10.1007@s13198 018 0734 9
10.1007@s13198 018 0734 9
10.1007@s13198 018 0734 9
https://doi.org/10.1007/s13198-018-0734-9
ORIGINAL ARTICLE
123
Int J Syst Assur Eng Manag
the control of the generator is very important. Many to the power grid directly by the stator through two static
techniques have been used to control the DSIG such as a converters. These latest are connected to a grid side
sliding mode control (SMC) associated to the field oriented through a DC bus and a filter.
control (Amimeur et al. 2012), Genetic Algorithm Opti- According to Fig. 1, two detailed controls are required
mized PI and Fuzzy Logic Speed Vector Control (Ameur to ensure the operation of the wind turbine.
and Kouzi 2013), model reference adaptive system based
reactive power controller (Basak et al. 2014), nonlinear
method based on the theory of fuzzy logic controlled tor- 3 Modeling of the wind turbine
que (Chekkal et al. 2014), control of the DSIG using the
instantaneous power theory and a control flux orientation at The wind, with speed V, applied to the blades of the wind
low speeds (Bu et al. 2015). turbine, causes its rotation and creates a mechanical power
The objective of this work is the simulation of the entire on the shaft of the turbine, denoted Pt and expressed by
wind conversion system using two different techniques to (Elmansouri et al. 2015; Guediri and Ben Attous 2015):
control the DSIG. The first technique uses a linear control
Pt ¼ 0:5Cp ðkÞqSV 3 ð1Þ
based on the PI controller and the second technique is a
relatively new nonlinear control based on backstepping Where the tip speed ratio k is defined by:
controller used in mode generation. These techniques are RXt
used in order to compare the performances of the system in k¼ ð2Þ
V
the two cases.
The mathematical model of the entire conversion system The power coefficient Cp represents the aerodynamic
should be studied and detailed. The obtained results using efficiency of the wind turbine and also depends on the
Matlab/Simulink software will be examined and interpreted. characteristic of the turbine. This coefficient has a theo-
retical limit, called the Betz limit, equal to 0.593 and is
never reached in practice (Kumar and Chatterjee 2016; Tria
2 Description of the wind system et al. 2017). We will use an approximate expression of the
0
power coefficient as a function of the relative velocity k
The variable speed wind system studied in this section is and of the blade angle b of which the following expression
based on a DSIG machine as it is shown in Fig. 1. The is (Tamaarat and Benakcha 2014):
turbine, via a gearbox, drives the DSIG, which is connected
Fig. 1 Bloc diagram of the whole wind power conversion system based on DSIG
123
Int J Syst Assur Eng Manag
Tt Xmec
β5=6 Tg ¼ ; Xt ¼ ð4Þ
0.3
G G
β6=8
For a given operating point, it is desired that the
0.2
mechanical power be maximum.This corresponds to the
maximum value of the power coefficient Cp. This is
obtained if the relative velocity k is equal to its optimal
0.1 value kopt for b ¼ 0 (Errami et al. 2015).
0
0 2 4 6 8 10 12
4 Modeling of DSIG
speed ratio λ
The dual stator induction generator model is composed of
two fixed three-phase stator windings displaced with an
Fig. 2 The graph of Cp in function of k electrical angle a ¼ 30 and a squirrel cage mobile rotor
winding (Pandit et al. 2016; Ameur et al. 2016) as shown in
151 18:4 (Fig. 4).
Cp ¼ 0:73 0 0:002b 13:2 exp 0 ð3Þ
k k Considering the choice of the reference (dq) linked to
the rotating field, a simplification of the DSIG expressions
with
is obtained by writing a following compact form (Ameur
0 1
k ¼ 1=ðk þ 0:08bÞ 0:035=ðb3 þ 1Þ and Kouzi 2013).
_ ¼ ½L1 ½B½U xgl ½C½I ½D½I
½I ð5Þ
For different values of the blade angle, the graph of the
power coefficient in function of tip speed ratio Cp ðkÞ, given where: xgl ¼ xs xr , xr ¼ p Xt ; ½U ¼ ½vds1 vqs1 vds
in Fig. 2, is plotted using expression (3). t
2vqs2 vdr vqr t ; ½I ¼ ids1 iqs1 ids2 iqs2 idr iqr ; _ ¼ d ½I;
½I dt
The gearbox adapts the rotation speed of the turbine
½B ¼ diag½111100,
(slow shaft) to the rotation speed of the DSIG (fast shaft)
cp
1 1 Ω
0.5 2 3 1
Ω +- +
∗
λ Ω∗
+
+-
MPPT
123
Int J Syst Assur Eng Manag
2 3
0 0 0 0 0 0
6 0 0 0 0 0 0 7
6 7
6 7
6 0 0 0 0 0 0 7
½C ¼ 6
6 0
7
7
6 0 0 0 0 0 7
6 7
4 0 Lm 0 Lm 0 ðLr þ Lm Þ 5
Lm 0 Lm 0 Lr þ Lm 0
2 3
ðLs1 þ Lm Þ 0 Lm 0 Lm 0
6 0 ðL þ L Þ 0 L 0 Lm 7
6 s1 m m 7
6 7
6 Lm 0 ðLs2 þ Lm Þ 0 Lm 0 7
½L ¼ 6
6
7
7
6 0 Lm 0 ðLs2 þ Lm Þ 0 Lm 7
6 7
4 Lm 0 Lm 0 ðLr þ Lm Þ 0 5
0 Lm 0 Lm 0 ðLr þ Lm Þ
2 3
Rs1 xs ðLs1 þ Lm Þ 0 xs Lm 0 xs Lm
6 x ðL þ L Þ Rs1 xs Lm 0 xs Lm 0 7
6 s s1 m 7
6 7
6 0 xs Lm Rs2 xs ðLs2 þ Lm Þ 0 xs Lm 7
½D ¼ 6
6
7
6 xs Lm 0 xs ðLs2 þ Lm Þ Rs2 xs Lm 0 7 7
6 7
4 0 0 0 0 Rr 0 5
0 0 0 0 0 Rr
123
Int J Syst Assur Eng Manag
8
6 Backstepping control of DSIG >
> e3 ¼ iqs1 iqs1
>
>
<e ¼ ids1 ids1
4
The application of the Backstepping technique to the DSIG ð14Þ
control consists in establishing a control law of the
>
>
> e5 ¼ iqs2 iqs2
>
:
machine via a chosen Lyapunov function to guarantee the e6 ¼ ids2 ids2
overall stability of the system (Herizi et al. 2015; Bouzidi
The final Lyapunov function is given by:
et al. 2016).
1
A. Step1 V2 ¼ ðe21 þ e22 þ e23 þ e24 þ e25 þ e26 Þ ð15Þ
2
In this step, we define the errors e1 and e2 representing Its derivative in function of time is:
the error between the reference speed and the actual speed
and the error between reference direct rotor flux and its V_ 2 ¼ e1 e_1 þ e2 e_2 þ e3 e_3 þ e4 e_4 þ e5 e_5 þ e6 e_6
actual value respectively. V_ 2 ¼ K1 e21 K2 e22 K3 e23 K4 e24 K5 e25 K6 e26
( :
e1 ¼ X Xmec 1
þ e3 K3 e3 r1 þ iqs1 vqs1
ð10Þ L
e2 ¼ udr udr
: 1
þ e4 K4 e4 r2 þ ids1 vds1 ð16Þ
The first Lyapunov function is chosen as: L
:
1 1
V1 ¼ ðe21 þ e22 Þ ð11Þ þ e5 K5 e5 r3 þ iqs2 vqs2
2 L
:
Its derivative is: 1
þ e6 K6 e6 r4 þ ids2 vds2
L
V_ 1 ¼ e1 e_1 þ e2 e_2
1
where:
¼ e1 ½X_ l iqs1 þ iqs2 udr Tr fr X ð12Þ 8
J > 1
>
> r1 ¼ r1 iqs1 xs L1 ids1 þ udr
þ u_ dr ½cudr þ cLm ðids1 þ ids2 Þ >
> L1
>
>
>
> 1
The tracking objectives V_ 1 \0 are achieved by choosing >
> r1 ids1 þ xs L1 iqs1 þ sr udr xsl
< r2 ¼
the references of the current components that represent the L1
stabilizing functions, as follows: >
> 1
>
> r3 ¼ r2 iqs2 xs L2 ids2 þ udr
8 >
> L2
J Tr fr X >
>
>
> i
þ i
¼ K e þ X_
þ þ >
> 1
< qs1 qs2
ludr
1 1
J J >
: r4 ¼ r2 ids2 þ xs L2 iqs2 þ sr udr xsl
h i ð13Þ L2
>
> 1 :
: ids1 þ ids2 ¼
K2 e2 þ udr þcudr
cLm The control voltages are selected as follows:
8 h : i
where: K1 ; K2 are positive constants; l ¼ LmLþL
m rr
, c ¼ Lr þL >
> v
¼ L K e r þ i
>
> qs1 1 3 3 1 qs1
r m
>
> h : i
>
>
B. Step2: < vds1 ¼ L1 K4 e4 r2 þ ids1
h : i ð17Þ
For this step, we eliminate the PI current regulators >
> v
¼ L K e r þ i
>
> qs2 2 5 5 3 qs2
(similar as in Fig. 5) by calculating the control voltages. >
> h
>
> : i
Other errors concerning the components of the stator cur- : vds2 ¼ L2 K6 e6 r4 þ ids2
rent and their references are defined as:
Fig. 5 Design of the current ids1 using PI controller Fig. 6 Design of the current ids1 using the backstepping controller
123
Int J Syst Assur Eng Manag
a 200 b 200
Ω* Ω*
Rotaon Speed (rad/s)
a b 5000
5000
T T*em1
Electromagnec torque (N.m)
0 T*em1 0
Tem1
Tem2 *
Tem2
-5000 T*em2 -5000 Tem2
T*em3
0 0
Tem3
-10000 -5000
* -10000 -5000
Tem3 Tem3
-10000
-10000
0 1 2 3 4 0 1 2 3 4
-15000 -15000
0 5 10 15 0 5 10 15
Time [s] Time [s]
Fig. 8 DSIG Torque and its reference for three constant speeds
The stability of the control is obtained if and only if a The grid reference active power is:
good choice of the gains K3 ; K4 ; K5 ; K6 is made. Figure 6 Pg ¼ Udc idc Udc idc ð19Þ
represents the current regulator configuration.
where the tuning of the DC link is carried out by a regu-
lation loop, using a PI corrector, which generates the
7 Grid side control capacitive current reference (idc ) in the capacitor:
idc ¼ PIðsÞðUdc
Udc Þ ð20Þ
The objective of the DC bus control loop is to set the
voltage constant. This voltage becomes a power source of Network reference currents expressions can be repre-
the DSIG to ensure the correct operation of the converters sented by:
and to impose an operation with an active reference power 8
>
Pg vdg þ Qg vqg
equal to that generated by the system of wind energy >
> i ¼
< dg
v2dg þ v2qg
conversion. ; ð21Þ
>
> P v Qg vdg
The value of the DC bus voltage is obtained from the > iqg ¼ g qg
:
integration of the capacitive current (Kammoun et al. v2dg þ v2qg
2017).
Z The reference voltages expressed in d-q frame reference
1 are given by:
Udc ¼ ic dt ð18Þ
C (
vd inv ¼ vdg þ vdg xs Lf iqg
where: ð22Þ
vq inv ¼ vqg þ vqg þ xs Lf idg
ic ¼ idc iinv
123
Int J Syst Assur Eng Manag
a b
2
2 φdr1
*
Rotor Flux (Wb)
φdr2 *
φ
1 2
1 dr
1.24
1.5 φdr3 1.22
φ
dr2
0.5 1 0.5 1.2
0.5
1.18
φ
0 1 2 3 4 1.16 dr3
0 1 2 3 4
0 0
0 5 10 15 0 5 10 15
Time [s] Time [s]
Fig. 9 Direct rotor flux and its reference for three constant speeds
Table 1 Summarization
V1 = 6 m/s V2 = 8 m/s V3 = 10 m/s
table of the response time and
overshoot of the two control Tr(s) D Tr(s) D Tr(s) D
strategies applied to the DSIG
PI control
Tem (Nm) 0.2425 1085 0.1652 2470 0.1444 4219
Flux (Wb) 0.2119 0.6080 0.1418 0.781 0.1146 0.866
Backstepping control
Tem (Nm) 0.1533 0.0000 0.1395 0.000 0.0488 0.000
Flux (Wb) 0.0322 0.0050 0.0321 0.014 0.0320 0.024
a 250 b 250
Ω Ω
Ω* Ω*
200 200
Wind Speed (m/s)
Wind Speed (m/s)
150 150
100 100
50 50
0 5 10 15 20 25 30 0 5 10 15 20 25 30
123
Int J Syst Assur Eng Manag
a b
5000 5000
T
Tem em
0 *
Tem
0 T*em
-5000 -5000
0
-10000 0 -10000
-5000
-5000
-15000 -15000
-10000
-10000 0 2 4
0 2 4
-20000 -20000
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Time [s] Time [s]
a 2
b
φdr 2 φdr
φ*dr φ*dr
Rotor Flux (Wb)
1.5 1.5
Rotor Flux (Wb)
1 2 1 2
1 1
0.5 0.5
0 0
0 1 2 3 4 5 0 1 2 3 4 5
0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Time [s]
Time [s]
Fig. 12 Direct and quadratic rotor flux
a 1500 b 1500
U Udc
dc
U*dc Udc
*
DC link voltage (V)
1130 1130
500 500
1125 1125
14 14.1 14.2 14.3 14.4 14 14.1 14.2 14.3 14.4
0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Time [s] Time [s]
medium and high wind speeds and for both controls, the meaning that the DSIG delivers faster with the backstep-
response time (Tr) is improved when the wind speed ping control than with the PI control and delivers faster
increases. It indicates also that the backstepping control when the wind speed increases. The synthesis of this
eliminates the overshoot (D). It does not require any torque analysis is summarized in the Table 1.
limiter. This is obviously not the case of the PI control The second test (Fig. 10) is conducted using the wind
where the overshoot increases with the wind speed, variable speed profile. The figures are specified (a) for the
123
Int J Syst Assur Eng Manag
6
a 1 x 10 b x 10
6
1
Pr Pg
* *
0 Pr Pg
0
-1 -1
-2 -2
-3 -3
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Time [s]
Time [s]
a 5
b 5
x 10
2 x 10 2
Grid reacve power (VAR)
0 0
-1 -1
-2 -2
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Time [s]
Time [s]
a b
2000 2000
ig1 ig1
Grid currents (A)
ig2 ig2
Grid currents (A)
1000 1000
ig3 ig3
0 0
-1000 -1000
-2000 -2000
11 11.01 11.02 11.03 11.04 11.05 11 11.01 11.02 11.03 11.04 11.05
PI control and (b) for the backstepping control. It is clear its reference of the DSIG are illustrated in (Fig. 12). The
from this figure that the rotation speed follows perfectly its DC link voltage is constant and follows its set level 1130 V
reference, which varies depending on the imposed wind (Fig. 13). It can be seen from (Figs. 14 and 15) that the
profile. The waveform of the electromagnetic torque gen- active and reactive grid powers, respectively, follow in an
erator follows its reference resulting from the MPPT acceptable way in accordance to their references at all
algorithm, as shown in (Fig. 11). The direct rotor flux and simulation time. In order to get a unit power-factor in
123
Int J Syst Assur Eng Manag
Mag (% of fundamental)
Mag (% of fundamental)
100 2 100
2
80 80
1
1
60 60
0
0 100 200 300 400 500 0
40 Ti 40 Ti 0 100 200 300 400 500
20 20
0 0
0 200 400 600 800 1000 0 200 400 600 800 1000
Frequency [Hz] Frequency [Hz]
network side, the reactive power reference Qg is fixed at Appendix: Parameters
zero value. For the proposed algorithms the currents in the
three phase’s grid, illustrated in Fig. 16, constitute a Radius of the turbine R = 36 m
sinusoidal and a balanced system of the frequency 50 Hz. Gear box gain G = 90
Figure 17 in turn, gives the THD (total harmonic dis- Maximum power coefficient Cpmax ¼ 0:44
tortion) of the grid current Ig1. This THD is defined as the Optimal relative wind speed kopt ¼ 7:05
ratio of total effective value of harmonics (their quadratic DSIG nominal power Pn ¼ 1:5 MW
sum) to the rms value of the fundamental component. We RMS voltage value U = 400 V
analyze a sequence of time. As we see, the harmonics Frequency F = 50 Hz
appearing in the grid current are minimized. We notice also Number of pole pairs p=2
that the THD of PI control is lower than the THD of Stator resistance Rs1 = Rs2 = 0.008 X
Backstepping control. Stator inductance Ls1 = Ls2 = 0.134
mH
Magnetizing inductance Lm = 0.0045 H
9 Conclusion Rotor resistance Rr = 0.007 X
Rotor inductance Lr = 0.067 mH
This paper discusses synthesis of linear (based on the PI Inertia J = 10 kg.m2
control) and nonlinear control (based on backstepping Viscous coefficient f = 2.5 Nm s/rd
control) of a DSIG dedicated to a wind turbine system. The Filter inductance Lf ¼ 0:001 H
two control laws are applied to the stator side inverters. We Filter resistance Rf ¼ 0:01 X
define the complete model of the wind system by inte- Capacitance of the DC link C ¼ 0:072 F
grating the previous process, adjusting the DC link voltage voltage
and controlling the active and reactive powers to the grid
side converter. All system is delivering into the grid.
Finally, the results for the two approaches have been dis-
cussed and compared. The responses of the system show a References
significant overshoot for the electromagnetic torque and the
rotor flux at the starting up for the PI control, while for the Ameur F, Kouzi K (2013) Genetic algorithm optimized PI and fuzzy
backstepping control, the trajectory tracking is done with- logic speed vector control of dual stator induction generator in
out any overshoot. wind energy conversion system. In: Proceedings of the 3rd
international conference on systems. Algiers, October 29–31
The observation of the different results obtained shows a Ameur F, Kouzi K, Ameur A, Kasbadji NM (2016) Robust control of
good behavior of the system for the PI control. However, dual stator induction generator used in wind conversion system
the backstepping strategy has better dynamic connectes to the grid usin direct torque control. In: 4th
performances. International conference on renewable energy: generation and
applications ICREGA’16, At Belfort, Franch
123
Int J Syst Assur Eng Manag
Amimeur H, Aouzellag D, Abdessemed R, Ghedamsi K (2012) Kammoun S, Sallem S, Kammoun MBK (2017) Backstepping control for
Sliding mode control of a dual-stator induction for wind energy low-voltage ride through enhancement of DFIG-based wind
conversion systems. Electr Power Energy Syst 42:60–70 turbines. Arab J Sci Eng. https://doi.org/10.1007/s13369-017-
Basak S, Chakraborty C, Sinha AK (2014) Dual stator induction 2606-z
generator with controllable reactive power capability. In: 23rd Kumar D, Chatterjee K (2016) A review of conventional and
International symposium on industrial electronics (ISIE), IEEE advanced MPPT algorithms for wind energy systems. Renew
Bekakra Y, Ben Attous D (2015) Optimizing of IP speed controller Sustain Energy Rev 55:957–970
using particle swarm optimization for FOC of an induction Lekhchine S, Bahi T, Soufi Y (2014) Indirect rotor field oriented
motor. Int J Syst Assur Eng Manag. https://doi.org/10.1007/ control based on fuzzy logic controlled double star induction
s13198-015-0391-1 machine. Electr Power Energy Syst 57:206–211
Bouzidi M, Benaissa A, Barkat S, Bouafia S, Bouzidi A (2016) Mahboub MA, Drid S, Sid MA, Cheikh R (2016) Sliding mode
Virtual flux direct power control of the three-level NPC shunt control of grid connected brushless doubly fed induction
active power filter based on backstepping control. Int J Syst generator driven by wind turbine in variable speed. Int J Syst
Assur Eng Manag. https://doi.org/10.1007/s13198-016-0433-3 Assur Eng Manag. https://doi.org/10.1007/s13198-016-0524-1
Bu F, Hu Y, Huang W, Zhuang S, Shi K (2015) Wide-speed-range- Mehammai C, Zidani F, Benaicha S, Nait-Said MS (2014) Research
operation dual stator-winding induction generator DC generating on improvement of FOC system for induction motor using fuzzy
system for wind power applications. IEEE Trans Power Electron logic. Int J Modell Identif Control 21(4):370–377
30(2):561–573 Miryousefi Aval SM, Ahadi A, Hayati H (2015) A novel method for
Chekkal S, Aouzellag Lahaçani N, Aouzellag D, Ghedamsi K (2014) reliability and risk evaluation of wind energy conversion systems
Fuzzy logic control strategy of wind generator based on the dual considering wind speed correlation. Front Energy. https://doi.
stator induction generator. Electr Power Energy Syst 59:166–175 org/10.1007/s11708-015-0384-4
Elmansouri A, El mhamdi J, Boualouch A (2015) Control by back Pandit JK, Aware MV, Levi RE (2016) Direct torque control scheme for
stepping of the DFIG used in the wind turbine. Int J Emerg a six-phase induction motor with reduced torque ripple. IEEE Trans
Technol Adv Eng 5(2):472–478 Power Electron. https://doi.org/10.1109/TPEL.2016.2624149
Errami Y, Ouassaid M, Maaroufi M (2015) Optimal power control Taheri A (2016) Harmonic reduction of direct torque control of six-
strategy of maximizing wind energy tracking and different phase induction motor. ISA Trans. https://doi.org/10.1016/j.
operating conditions for permanent magnet synchronous gener- isatra.2016.02.014i
ator wind farm. Energy Procedia 74:477–490 Tamaarat A, Benakcha A (2014) Performance of PI controller for
Ghoudelbourk S, Dib D, Omeiri A, Azar AT (2016) MPPT control in control of active and reactive power in DFIG operating in a grid-
wind energy conversion systems and the application of fractional connected variable speed wind energy conversion system. Front
control (PIa) in pitch wind turbine. IntJ Model Identif Control Energy 8(3):371–378
26(2):140–151 Taraft S, Rekioua D, Aouzellag D (2013) Wind power control system
Guediri AK, Ben Attous D (2015) Modeling and fuzzy control of a associated to the flywheel energy storage system connected to
wind energy system based on double- fed asynchronous machine the grid. Energy Procedia 36:1147–1157
for supply of power to the electrical network. Int J Syst Assur Tiwari R, Babu NR (2016) Recent developments of control strategies
Eng Manag. https://doi.org/10.1007/s13198-015-0367-1 for wind energy conversion system. Renew Sustain Energy Rev
Herizi A, Balla R, Ahmani SA (2015) Backstepping control of 66:268–285
induction motors. El Wahat pour les Recherches et les Etudes Tria FZ, Srairi K, Benchouia MT, Benbouzid MEH (2017) An
8:132–145 integral sliding mode controller with super-twisting algorithm
Hossain MM, Mohd HA (2015) Future research directions for the for direct power control of wind generator based on a doubly fed
wind turbine generator system. Renew Sustain Energy Rev induction generator. DOI, Int J Syst Assur Eng Manag. https://
49:481–489 doi.org/10.1007/s13198-017-0597-5
123