Design of Doubly Fed Induction Generator Driven by Wind Turbine ...

IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498, 

Vol.2, No. 4, August 2012 

Design of Doubly Fed Induction Generator Driven by Wind Turbine under 

Unbalanced Voltage Conditions with Series Grid-Side Converter 

K.V.RAMANA 

ramana.kondangi@gmail.com, 

VITAM College of Engg, Visakhapatnam, India 

D.V.N.ANANTH 

nagaananth@gmail.com, 

VITAM College of Engg, Visakhapatnam, India 

ABSTRACT: The aptness of Wind Turbines with Doubly Fed Induction 

Generator for original grid operator that require low Voltage Ride Through 

operation is considered. This sophisticated grid technique require the wind 

turbines to remain connected even under voltage dips in the grid due to fault or 

any unpredicted perturbations. An analysis of such problems associated with 

voltage dips and methods for Ride Through operation of such system are 

presented. Simulation results of Power Error Vector Control method for the 

Voltage Source Inverter connected to the rotor during a voltage dip are shown. 

The simulated results show that this type of control may eliminate the need for a 

crowbar, or at least a great reduction in the rotor currents is obtained and 

activation of crowbar may be limited to very extreme cases when very strong wind 

gusts and very deep voltage dips take place at the same time. A DFIG system with 

series grid=side converter (SGSC) has an excellent potential for voltage dips 

tolerance. This paper analyzes the reasons of a DFIG system with SGSC for 

ride=through and presents a control scheme for operation under unbalanced grid 

faults conditions. During grid faults, the each component of the generator’s 

stator flux is effectively controlled through the SGSC. Also, the stator and rotor 

currents are further restricted by controlling the rotor=side converter (RSC) to 

reduce the absorbing energy from wind turbine. Then, successful ride-through of 

a DFIG system with SGSC under all types of severe unbalanced grid faults at the 

point of common coupling (PCC) are achieved with reduced electromagnetic 

torque oscillation. The proposed control scheme is validated by means of 

simulations. 

Keywords: Doubly fed induction generator (DFIG), series grid-side Converter 

(SGSC), unbalanced grid fault, wind turbine, low voltage ride-through (LVRT), 

wind power generation, voltage dips, voltage sags. 

1 INTRODUCTION 

Wind power has become one of the most important and 

promising sources of renewable energy that can partially solve the energy 

crisis and environmental dilemma we are confronting today [1]. With the 

increased penetration of wind power into power grids, its impact on a 

power grid cannot be neglected as before. Grid codes now usually require 

wind turbines to remain connected to the grid even during extreme 

voltage dips such as a short-term low or even a zero voltage event at the 

point of common coupling (PCC). This is the so-called “low voltage ridethrough 

(LVRT)” requirement adopted by most European countries [2,3]. 

The majority of wind turbines above 1MW are Doubly Fed Induction 

Generators (DFIGs) with a partially rated back=to=back converter used 

between the rotor circuit and the grid [4=8]. While a saving is made on 

the size of the power electronic converter, it is well known that a DFIG 

system with this configuration is very sensitive to grid disturbances, 

especially to voltage dips. During a fault, the transient current on the 

stator is reflected on the rotor windings. The resulting rotor current may 

exceed the rotor=side converter (RSC) current rating and destroy the 

converter if no protection elements are included. The mainstream scheme 

adopted by manufacturers to ride through grid faults is the so-called 

“crowbar protection” [9,10]. When the crowbar is in operation, the high 

fault current will flow through the crowbar instead of the converter. 

Thus, the RSC is well preserved during the fault. However, 

under such a scenario, the DFIG behaves as a squirrel=cage induction 

machine and consumes lots of reactive power from the grid for 

magnetization, which may further exacerbate the stability of the 

connected grid. Also, the operation of a crowbar is associated with large 

transient electromagnetic torques which will increase the fatigues on the 

mechanical components of the wind turbine generator system (WTGS), 

especially on the gearbox. Another LVRT technology is the rotor flux 

control technique [11], which controls the RSC to counteract the 

undesired components in the stator flux. However, the active and reactive 

power control over the machine is lost during the fault and the DFIG 

system also consumes reactive power during the grid recovery. 

Furthermore this method is unable to ride through the severe grid faults 

such as the voltage dips down to zero at PCC which means it is not 

suitable for the new grid codes. In order to fulfill the new grid codes 

which require wind turbines to ride through a short=term low or even a 

zero voltage event at PCC, modifications to he traditional DFIG 

configuration for ride=through have become necessary. Petersson [12] 

has proposed a new configuration with a grid=side converter in series 

with the stator windings of the DFIG and has revealed that DFIG system 

with this configuration has an excellent potential for voltage dips 

tolerance, but this scheme is not efficient in the power processing 

capability. 

This paper describes a technique to control DFIG under unbalanced 

voltage sags. In comparison with [15] and [16], it uses the 

approach based on separating the positive and negative components of all 

the currents and voltages, as suggested in [12] and [13] for dc/ac 

converters and applied to the DFIG as in in [14]. This paper introduces 

the following contributions: 1) The whole system is analyzed, 

considering both the grid-side and rotor-side converters. The grid-side 

converter control is not considered in in [14] or [16]. 2) A technique to 

keep the dc bus stable is proposed, based on compensating the rotor 

power delivered by the rotor-side converter in the grid-side converter. 3) 

The objective of the technique is to ride through voltage sags; hence, the 

main analyzed quantities are the generator torque and the dc voltage bus. 

4) Since this paper deals with ride through voltage sags, the crowbar 

protection is considered (it is not considered in [14] or in [16]). 

The compared to the balanced faults, due to the negative=sequence 

component injected by the unbalanced grid faults, a more severe current 

and torque oscillation will appear when unbalanced voltage dips occur 

[16]. Therefore, it is necessary to study the behavior of DFIG with this 

new configuration under unbalanced grid faults. This paper proposes an 

unbalanced-grid-fault ride-through control scheme for a DFIG wind 

turbine with the configuration explored in [14, 15, 17]. The paper has 

been organized as follows. Section 2 reviews the configuration of the 

DFIG wind turbine with SGSC. The behaviors of DFIG system with 

SGSC under unbalanced grid faults are analyzed in Section 3. The 

proposed control scheme is designed in Section 4 and validated by means 

of simulations in Section 5. Discussion is provided in Section 6 and 

finally Section 7 draws the conclusions. 

2 CONTROL SCHEME - BALANCED CONDITIONS 

A DFIG wind turbine primarily consists of three parts: a wind turbine 

drive train, an induction generator, and a power electronic converter (Fig. 

1) [2], [4]. In the wind turbine drive train, the rotor blades of the turbine 

catch wind energy that is then transferred to the induction generator 

through a gearbox. The induction generator is a standard, wound rotor 

induction machine with its stator windings directly connected to the grid 

and its rotor windings connected to the grid through a frequency 

converter. The frequency converter is built by two self-commutated 

voltage-source converters, the RSC and the GSC, with an intermediate dc 

voltage link. The control in a DFIG wind power plant has three levels: 

the generator level, the wind turbine level, and the wind power plant 

level (Fig. 1) [4]. At the generator level, the RSC controller regulates the 

DFIG to achieve one of the following two goals: 1) maximum energy 

extraction from the wind or 2) compliance with a wind plant control 

demand; the GSC controller maintains a constant dc-link voltage and 

adjusts reactive power absorbed from the grid by the GSC. At the turbine 

level, there are a speed controller and a power limitation controller. At a 

681



low wind speed, the speed controller gives a power reference to the RSC 

based on the principle of maximum energy extraction. At a high wind 

speed, the power limitation controller increases or decreases the pitch 

angle of the turbine blades to prevent the wind turbine from going over 

the rated power. At the wind power plant level, the power production of 

the entire plant is determined based on a grid requirement. The central 

control system sends out power references to each individual turbine, 

while the local turbine control system ensures that the power reference 

from the central control level is reached. 

A. GSC Transient and Steady-State Models 

Fig. 2 shows the schematic of the GSC, in which a dc-link capacitor is on 

the left and a three-phase voltage source, representing the voltage at the 

point of common coupling (PCC) of the ac system, is on the right. In the 

grid-side converter, the dc bus voltage and reactive power references 

determine the current references, which determine the voltages to be 

applied in the grid side.1) System Equations: In a synchronous reference 

frame, the grid-side voltage equations can be written as 

B. Rotor-Side Converter 

In the rotor-side converter, the referenced torque and reactive power 

determine the current references, which determine the voltages to be 

applied in the rotor side. 

1) Machine Equations: It is usually assumed that when the stator 

and rotor windings are placed sinusoidally and symmetrically, 

the magnetical saturation effects and the capacitance of all the 

windings are negligible. The relation between voltages and 

currents on a synchronous reference qd can be written as 

Linkage fluxes can be written as 

The torque can expressed as 

The reactive power yields 

Fig. 1. General system scheme. 

Active and reactive power provided by the grid-side converter can be 

written as Pz =3/2(vzq ilq + vzd ild) and Qz =3/2(vzq ild − vzd ilq ).The 

dc bus voltage can be expressed as 

2) Reference Quantities: The grid-side converter controls the reactive 

power and dc bus voltage. The q-axis may be aligned to the grid voltage 

allowing active and reactive decoupled control. To control the reactive 

power, a i ld reference is computed as 

2) Reference Quantities: Orientating the synchronous reference 

qd with the stator flux vector so that λ sd =0,therotor current 

references can be computed as 

Current Loops Implementation: The control of the current is done by 

linearizing the current dynamics using the following state feedback. 

By neglecting stator current transients, the decoupling leads to 

C. Current Controllers Tuning 

Controllers have been designed using the so-called internal mode control 

(IMC) methodology detailed in [19]. The parameters of a PI controller to 

obtain a desired time constant τ are obtained as 

The active power, which is responsible for the evolution of the dc bus 

voltage, is controlled by the i lq component. A linear controller is usually 

designed to control the dc bus voltage. 3) Current Loops Implementation: 

The current control is done by the following state linearization feedback 

[18]: 

where the ˆvlq and ˆvld are the output voltages of the current controller. 

The decoupling leads to 

The currents and voltages have been limited according to the converter 

operating limits. PI controllers have been designed with anti-windup in 

order to prevent control instabilities when the controllerexceed the limit 

values. 

D. Crowbar Protection 

The so-called crowbar is connected to avoid over voltages in the dc bus 

due to excessive power flowing from the rotor inverter to the grid-side 

converter, guaranteeing ride through operation of the generator when 

voltage sags or other disturbances occur. The crowbar is triggered when 

682



the dc voltage reaches a thresh-old vcrow−c and disconnects when it goes 

below another thresh-old vcrow−d . 

During its operation, the rotor-side converter may be disconnected, as 

described in [20], or be kept connected [9] to avoid losing control over 

the machine. In this paper, the rotor-side converter is kept connected. 

Crowbar operation is a temporary measure developed as a 

rotor circuit protection device, long before the advent of wind turbine 

grid-code regulations. The term describes any device that connects 

together the rotor windings of a wound-rotor machine in a closed circuit 

through a designated resistance, diverting current from the rotor-side 

converter, and rapidly deenergizing the rotor. The crowbar absorbs the 

initial energy outflow from the machine, while the resistance shortens the 

effective decay timescale of the rotor flux decay, hence hastening the 

demagnetization process. Conventionally, the crowbar is applied for an 

extended duration to fully demagnetize the rotor [24], [25]. A crowbar 

activation period typically lasts about 120 ms. Unfortunately, this can 

lead to a 100-ms postfault period of at least 50% reactive power 

absorption and associated voltage suppression, and thus, failure to meet 

grid-code FRT requirements. 

systems are concerned, it is useful to express three-phase quantities xabc 

= {xa ,xb ,xc }T in direct and inverse components as 

Where x =2/3(xa + axb + a 2 xc), a = e j2π/3 , x p = x p d +jx p q , and x n = x n d + jx n q 

. In this section, voltages, currents, and fluxes are regarded as a 

composition of such positive and negative sequences. 

A. Grid-Side Converter Analysis 

1) Voltage Equations: Considering two rotating reference frames 

at +ωe and −ωe , the voltage equations for the positive and 

negative sequences yield 

2) Active and Reactive Power: Active and reactive power can be 

written as [13] 

Where 

Fig. 3. Crowbar circuit 

When the crowbar is engaged, vector control is lost and the 

DFIG resembles a high-resistance singly fed machine, operating at very 

high slip [28]. This temporary configuration is a worstcase combination 

of poor torque output and very high reactive-current demand. This 

reactive power absorption can only occur on the stator, acting to suppress 

the local fault voltage. The poor torque output worsens the mechanical 

stability. It is therefore necessary that crowbar application periods are 

minimized. The failure of a DFIG system to meet FRT requirements 

when using a crowbar for periods of 100 ms or more has been reported 

[29]. Other more complex FRT techniques have also been suggested. 

For example, Kasem et al. [30] presented a simulation study 

of a scheme employing a crowbar circuit in conjunction with an 

additional series static switch to reconfigure the rotor circuit on fault 

detection. In this arrangement, the rotor-side converter is totally isolated 

from the rotor circuit for the duration of the fault and connected in 

parallel with the supply-side converter to inject reactive power into the 

supply. Because of the isolation of the rotor converter, however, the 

DFIG cannot provide active power in proportion to the retained balanced 

fault voltage. Furthermore, the crowbar is applied and the machine is left 

uncontrolled for the duration of the fault. This means that the potential 

benefits of the reactive power export from both converters would, to a 

large extent, be cancelled by the reactive power requirement of the 

crowbar-activated induction machine operating at high slip. Any practical 

implementation of such a scheme would also suffer from higher on-state 

losses and higher costs compared with a standard crowbar system 

because of the inclusion of the static series switch and the current-sharing 

reactances needed to reduce circulating currents during the periods in 

which the two converters are connected in parallel. 

It can be noted that both active and reactive power have three different 

components each, and hence with the four regulatable currents i p ld , i p lq , 

i n ld , and i n lq , only four of such six powers can be controlled. 

B. Machine-Side Converter Analysis 

1) Voltage Equations: Considering two rotating reference frames 

at +ωe and −ωe , the voltage equations for the positive and negative 

sequences can be obtained as 

2) Stator Power Expression: The apparent stator power can be 

expressed as 

Using(14),wehave 

III. CONTROL SCHEME UNDER UNBALANCED CONDITIONS 

In this section, nonsymmetrical voltage sags are considered. Such 

unbalanced sags imply negative sequence components in all the relevant 

quantities. Therefore, important oscillations appear in torque, active and 

reactive power. Such oscillations have a pulsation of 2ωe . In order to 

mitigate such oscillations, an approach taking into account the negative 

sequence quantities is required. Such an approach has been discussed in 

[12] and [13], and has been applied to the rotor-side converter of a DFIG 

in [14]. This section analyzes a whole back-to-back converter taking into 

account both the positive and negative sequence components, and 

proposes a technique to control optimally both the dc bus voltage and the 

torque when unbalanced voltage sags occur. As far as unbalanced 

Taking into account x i s= x i sd+ jx i sq, and rearranging it gives 

where 

683



Substituting stator currents in (27) 

and Q∗l0 for the grid-side converter. It can be noted that P∗l0 , P∗lcos, 

and P∗lsin are directly linked to the dc bus voltage. 

The dc voltage E is regulated by means of a linear controller whose 

output is the power demanded by the grid-side converter. Considering the 

power terms Pr0 , Prcos , and Prsin in the rotor side converter, Pr0 can be 

regarded as the average power delivered, while Prcos and Prsin are the 

rotor power oscillating terms. Such terms will cause dc voltage 

oscillations, and hence they can be canceled by choosing 

Pl0 can be computed as 

where P∗E is the output of the dc voltage linear controller. 

it can be noted that both active and reactive power quantitities 

have three different components each, and therefore, with the 

four regulable currents i p rd , i p rq , i n rd , and i n rq , only four of the 

six power quantities can be controlled. 

3) Rotor Power Expression: The apparent rotor power can be 

expressed as 

Using(14),wehave 

3 Behaviors of DFIG System with SGSC under 

Unbalanced Grid Faults 

3.1 Dynamic Response of Traditional DFIG System under 

Unbalanced Grid Faults 

In the traditional DFIG configuration, the stator is directly connected to 

the grid. An abrupt drop of the 

grid voltage will directly transmit to the stator terminals of DFIG. 

According to the law of constant flux, additional negative=sequence and 

transient DC flux components in the stator flux will appear to guarantee 

the total stator flux constancy. Therefore, under unbalanced grid faults, 

the generator stator flux will contain positive=sequence, negativesequence 

and transient DC flux components. The relationship of each 

component between stator flux and stator voltage can be expressed as 

(37) 

Taking into account x i s = x i sd + jx i sq , and rearranging and 

analyzing the active rotor power 

where 

where subscripts 1, 2, DC denote positive=, negative-and zero= (DC) 

sequence components in the flux or voltage vectors, respectively. Ψs, ω0 

and τs are the stator flux vector, synchronous angular speed and stator 

time constant, respectively. Us and U's are the stator voltage vectors 

just before and after the grid fault. u's represents the stator voltage 

vector after the grid fault. 

Each component of the stator flux as shown in (1) will induce a voltage 

in the rotor according to its amplitude and its relative speed with respect 

to the rotor. The induced rotor voltages can be expressed in the stationary 

reference frame as [18] 

3) Torque Expression: Analogously, electrical torque can be 

expressed as 

Where 

(38) 

where Lm, Ls, ur and ωr are the mutual inductance, stator self 

inductance, rotor voltage vector and rotor speed, respectively. 

Substituting (1) into (2), the induced rotor voltages can be expressed in 

the rotor reference frame as Table 1 Amplitudes of stator flux 

components and their induced voltages in the rotor side under phase-tophase 

grid fault. 

C. Reference Current Calculation 

Since there are eight degrees of freedom (the rotor-side currents iprd , 

iprq , inrd , and inrq , and the grid-side currents i p rd , i p rq , i n rd , and i n rq, 

eight control objectives may be chosen. This implies that it is not 

possible to eliminate all the oscillations provoked by the unbalance. In 

this paper, the main objective is to ride through voltage dips. Hence, it is 

important to keep the torque and dc bus voltage as constant as possible 

and to keep reasonable values of reactive power. To this end, it has been 

chosen to determine the currents to keep certain valuesof Γ∗0 , Γ∗cos , 

Γ∗sin , and Q∗s0 for the rotor-side converter and P∗l0 , P∗lcos, P∗lsin 

(40) 

684



(42) 

Splitting the terms R r/s and u r /s into R r+(1-s)R r /s and u r /s+(1-s)u r /s, 

(42) can be rewritten as 

where s is the slip and superscript r denotes the rotor reference frame. 

Table 1 shows the approximate amplitudes of each stator flux component 

and their induced voltages in the rotor side under phase=to=phase grid 

fault. The symbol p denotes the depth of the grid fault. As can be seen in 

Table 1 and (3), relative large voltages are induced in the rotor side. For 

the worst case (s=–0.3 and p=1), the maximum voltage amplitudes 

induced by the negative=sequence and the transient DC flux components 

can reach 1.15pu and 1.3pu, respectively. Considering that the rotor 

resistance and the transient inductance are usually very small, if the RSC 

cannot generate a voltage equal to the sum of the three voltages induced 

by the stator flux components, the RSC will lose current control 

transitorily and rotor over=current will appear. Fig. 2 shows the 

simulation results of a 2MW traditional DFIG detailed in the Appendix 

under phase=to=phase grid fault at PCC. The machine initially operates 

with full load and is at 30% super=synchronous speed. The control 

scheme implemented in the simulation model is the vector control 

algorithm to achieve the decoupling control of stator active and reactive 

powers and to keep the common dc link voltage constant [4]. No 

additional action or control limit is included to constrain the fault current. 

As shown in Fig. 2, due to the RSC cannot generate a voltage to 

counteract the induced voltage by the stator flux components, high rotor 

over=current appears and further causes severe oscillation of the 

electromagnetic torque. 

Thus, to avoid the appearance of rotor over=current and guarantee the 

DFIG system to ride through the grid faults, the stator flux each 

component should be effectively restricted, especially for the negative= 

sequence and the transient DC flux components due to the relative 

speeds of which with respect to the rotor windings are rather high. 

Besides, the severe torque oscillation, which will increase the fatigues on 

the mechanical components of WTGS, should also be considered. 

3.2 Principle for Suppressing the Rotor Over- Current 

As shown in Fig. 1, a SGSC is introduced between the grid and the stator 

terminals of DFIG. The generator’s stator terminal voltage us becomes 

the sum of the grid voltage vector u i and SGSC series injected voltage 

vector u g which can be expressed in the stationary reference frame as 

(41) 

Equation (41) indicates that the generator’s stator terminal 

voltage can be changed due to the existence of the SGSC. Also, the 

transition of the generator stator flux imposed by the stator terminal 

voltage during the grid faults can be changed through the output voltage 

of SGSC. As analyzed in Section 3.1, under unbalanced grid faults, the 

negative-sequence and the transient DC flux components will induce 

relative large voltages in the rotor side. If the SGSC can generate a 

voltage to counteract the negative-sequence and the transient DC flux 

components in the stator side, the rotor over=current caused by these two 

components can be eliminated. On the other hand, if the DFIG equivalent 

mechanical output power P mec imported from the rotor shaft is still large, 

an abrupt drop of the grid voltage positive=sequence component will 

cause the increase of the positive=sequence current which also maybe 

cause the rotor over=current. So the positive-sequence current should 

also be restricted. 

On the other hand, if the DFIG equivalent mechanical output 

power Pmec imported from the rotor shaft is still large, an abrupt drop of 

the grid voltage positive=sequence component will cause the increase of 

the positive=sequence current which also maybe cause the rotor 

over=current. So the positive= sequence current should also be restricted. 

Neglecting the impedance of the series transformer, the electromagnetic 

power of DFIG can be written as 

(43) 

In (43), the first term is the power loss in the rotor windings. The third 

term is the power input from the excitation power supply to the rotor 

side. The remaining terms are the DFIG equivalent mechanical output 

power Pmec imported from the rotor shaft, which can be expressed as 

(44) 

Fig. 3 shows the equivalent circuit of DFIG with Rs, Xls, Rr, Xlr, im 

and Xm the stator resistance and leakage reactance, rotor resistance and 

leakage reactance, magnetizing current and reactance based on (42)-(44). 

All the quantities in Fig. 3 are referred to the stator side. 

Fig 4. Equivalent circuit of DFIG 

Neglecting the power loss in the DFIG, the stator active power output Ps 

is approximately equal to the electromagnetic power Pem. If Ps is 

controlled to zero, Pem is also approximately equal to zero. So do both 

the rotor active power output Pr and the DFIG equivalent mechanical 

output power Pmec imported from the rotor shaft according to (5) and 

(7). Also, if Pem is zero, the relative positive=sequence stator and rotor 

currents can also be restricted according to (5), which can increase the 

possibility of the DFIG system to ride through the grid faults. It is also 

important to note that the rest power will be stored in the mechanical 

inertia of the wind power generation system, causing the speedup of the 

DFIG. If Pmec is reduced to zero in the super=synchronous speed 

operation, the energy absorbed by the turbine rotor for a 2MW wind 

turbine system with a grid fault is approximately equal to 

(45) 

where ∆t is the duration of the grid fault. The relationship between the 

energy and generator speed can be expressed as [19]. 

(46) 

where J is the moment of inertia, and ω1 and ω2 are the speed of 

generator before the grid faults instant and after the recovery voltage 

instant, respectively. Also, the relationship between the moment of 

inertia J and the inertia constant H is 

(47) 

where P nom is the rated power of generator. Assuming the speed of 

generator before the grid fault is 1.3pu. According to (8)=(10), and with 

the parameters provided in the appendix, the calculated speedup of the 

generator with a 150ms grid fault is approximately 31r/min, which is 

about 1.5% of the generator speed. Thus, the speedup of generator is 

quite small, and the scheme of controlling the DFIG active power output 

Ps to zero is feasible during a short-time grid fault. However, for a longer 

time grid fault, the emergency pitch angle regulation scheme should be 

used to reduce the incoming torque. Otherwise, over=speed will occur 

and the over=speed protection will trip the turbine. 

3.3 Active Power Flow of DFIG System with SGSC under Grid 

Faults 

685



Fig. 4 shows the active power flow of the DFIG system with 

SGSC under grid faults during supersynchronous speed operation. As 

shown in Fig. 4, if the output power is still large enough during the grid 

faults, the surplus active power will flow into the common dc link. 

Meanwhile, the active power from the RSC to the common dc link is also 

quite large. Considering the limitation of the PGSC rating, these active 

powers cannot be delivered to the grid through the PGSC. The common 

dc link voltage will rise sufficiently and exceed its safe limitation. 

Therefore, during the grid faults, the active power output of DFIG should 

be scaled with the remaining voltage or reduced to an even lower value to 

protect the dc link capacitor. Otherwise, an additional dc link crowbar is 

needed. Thus, the reduction of DFIG active power output not only 

increases the possibility of the DFIG system to ride through the grid 

faults, but also can effectively protect the dc link capacitance. 

Note that when the DFIG system operates in the 

sub=synchronous speed, the surplus active power through the SGSC to 

the common dc link can be partially or completely consumed by DFIG 

through RSC. Therefore the regulation stress of the PGSC in the 

sub=synchronous speed is smaller than the one in the super=synchronous 

speed. Similarly, when DFIG system operates in the synchronous speed, 

although the active power from the SGSC cannot be consumed by DFIG 

through RSC anymore, the active power from the RSC to the common dc 

link is zero. So the regulation stress of the PGSC in the synchronous 

speed is also smaller than the one in the super= synchronous speed. So 

the super=synchronous speed operation condition is the worst case 

scenario for the PGSC to keep the common dc link voltage constant. And 

for this reason, we only show the simulations of DFIG in the supersynchronous 

speed operation for the sake of brevity in Section 5. 

the rotor side and accelerate its attenuation and the positive=sequence 

stator flux controller is only used to control the generator stator 

positive=sequence flux component to match the remaining grid voltage. 

A detailed description of the proposed control scheme of DFIG system 

with SGSC for the grid unbalanced faults ride=through is given as 

follows: For RSC and PGSC, the control schemes are still the same as the 

normal grid conditions. But for the RSC, the active power output 

command of DFIG is set to zero to increase the possibility of the DFIG 

system to ride through the grid faults and effectively protects the dc link 

capacitance as mentioned in Section 3. For the PGSC, an area of concern 

with unbalanced grid faults is the stabilization of the common dc link 

voltage. This was checked in simulation and it was found that the gridside 

converter could be controlled to avoid over=current during the 

unbalanced grid faults [11, 20]. Also, with other improved control 

scheme of PGSC such as the one mentioned in [21] the common dc link 

voltage can be more readily stabilized. We have focused our 

investigations on the control scheme of SGSC, and the control scheme of 

PGSC isn’t discussed too much for the sake of brevity. The SGSC is 

activated to control the each component of stator flux as shown in Fig. 5 

when a grid fault is detected. 

Fig. 5 shows the proposed controller of SGSC including three 

parts: 1) DFIG stator flux estimation and decomposition; 2) Generation 

of stator positive= sequence flux component command and the setting of 

stator negative=sequence and transient DC flux components command; 

3) Implementation of SGSC output series voltage control. The 

generator’s total stator flux can be calculated as follows using the 

measured stator terminal voltage and current in the stationary reference 

frame 

(48) 

Based on the analysis in Section 3.1, Ψs contains Ψs1, Ψs2 

and ΨsDC. In the positive reference frame (PRF) rotating at the speed of 

ω0, Ψs1 will be changed to a dc component while Ψs2 and ΨsDC will 

be changed to ac components at the frequencies of 2ω0 and ω0, 

respectively. Similarly, in the negative reference frame (NRF) rotating at 

the speed of –ω0, Ψs2 will be changed to a dc component while Ψs1 and 

ΨsDC will be changed to ac components at the frequencies of 2ω0 and 

ω0, respectively. Thus, Ψs in the PRF and NRF can be expressed as 

Fig 5. Active power flow of DFIG system with SGSC under grid faults 

during supersynchronous speed operation 

4 Proposed Control Scheme for Grid Unbalanced 

Faults Ride-through 

During normal grid condition, the RSC and PGSC are 

controlled in conventional manners to achieve the decoupling control of 

stator active and reactive powers and to keep the common dc link voltage 

constant, respectively [4]. And for the SGSC, it is maintained in a zero 

voltage vector switch state to eliminate harmonic losses as shown in [14, 

17]. During the severe unbalanced grid faults, Flannery and 

Venkataramanan [17] proposed that the negative-sequence component of 

the stator flux is kept at zero to reject the undesirable effect to the rotor 

side while the positive=sequence component of the stator flux is to scale 

in proportion to the positive-sequence component of the remaining grid 

flux to eliminate the transient DC flux component and to restrain the 

surplus active power flowing into the common dc link. As a result, in 

order to avoid the appearance of the transient DC flux component and 

provide the required system response, the positive=sequence stator flux 

controller must be carefully tuned. On the other hand, since the transient 

DC flux component is not considered as the control target to be 

restricted, once the transient DC flux component appears, it will only be 

damped with the stator time constant τs. However, τs is relatively big 

(about a few seconds) for a MW-class DFIG, thus the attenuation speed 

of the transient DC flux component is rather slow. 

Therefore, in order to overcome the problems highlighted 

above, we propose that the transient DC component is controlled in the 

negative=sequence stator flux controller to suppress the adverse effect on 

(49) 

Applying two notch=filters [16] at 2ω0 and ω0 to remove the 

negative=sequence and transient DC flux components, the 

positive=sequence flux component in the PRF can be obtained. 

Analogously, applying a notch=filter at 2ω0 to remove the 

positive=sequence flux component in the NRF, the negative=sequence 

and the transient DC flux components can be obtained. Thus, after 

filtering, the flux components in the PRF and NRF reduce to 

(50) 

In addition, during unbalanced grid faults, the grid flux Ψg 

(including stator ohmic drop) can be expressed in the PRF as 

(51) 

Equation (16) indicates Ψg also contains positive-sequence, 

negative=sequence and transient DC flux component, which can be 

expressed in the PRF as 

(52) 

Similarly, applying two notch-filters at 2ω0 and ω0 to remove the 

negative=sequence and transient DC flux components, the positivesequence 

component of the grid flux can be obtained. This is the 

686



command for the positive-sequence stator flux component controller, 

viz.: 

(52) 

For the negative=sequence and transient DC flux components, they are 

undesirable components in the stator flux. Thus the command for these 

two flux components is zero, viz.: 

(53) 

The proportional regulators are driven by the error between the command 

and their respective estimated stator flux components expressed in (14) 

and (15). Note that the command for the positive=sequence stator flux 

component controller has some important characteristics which are 

summarized as follows: 1) Equation (18) also can indicate the grid flux 

during normal operation. Thus when the grid fault is cleared, this 

command can bring the stator flux back to the normal value. 2) A 

reduction of the stator positive=sequence flux component also causes a 

reduction of corresponding induced rotor voltage. This means that there 

is more rotor voltage margin to generate a voltage to control the stator 

active power output of DFIG to zero and to counteract the remaining 

negative=sequence and transient DC flux components which cannot be 

rejected completely by SGSC due to the proportional regulator. 

modulation (SVPWM) techniques. The technique is exemplified for the 

rotor-side converter case in Fig. 7. 

5 SIMULATION RESULTS 

In order to evaluate the ride=through capability of the proposed control 

scheme, the system has been simulated on a 2MW DFIG wind turbine 

with SGSC. 

In order to evaluate the ride-through capability of the proposed scheme, 

the system has been simulated with severe voltage sags, making the 

control work at the maximum output voltage and dealing with the 

triggering of the crowbar protection. The system under study is a 2- MW 

DFIG-based wind turbine, where a two-phase 50% type E [22] voltage 

sag of 2s has been applied. The data of the simulated system may be seen 

in Table I. In order to compare the presented control scheme and some 

existing techniques, the following three cases have been studied. 

In this study, there are two cases analyzed, 1 st is without controller and 

2 nd is with controller techniques. 

Fig. 8 shows the simulation results of the DFIG system for 

double phase to ground fault near PCC. It is of unbalanced grid faults 

double phase-to-ground fault. As analyzed in [22], the propagation of 

voltage dips through different types of transformer connections results in 

a different performance of voltage dips on the secondary side of the 

transformers. 

The different voltage dips types through a delta/Yg 

transformer seen at the DFIG stator terminal compared to all types of 

unbalanced grid faults at PCC are shown in Fig6. The proposed control 

schemes have been evaluated by means of simulations with one balanced 

and one unbalanced voltage sags. 

Unbalanced voltage sag 

Fig 6. Positive and Negative Components Calculation 

1) Positive and Negative Components Calculation: The positive and 

negative sequence components calculation is done by using the Clarke 

transformation, rotating either e jωet or e −jωet , and finally, applying a notchfilter 

at 2ω e to eliminate the opposite sequence. The technique is 

exemplified in Fig. 6. For the rotor voltages and currents, the rotation 

applied is either e j (ωe −ωr )t or e j (−ωe −ωr )t . 

2) Reference Orientation: The rotating references have been aligned with 

the stator voltage so that v p sq =0. Nevertheless, v p sq has not been 

substituted in previous expressions for the sake of describing general 

results. Orientation may be done computing the required θ0 assuming a 

constant ωe or using a PLL [21] to determine both ω e and θ 0 . 

Fig 7. Control technique 

3) Output Voltage Calculation 

The output voltages calculation is done by summing the resulting 

positive and negative sequence voltages in the stationary reference frame. 

For the line side 

For the rotor side 

(54) 

The MATLAB based proposed test system is as shown in Fig 6. The 

DFIG wind turbine system is connected to the grid by using a step-up 

transformer. The subsystem with DFIG wind turbine set with and without 

controller was shown in Fig 9 and 10a. The back to back converter 

subsystem for Fig 8a is shown in Fig 10b and crowbar control scheme is 

as shown in Fig 10c. This converter converts Ac to DC at different 

voltages and frequency to DC supply and then to a constant magnitude 

and frequency AC voltage and current by using an inverter. As wind 

speed is varying with time and with season, the generator which is 

connected to the turbine will also have different speeds of rotation, so 

mechanical input varies, so voltage and current varies. Also because of 

generator speed varies with wind speed, output electrical power 

frequency from the generator also varies. Therefore the necessity to 

convert AC to DC and finally to AC has come to light. 

A50% voltage sag has been applied to two phases leaving the 

third phase undisturbed. The disturbance is produced by fault generator 

block as shown in Fig 8. The disturbance has been analyzed in a 2MW 

wind turbine was generating with a wind of 20m/s. The DFIG stator and 

rotor side voltages are shown in Fig 11 without controller. There is a 

fault between 0.1to0.2sec. The generator torque and DC bus voltage 

response are illustrated in Fig. 12 and 17. It can be seen that although the 

inverse sequence provokes an oscillating flux, an almost constant torque 

can be achieved after a transient. As it has been stated, the constant 

torque implies oscillating rotor power which can be compensated with 

oscillating grid-side converter power. Using the proposed technique, the 

resulting DC bus voltage has minimized the oscillations. 

The stator voltages in positive and negative sequence and the 

abc stator voltages and currents are illustrated in Fig. 13. Active and 

reactive power is illustrated in Fig.14. It can be seen that while the total 

power (depending on the torque) is almost constant, stator and rotor 

active power are of oscillating nature. 

(55) 

The resulting voltages are limited according to the converter rating. The 

final voltages can be applied using standard space vector pulse width 

687



Fig 8, MATLAB/ SIMULINK based test system, DFIG wind turbine 

system connected to Grid 

Fig 10c,Crowbar Controller subsystem 

Case A: DFIG system without crowbar and back-to-back 

controller schemes 

Fig 9, DFIG wind turbine subsystem block without controller 

Fig 11a, DFIG stator and rotor side voltages without crowbar controller 

and back-to-back controller 

Fig 10a, DFIG wind turbine subsystem block with back-to-back 

controller and crowbar controller scheme designed using MATLAB 

Fig 11b, zoomed view of DFIG stator and rotor side voltages without 

crowbar controller and back-to-back controller 

Fig 10b, Controller subsystem block – back to back AC converter 

Fig 11c, dqo voltages for stator side voltage of induction generator 

688



It can also be identified that, voltage harmonics are very high, it means, 

there is some distortions in voltage waveforms even under steady-state 

conditions. 

Case B: DFIG system with crowbar and without back-toback 

controller scheme 

Fig 12 Torque response to a un-balanced voltage sag 

Fig 13 Stator voltage in positive and negative sequence 

The DFIG system connected to grid is considered in this case with 

crowbar protection only. The crowbar works as a short circuit of rotor by 

using diode rectifier with resistance connection and IGBT to control the 

current through the resistance. The circuit is shown in figure 10c. 

A symmetrical fault occurred at 0.1seconds and cleared in 0.5seconds 

with crowbar protection for second case and the waveforms for stator and 

rotor voltages for DFIG system is as shown in Fig 17a and 17b. 

Comparing with Fig 11a and 11b, the voltages also decreased as in with 

and without crowbar protection. With crowbar, the system will be 

protected, so no voltage surge is produced in post-fault condition. Hence 

we can strongly say that crowbar protection is highly essential for basic 

backup protection for DFIG system connected to grid. 

In Fig 19, the DC voltage and grid voltage waveforms were shown. 

During the fault, excess voltage is reached to dc supply and is stored in 

the capacitor bank. It can also be identified that, with the proposed 

controller, the stator voltage and rotor voltage phase sequence is changed 

as in Fig 14, there-by maintaining nearly constant voltage near the grid. 

Fig 14 Rotor, stator and total active power 

Fig 17a. DFIG stator and rotor side voltages with crowbar controller as 

shown in Fig 10c and without back-to-back controller 

Fig 15 stator reactive power 

When comparing figures 11 and 16, during a double phase to ground 

fault, there has been severe change in the voltage waveforms. Without 

controller, the system after post-fault is very unstable as seen in Fig 8, 

voltage has never reached its rated value amd is decreasing in nature. But 

in Fig 16, which is same system with same fault conditions, two 

unhealthy phases got disturbed, and the system after post- fault is very 

stable. It has reached its pre fault state very immediately. 

In Fig 13, positive and negative sequence components were shown. The 

positive sequence voltage has changed very drastically and there been 

some surges for negative sequence components. It can be observed from 

Fig 14 and 15, during the fault conditions, the real and reactive powers 

were disturbed very drastically. The real power from rotor, stator and the 

composition of both were changed as in Fig 14. I this, rotor real power 

which is nearly -10KW, has become oscillating and is sinusoidal during 

the fault and many oscillations in stator real power. 

The system without crowbar and back-to-back controller, the system will 

slowly becomes unstable even after the fault has been cleared. It can be 

observed from Fig 11a and 11b, slowly voltages on rotor and stator side 

degrades and current increases, torque also changes its polarity, hence it 

leads to system collapse. Hence, crowbar has to be used as a primary 

requirement for system protection. 

Fig 17b. Grid side voltages with crowbar controller and without back-toback 

controller 

The fault moves in the direction of DFIG side than towards grid side, so 

there is a very high dip in the voltage waveforms near the generator side 

than on grid side. This is due to the fact that grid is a very strong system 

and disturbance effects it very little compared to a wind generator 

system. 

Case C: DFIG system with crowbar protection and with 

back-to-back controller schemes 

The system with DFIG connected to grid is now protected with crowbar 

protection and back-to-back controller schemes. 

689



Fig 18, DFIG stator and rotor side voltages 

been considered, detailing the control scheme of each converter while 

considering the effect of the crow-bar protection. The control strategy has 

been validated by means of simulations for balanced and unbalanced 

voltage sags. 

In this paper the modeling of a DFIG, considering the behavior of the 

generator when transients in the stator voltage occur, was developed. 

Thanks to this model, that permits to predict the performance of a DFIG 

under faulty scenarios, a novel control strategy for the rotor side 

controllers, oriented to enhance its response during severe voltage sags, 

was proposed. This strategy is based on using the measured stator current 

values as the set-point for the rotor current controller during the fault. As 

it has been demonstrated, in this way, it is possible to synthesize a current 

in the stator in opposition to the currents generated during the fault, 

preventing thus the stator/rotor windings from suffering over-currents, 

with no need of using crowbar circuits. 

The performance of this strategy was tested and compared with and 

without control algorithm, already presented in previous works, which 

makes the active and reactive power references equal to zero during the 

fault. The analytical equations and the simulation models based on 

MATLAB have shown that the fluctuations in the stator and rotor 

currents, as well as in the electromagnetic torque were reduced by the 

half, when using the new proposed strategy in case of severe balanced 

and unbalanced voltage sags at the PCC. 

Therefore the results presented in this paper, show that it is possible to 

control the stability of a DFIG during severe contingencies in the power 

network, without the need of external auxiliary circuits. 

7 REFERENCES 

Fig 19, C voltage and grid voltages 

Compared to the system with crowbar alone protection and the system 

with crow bar and rotor-side and grid-side back-to-back controllers, post 

fault condition is very stable. The voltage of DFIG doesn’t degrade after 

the fault has cleared and reaches its pre-fault state immediately by 

observing Fig 17 and 18. The grid side voltage also remains same during 

and post-fault condition. The DC output voltage from the back-to-back 

controller is decreased because it supplies power to the system and 

maintains voltage balance during the total period of action. 

Comparing Fig 11c and Fig 20, the direct and quadrature axis voltages 

has improved a lot. In this advanced control schemes, d and q axis 

voltages are maintained nearly constant unlike in case without 

controllers. 

Fig 20, dqo voltages for stator side voltage of induction generator with 

crowbar and b2b controller 

6 Conclusions 

The present chapter has presented presents a control technique for doubly 

fed induction generators under different voltage disturbances. The current 

reference is chosen in the positive and negative sequences so that the 

torque and the DC voltage are kept stable during balanced and 

unbalanced conditions. Both rotor-side and grid-side converters have 

[1] C. Luo, H. Banakar, S. Baike, and O. Boon-Teck, “Strategies to smooth wind 

power fluctuations of wind turbine generator,” IEEE Trans. Energy Convers., vol. 

22, no. 2, pp. 341–349, Jun. 2007. 

[2] M.Kayikci and J.Milanovic, “Reactive power control strategies forDFIG-based 

plants,” IEEE Trans. Energy Convers., vol. 22, no. 2, pp. 389–396, Jun. 2007. 

[3] C. Eisenhut, F. Krug, C. Schram, and B. Klockl, “Wind-turbine model for 

system simulations near cut-in wind speed,” IEEE Trans. Energy Convers., vol. 22, 

no. 2, pp. 414–420, Jun. 2007. 

[4] K. Seul-Ki and K. Eung-Sang, “PSCAD/EMTDC-based modeling and analysis 

of a gearless variable speed wind turbine,” IEEE Trans. Energy 

Convers., vol. 22, no. 2, pp. 421–430, Jun. 2007. 

[5] R. Pena, J. J. C. Clare, and G. Asher, “Doubly fed induction generator using 

back-to-back PWM converters and its application to variable-speed wind-energy 

generation,” Proc. Inst. Elect. Eng. Electric Power Appl., vol. 143, no. 3, pp. 231– 

241, 1996. 

[6] P. Ledesma and J. Usaola, “Doubly fed induction generator model for 

transient stability analysis,” IEEE Trans. Energy Convers., vol. 20, no. 2, 

pp. 388–397, Jun. 2005. 

[7] Y. Lei, A. Mullane, G. Lightbody, and R. Yacamini, “Modeling of the 

wind turbine with a doubly fed induction generator for grid integration studies,” 

IEEE Trans. Energy Convers., vol. 21, no. 1, pp. 257–264, Mar. 2006. 

[8] D. Xiang, L. Ran, P. Tavner, and S. Yang, “Control of a doubly fed induction 

generator in a wind turbine during grid fault ride through,” IEEE Trans. Energy 

Convers., vol. 21, no. 3, pp. 652–662, Sep. 2006. 

[9] J. Morren and S. de Haan, “Ridethrough of wind turbines with doubly fed 

induction generator during a voltage dip,” IEEE Trans. Energy Convers., 

vol. 20, no. 2, pp. 435–441, Jun. 2008. 

[10] J. Morren and S. W. H. de Haan, “Short-circuit current of wind turbines 

with doubly fed induction generator,” IEEE Trans. Energy Convers., 

vol. 22, no. 1, pp. 174–180, Mar. 2007. 

[11] A. Junyent-Ferr´ e, A. Sumper, O. Gomis-Bellmunt,M. Sala, andM.Mata, 

“Digital simulation of voltage dip characteristics ofwind turbine systems,” in proc. 

9th Int. Conf. Elect. Power Quality Utilization, Barcelona, Spain, 2007, pp. 1–6. 

[12] P. Rioual, H. Pouliquen, and J.-P. Louis, “Regulation of a PWM rectifier in the 

unbalanced network state using a generalized model,” IEEE Trans. Power Electron., 

vol. 11, no. 3, pp. 495–502, May 1996. 

[13] H.-S. Song and K. Nam, “Dual current control scheme for PWMconverter 

under unbalanced input voltage conditions,” IEEE Trans. Ind. Electron., vol. 46, no. 

5, pp. 953–959, Oct. 1999. 

[14] L. Xu and Y.Wang, “Dynamic modeling and control of DFIG-based wind 

turbines under unbalanced network conditions,” IEEE Trans. Power Syst., vol. 22, 

no. 1, pp. 314–323, Feb. 2007. 

[15] S. Seman, J. Niiranen, and A. Arkkio, “Ride-through analysis of doubly fed 

induction wind-power generator under unsymmetrical network disturbance,” IEEE 

Trans. Power Syst., vol. 21, no. 4, pp. 1782–1789, Nov.2006. 

[16] T. K. A. Brekken and N. Mohan, “Control of a doubly fed induction wind 

generator under unbalanced grid voltage conditions,” IEEE Trans. Energy Convers., 

vol. 22, no. 1, pp. 129–135, Mar. 2007. 

690



[17] R. Pena, R. Cardenas, E. Escobar, J. Clare, and P. Wheeler, “Control system 

for unbalanced operation of stand-alone doubly fed induction generators,” IEEE 

Trans. Energy Convers., vol. 22, no. 2, pp. 544–545, Jun.2007. 

[18] A. Junyent-Ferr´ e, “Modelitzaci´ o i control d’un sistema de generaci´ oel` 

ectrica de turbina de vent,” Master’s thesis, ETSEIB-UPC, Barcelona, 

Spain, 2007. 

[19] L. Harnefors and H.-P. Nee, “Model-based current control of ac machines 

using the internalmodel controlmethod,” IEEE Trans. Ind. Appl., vol. 34, no. 1, pp. 

133–141, Jan./Feb. 1998. 

[20] A. D. Hansen and G. Michalke, “Fault ride-through capability of DFIG 

wind turbines,” Renew. Energy, vol. 32, no. 9, pp. 1594–1610, Jul. 2007. 

[21] P. Rodriguez, J. Pou, J. Bergas-Jan´ e, J. Candela, R. Burgos, andD. Boroyevich, 

“Decoupled double synchronous reference frame pll for power converters 

control,” IEEE Trans. Power Electron., vol. 22, no. 2, pp. 584–592, 

Mar. 2007. 

[22] M. H. Bollen, Understanding Power Quality Problems: Voltage Sags and 

Interruptions. Piscataway, NJ: IEEE Press, 2000. 

[23] Texas Instruments Inc., “TMS320F2812 data manual,” Tech. Rep.SPRS174O, 

2007. 

[24] I. Erlich, H. Wrede, and C. Feltes, “Dynamic behaviour of DFIG-based wind 

turbines during grid faults,” presented at the IEEE Power Convers. Conf. (PCC 

2007), Nagoya, Japan, Apr. 

[25] A. Hansen and G.Michalke, “Fault ride-through capability of DFIG wind 

turbines,” Renewable Energy, vol. 32, pp. 1594–1610, 2007. 

[26] O. Gomis-Bellmunt, A. Junyent-Ferre, A. Sumper, and J. Bergas-Jan, “Ridethrough 

control of a doubly fed induction generator under unbalanced voltage 

sags,” IEEE Trans. Energy Convers., vol. 23, no. 4, 

pp. 1036–1045, Dec. 2008. 

[27] B. Xie, B. Fox, and D. Flynn, “Study of fault ride-through for DFIG 

based wind turbines,” presented at the IEEE Int. Conf. Electric Utility eregulation, 

Restruct. Power Technol., DWT 2004, Hong Kong, Apr. 

[28] J. Morren and S. W. H. de Haan, “Ridethrough of wind turbines with doublyfed 

induction generator during a voltage dip,” IEEE Trans. Energy Convers., vol. 

20, no. 2, pp. 435–441, Jun. 2005. 

[29] A. D. Hansen,G.Michalke, P. Sørensen,T.Lund, and F. Iov, “Co-ordinated 

voltage control of DFIG wind turbines in uninterrupted operation during grid 

faults,” Wind Energy, vol. 6, pp. 22–26, 2006. 

[30] G. Pannell, D. J. Atkinson, R. Kemsley, L. Holdsworth, P. Taylor, and O. 

Moja, “DFIG control performance under fault conditions for offshore wind farms,” 

presented at the CIRED: 18th Int. Conf. Elect. Distrib., Turin, Italy, Jun. 2005. 

[31] A. H. Kasem, E. F. El-Saadany, H. H. El-Tamaly, and M. A. A.Wahab, “An 

improved fault ride-through strategy for doubly fed induction generatorbased wind 

turbines,” IET Renewable Power Generation, vol. 2, no. 4, pp. 201–214, 2008. 

691

Design of Doubly Fed Induction Generator Driven by Wind Turbine ...

Create successful ePaper yourself

Delete template?

Save as template?