IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 8, NO.
1, JANUARY 2017
Grid-Interfaced DFIG-Based Variable Speed Wind
Energy Conversion System With Power Smoothening
N. K. Swami Naidu, Member, IEEE, and Bhim Singh, Fellow, IEEE
Abstract—This paper deals with the analysis, design, and control
of grid-interfaced doubly fed induction generator (DFIG) based
variable speed wind energy conversion system (WECS) for power
smoothening with maximum power point tracking (MPPT) capa-
bility. This DFIG uses rotor position computation algorithm for the
sensorless control through rotor position estimation. Power fluctu-
ations due to the unpredictable nature of the wind are eliminated
by introducing battery energy storage system (BESS) in the dc
link between two back-to-back connected voltage source convert-
ers. The design of BESS is presented for feeding regulated power
to the grid irrespective of the wind speeds. The control algorithm
of the grid-side converter is modified for feeding regulated power
to the grid. Rotor-side converter is controlled for achieving MPPT
and unity power factor operation at the stator terminals. A proto-
type of the proposed DFIG-based wind energy conversion system is
developed using a digital signal processor (DSP-dSPACE DS1103).
This developed DFIG is tested extensively at different wind speeds
and also presented some of the steady-state test results. Dynamic
performance of this DFIG is also demonstrated for the variable
wind speed operation.
Index Terms—Battery energy storage system (BESS), doubly
fed induction generator (DFIG), maximum power point track-
ing (MPPT), power leveling, sensorless, wind energy conversion
system (WECS).
I. INTRODUCTION
ARTH’s fossil energy resources such as oil, gas and coal
E are limited in production and are expected to use beyond
their peak in the next decades, so the price of energy may con-
tinue to rise. Ever rising need of the energy in the future can
be met by contributing more renewable energy sources. Re-
newable sources are climate-friendly because of the absence
of detrimental emissions to the environment [1]. Out of all re-
newable energy technologies, the wind energy is one of the
lowest-priced because of advancements in power electronics
[2]. Fixed speed wind energy conversion systems (WECSs) us-
ing squirrel cage induction machines are mostly used because
of their simplicity and low cost. From the wind turbine charac-
teristics, one can clearly observe that for different wind speeds,
the machine should run at different rotor speeds for achieving
maximum power. As fixed speed induction generators (FSIG)
are running at same speed for different wind speeds, FSIG are
Manuscript received August 15, 2015; revised April 9, 2016; accepted May
11, 2016. Date of publication June 20, 2016; date of current version December
14, 2016. Paper no. TSTE-00682-2015.R2.
The authors are with the Department of Electrical Engineering, In-
dian Institute of Technology, New Delhi 110016 India (e-mail: nkswami-
naidu@gmail.com; bsingh@ee.iitd.ac.in).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TSTE.2016.2582520
1949-3029 © 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications standards/publications/rights/index.html for more information.
51
less efficient [3]. As the stator is directly connected to the grid,
these fixed speed generators draw large lagging reactive power.
Now a days, variable speed induction generators are used to run
at desired speed using power electronic converters to improve
the energy output substantially and also to achieve unity power
factor [4]. Doubly fed induction generator (DFIG) is the mostly
preferred variable speed WECS topology due to the reduction
in the size of the power converters and also the converter losses
[5]–[7]. DFIG with single-stage gearbox seems the most inter-
esting choice due to its energy yield divided by cost [7]. Vector
control methods have been successfully employed to control
both active and reactive powers of DFIG independently [8], [9].
Due to the intermittent nature of wind, the change in power
generated by the wind turbine is very high. This problem pre-
vails when the wind energy penetration increases in the grid. So
many attempts have been made for reducing the volatile effect
of wind power generation on the grid [10]–[15]. Different strate-
gies like tip speed ratio (TSR) control and Pref versus ω m are
used for smoothening the wind power fluctuations [10]. But in
this case, the power yield is reduced because it is not operating
at MPPT. This can be overcome by integrating energy storage
to the WECS. The performance analysis of different hybrid en-
ergy storage systems connected to wind turbines is analyzed for
improving power quality [11]. Different energy storage tech-
nologies are compared for their operational suitability for dif-
ferent time scales. The appropriate sizing of storage and energy
capacity needed for specific power rating is also investigated in
[12]. Short term energy storage devices such as super capacitors
are introduced for smoothening the fast wind induced power
variations [13], [14]. Li-ion capacitors are used as a potential
solution for filtering power variations at the scale of tens of sec-
onds in [15]. The voltage fluctuations and power smoothening
are achieved by integrating flywheel with DFIG. This conse-
quently increases the wind energy penetration in the grid [16],
[17]. However, the flywheel storage is normally preferable for
the time scale of 20 minutes [10]. Rapid advancements in bat-
tery energy storage system (BESS) have completely changed its
applications in high power ratings [18]. Large installations of
BESS for high power, applications have emerged, which also in-
clude features such as transmission line stability, load leveling,
power quality, black start, and local VAR control [18]. Nowa-
days researchers have started introducing BESS in the WECS
for reducing the power variations [19]–[23]. In [19], [20], au-
thors have used separate inverter for the integration of BESS.
BESS is also integrated with STATCOM for improving transient
stability and also to control the power output [21]. However, the
same functionalities can be achieved by integrating BESS at
the DC link of DFIG. So, one converter can be reduced and
52
Fig. 1. Proposed system configuration.
the system becomes simple and efficient. In [22], [23], authors
have placed the battery in the DC link of two back to back con-
nected voltage source converters (VSCs) for coordinated tuning
and control.
This work proposes the grid interfaced DFIG based WECS for
power smoothening. The rotor position computation algorithm
(RPCA) is used for the estimation of rotor position [24]. The
novelty of the work lies in the control of grid side converter
(GSC). So, the authors have clearly demonstrated the control
algorithm for feeding regulated power to the grid. The selection
of BESS is the other important aspect of DFIG based WECS
for power smoothening. The variations in the powers with the
increase in wind speeds are demonstrated by comparing the
conventional DFIG and the proposed DFIG. The operation of
the system is experimentally validated for regulating the power
of DFIG even under variable wind speed conditions.
II. SYSTEM CONFIGURATION AND CONTROL PRINCIPLE
The schematic diagram of proposed grid interfaced DFIG
based WECS is shown in Fig. 1. BESS is connected in the DC
link of two back to back connected VSCs. Here the stator is
directly connected to the grid. RSC is controlled in voltage ori-
ented reference frame. The d-axis of the synchronously rotating
reference frame is oriented along voltage axis using enhanced
phase locked loop (EPLL). Here, the rotor position computation
algorithm is used for the position estimation. GSC is controlled
in such a way that the regulated power is fed to the grid. When
the power produced is more than the regulated power, and then
the remaining power is stored in the BESS. If power produced
is less than regulated power, then the BESS supplies the re-
maining power to the grid. The control algorithms are presented
in Fig. 2.
III. DESIGN OF VARIABLE SPEED WECS
Selection of ratings of converters, BESS voltage, and BESS
rating are very much important for the successful operation of
the WECS. If the battery rating is very high, then the system is
reliable but the cost is very high. So the selection of rating of
BESS is very critical from economical point of view. The selec-
tion of battery voltage and also the suitable rating are presented
for the system specifications given in Appendix.
A. Selection of Battery Voltage
In general, DC link voltage of VSC must be greater than twice
the peak of phase voltage [25]. The battery voltage is selected
by considering voltages at both GSC and RSC.
IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 8, NO. 1, JANUARY 2017
Normally DFIG operating speed range for WECS is between
0.7 p.u. to 1.3 p.u. So the maximum operating slip (S m ax ) is 0.3.
The possible maximum RMS phase voltage of the rotor (vr ) is
given as,
 
Nr
vr = Sm ax vp (1)
Ns
where, vp is the stator phase voltage = 230 V, Sm ax = 0.3,
Nr /Ns = rotor to stator turns ratio = 1/2.
By placing the above values, rotor phase voltage (vr ) is
obtained as 34.64 V.
There is a star delta transformer between GSC and
grid with 2:1 turn’s ratio. So, the phase voltage at GSC
(V GSC ) = 66.67 V .
As vGSC is more than vr , the DC link voltage should be
selected according to vGSC .
The DC link voltage is estimated as,
√
Vdc ≥ (2 2/m) vGSC . (2)
Maximum modulation index of VSC is selected as 1 for linear
region. So the value of Vdc by (2) is 188.57 V. By considering
the battery voltage for the minimum state of charge, and also
considering the availability of the battery in the system, the
nominal battery voltage is selected as 240 V.
B. Design of BESS
Large size of the battery increases the reliability but also in-
creases the initial investment. However, small rating of the bat-
tery affects the reliability [18]. So the proper design of BESS is
necessary for the satisfactory operation of the proposed WECS.
The storage capacity of the BESS depends upon the approxi-
mate wind profile at the site. The rating of the BESS is decided
by the total energy stored in to the battery. In this proposed
DFIG, the constant value of the power feeding to the grid is
selected as the average value of the power generation from the
previous wind data. So the energy stored in to the BESS, when
the power generated from the DFIG is more than the regulated
power. On the other hand, the energy is taken from the BESS,
when the power generation is less than the regulated power. The
average power is selected from a past wind data of a particular
wind site. In this work, the wind speed of Perambalur, Tamil
Nadu, India is taken from the website [26] for the calculation
of average power. The wind data is given at 20 meters height in
the website for every one hour throughout the day is listed in
Table I. However, practically the turbine is installed at much
higher height. The wind speed at turbine height is calculated as
[23],
 n
v h
= (3)
v0 h0
where, v is the new wind speed at a height h, v0 is the old wind
speed at a height of h0 and n is terrain factor. Here in this work,
a terrain factor of 0.13 is selected for the wind speed calibra-
tion. So the wind speeds are calculated at 50 m height and the
power outputs of the DFIG at that wind speeds are provided in
Table I. The average power (Pavg ) generated for a day is
NAIDU AND SINGH: GRID-INTERFACED DFIG-BASED VARIABLE SPEED WECS WITH POWER SMOOTHENING 53

Fig. 2. Control algorithm of the proposed WECS.

TABLE I Pm i is calculated as,

POWER GENERATED AT DIFFERENT WIND SPEEDS
Pm i = Pinst − Pavg (5)
S. No Time Wind Speed DFIG developed Power fed to Power fed to
(m/sec) power the grid battery where Pinst is the instantaneous power at any instant and Pavg
(P inst ) (P avg ) (P mi = P inst
(kW) (kW) − Pav g )
is the average power fed to the grid.
(kW) The rating of the battery bank calculated from (4), is estimated
as 20 kWh. However, in this proto-type, as the battery voltage
1 12–1 AM 11.59 5.85 3.85 2
2 1–2 AM 7.24 1.42 3.85 –2.43
is selected as 240 V, total 20 batteries of 12 V, 7 Ah are con-
3 2–3 AM 8.69 2.46 3.85 –1.39 nected in series. For increasing the current rating, three parallel
4 3–4 AM 13.04 6.5 3.85 2.65 branches of 7 Ah are connected. So the total battery capacity of
5 4–5 AM 14.49 6.5 3.85 2.65
6 5–6 AM 14.49 6.5 3.85 2.65
5.04 kVAh is used due to the easy availability for validating the
7 6–7 AM 12.31 6.5 3.85 2.65 control approach.
8 7–8 AM 10.86 4.82 3.85 0.97
9 8–9 AM 10.86 4.82 3.85 0.97
10 9-10 AM 10.14 3.92 3.85 0.07 C. Selection of VSC Rating
11 10–11 AM 10.14 3.92 3.85 0.07
12 11-12 AM 8.69 2.46 3.85 –1.39 The power flows through RSC and GSC are different because
13 12–1 PM 6.52 1.04 3.85 –2.81 of BESS in the DC link. The maximum power flows through
14 1–2 PM 7.24 1.42 3.85 –2.43
15 2–3 PM 8.69 2.47 3.85 –1.38
the GSC, when the wind turbine is in shut down condition and
16 3–4 PM 9.42 3.14 3.85 –0.71 the BESS is feeding the whole power to the grid. This constant
17 4–5 PM 10.14 3.92 3.85 0.07 power generation is normally selected as average power which
18 5–6 PM 8.69 2.47 3.85 –1.38
19 6–7 PM 7.97 1.9 3.85 –1.95
is 3.85 kW in this case. Therefore, the GSC rating is selected as
20 7–8 PM 8.69 2.47 3.85 –1.38 3.85 kVA.
21 8–9 PM 8.69 2.47 3.85 –1.38 RSC rating depends upon the reactive power supplied from
22 9–10 PM 8.69 2.47 3.85 –1.38
23 10–11 PM 13.04 6.5 3.85 2.65
the rotor side for realizing unity power factor at the stator side
24 11–12 PM 13.04 6.5 3.85 2.65 and also the rotor active power. The DFIG draws a lagging
volt-ampere reactive (VAR) for its excitation to build the rated
air gap voltage. It is calculated from the machine parameters
calculated as 3.85 kW from Table I. This average power may be that the lagging VAR of 2 kVAR is needed when it is run-
calculated by considering every day or a month. ning as a motor. In DFIG case, the operating speed range is
The rating of the battery bank is calculated as, 0.7 p.u. to 1.3 p.u. So the maximum slip (sm ax ) is 0.3. For

n achieving unity power factor at the stator side, reactive power of
Eb = (Pm i ∗ ti ) (4) 600 VAR (S m ax ∗ Qs = 0.3∗ 2 kVAR) is needed from the rotor
i=1 side (Qr m ax ). Maximum rotor active power is Sm ax ∗ P . The
where, Pm i is the excess power stored into the battery at every power rating of the DFIG is 5 kW. Therefore, the maximum
instant and ti is the time period taken as 5 min in this case. rotor active power (P r m ax ) is 1.5 kW (0.3∗ 5kW = 1.5 kW).
54 IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 8, NO. 1, JANUARY 2017

So the rating of the VSC used as RSC, Srated is estimated as


Srated = Pr2m ax + Q2r m ax . (6)
Thus kVA rating of RSC, Srated is calculated as 1.615 kVA.

IV. CONTROL STRATEGY

Control algorithms of GSC, RSC and sensorless operation
of DFIG are discussed in detail in this Section. A schematic
diagram of control for GSC and RSC are presented in Fig. 2.
Fig. 3. Enhanced phase locked loop.
A. Control of RSC
An independent control of both active and reactive powers The direct and quadrature rotor current errors (Ider ,Iqer )
is achieved by the RSC. Through this RSC control, the maxi- between estimated (Idr , Iqr ) and reference (Idr ∗ ∗
, Iqr ) direct,
mum power point is achieved operationally. RSC is controlled quadrature axis rotor currents are processed through PI con-
in voltage oriented reference frame. So the active and reactive trollers for achieving reference direct and quadrature axis rotor
powers are controlled by controlling d and q axis rotor reference voltages (Vdr and Vqr ) as
currents (Idr and Iqr ), respectively.
∗
Direct axis rotor reference current (Idr ) is obtained through Vdr (n) = Vdr (n − 1) + kp dv {Ider (n) − Ider (n − 1)}
proportional integral (PI) speed controller by processing the + kidv Ider (n) (11)
speed error (ωer ) between reference and estimated rotor speeds
(ωr∗ and ωr ) as Vqr (n) = Vqr (n − 1) + kp qv {Iqer (n) − Iqer (n − 1)}
∗ ∗ + kiqv Iqer (n) (12)
Idr (n) = Idr (n − 1) + kp d [ωer (n) − ωer (n − 1)]
+ kid ωer (n) (7) where kp dv , kidv are the proportional and integral gains of di-
rect axis current controller. kp qv ,kiqv are the proportional and
where kp d , kid are the proportional and integral gains of speed integral gains of quadrature axis current controller. Ider (n) and
controller. ωer (n) and ωer (n − 1) are the speed errors at nth Ider (n − 1) are the direct axis current errors at nth and (n − 1)th
∗ ∗
and (n − 1)th instant. Idr (n) and Idr (n − 1) are the direct axis instant. Vdr (n) and Vdr (n − 1) are direct axis rotor voltages
rotor reference currents at nth and (n − 1)th instant. at nth and (n − 1)th instant. Iqer (n) and Iqer (n − 1) are the
The rotor speed (ω r ) of the DFIG is estimated using RPCA quadrature axis rotor current errors at nth and (n − 1)th instant.
sensorless algorithm discussed in Section IV-C. Vqr (n) and Vqr (n − 1) are quadrature axis rotor voltages at nth
Reference speed (ωr∗ ) is selected for achieving MPPT. In this and (n − 1)th instant.
algorithm, the reference rotor speed is estimated using TSR Direct and quadrature axis rotor voltages (Vdr and Vqr ) are
control using wind speed. The reference quadrature axis rotor added with the compensation terms for achieving reference rotor
∗
current (Iqr ) is selected to control the reactive power to zero at voltages (Vdr∗ and Vqr∗ ) as,
stator terminals.
The direct and quadrature axis rotor currents (Idr and Iqr ) are Vdr∗ = Vdr − (ωe − ωr )σLr Iqr (13)
computed from sensed rotor currents (ira , irb and irc ) as, Vqr∗ = Vqr − (ωe − ωr )(Lm im s + σLr Idr ). (14)
2  
These reference direct and quadrature voltages (Vdr∗ , Vqr∗ )
Idr = ira sin θslip + irb sin θslip − 2π
3
3 are converted into three phase reference rotor voltages
  ∗ ∗ ∗
+ irc sin θslip + 2π3 (8) (vra , vrb , vrc ) as [27],
2   ∗
vra = Vdr∗ sin θslip + Vqr∗ cos θslip (15)
Iqr = ira cos θslip + irb cos θslip − 2π
3
3 ∗
  vrb = Vdr∗ sin(θslip − 2π
3 ) + Vqr∗ cos(θslip − 2π
3 ) (16)
+ irc cos θslip + 2π3 (9)
∗
vrc = Vdr∗ sin(θslip + 2π
3 ) + Vqr∗ cos(θslip + 2π
3 ). (17)
where slip angle (θslip ) is calculated as,
∗ ∗ ∗
These three phase rotor reference voltages (vra , vrb , vrc ) are
θslip = θe − θr. (10) compared with triangular carrier wave of switching frequency
for generating the PWM signals for the insulated gate bipolar
The voltage angle θe is calculated using EPLL as shown
transistors of the RSC.
in Fig. 3. The rotor position (θr ) is calculated using RPCA
sensorless algorithm as discussed in Section IV-C.
B. Control of GSC
The direct and quadrature axis rotor currents (Idr and Iqr )
are controlled close to reference direct and quadrature axis rotor The novelty of this work lies in the control of GSC.
∗ ∗
currents (Idr and Iqr ) by using two current controllers. The control of this GSC is realized on voltage oriented
NAIDU AND SINGH: GRID-INTERFACED DFIG-BASED VARIABLE SPEED WECS WITH POWER SMOOTHENING 55

reference frame. The active and reactive powers (P & Q) fed

to the grid are controlled by controlling direct and quadrature
grid currents (Idg & Iqg ) respectively. Direct axis grid current
(Idg ) is obtained from the reference power (P ∗ ) as,

∗
2 
Idg = 3 (P ∗ /Vdg ). (18)

∗
For realizing unity power factor at ac mains, Iqg is selected
as zero. Actual direct and quadrature grid currents (Idg and
Iqg ) are estimated from the sensed grid currents. The error
(Ideg and Iqeg ) between actual (Idg and Iqg ) and reference
∗ ∗
(Idg and Iqg ) direct and quadrature currents are processed
through a PI controller as shown as, Fig. 4. Rotor position computation algorithm.

 
Vdg (n) = Vdg (n − 1) + kp dv {Ideg (n) − Ideg (n − 1)} The unit templates of rotor currents aligned to the stator
co-ordinate system (cos θs , sin θs ) are calculated as
+ kidv Ideg (n) (19)
irsα

Vqg (n) = 
Vqg (n − 1) + kp qv {Iqeg (n) − Iqeg (n − 1)} cos θs =
(23)
i2rsα + i2rsβ
+ kiqv Iqeg (n) (20)
irsβ
sin θs =
(24)
where kp dv , k idv are the proportional and integral gains of di- i2rsα + i2rsβ
rect axis current controller. kp qv , k iqv are proportional and in-
tegral gains of quadrature axis current controller. Ideg (n) and where irsα and irsβ are the rotor currents aligned in stator
Ideg (n − 1) are direct axis current errors at nth and (n − 1)th co-ordinate system is calculated as,
instant. Vdg 
(n) and Vdg (n − 1) are the direct axis grid voltages vsβ − Rs isβ − ωe Ls isα
irsα = (25)
at nth and (n − 1)th instant. Iqeg (n) and Iqeg (n − 1) are the ωe Lm
quadrature axis grid current errors at nth and (n − 1)th instant. Rs isα − vsα − ωe Ls isβ
 
Vqg (n) and Vqg (n − 1) are the quadrature axis grid voltages at irsβ = . (26)
ωe Lm
nth and (n − 1)th instant.
  Stator currents (isa , isb ), stator voltages (vsa , vsb ) are trans-
Direct and quadrature axis grid voltages (Vdg and Vqg )
are added with the compensation terms for achieving formed into two phase system using Clarke’s transformation
∗ (isα , isβ , vsα and vsβ ).
reference direct and quadrature axis grid voltages (Vdg and
∗ ∗ ∗ ∗ The unit vectors of rotor position angle are computed as,
Vqg ). Three phase reference grid voltages (vga , vgb , vgc ) are
calculated from the reference direct and quadrature voltages cos (θm )est = cos(θs ) cos(θr ) + sin(θs )sin(θr ) (27)
∗ ∗ ∗ ∗ ∗
(Vdg , Vqg ). These reference grid voltages (vga , vgb , vgc ) are
compared with PWM signals and then these pulses are fed to sin (θm )est = sin(θs ) cos(θr ) − cos(θs )sin(θr ). (28)
the GSC.
V. IMPLEMENTATION AND OPERATIONAL SEQUENCE
A prototype of proposed DFIG based WECS is developed
C. Rotor Position Computation Algorithm using a digital signal processor (DSP-dSPACE DS1103) in the
In this algorithm, rotor current (ir ) makes an angle θs from the laboratory. In this prototype, Hall-Effect voltage sensors (LEM
stator co-ordinate system and the same rotor current (ir ) makes LV25P) are used to sense voltages and Hall-Effect current sen-
an angle θr from the rotor co-ordinate system [24]. The angle sors (LEM LA55P) are used for current sensing. In this DFIG, a
between stator and rotor is calculated as (θm )est = (θs − θr ). wound rotor induction machine (WRIM) is coupled with a DC
The schematic diagram of sensorless scheme is shown in Fig. 4. machine. Wind turbine characteristics are emulated using Type
Rotor currents (ira , irb ) are sensed and transformed into two A chopper and the DC machine. Before starting of the system,
phases using Clarke’s transformation (ir α & ir β ). Unit templates the stator of the DFIG is kept isolated from the grid using switch
of rotor currents aligned to rotor axis are calculated as, S1 in Fig. 2. Initially the DFIG runs at fixed speed by giving
fixed duty ratio to Type-A chopper. The slip angle (θslip ) is
ir α obtained from the grid voltage and sensorless algorithm. Open
cos θr =
(21) loop DFIG voltage is matched with the grid voltage by selecting
i2r α + i2r β the reactive power component of rotor current (idr ). An active
ir β power component of rotor current (idr ) is kept zero for realiz-
sin θr =
. (22) ing the condition that the stator voltage and the grid voltage are
i2r α + i2r β in same phase. Now synchronization of DFIG with the grid is
56 IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 8, NO. 1, JANUARY 2017

Fig. 5. Steady state performance of the proposed DFIG-based WECS at Fig. 6. Steady state performance of the proposed DFIG-based WECS at fixed
fixed wind speed of 7 m/sec (rotor speed of 1100 rpm), (a) grid power (P G ), wind speed of 8.5 m/sec (rotor speed of 1300 rpm), (a) grid power (P G ),
(b) stator power (P S ) and (c) GSC power (P GSC ). (b) stator power (P S ) and (c) GSC power (P GSC ).

achieved by making switch S1 on. Then the DC machine con-

trol is changed from fixed duty mode to the wind turbine mode.
As the DFIG is under no load condition, the speed ramps up
to maximum depending upon the inertia and the field current
of the DC machine. Now the speed controller is made on. The
DFIG speed settles to the reference speed given by the MPPT
controller and starts feeding power to the grid. Now the GSC Fig. 7. Steady state performance of the proposed DFIG-based WECS at
control is made on to operate it under wide speed range. fixed wind speed of 9 m/s (rotor speed of 1500 rpm, (a) grid power (P G ),
(b) stator power (P S ) and (c) GSC power (P GSC ).

VI. RESULTS AND DISCUSSION

The steady state and dynamic behaviors of regulated power
DFIG based WECS is presented in this section. Test results are
recorded in terms of line voltage (vab ), grid currents (iga ,igb and
ig c ), stator currents (isa ,isb and isc ), GSC currents (iGSCa ,iGSCb
and iGSCc ), rotor currents (ira ,irb and irc ), stator power (PS ),
grid power (PG ), GSC power (PGSC ), battery voltage (Vb ),
battery current (Ib ), quadrature axis rotor current (Iqr ), direct
axis rotor current (Idr ), quadrature axis reference rotor current
∗ ∗
(Iqr ), direct axis reference rotor current (Idr ), rotor speed (ω r ),
∗
reference rotor speed (ω r ) and wind speed (vw ). The power
that is discharging from the battery through GSC and RSC are
considered as positive. Conversely, the battery charging is taken
as negative through both GSC and RSC.

A. Steady State Performance of Proposed WECS

Test results are presented for a specific wind speed in all
three cases such as sub-synchronous, synchronous and super-
synchronous speeds as shown in Figs. 5–7. In this study, a
regulated power is selected as 1.25 kW. So, the grid power is
maintained as 1.25 kW irrespective of the wind speed.
Fig. 5(a)–(c) shows the test results of proposed DFIG based
WECS for a fixed wind speed of 7 m/s. For achieving maxi-
Fig. 8. Steady state performance of the sensorless algorithm of constant
mum power from the wind turbine, the reference rotor speed is power DFIG-based WECS at fixed wind speeds (a)–(b) rotor speed (ω r ), unit
selected as 0.7 p.u. Fig. 5(a)–(c) shows PG , PS , PGSC respec- templates of rotor currents aligned to rotor axis (sin(θ r )), unit templates of
tively. From these results, one can observe that the grid power rotor currents aligned to stator axis (sin(θ s )) and unit vectors of rotor position
angle (sin(θ m )).
is maintained at 1.25 kW and stator power is 0.86 kW due to
the low wind speed. So the remaining power is coming from the
battery as shown in Fig. 5(c). at 1.25 kW as shown in Fig. 6(a). However, the stator power is
Test results of proposed WECS for a fixed wind speed of 1.41 kW as shown in Fig. 6(b). The remaining power is going
8.5 m/s, are shown in Fig. 6(a)–(c). The reference rotor speed is to GSC to charge battery as shown in Fig. 6(c).
selected a value of 0.86 p.u for operating at MPPT. Even for dif- Test results are shown in Fig. 7(a)–(c) at fixed wind speed
ferent wind speeds, the power feeding to the grid is maintained of 9.15 m/s (1500 rpm). According to MPPT, the reference
NAIDU AND SINGH: GRID-INTERFACED DFIG-BASED VARIABLE SPEED WECS WITH POWER SMOOTHENING 57

Fig. 9. Dynamic performance of proposed DFIG-based WECS under rise in wind speed, (a) rotor speed (ω r ), grid power (P G ), stator power (P S ) and GSC
power (P G S C ), (b) rotor speed (ω r ), rotor current (ira ), stator power (P s ) and grid power (P G ), (c) rotor speed (ω r ), battery current towards GSC (ib g ), battery
current towards RSC (ib r ) and grid power (P G ).

speed (ωr∗ ) is selected as 1500 rpm. Here also the grid power is
maintained at 1.25 kW as shown in Fig. 7(a). As the stator power
is more than the reference power command and the battery is
charging through GSC as shown in Fig. 7(c).
Fig. 8 shows the effective working of sensorless algorithm
at different rotor speeds. Where sin (θr ), sin (θs ), sin (θm )
represent unit templates of rotor currents aligned to rotor axis,
unit templates of rotor currents aligned to stator axis, and unit
vectors of rotor position angle respectively.

B. Dynamic Performance of WECS at Change in Wind Speed

Test results are also presented for dynamic changes in wind
speed as shown in Fig. 9. Fig. 9 demonstrates the successful
operation of proposed DFIG based WECS for the increase in
wind speed. With the increase in wind speed, reference rotor
speed is increasing for achieving MPPT operation. So the ac-
tual rotor speed is increasing because of the speed PI controller
as shown in Fig. 9(a). So the power generation increases from
the stator (PS ) as shown in Fig. 9(a). However, the grid power
(P G ) is maintained at constant value. At reduced wind speeds,
the stator power is less. So, the battery is discharging through
GSC. As the rotor speed increases, stator power increases, so the
GSC power is reversed and the battery is charging. As shown in
Fig. 9(b), the rotor currents have changed their phase sequence as
the speed varying from sub-synchronous to super-synchronous
speeds. Battery current flowing through GSC and RSC are get-
ting reversed with an increase in the wind speed as shown
in Fig. 9(c).
Fig. 10 shows the variation of stator power, rotor power, GSC
power fed to the grid in the proposed DFIG and power fed to
the grid in the conventional DFIG for different wind speeds.
The stator power and the power fed to the grid in a conventional
DFIG is increasing with the wind speed. However, the power
fed to the grid in the proposed DFIG is maintained constant as
shown in Fig. 10. In a conventional DFIG without BESS, the
rotor power is same as GSC power. However, in this proposed
DFIG, both GSC power and RSC powers are different. Because,
at low wind speeds the battery is discharging through GSC as
the power produced by the DFIG is less than the average power.
At high wind speeds, the BESS is charging through GSC as the
power produced by the DFIG is more than the demanded power.
Fig. 10. Variation in all powers with the increase wind speed.
VII. CONCLUSION
The proposed DFIG based WECS has been found capable
of supplying regulated power in all wind speeds by introducing
BESS in the DC link. The design of BESS in this DFIG based
WECS has been presented in detail. The control algorithm of
RSC has been implemented for sensorless and MPPT. The con-
trol algorithm of GSC has been modified for feeding regulated
power to the grid. The performance of proposed DFIG has been
validated through test results for the both fixed and variable wind
speeds at all possible rotor speeds. The satisfactory dynamic per-
formance of this proposed DFIG has been demonstrated through
test results. This proposed WECS has been found advantageous
for supplying power even at reduced wind speeds. Even for short
term, power demands can be supplied with this proposed DFIG.
APPENDIX
A. WRIM 3.7 kW, Rs = 1.32 Ω, Lls= 6.832 mH,R2 =
1.708 Ω, Llr = 6.832 mH, Rc = 419.646 Ω, Lm =
0.219 H, J =
0.1878 kg - m2 , stator to rotor turns ratio Nr /N s = 1/2,
stator rated rms current Is = 12 A, rotor rated rms current
Ir = 18 A.
B. DC Machine Ra = 1.3 Ω, Rf = 220 Ω, La= 7.2 mH, Lf =
7.5 mH, KΦ = 1.3314.
58
REFERENCES
[1] Manfred Stiebler, Wind Energy Systems for Electric Power Generation.
New York, NY, USA: Springer-Verlag, 2008.
[2] H.-J. Wagner and J. Mathur, Introduction to Wind Energy Systems Basics,
Technology and Operation. New York, NY, USA: Springer-Verlag, 2009.
[3] S. S. Murthy, B. Singh, P. K. Goel, and S. K. Tiwari, “A comparative study
of fixed speed and variable speed wind energy conversion systems feeding
the grid,” in Proc. IEEE 7th Int. Conf. Power Electron. Drive Syst., Nov.
27–30, 2007, pp. 736–743.
[4] M. Mansour, M. N. Mansouri, and M. F. Mimouni, “Comparative study of
fixed speed and variable speed wind generator with pitch angle control,”
in Proc. Int. Conf. Commun., Comput. Control Appl., Mar. 3–5, 2011,
pp. 1–7.
[5] R. Datta and V. T. Ranganathan, “Variable-speed wind power gener-
ation using doubly fed wound rotor induction machine-a comparison
with alternative schemes,” IEEE Trans. Energy Convers., vol. 17, no. 3,
pp. 414–421, Sep. 2002.
[6] L. Holdsworth, X. G. Wu, J. B. Ekanayake, and N. Jenkins, “Comparison
of fixed speed and doubly-fed induction wind turbines during power sys-
tem disturbances,” IEE Proc. Gener., Transmiss. Distrib., vol. 150, no. 3,
pp. 343–352, 2003.
[7] H. Polinder, F. F. A. Van Der Pijl, G. J. de Vilder, and P. J. Tavner, “Com-
parison of direct-drive and geared generator concepts for wind turbines,”
IEEE Trans. Energy Convers., vol. 21, no. 3, pp. 725–733, May 2005.
[8] S. Muller, M. Deicke, and R. W. De Doncker, “Doubly fed induction
generator systems for wind turbines,” IEEE Ind. Appl. Mag., vol. 8, no. 3,
pp. 26–33, May/Jun. 2002.
[9] R. Pena, J. C. Clare, and G. M Asher, “Doubly fed induction genera-
tor using back-to-back PWM converters and its application to variable-
speed wind-energy generation,” Proc. Elect. Power Appl., vol. 143, no. 3,
pp. 231–241, May 1996.
[10] C. Luo, H. Banakar, B. Shen, and B.-T. Ooi, “Strategies to smooth
wind power fluctuations of wind turbine generator,” IEEE Trans. Energy
Convers., vol. 22, no. 2, pp. 341–349, Jun. 2007.
[11] Y. Kim and R. Harrington, “Analysis of various energy storage systems
for variable speed wind turbines,” in Proc. IEEE Conf. Technol. Sustain.,
Ogden, UT, USA, 2015, pp. 7–14.
[12] J. P. Barton and D. G. Infield, “Energy storage and its use with inter-
mittent renewable energy,” IEEE Trans. Energy Convers., vol. 19, no. 2,
pp. 441–448, Jun. 2004.
[13] C. Abbey and G. Joos, “Super capacitor energy storage for wind energy ap-
plications,” IEEE Trans. Ind. Appl., vol. 43, no. 3, pp. 769–776, May/Jun.
2007.
[14] L. Qu and W. Qiao, “Constant power control of DFIG wind turbines with
super capacitor energy storage,” IEEE Trans. Ind. Appl., vol. 47, no. 1,
pp. 359–367, Jan./Feb. 2011.
[15] G. Mandic, A. Nasiri, E. Ghotbi, and E. Muljadi, “Lithium-Ion capacitor
energy storage integrated with variable speed wind turbines for power
smoothing,” IEEE J. Emerg. Sel. Topics Power Electron., vol. 1, no. 4,
pp. 287–295, Dec. 2013.
[16] R. Cardenas, R. Pena, G. Asher, and J. Clare, “Power smoothing in wind
generation systems using a sensorless vector controlled induction Ma-
chine driving a flywheel,” IEEE Trans. Energy Convers., vol. 19, no. 1,
pp. 206–216, Mar. 2004.
[17] S. Ghosh and S. Kamalasadan, “An energy function-based optimal con-
trol strategy for output stabilization of integrated DFIG-Flywheel energy
storage system,” IEEE Trans. on Smart Grid, vol. PP, no. 99, pp.1–10, to
be published.
[18] D. H. Doughty, P. C. Butler, A. A. Akhil, N. H. Clark, and J. D. Boyes,
“Batteries for large-scale stationary electrical energy storage,” in Proc.
Electrochem. Soc. Interface, Fall 2010, pp. 49–53.
[19] T. Kai and A. Tanaka, “A new smooth scheme for power fluctuation using
inverter of wind power generation with doubly fed induction generator,”
in Proc. Int. Conf. Elect. Mach. Syst., Oct. 17–20, 2008, pp. 2390–2395.
[20] S. Rahmani, A. Hamadi, A. Ndtoungou, K. Al-Haddad, and H. Y. Kanaan,
“Performance evaluation of a PMSG-based variable speed wind generation
system using maximum Power Point Tracking,” in Proc. IEEE Elect.
Power Energy Conf., Oct. 10–12, 2012, pp. 223–228.
[21] A. Kanchanaharuthai, V. Chankong, and K. A. Loparo, “Transient sta-
bility and voltage regulation in multimachine power systems vis-à-vis
STATCOM and battery energy storage,” IEEE Trans. Power Syst., vol. 30,
no. 5, pp. 2404–2416, Sep. 2015.
IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 8, NO. 1, JANUARY 2017
[22] Y. Mishra, S. Mishra, and Li Fangxing, “Coordinated tuning of DFIG-
based wind turbines and batteries using bacteria foraging technique for
maintaining constant grid power output,” IEEE Syst. J., vol. 6, no. 1,
pp. 16–26, Mar. 2012.
[23] V. C. Ganti, B. Singh, S. K. Aggarwal, and T. C. Kandpal, “DFIG-based
wind power conversion with grid power leveling for reduced gusts,” IEEE
Trans. Sustain. Energy, vol. 3, no. 1, pp. 12–20, Jan. 2012.
[24] A. Karthikeyan, C. Nagamani, and G. S. Ilango, “A versatile rotor position
computation algorithm for the power control of a grid-connected doubly
fed induction generator,” IEEE Trans. Energy Convers., vol. 27, no. 3,
pp. 697–706, Sep. 2012.
[25] S. Sharma and B. Singh, “Voltage and frequency control of asynchronous
generator for stand-alone wind power generation,” IET Power Electron.,
vol. 4, no. 7, pp. 816–826, Aug. 2011.
[26] Hourly Wind Energy Data. [Online]. Available: http://www.imd.
gov.in/section/nhac/aws/aws.htm
[27] G. Abad, J. López, M. Rodrı́guez, L. Marroyo, and G. Iwanski, Doubly Fed
Induction Machine: Modeling and Control for Wind Energy Generation
Applications. New York, NY, USA: Wiley-IEEE Press, 2011.
N. K. Swami Naidu received the B.Tech. degree in
electrical and electronics engineering from Jawahar-
lal Nehru Technological University, Hyderabad, In-
dia,, the M.Tech. degree in power electronics and
drives from the National Institute of Technology,
Kurukshetra, India, and the Ph.D. degree from the
Electrical Engineering Department, Indian Institute
of Technology Delhi, New Delhi, India, in 2007, 2009
and 2015, respectively.
He is currently working as a Research Fellow in
Nanyang Technological University, Singapore. His
research interests include power electronics, wind energy conversion systems,
power quality, and microgrid-based power systems.
Bhim Singh received the Bachelor of Engineering
(Electrical) degree from the University of Roorkee,
Roorkee, India, in 1977, and the M.Tech. degree in
power apparatus and systems and Ph.D. degree from
the Indian Institute of Technology Delhi (IITD), New
Delhi, India, in 1979 and 1983, respectively.
In 1983, he joined the Department of Electrical
Engineering, University of Roorkee, as a Lecturer,
and became a Reader in 1988. In December 1990,
he joined the Department of Electrical Engineering,
IITD, as an Assistant Professor, where he became an
Associate Professor and a Professor in 1994 and 1997, respectively. His research
interests include power electronics, electrical machines, electric drives, power
quality, renewable energy, flexible ac transmission systems, and high-voltage
direct current transmission systems.
Prof. Singh is a Fellow of the Indian National Science Academy, the Indian
National Academy of Engineering, The National Academy of Science, India,
The Indian Academy of Sciences, India, the World Academy of Sciences, In-
stitute of Electrical and Electronics Engineers, the Institute of Engineering and
Technology, Institution of Engineers (India), and Institution of Electronics and
Telecommunication Engineers, and a Life Member of the Indian Society for
Technical Education, System Society of India, and the National Institution of
Quality and Reliability.