Low voltage ride through of doubly-fed induction generator connected to the grid using sliding mode control strategy

Renewable Energy 80 (2015) 583e594

Contents lists available at ScienceDirect

Renewable Energy journal homepage: www.elsevier.com/locate/renene

Low voltage ride through of doubly-fed induction generator connected to the grid using sliding mode control strategy Naggar H. Saad a, Ahmed A. Sattar a, Abd El-Aziz M. Mansour b, * a b

Electrical Power and Machines Department, Faculty of Engineering, Ain Shams University, Cairo 11517, Egypt Department of Electrical Engineering, Police Academy, Cairo, Egypt

a r t i c l e i n f o

a b s t r a c t

Article history: Received 29 April 2014 Accepted 22 February 2015 Available online

Wind Energy Conversion System (WECS) based on Doubly Fed Induction Generator (DFIG) connected to the grid is subjected to high transient currents at rotor side and rise in DC-link voltage during voltage sag at stator/grid side. To secure power system operation wind turbines have to meet grid requirements through the Low voltage ride through (LVRT) capability and contribute to grid voltage control during severe situations. This paper presents the modeling and control designs for WECS based on a real model of DFIG taking into account the effect of stator resistance. The non-linear control technique using sliding mode control (SMC) strategy is used to alter the dynamics of 1.5 MW wind turbine system connected to the grid under severe faults of grid voltage. The paper, also discusses the transient behavior and points out the performance limit for LVRT by using two protection circuits of an AC-crowbar and a DC-Chopper which follow a developed flowchart of system protection modes under fault which achieved LVRT requirements through results. The model has been implemented in MATLAB/SIMULINK for both rotor and grid side converters. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Doubly fed induction generator Sliding mode control Low voltage ride through Active AC-crowbar DC-chopper

1. Introduction DFIG has the largest world market share of wind turbine concepts since the year 2002, because of its ability to provide variable speed operation and independent active and reactive control in a cost-effective way. The power production of variable speed wind turbines is higher than fixed-speed turbines, as they can rotate at the optimal rotational speed for each wind speed. Other advantages of the variable speed wind turbines are that they reduce mechanical stresses, they improve power quality, and they compensate for torque and power pulsations. A large number of papers describe the modeling of DFIGs [1e4]. The major drawback of variable speed wind turbines, especially for turbines with DFIGs, is their operation during faults. Faults in the power system even far away from the location of the turbine, can cause a voltage dip at the connection point of the wind turbine. The dip in the grid voltage will result in an increase in the current in stator windings of the DFIG. Because of the magnetic coupling between stator and rotor, this current will also flow in the rotor

* Corresponding author. Tel.: þ20 011 11118227. E-mail addresses: [email protected] (N.H. Saad), [email protected] (A.A. Sattar), [email protected] (A.E.-A.M. Mansour). http://dx.doi.org/10.1016/j.renene.2015.02.054 0960-1481/© 2015 Elsevier Ltd. All rights reserved.

circuit and the power electronic converter; this can lead to destroy the converter. The reaction of DFIG to grid voltage disturbances is sensitive, as described in Refs. [5,6] for symmetrical and unsymmetrical voltage dips and requires additional protection for the rotor side power electronic converter. Today, there is a need to control wind power, both in active and reactive powers, and to be able to stay connected to the grid when grid faults happen. The necessity to establish a grid code was raised due to the fact of the high amount of wind power targeted in Egypt [7]. The low voltage ride through is the most important requirement regarding the wind farm operation that has been recently introduced in the grid codes. It is vital for a stable and reliable operation of power supply networks, especially in regions with high penetration of wind power generation. Faults in the grid can cause a large voltage dip across wide regions and some generation units can be lost as a consequence. SMC is a nonlinear control technique derived from variable structure control system theory which used for torque and pitch control in Ref. [8] and for active and reactive power control in Refs. [9,10]. Ride through of wind turbine with DFIG during voltage dips and using a PI controller is discussed in Ref. [11]. The behavior of the DFIG based wind turbine during grid faults using DC chopper and AC crowbar is explained in Ref. [12]. In Ref. [13] DFIG behavior

584

N.H. Saad et al. / Renewable Energy 80 (2015) 583e594

under unsymmetrical voltage dips is studied. A crowbar protection with a stator current feed-back solution is used in Ref. [14]. Improved grid voltage control strategy is discussed in Ref. [15]. In our work, contrary to the previous work carried out on the DFIG where the researchers always neglect the stator resistance to facilitate its control, this resistance was considered in order to return the system studied near to reality. This paper presents LVRT of DFIG connected to the grid using the SMC strategy without/with an active AC crowbar and DC-Chopper circuits that follow a developed flowchart of system protection modes under fault. 2. Grid code requirements Recent grid codes require wind farms to remain connected and support the grid during and after a fault. They must withstand voltage dips of a certain percentage of the normal voltage for the specified time durations, as shown in the LVRT voltageetime profiles of Fig. 1. Disconnection is not allowed above the borderline and the turbine stay connected even when the voltage at the point of common coupling drops to zero. Below the border line wind turbines are not required to contribute the grid and they can be tripped by circuit breakers after a 150 ms delay. In Fig. 1, region 1 indicates no tripping and WECS stay connected to the grid even the voltage of Point of Common Coupling to the grid (PCC) dropped to zero. Region 2 indicates tripping of WECS.

vdr ¼ Rr Idr þ

lqs ¼ Ls iqs þ Lm iqr

(5)

lds ¼ Ls ids þ Lm idr

(6)

lqr ¼ Lr iqr þ Lm iqs

(7)

ldr ¼ Lr idr þ Lm ids

(8)


 lqm ¼ Lm iqs þ iqr

(9)

ldm ¼ Lm ðids þ idr Þ

(10)

where Ls, Lr, and Lm are the stator inductance, rotor inductance and mutual inductances, respectively. The stator active and reactive power can be expressed as:

Ps ¼

 3
v i þ vqs iqs 2 ds ds

(11)

Qs ¼

 3
vqs ids  vds iqs 2

(12)

And the active and reactive rotor powers are given by

Pr ¼

 3
v i þ vqr iqr 2 dr dr

(13)

Qr ¼

 3
vqr idr  vdr iqr 2

(14)

The electromagnetic torque equation is given by:

dlqs þ ðws lds Þ ¼ Rs Iqs þ dt

(1)

Te ¼

vds ¼ Rs Ids þ

 dlds
 ws lqs dt

(2)

where p is the pole number.

vqr ¼ Rr Iqr þ

dlqr þ ðws  wr Þldr dt

(3)

vqs

(4)

where ws is the rotational speed of the synchronous reference frame, wr is the generator rotor speed. The flux linkages are given by:

3. Doubly fed induction generator DFIG is currently the most widely used types of electrical generators for wind turbine systems in the Megawatt range [4]. Schematic diagram of on a grid connected DFIG is shown in Fig. 2. A synchronous rotating deq reference frame is used to model the DFIG with the direct -axis oriented along the stator flux position. In this way, decoupled control between the electrical torque and the rotor excitation current is obtained. The reference frame is rotating with the same speed as the stator voltage. The voltages equations are [3]:

dldr  ðws  wr Þlqr dt

 3p
 3p Lm
Lm iqs idr  ids iqr lqs idr  lds iqr ¼ 4 Ls 4

(15)

4. Sliding mode controller SMC is a nonlinear control technique derived from Variable Structure Control (VSC) system theory and developed by Vladim

Fig. 1. LVRT requirements of Egypt grid code.

N.H. Saad et al. / Renewable Energy 80 (2015) 583e594

PCC

585

DFIG

Ps+jQs

Grid

Pf+jQf

GSC

DC Chopper

RSC

G.B

Pr+jQr

C Lf

R

rf

Vdc SMC

SMC

Grid Side Controller

Rotor Side Controller

Rc Craw-bar circuit

Fig. 2. Schematic diagram of on grid connected DFIG.

UTKIN. Basically, a VSC includes several different continuous functions that can map plant state to a control surface; whereas switching among different functions is determined by plant state represented by a switching function [8]. Such control solution has several advantages such as simple implementation, robustness and good dynamical response. In order to easily control the production of electricity by the wind energy conversion system the stator flux vector is aligned with d-axis. This orientation will be made with a real model of DFIG; taking into account the stator resistance.

lqs ¼ 0; lds ¼ ls

3 Qs ¼ 2

ws ls Lm ws l2s idr þ  Ls Ls

! (24)

From the analysis of equations (23) and (24), we lead to uncoupled power control; where the component iqr of the rotor current control the active power. The reactive power is imposed by the direct component idr. By rearranging equations (3) and (4) we can deduce a formula of rotor currents as a function of rotor voltages in equations (25) and (26).

(16)

vqr ¼ Rr Iqr þ sLr

dIqr Lm ls þ gws sIdr þ gws dt Ls

(25)

(17)

vdr ¼ Rr Idr þ sLr

dIdr  gws sIqr dt

(26)

(18)

dIqr 1 ¼ sLr dt

And the stator voltages are reduced to

vqs ¼ Rs Iqs þ ws ls vds ¼ Rs Ids þ

dls dt

Suppose that electrical supply network is stable, which leads to constant stator flux and a new value stator voltage will be:

vds ¼ Rs Ids ; vqs ¼ Rs Iqs þ ws ls

(19)

Substitute (16) in equations (5), (6) and (15) we get:

iqs ¼ 

Lm iqr Ls

(20)

Lm ls ids ¼  idr þ Ls Ls

(21)

3p Lm ls iqr Te ¼  4 Ls

(22)

Substitute by equations (20) and (21) into equations (11) and (12) we get:

3 ws ls Lm V 2 w2 l2 Ps ¼ iqr  s þ s s 2 Ls Rs Rs

!

  Lm ls vqr  Rr Iqr  gws sIdr  gws Ls

 dIdr 1
¼ v  Rr Idr þ gws sIqr sLr dr dt

(27)

(28)

At steady state the second derivative terms of the two equations (25) and (26) are nil, and we can write

vqr ¼ Rr Iqr þ gws sIdr þ gws vdr ¼ Rr Idr  gws sIqr

Lm ls Ls

(29) (30)

where s ¼ 1  L2m =Lr Ls , is the leakage factor and g ¼ 1  wr =ws , is the slip. The design of the SMC will be demonstrated mainly in four steps for a nonlinear system [9]: 4.1. State space equation Rearrange the system equations and variables in the form of:

(23)

x_ ¼ Zðx; tÞ þ Yðx; tÞ$uðx; tÞ

(31)

586

N.H. Saad et al. / Renewable Energy 80 (2015) 583e594

where x is the state vector, Z(x, t), Y(x, t) are two continuous and uncertain non-linear functions, u is the control vector. 4.2. Choice of the switching surface From the system above in (31), it is possible to determine the sliding surface, proposed by J.J Slotine, given by:

 Sðx; tÞ ¼

d þt dt

n1 $e

(32)

_ …; xn1 T is the state vector, e ¼ xd  x; x ¼ ½x; x; T  is the desired state vector. xd ¼ ½xd ; x_d ; …; xn1 d _ …; en1 T , is the error vector, t is a positive coefficient e ¼ ½e; e; and n is the system order. where

4.3. Convergence condition The convergence condition is satisfied by the Lyapunov equation that makes the surface attractive and invariant. Consider a Lyapunov function

1 V ¼ SðxÞ2 2

(33)

From Lyapunov theorem we know that if V_ is negative definite, the system trajectory will be driven and attracted towards the sliding surface and remain sliding on it until the origin is reached asymptotically

_ SðxÞSðxÞ 0

(34)

 satð∅Þ ¼

j∅j  0 j∅j < 0

sgnð∅Þ; ∅;

(39)

To control DFIG using sliding mode theory, the surfaces are chosen as functions of the error between the desired input signals and the measured signals.

4.5. Rotor side converter control The rotor currents which are linked to active and reactive powers by equation 22 and 23 have to track appropriate current references, so a sliding mode control based on the above Park reference frame is used. The sliding surfaces representing the error between the measured and references rotor currents.

sd ðidr Þ ¼ iddr  idr

(40)


 sq iqr ¼ idqr  iqr

(41)

d

s_d ðidr Þ ¼ i_dr  i_dr

(42)


 d s_q iqr ¼ i_qr  i_qr

(43)

Equations (42) and (43) can be simulated by taking into account the following invariance conditions

(

  sd ðidr Þ ¼ iddr  idr ¼ 0 s_d ðidr Þ ¼ 0

(44)

eq

n We take Vdr ¼ Vdr þ Vdr

4.4. Control law calculation

eq Vdr ¼

2sLr Ls _ sLr ls Q þ Rr Idr  gws sLr iqr þ 3ws ls Lm sref Lm

The control law satisfies the precedent conditions is presented in the following form:



where idr

u ¼ ueq þ un n u ¼ kf $sgnðsðx; tÞÞ

Qs

ref



3 ws l2s 2 Ls

! (46)

And for quadrature-rotor current

where u is the control vector and u is the equivalent control vector, un is the switching part of the control (the correction factor), kf is the controller gain. ueq can be obtained by considering the conditions for the sliding regiment, sðx; tÞ. The equivalent control keeps the state variables on sliding surface, once they reach it.


 d s_q iqr ¼ i_qr  i_qr eq

n We take Vqr ¼ Vqr þ Vqr During the sliding mode and in permanent regime, we have



 n s iqr ¼ 0; s_ iqr ¼ 0; Vqr ¼0

(36) eq

Vqr ¼

The controller described by the equation (36) presents high robustness, insensitive to parameter fluctuations and disturbances, but it will have high-frequency switching (chattering phenomena) near the sliding surface due to sgn function involved. These drastic changes of input can be avoided by introducing a boundary layer with width ε [9]. Thus, replacing sgnðsðtÞ=εÞ by satðsðtÞ=εÞ (saturation function), in (35), we have:

u ¼ ueq þ un n u ¼ kf $satðsðx; tÞÞ

(37)

u ¼ ueq  kf $satðsðx; tÞÞ

(38)

where ε > 0

2 Ls 3 ws ls Lm

(35) eq

8 < 1; if ∅ > 0 sgnð∅Þ ¼ 0; if ∅ ¼ 0 : 1; if ∅ < 0

ref

¼

(45)

iqr

ref

2sLr Ls _ Lm ls P s þ Rr Iqr þ gws sLr idr þ gws 3ws ls Lm ref Ls   2 2 2 Ls Vs  ws ls þ ws ls Lm Rs ¼

Ls Lm ls ws

2 P 3 s

ref

þ

Vs2 w2s l2s  Rs Rs

(47)

! (48)

To obtain good dynamic performance, the control vector is imposed as follows:

  eq Vdq ¼ Vdq  kdq sign sdq ; kdq is positive Fig. 3 shows the block diagram of the proposed sliding mode control applied to the rotor side converter.

N.H. Saad et al. / Renewable Energy 80 (2015) 583e594

587

Fig. 3. Block diagram of the proposed sliding mode applied to the RSC.

Qf ¼

  3 3 vqg idf  vdg iqf ¼  vdg iqf 2 2

(52)

Applying Kirchhoff's voltage law to the grid filter model in Fig. 4 we get: Fig. 4. Grid filter model in stator coordinates.

  dif þ vf vg ¼  rf þ juLf if  Lf dt

4.6. Gird side converter control The control objective of the grid side converter is to keep the DC-link voltage constant regardless of the magnitude and the direction of the rotor power. Direct axis current is controlled to keep the DC-link voltage constant, and quadrature axis current component can be used to regulate the reactive power flow between the grid side converter and the grid. The d-axis of the reference frame is aligned with the grid voltage angular position. Since the amplitude voltage of the grid is constant, vqg is zero and vdg is constant.

Vdg ¼ Vg

(49)

Vqg ¼ 0

(50)

The active and reactive powers injected to the grid from the GSC are given by:

Pf ¼

 3  3 v i þ vqg iqf ¼ v i 2 dg df 2 dg df

(51)

(53)

where Vg is the grid voltage, if is grid-filter current, and Vf is the grid-filter voltage supplied from the grid-side converter.

vqf ¼ rf iqf þ Lf

diqf þ us Lf idf þ vqg dt

(54)

vdf ¼ rf idf þ Lf

didf  us Lf iqf þ vdg dt

(55)

where Vqf, Vdf are the voltages at the output of GSC and iqf, idf are the corresponding currents injected into the grid. It is observed from equations (51) and (52), that the active and reactive power flow between the supply side converter and the supply will be proportional to idf and iqf respectively. The amount of energy stored in the DC-link capacitor is given by:

 dV 2 dWdc 1 3 ¼ Cdc dc ¼ Pf  Pr ¼  vdg idf  Pr 2 2 dt dt

Fig. 5. Block diagram of the proposed sliding mode applied to the GSC.

(56)

588

N.H. Saad et al. / Renewable Energy 80 (2015) 583e594

Fig. 6. Behavior of DFIG wind turbine during three phase grid short circuit with PI controller without protection schemes.

N.H. Saad et al. / Renewable Energy 80 (2015) 583e594

589

Fig. 7. Behavior of DFIG wind turbine during three phase grid short circuit with SMC controller without protection schemes.

dvdc 1 ¼ Cdc vdc dt



3  vdg idf  Pr 2

 (57)

where Vdc is the dc-link voltage, Cdc is the dc-link capacitance, Pf is the power delivered to the grid filter and Pr is the power delivered to the rotor circuit of DFIG. So we can define the two switching surfaces s1 and s2 as:

s1 ¼ idf

ref

 idf

(58)

s2 ¼ iqf

ref

 iqf

(59)

2 Po ref 3 vdg

(60)

idf

ref

¼

590

N.H. Saad et al. / Renewable Energy 80 (2015) 583e594

Fig. 5 shows the block diagram of the proposed sliding mode control applied to the GSC. 5. Results and simulation 5.1. Voltage sag behavior of DFIG without protection Fig. 8. Active craw-bar circuit.

iqf

ref

2 Qo ref ¼ 3 vdg

(61)

s_1 ¼ i_df

ref

 i_df

(62)

s_2 ¼ i_qf

ref

 i_qf

(63)

 didf 1 v  rf idf þ us Lf iqf  vdg ¼ Lf df dt

(64)

 diqf 1 vqf  rf iqf  us Lf idf ¼ Lf dt

(65)

s_1 ¼ 0; i_df eq

vdf ¼

eq

ref

 i_df ¼ 0; s_2 ¼ 0; i_qf

ref

 i_qf ¼ 0

2P_ o ref Lf þ rf idf  uLf iqf þ vdg 3vdg

vqf ¼ 

2Q_ o ref Lf rf iqf þ uLf idf 3vdg eq

Vdf ¼ Vdf  k1 signðs1 Þ; k1 is positive eq  k2 signðs2 Þ; k2 is positive Vqf ¼ Vqf

(66)

(67)

A detailed model of DFIG-based wind turbine system was developed in Matlab/Simulink in order to investigate its behavior in case of severe low voltage faults. In this paper, we dealt with voltage sag phenomena. Voltage sag is a sudden reduction of the voltage at a point in the electrical system between 10% and 100% and lasts for half cycle to 1 min. Three-phase symmetrical faults represent one of the most common causes of voltage sags. The dynamics of the DFIG connected to the grid under voltage fault of 100%, implying that no grid voltage remains from 0.85 s to 1 s are provided in Fig. 6 using PI controller and in Fig. 7 using SMC. As shown in Fig. 6 the stator voltage drops to 0% of nominal value and as a result, a high transient stator current at the instant of fault reaches 3.8 p.u and 4.1 p.u at the instant of clearing fault with high oscillations. It reaches a steady state at 1.2 s; it means that after clearing fault by 0.2 s. A high transient rotor current at the instant of fault reaches 4 p.u and 3.9 p.u at the instant of clearing fault with high oscillations. It reaches a steady state at 1.25 s; it means that after clearing fault by 0.25 s. DC link voltage rises to 2 p.u at the instant of fault and then it starts to follow reference value after recovery grid voltage at 1.2 s. The electromagnetic torque due to the transient in the rotor and stator, first it increases in the negative direction down to 2.6 p.u then rapidly rises in the positive direction to 1.3 p.u. It starts to oscillate at 0.9 s and finally reaches steady state after the grid voltage recovery at 1.3 s. Active power at the instant of fault reaches a peak value of 1.2 p.u and then starts to decrease rapidly to zero until the instant of grid voltage recovery, it comes to increasing to pre-fault value with overshoot reaches 30%. Reactive power starts to increase with the instant of fault to 0.6 p.u and supports grid voltage decrease, and then comes back to prefault value after voltage recovery. As shown in Fig. 7 a high transient stator current at the instant of fault reaches 4 p.u and 4.2 p.u at the instant of clearing the fault. It reaches a steady state at 1.16 s; it means that after clearing fault by 0.16 s with less number of oscillations and a high transient rotor current at the instant of fault reaches 4.5 p.u and 4 p.u at the instant of clearing the fault. It reaches a steady state at 1.20 s; it means that after clearing fault by 0.2 s with less number of oscillations. DC link voltage is varying between 0.85 p.u and 1.15 p.u at the instant of fault and the instant of recovery grid voltage and reaches reference value at 1.3 s. The SMC controls both the active and reactive power flow between the two converters which keeps zero reactive power during the fault. This decrease in reactive power flow decreases the DC voltage as compared with PI controller. The electromagnetic torque due to the transient in the rotor and stator, first it increases in the negative direction down to 2.8 p.u then rapidly rises in the positive direction to 1.8 p.u. It starts to oscillate at 0.9 s and finally reaches a steady state after grid voltage recovery at 1.2 s with low number of oscillations. Active power reaches a peak value of 1.1 p.u and then starts to decrease rapidly until the instant of grid voltage recovery, it comes to increasing to pre-fault value. Reactive power starts to increase with the instant of fault to support grid voltage decrease and then comes back to pre-fault value after voltage recovery by 0.1 s. 5.2. Voltage sag behavior of DFIG with protection

Fig. 9. Flowchart of system protection modes under fault.

In this section two protection circuits are used. First is active ac crowbar which is used to limit the rotor current during fault

N.H. Saad et al. / Renewable Energy 80 (2015) 583e594

between the rotor and the RSC as shown in Fig. 2. This consists of a full-wave bridge rectifier, a power resistor and IGBT switch. The second is a DCechopper which is used to limit the DC-link overvoltages during the fault. Fig. 8 shows the active craw-bar circuit; during normal operation the switch is open. The switch is activated on detection of rotor over currents or DC-link over voltage in order to redirect the rotor

591

current into the crowbar circuit where the energy is being dissipated through the crowbar resistor. Fig. 9 indicates four categories of wind turbine operation modes during fault ride-through, which are normal operation (Mode 0), converter deactivation (Mode 1), crowbar and DC chopper activation (Mode 2), and rotor-side converter reactivation (Mode 3).

Fig. 10. Behavior of DFIG wind turbine during three phase grid short circuit using PI controller with different modes of protection schemes activation.

592

N.H. Saad et al. / Renewable Energy 80 (2015) 583e594

Fig. 11. Behavior of DFIG wind turbine during three phase grid short circuit with SMC controller with different modes of protection schemes activation.

Fig. 10 shows the behavior of a DFIG wind turbine during three phase grid short circuit with a PI controller with protection schemes activation. A high transient stator current at the instant of fault reaches 3.8 p.u and 4.1 p.u at the instant of clearing fault with low oscillations. It reaches a steady state at 1.2 s; it means that after clearing fault by 0.2 s and a high transient rotor current at the

instant of fault reaches 3.4 p.u and 3.9 p.u at the instant of clearing fault with low oscillations. It reaches a steady state at 1.3 s; it means that after clearing fault by 0.3 s. Considering Stator resistance in calculations helps to reduce overshoot in stator currents, rotor currents and improve the electromagnetic torque during zero voltage of the grid, if it compared with simulation results in

N.H. Saad et al. / Renewable Energy 80 (2015) 583e594

Ref. [12]. Fig. 10 shows the highest transient stator currents reaches at the instant of fault 3.8 p.u while in Ref. [12] it reaches 4 p.u, the highest transient rotor currents reaches at the instant of fault 3.2 p.u while in Ref. [12] it reaches 5 p.u. DC link voltage rises to 1.1 p.u at the instant of fault and then it starts to follow reference value after grid voltage recovery at 1.2 s. The electromagnetic torque due to the transient in the rotor and stator, first it increases in the negative direction down to 3.6 p.u then rapidly rises in the positive direction to 0.5 p.u. It starts to oscillate at 0.9 s and finally reaches a steady state after grid voltage recovery at 1.4 s. Active power reaches a peak value of 1.1 p.u at the instant of fault and then starts to decrease rapidly to zero until the instant of grid voltage recovery, it comes to increasing to pre-fault value with overshoot reaches 30%. Reactive power starts to increase with the instant of fault to 0.6 p.u and supports grid voltage decrease, and then comes back to zero after voltage recovery. Fig. 11 shows the behavior of a DFIG wind turbine during three phase grid short circuit with a SMC controller with protection schemes activation. The different modes of protection scheme are marked in Figs. 10 and 11; however mode 1 exists for a very short time at the instant of fault between mode 0 and mode 2. A high transient stator current at the instant of fault reaches 4 p.u and 4.2 p.u at the instant of clearing the fault. It reaches a steady state at 1.16 s; it means that after clearing fault by 0.16 s with less number of oscillations. A high transient rotor current at the instant of fault reaches 4.5 p.u and 4 p.u at the instant of clearing the fault. It reaches a steady state at 1.20 s; it means that after clearing fault by 0.2 s with less number of oscillations. DC link voltage is varying between 0.80 p.u and 1.2 p.u at the instant of fault and the instant of recovery grid voltage and it reaches reference value at 1.2 s. The

593

Crowbar is activated as it detects the increase in the rotor current while the DC chopper is not activated. The SMC controller keeps RSC connected to the system so the DC voltage decreases. The electromagnetic torque due to the transient in the rotor and stator, first it increases in the negative direction down to 1.8 p.u then rapidly rises in the positive direction to 1.9 p.u. It starts to oscillate at 0.9 s and finally it reaches a steady state after grid voltage recovery at 1.2 s with low number of oscillations. Active power reaches a peak value of 1.1 p.u at the instant of fault and then it starts to decrease rapidly until the instant of grid voltage recovery; it comes to increasing to pre-fault value. Reactive power starts to increase with the instant of fault to support grid voltage decrease and then it comes back to pre-fault value after voltage recovery by 0.1 s. Figs. 12 and 13 show the behavior of protection schemes and rotor side converter with PI controller and SMC including rotor side converter currents, crowbar currents and chopper currents. The RSC current is zero in Fig. 12, using PI controller, due to the increase in DC voltage and rotor current that reach the values to trigger the Crowbar and the DC chopper protection schemes. These protection schemes disconnect the RSC. The RSC current in Fig. 13 is non zero because the DC voltage at the instant of the fault using SMC controller decreases, so the DC chopper is not triggered and the RSC is still connected to the system. 6. Conclusion This paper presents the modeling and control of a DFIG taking into consideration the stator resistance to bring the system to reality for driving of active and reactive power equations. It also helps to reduce overshoot in stator currents, rotor currents and improve

Fig. 12. Behavior of protection schemes and rotor side converter with PI controller.

594

N.H. Saad et al. / Renewable Energy 80 (2015) 583e594

Fig. 13. Behavior of protection schemes and rotor side converter with SMC controller.

the electromagnetic torque during zero voltage of the grid. This modification gives accurate and accepted results in simulation not as many researchers did before that they used to neglect it. A non-linear robust control technique using sliding mode control is implemented for active and reactive power control of 1.5 MW wind turbine system connected to the grid. Complete control of both rotor side and grid side converters is implemented using Matlab/Simulink with complete simulation under severe faults of grid voltage. The results show superior behavior of the SMC than the PI controller; it improves performance of the system, and reduces overshoots and transient time. Two circuits of protection schemes are used; crow bar and DC chopper circuit to protect the rotor side converter and DC link capacitor from failure or damage during high transient currents through fault. The results show that the protection schemes were capable of reducing the rotor transient current and reduce DC-link voltage below threshold values during the fault. LVRT requirements of Egyptian grid are achieved using a developed flowchart of system protection modes under fault which execute the control processes and improve the performance of protection schemes. References [1] Pena R, Clare JC, Asher GM. Doubly fed induction generator using back-to-back PWM converters and its application to variable speed wind energy generation. IEE Proc Electr Power Appl May 1996;143(3):231e41. [2] Ekanayake Janaka B, Holds worth Lee, Wu Xue Guang, Jenkins Nicholas. Dynamic modeling of doubly fed induction generator wind turbines. IEEE Trans Power Syst May 2003;18(2):803e9.

[3] Qiao Wei, Member IEEE. Dynamic modeling and control of doubly fed induction generator driven by wind turbines. In: IEEE/PES Power Systems Conference and Exposition; 2009. p. 1e8. PSCE'09. [4] Jou Sung-Tak, Lee Sol-Bin, Park Yong-Bae, Lee Kyo-Beum. Direct power control of a DFIG in wind turbines to improve dynamic responses. JPE September 2009;9(5):781e90. [5] Lopez J, Sanchis P, Roboam X, Marroyo L. Dynamic behavior of the doubly fed induction generator during three-phase voltage dips. Energy Convers IEEE Trans 2007;22(3):709e17. [6] Seman S, Niiranen J, Arkkio A. Ride-through analysis of doubly fed induction wind power generator under unsymmetrical network disturbance. Power Syst IEEE Trans 2006;21(4):1782e9. [7] www.slideshare.net/rcreee/boo-wind-power-in-egypt May 12, 2013. [8] Lee Sung-Hun, Joo Youngjun, Back Juhoon, Seo Jin-Heon, Choy Ick. Sliding mode controller for torque and pitch control of PMSG wind power systems. J Power Electron May 2011;11(3):342e9. [9] Bekakra Youcef, attous Djilani. Active and reactive power control of a DFIG with MPPT for variable speed wind energy conversion using sliding mode control. World Academy of Science, Engineering and Technology; 2011. p. 1543e9. [10] Belgacem kheira, Mezouar A, Massoum A. Sliding mode control of doubly fed induction generator for wind energy conversion. Int J Energy Eng 2013;3(1): 30e6. [11] Morren Johan, Student Member IEEE, de Haan Sjoerd WH, Member IEEE. Ride through of wind turbines with doubly-fed induction generator during a voltage dip. IEEE Trans Energy Convers June 2005;20(2):435e41. [12] Erlich I, Wrede H, Feltes C. Dynamic behavior of DFIG-based wind turbines during grid faults. IEEE 2007:1195e200. [13] Zhou Yi, Student Member IEEE, Bauer Paul, Senior Member IEEE, Ferreira Jan A, Fellow IEEE, et al. Operation of grid-connected DFIG under unbalanced grid voltage condition. IEEE Trans Energy Convers March 2009;24(1):240e6. [14] Wessels Christian, Student Member IEEE, Fuchs Friedrich W, Senior Member IEEE. Fault ride through of DFIG wind turbines during symmetrical voltage dip with crowbar or stator current feedback solution. IEEE Sept. 2010;12:2771e7. [15] Zhang Xueguang, Pan Weiming, Liu Yicheng, Xu Dianguo. Improved grid voltage control strategy for wind farms with DFIGs connected to distribution networks. JPE May 2012;12(3):495e502.