Beruflich Dokumente
Kultur Dokumente
AbstractIn this paper a Backstepping Controller (BC) and a Maximum Torque Per Ampere
(MTPA) strategy have been designed in order to
control a permanent magnet synchronous generator
(PMSG) at below and above rated wind velocity.
At low wind speed, the wind turbine operates in
partial load (PL) and the maximum amount of
power is extracted. When the wind speed reaches
the nominal value, the wind turbine operates in full
load (FL). In this case, the pitch angle is controlled
to limit the power. The proposed method is compared to the traditional PI via Matlab/Simulink
software. The robustness tests show that the system
has a faster and a more robust response.
I.
Introduction
2015 IEEE
(7)
+ 0.1
)
III. Control Strategy
14.8 0.3( 2)
The wind turbines with variable speed, variable pitch
0.00184( 3)( 2) (3)
can operate in dierent areas. In this paper two regions
where t is the wind turbine rotational speed.
can be identied as shown in Fig 2 [9]. In the low wind
The aerodynamic torque is obtained as in [7],
speed (Region 1), the maximum power is captured
from the wind. In this region the pitch is kept at the
aer = Paer
t
optimal value, and the rotational speed is controlled
(4)
aer = 12 Ct (, ) Rt3 V 2
via the generator torque to reach the optimal speed
dened by MPPT strategy [10]. When the wind speed
Cp (, )
where : Ct (, ) =
is the torque coecient.
exceed the rated value (Region 2), it is necessary to
Power Coefficient (C )
Cp (, ) = 0.50.0167( 2) sin(
0.2
0.1
0
0
4
6
8
TipSpeed Ratio ()
10
12
+
vq
(5)
Lq
Lq
Lq
Lq
where R, Ld and Lq are respectively the stator resistance and inductances according to the (dq) axis, id ,
iq are the stator currents, p is the pole pairs number,
f is the rotor magnetic eld and J is the total inertia.
The mechanical equation is given by
1
= (m em fv )
(6)
J
Figure 3.
z = ref .
J
fv
1
[K z + ref + + m ].
p(f + (Ld Lq )id )
J
J
(11)
Figure 4.
iq
pLd id pf
R
+
]. (13)
+ iq +
dt Lq
Lq
Lq
Kp = 2 n J
Ki = n2 J
(10)
(9)
5.8
, ( = 0.7, Ts = 0.02).
Ts
n , Ts present respectively the system dynamic and
the desired closed loop time constant [12].
where; n =
The maximum power is directly related to the Maximum Torque per Ampere control strategy. This strategy is used to optimize the eciency of the synchronous generator [15]. To maximize the eciency
and minimize the losses the authors in [16], had
expressed the d-axis current to reduce the stator ux.
Is2
i2d
i2q
i2q
By substituting iq in (7);
Is2
i2d .
f
2(Ld Lq )
(14)
(15)
em
=0.
id
2f
+ i2q .
4(Ld Lq )2
(16)
= Ld [Kd zd +
V.
The proposed method control is simulated using Matlab/Simulink. The parameters of the machine are
taken from the IRCCyN Lab Test bed [17]. The gains
K , Kd , Kq represents respectively the closed loop
responses of mechanical and electrical dynamic. The
principal simulation results are shown in Fig 7 to Fig
9. A good rotor speed tracking is obtained (Fig 7).
The maximum of power coecient is obtained at the
optimal pitch angle (Fig 9). The change in the operating area results an overshooting of the rotational
speed, especially when the wind turbine is controlled
by the PI. A comparison is given in Fig 10 between
the proposed method and the PI controller. It can be
shown that the control with backstepping has faster
and more robust response.
The objective of the robustness tests is to check the
eciency of the proposed control strategy. In the
rst time, the d, q-axis stator inductance (Ld , Lq ) are
changed by +20% (Fig 11-a) and 20% (Fig 11-b) of
their initial values. The results are similar to those
obtained with nominal parameters. In the second,
Ld , Lq are maintained at initial values, and the stator
resistance is changed by +50% (Fig 11-c). The similar
results to those obtained when the d, q-axis stator
inductance are changed. A good tracking of the rotor
speed is obtained with all the tests. This conrms
the robustness of the Backstepping controller. Another
test on total inertia J by the variation of +50% (Fig
11-d) shows that the rotor speed is not aected with
BC, which displays the robustness of the proposed
method.
14
Wind Speed (rad/s)
13
12
11
10
0
6
Time (s)
Figure 6.
Figure 5.
Wind speed
10
250
200
Rotor speed
tracking (rad/s)
2000
1500
1000
500
150
ref
BC
PI
100
50
0
0
10
Time (s)
0
0
4 Time (s) 6
10
Full Load
200
0
0
5
c
10
250
10
200
5
Speed error
tracking (rad/s)
Speed tracking
error (rad/s)
ref
Partial
Load
100
Rotor speed
tracking (rad/s)
Rotational
speed (rad/s)
300
150
5
0
Time (s)
10
ref
100
BC
PI
0
0
dref
10
id
Speed tracking
error (rad/s)
Rotor speed
tracking (rad/s)
iq
100
0
0
10
10
0
10
10
10
c
250
10
2.21
2.22
2.23
Time (s)
2.24
2.25
200
5
Figure 8.
current
10
Speed tracking
error (rad/s)
2.2
150
100
0
0
0.5
10
10
0
d
10
10
Speed tracking
error (rad/s)
Rotor speed
tracking (rad/s)
200
150
100
50
4
2
0
250
0.4
0.35
0
300
Partial
Load
Full
Load
0.45
50
Pitch
Angle ( )
10
50
10
10
Power
Coefficient
10
150
Rotor speed
tracking (rad/s)
qref
10
0
10
200
10
20
0
6
b
250
BC
PI
50
0
0
4
6
Time (s)
10
10
0
4
6
Time (s)
10
Time (s)
Figure 9.
Figure 11.
Rotor speed and speed tracking error with
the robustness tests: a) +20% on Ld,q . b) 20% on Ld,q . c)
+50% on Rs . d) +50% on J
VI. Conclusion
A Backstepping Controller with a Maximum Torque
Per Ampere (MTPA) strategy is designed in this
paper to control the Permanent Magnet Synchronous
Generator in partial load and full load. In the framework of wind turbine control, it allows the maximum
extraction of the wind power by applying the MPPT
strategy. The pitch angle is controlled when the wind
speed exceeds the rated speed in order to limit the
power output at a rated value. The eectiveness and
robustness of the proposed method is checked by a
comparison with PI controller.
In future work, we plan to establish a model of Twin
Wind Turbine to track the maximum power by following the wind direction variation, thanks to the special
structure.
[5]
[6]
[7]
[8]
[9]
Appendix
Machine Parameters
Rated Power Pn
2 kW
Rated Rotor Speed n
2100 rpm
Stator resistance R
1.2
Pole pairs number p
5
Rotor magnetic eld f
0.17 W b
d-axis stator inductance Ld
5.7 mH
q-axis stator inductance Lq
9.2 mH
Viscous friction coecient fv
0.0034 Kg.m2 .s1
Turbine Parameters
Rated wind speed
11 m/s
opt
2
opt
9.1
Blade radius Rt
2m
Air density
1.225 Kg.m3
Gear ratio G
3.9
Total inertia J
0.036 Kg.m2
Control Parameters
K
20
Kd
1000
Kq
1000
[10]
[11]
[12]
[13]
[14]
References
[1]
[2]
[3]
[4]
[15]
[16]
[17]