann, Wind Power in Power
Systems. New York, UK: John Wiley &
Sons, 2005.
[2] R. Datta and V. T. Ranganathan, A
simple position sensorless algorithm for
rotor side field oriented control of wound
rotor induction machine, IEEE Trans.
Ind. Electron. Soc., vol. 48, pp. 786793,
Aug. 2001.
[3] R. Datta and V.T. Ranganathan," A
Method of Tracking the Peak Power
Points for a Variable Speed Wind Energy
Conversion System", IEEE Trans, energy
conversion, VOL. 18, NO. 1, MARCH
2003
[4] Ion Boldea, Variable Speed Generators.
Boca Raton, FL: Taylor & Francis, 2006.
[5] Hee-Sang Ko, Gi-Gab Yoon, and Won-
Pyo Hong, "Active Use of DFIG-Based
Variable-Speed Wind-Turbine for Voltage
Regulation at a Remote Location," IEEE
Trans, Power systems, VOL. 22, NO.
4,pp. 1916-1925, NOVEMBER 2007
[6] B. Beltran, M.E.H. Benbouzid

and T.
Ahmed-Ali, " High-Order Sliding Mode
Control of a DFIG-Based Wind Turbine
for Power Maximization and Grid Fault
Tolerance " Electric Machines and Drive
Conference, 2009
[7] Xuemei Zheng, Lin Li, Dianguo Xu and
Jim Platts, "Sliding Mode MPPT Control
of Variable Speed Wind Power System"
Power and energy engineering conference,
APPEEC, 2009
[8] Changhong Shao, Xiangjun Chen and
Zhonghua Liang, " Application Research
of Maximum Wind-energy Tracing
Controller Based Adaptive Control
Strategy in WECS", IPEMC 2006
[9] Baike Shen, , Bakari Mwinyiwiwa,
Yongzheng Zhang, , and Boon-Teck Ooi,"
Sensorless Maximum Power Point
Tracking of Wind by DFIG Using Rotor
Position Phase Lock Loop (PLL)" IEEE
Trans, power electronics, VOL. 24, NO.
4, APRIL 2009
[10] A. Petersson " Analysis, Modeling and
Control of Doubly-Fed Induction
Generators for Wind Turbines", Ph.D.
dissertation, Dept. of Energy and
Environment, Univ. Chalmers, Goteborg,
Sweden 2005
[11] E. Hau, Wind turbines. Berlin,
Germany: Springer, 2006
[12] F.B. Bianchi, H. De Battista, Wind
Turbine Control Systems. Berlin,
Germany: Springer, 2007
[13] F. Iov, A. D. Hansen, P. Srensen, F.
Blaabjerg, "Wind Turbine Blockset in
Matlab Simulink". Aalborg University,
Denmark: March 2004