Direct model reference adaptive internal model controller for better voltage sag ride through in doubly fed induction generator wind farms

Electrical Power and Energy Systems 47 (2013) 255–263

Contents lists available at SciVerse ScienceDirect

Electrical Power and Energy Systems journal homepage: www.elsevier.com/locate/ijepes

Direct model reference adaptive internal model controller for better voltage sag ride through in doubly fed induction generator wind farms N. Amuthan a,⇑, P. Subburaj b, P. Melba Mary c a

Department of Electrical and Electronics Engineering, PET Engineering College, Valliyoor 627 117, Tamil Nadu, India National Engineering College, Kovilpatti, Tamil Nadu, India c V V College of Engineering, Tisaiyanvilai (Via), Tirunelveli 627 657 (Dist.), Tamil Nadu, India b

a r t i c l e

i n f o

Article history: Received 7 April 2011 Received in revised form 28 June 2012 Accepted 27 October 2012

Keywords: Direct model reference adaptive internal model controller Fuzzy sets Adaptive network based fuzzy inference system Doubly fed induction generator Wind farms

a b s t r a c t This paper presents the function of a direct model reference adaptive internal model controller (DMRAIMC) with variable gain adjustment mechanism for a doubly fed induction generator (DFIG) in a wind farm. Rotor current is controlled using the above controller with variable gain adjustment mechanism achieved using fuzzy sets and ANFIS to improve the voltage sag ride. Performance of the variable gain and linearly variable gain adjustment mechanisms are compared and their improvements are explored. Simulation results are used to demonstrate the effectiveness and robustness of the proposed control strategy, during variations in the variable gain adjustment mechanism. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction Wind energy is an important source of green electricity in our country. Lot of research is still going on, to improve the generation of energy from wind. Conversion of wind energy technology is a rapidly developing area. Commercialization of wind energy depends mainly on technology development, making it economically viable. A variable speed generator is one step ahead to increase the efficiency of wind energy conversion. Recently, doubly fed induction generator (DFIG) is a highly energy efficient variable speed wind generator used in the wind farms, as the variable-speed operation will improve the wind energy production further [1]. Dynamic and transient behavior of the DFIG-based WTs under voltage dips and wind speed fluctuations and rotor current controller is presented in [2]. Alan Mullane et al., says that nonlinear controller design results in considerable improvement in the ‘ride-through faults’ capability of wind turbines [3]. Boukhezzar and Siguerdidjane [4] has also confirmed nonlinear control strategies bring more performance in the exploitation of wind energy conversion systems. An internal model controller (IMC) is suggested in [5,6]. A direct ⇑ Corresponding author. Tel.: +91 04637 220999; fax: +91 04637 222205. E-mail address: [email protected] (N. Amuthan). 0142-0615/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijepes.2012.10.064

model reference adaptive internal model controller being a nonlinear controller will be a better option [7]. Performance improvement of wind energy conversion system using matrix converter is described in [8]. Impact of FACTS controllers on the stability of power systems connected with wind energy conversion systems is cited in [9]. Accurate monitoring and estimating the state of power system to identify the voltage collapses is presented in [10]. This paper aims at investigating the function of a direct model reference adaptive internal model controller with variable gain adjustment mechanism to improve the voltage sag ride through performance in a doubly fed induction generator. To compare the effectiveness of variable gain adjustment mechanism, simulation is done using Matlab/Simulink. The comparison in improvement using different adjustment mechanisms is discussed. 2. Modeling of doubly fed induction generator Fig. 1 shows the general configuration for shunt connected DFIG system. Normally, induction generators are very sensitive to sudden changes in voltage. DFIG is a complex nonlinear and variable speed generator, separately controllable on rotor and stator side. This generator is highly supportive for voltage sag ride through problems due to its high reactive power capabilities, also series and shunt grid connections are possible.

256

N. Amuthan et al. / Electrical Power and Energy Systems 47 (2013) 255–263

Nomenclature Vs Rs is ks Vr Rr ir kr Vds ids kds Vqs iqs kqs Vdr idr kdr

Vqr iqr kqr Lr Lm Lsl Wo Wk Wm C J D p ¼ dtd R(S) U(S) Y(S)

stator voltage (V) stator resistance (X) stator current (A) stator flux (Wb turns) rotor voltage (V) rotor resistance (X) rotor current (A) rotor flux (Wb turns) stator direct axis voltage (V) stator direct axis current (A) stator direct axis flux (Wb turns) stator quadrature axis voltage (V) stator quadrature axis current (A) stator quadrature axis flux (Wb turns) rotor direct axis voltage (V) rotor direct axis current (A) rotor direct axis flux (Wb turns)

rotor quadrature axis voltage (V) rotor quadrature axis current (A) rotor quadrature axis flux (Wb turns) rotor leakage Inductance (H) magnetizing Inductance (H) stator leakage Inductance (H) base speed (rad/s) speed of the reference frame (rad/s) rotor speed (rad/s) capacitance (F) inertia of the rotor (kg-m2) active damping torque (N-m) (derivative function) set point manipulated input of the process output

2.1. Mathematical model of doubly fed induction generator

pkds ¼ f ðV ds ; V qs ; V dr ; V qr Þ

ð7Þ

Important aspects of the modeling are presented below. The system is modeled and simulated using the Matlab/Simulink toolbox. The ode23tb variable step-size solver of Matlab is used to perform the simulations.

pkqs ¼ f ðV ds ; V qs ; V dr ; V qr Þ

ð8Þ

pkdr ¼ f ðV ds ; V qs ; V dr ; V qr Þ

ð9Þ

pkqr ¼ f ðV ds ; V qs ; V dr ; V qr Þ

ð10Þ

V s ¼ Rs is þ 1=W O

dks þ W k Mðpi=2Þks dt

ð1Þ The machine parameters used for simulation is presented in Table 1.

dkr V r ¼ Rr ir þ ð1=W O Þ þ ðW k  W m ÞMðpi=2Þkr dt

ð2Þ

Mðpi=2Þ ¼ 90 space rotatir i:e:;Mðpi=2Þ ¼ ½; 1; 1 0

ð3Þ

Flux linkage equations:

kr ¼ Lm is þ Lr ir

ð4Þ

ks ¼ Ls is þ Lm ir

ð5Þ

3. Controller design Fig. 2 shows the general internal model controller and the design algorithm is provided in [11] for a first order system. The linearized first order Plant model is

Gp ðsÞ ¼

where

Ls ¼ Lm þ Lsl and Lr ¼ Lm þ Lrl :

ð6Þ

By applying d–q theory in Eqs. (1)–(5) and rearranging the equations with variables Vds, Vqs, Vdr and Vqr becomes,

K 1 þ ss

ð11Þ

Internal model

Ginv ðsÞ ¼

1 þ ss K

ð12Þ

Grid 3

Wind Turbine

3

DFIG

3 3

Rotor side converter control

Grid side converter control

Fig. 1. General configuration for shunt connected DFIG wind turbine system.

257

N. Amuthan et al. / Electrical Power and Energy Systems 47 (2013) 255–263 Table 1 Parameters for simulated DFIG. 1.5 MW 590 V 0.0071 P.U. 0.171 P.U. 2.9 P.U. 3 0.5 P.U. 0.01 P.U. 0.1761 P.U. 12 m/s 4 m/s 20 m/s 14 m/s 1400 rpm

1 s

g ( s)

Mu x

Rated power Stator voltage Rs Ls Lm Number of pole pairs J D r(Leakage coefficient) Base wind speed Cut in wind speed Cut out wind speed Rated wind speed Rated rotor speed

Ym( s )

Rm ( s )

e (s)

θ

R(s)

E(s)

U (s)

Gimc

Y (s)

Gp

G inv d (s) Fig. 4. General direct model reference adaptive IMC with FuzzyMIT adjustment mechanism.

Gc ðsÞ ¼

HðsÞ 1 þ ss ¼ 1  HðsÞGp ðsÞ K/s

ð15Þ

4. Direct model reference adaptive internal model controller Fig. 2. General IMC.

4.1. Adjustment mechanism using Fuzzy Rules

Low pass filter F(s) is used to avoid model mismatch:

FðsÞ ¼

1 ð1 þ /sÞn

ð13Þ

/ = Speed response tuning parameter, n = order of low pass filter (used to add right number of poles to compensate the zeros). For a first order system, normally n = 1,

HðsÞ ¼ FðsÞGinv ðsÞ ¼

1 þ ss Kð1 þ /sÞn

ð14Þ

The controller is

Table 2 Inference rules of fuzzy adjustment mechanism. e(s) ym(s) NB NS Z PS PB

The Fuzzy Rules are proposed in [12]. Each variable is divided into a set of fuzzy regions, which are called as ‘labels’. The most popular labels are Negative Big (NB), Negative Small (NS), Zero (Z), Positive Small (PS) and Positive Big (PB). Based on experience and understanding of the system characteristics, functions of the primary and consequent parts are defined as triangular membership functions. Inference rules table of fuzzy adjustment mechanism can be found in Table 2. Also, the fuzzy triangular membership function is as shown in Fig. 3. Online parameter identification technique using fuzzy logic is mentioned in [13]. Rule based adjustment mechanism used is FuzzyMIT in [14],

h¼

NB

NS

Z

PS

PB

PB PB PB PS Z

PB PB PS Z NS

PB PS Z NS NB

PS Z NS NB NB

Z NS NB NB NB

g(s)

1 eðsÞymðsÞgðsÞ; s

ð16Þ

g(s) = Fuzzy Rulesh = variable gain output 4.1.1. Reference model The reference model of this controller is described in [15]. A first order reference model is considered for a first order plant and this model is used to cater the desired behavior of the closed loop system.

Fig. 3. Fuzzy triangular membership function.

258

N. Amuthan et al. / Electrical Power and Energy Systems 47 (2013) 255–263

Fig. 5. Surface plot of Fuzzy Rules.

Fig. 7. Surface plot of ANFIS rules.

Ym ( s )

Rm( s )

a( s)

Mux

1 s

e( s )

θ

E (s)

R( s)

U (s)

Gimc

Y (s)

Gp

Fig. 8. ANFIS Error training.

Ginv d (s) Fig. 6. General direct model reference adaptive IMC with ANFISMIT adjustment mechanism.

Rm ðsÞ ¼

1 þ k0 ð/sÞn

ð17Þ

similar to the classical way of Neural Network training method. A fault tolerant fuzzy IMC controller and ANFIS training methods are presented in [16]. 25 Gaussian type linear membership functions are used to generate Fuzzy Rules with Sugeno fuzzy system. The ANFIS adjustment mechanism used is ANFISMIT, which is linearly varying and defined as:

h¼

1 eðsÞymðsÞaðsÞ s

ð20Þ

n

(ð/sÞ þ k0 ) is a factor of KðsÞ (i.e.,)

KðsÞ ¼ ðð/sÞn þ k0 ÞKq ðsÞ Kq ðsÞ ¼ s

n1

þ qn2 s

n2

þ  þ

aðsÞ ¼ ANFIS ð18Þ

qsþ1 1

ð19Þ

Fig. 4 shows the direct model reference adaptive internal model controller with Fuzzy Rules. The gain in adjustment mechanism is varied using Fuzzy Rules and it is added to the controller. Fuzzy logic based adjustment mechanism used in this paper is very simple with less mathematical computation. Fig. 5 shows the surface plot of Fuzzy Rules, for an adaptation gain between 1 and 1. 4.2. Adjustment mechanism using ANFIS To linearly vary the output of the adjustment mechanism, adaptive network based fuzzy inference system (ANFIS) is proposed as an adjustment mechanism for the DRAIMC. Fig. 6 shows the direct model reference adaptive internal model controller with ANFIS. ANFIS adjustment mechanism is formulated based on the knowledge of the error signal. The error signal between the reference model and plant is used to adjust the performance of the controller

h = linearly varying gain output. Fig. 7 shows the surface plot of ANFIS for a gain variation of approximately 3. Fig. 8 shows the training error of the ANFIS editor for the hybrid training method using 40 epochs. 4.3. Rotor current control The proposed control diagram simulated on a system is as shown Fig. 9 idr is used to control the stator reactive power (Qs) and iqr is used to control stator active power(Pr), both reactive power and active power are controlled simultaneously in the generator and disturbance considered is small, id and iq are also used to control the oscillation. 5. Simulation results Fig. 10 shows a schematic diagram of a wind farm consisting of 9-MW DFIG wind turbine connected to a 25-kV distribution system exporting power to a 120-kV grid through a 25-km, 25-kV cable.

259

N. Amuthan et al. / Electrical Power and Energy Systems 47 (2013) 255–263

Fig. 9. Block diagram of control loop schematic.

Bus 1

Bus 2

120kv Wind Farm

25km 120kv/25kv

9MW

25kv/575v

Fig. 10. Schematic diagram of simulated wind power system.

The behavior of the DFIG during voltage sag of 50% and 150 ms is stimulated. The generator can initially operate at variable speeds with full load, with a constant grid frequency. The effect of wind speeds variation and performance during frequency droops is presented in [17]. 1.6

FuzzyMIT AnfisMIT

1.4

Stator flux (p.u.)

1.2 1 0.8 0.6 0.4 0.2 14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

Time (s) Fig. 11. Stator flux during voltage sag in DFIG.

Sudden changes in grid voltage affect the behavior of the machine causing voltage sag in the system. This behavior is simulated and the performance is shown below. The wind speed varies from 4 m/s to 20 m/s is considered in a wind farm location. The average wind speed of 12 m/s is used for analysis. The effect of FuzzyMIT and AnfisMIT is compared and it is given below. Fig. 11 shows the stator flux during voltage sag. Stator flux is nearly the same in variable gain adjustment mechanism (FuzzyMIT) and linearly varying adjustment mechanism (AnfisMIT). The transient response is damped using the controller. The beginning and end of the voltage sag the transient response is high because crowbar protection circuit was not provided. From Fig. 12 shows the rotor fluxes of the DFIG during voltage sag. In the AnfisMIT adjustment mechanism the rotor flux is within the permitted per unit value and the sag duration is increased in the linearly varying adjustment mechanism. Fig. 13 shows the stator current during voltage sag. The current per unit level is increased in the AnfisMIT adjustment mechanism and the initial current transient peak is reduced and final transient peak is slightly increased. The linearly varying adjustment mechanism quickly clears the voltage sag compared to FuzzyMIT. Fig. 14 shows the rotor current during voltage sag. The current level is increased in the AnfisMIT adjustment mechanism. Initial

4

1.1

FuzzyMIT AnfisMIT

FuzzyMIT AnfisMIT

3.5

1

Stator current (p.u.)

3

Rotor flux (p.u.)

0.9 0.8 0.7 0.6

2 1.5 1

0.5 0.4 14.9

2.5

0.5

14.95

15

15.05

15.1

15.15

15.2

Time (s) Fig. 12. Rotor flux during voltage sag in DFIG.

15.25

15.3

0 14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

Time (s) Fig. 13. Stator current during voltage sag in DFIG.

15.3

260

N. Amuthan et al. / Electrical Power and Energy Systems 47 (2013) 255–263 4

6

FuzzyMIT AnfisMIT

3

Rotor current (p.u.)

FuzzyMIT AnfisMIT

5

Stator Reactive Power (p.u.)

3.5

2.5 2 1.5 1

4 3 2 1 0 -1 -2

0.5 0 14.9

-3 14.9 14.95

15

15.05

15.1

15.15

15.2

15.25

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

Time (s)

15.3

Time (s)

Fig. 17. Stator reactive power during voltage sag.

Fig. 14. Rotor current during voltage sag in DFIG.

FuzzyMIT AnfisMIT

Rotor Active Power (p.u.)

0.03 0.02 0.01 0 -0.01

FuzzyMIT AnfisMIT

Rotor Reactive Power (p.u.)

0.04

0.04

0.02

0

-0.02

-0.04 -0.02 14.9 -0.03 14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

Time (s)

15.3

Time (s)

Fig. 18. Rotor reactive power during voltage sag.

Fig. 15. Rotor active power of DFIG during voltage sag.

1.035 2

FuzzyMIT AnfisMIT

1.025

1

Rotor Speed (p.u.)

Stator Active Power (p.u.)

1.5

FuzzyMIT AnfisMIT

1.03

0.5 0 -0.5 -1

1.02 1.015 1.01 1.005 1

-1.5

0.995

-2 -2.5 14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

Time (s) Fig. 16. Stator active power of DFIG during voltage sag.

transient peak is reduced and the final transit limit is high in the AnfisMIT, nearly four times higher without protection circuit. Fig. 15 shows the rotor active power. The rotor active power generation in the linearly varying adjustment mechanism increases in the initial and final voltage sag transient. Fig. 16 shows the stator Active Power of DFIG during voltage sag. The negative sign in the p.u. indicate the active power is fed to the grid. The peak overshoot during fault is nearly two times in the final voltage sag. Fig. 17 shows the stator reactive power. The reactive power varies proportionally with the voltage, and the actual reactive power

0.99 14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

Time (s) Fig. 19. DFIG rotor speed during voltage sag.

absorption depends upon the power flow level. AnfisMIT adjustment mechanism produces high reactive power. Increasing the reactive power pulsation increases the imbalance of the stator current. During voltage sag, the active power drops and the generator supplies reactive power to the grid depending on the rotor circuit time constants as can be seen in Fig. 18. The improvements in the variable gain adjustment mechanism is shown in figure. In Fig. 19, during voltage sag the active power decreases and the rotor speed increase. The rotor speed increases can be minimized using AnfisMIT.

261

N. Amuthan et al. / Electrical Power and Energy Systems 47 (2013) 255–263 0.012

FuzzyMIT wind speed 6.45 m/s FuzzyMIT wind speed 9.45 m/s FuzzyMIT wind speed 12.45 m/s AnfisMIT wind speed 6.45 m/s AnfisMIT wind speed 9.45 m/s AnfisMIT wind speed 12.45 m/s

1.4

FuzzyMIT AnfisMIT

1.2 0.008

Rotor flux (p.u.)

Rotor Voltage (p.u.)

0.01

0.006 0.004

1

0.8

0.002 0.6 0 14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

Time (s)

0.4 14.9

14.95

15

15.05

Fig. 20. Rotor voltage during voltage sag.

15.1

15.15

15.2

15.25

15.3

Time (s) Fig. 23. Rotor flux during voltage sag in DFIG with various wind speed condition.

0.5

0

6

-0.5

5

Rotor current (p.u.)

Torque (p.u.)

FuzzyMIT AnfisMIT

-1

-1.5

-2

FuzzyMIT wind speed 6.45 m/s FuzzyMIT wind speed 9.45 m/s FuzzyMIT wind speed 12.45 m/s AnfisMIT wind speed 6.45 m/s AnfisMIT wind speed 9.45 m/s AnfisMIT wind speed 12.45 m/s

4 3 2 1

-2.5 14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3 0 14.9

Time (s)

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

Time (s)

Fig. 21. Torque of DFIG during voltage sag.

Fig. 24. Rotor current during voltage sag in DFIG with various wind speed condition.

FuzzyMIT wind speed 6.45 m/s FuzzyMIT wind speed 9.45 m/s FuzzyMIT wind speed 12.45 m/s AnfisMIT wind speed 6.45 m/s AnfisMIT wind speed 9.45 m/s AnfisMIT wind speed 12.45 m/s

2 1.8

1.4 1.2 1 0.8 0.6

FuzzyMIT wind speed 6.45 m/s FuzzyMIT wind speed 9.45 m/s FuzzyMIT wind speed 12.45 m/s AnfisMIT wind speed 6.45 m/s AnfisMIT wind speed 9.45 m/s AnfisMIT wind speed 122.45 m/s

6 5

Stator current (p.u.)

Stator flux (p.u.)

1.6

4 3 2

0.4 1 0.2 14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

Time (s)

0 14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

Time (s) Fig. 22. Stator flux during voltage sag in DFIG with various wind speed condition. Fig. 25. Stator current during voltage sag in DFIG with various wind speed condition.

Fig. 20 shows the variation of rotor voltage for different adjustment mechanisms. The injection of reactive power to a node tends to raise the voltage level. The rotor voltage is increased in the AnfisMIT adjustment mechanism. Torque varies inversely proportional to the rotor speed as seen from Fig. 21. The torque level is minimized slightly in the AnfisMIT adjustment mechanism. Fig. 22 shows the stator fluxes of DFIG with various wind speed conditions. As the wind speed increases the stator flux level also increases.

Fig. 23 shows the Rotor fluxes of DFIG with various wind speed conditions. The improved performance with wind speed increase is shown in figure. Fig. 24 shows the rotor current of DFIG during voltage sag. As the wind speed increases rotor current also increases and it is high in the linearly varying adjustment mechanism. Fig. 25 shows the Stator current during voltage sag in DFIG with various wind speed conditions. Stator current increases with wind speed increases.

N. Amuthan et al. / Electrical Power and Energy Systems 47 (2013) 255–263

FuzzyMIT wind speed 6.45 m/s FuzzyMIT wind speed 9.45 m/s FuzzyMIT wind speed 12.45 m/s AnfisMIT wind speed 6.45 m/s AnfisMIT wind speed 9.45 m/s AnfisMIT wind speed 12.45 m/s

Stator Active Power (p.u.)

6

4

2

0

12

FuzzyMIT wind speed 6.45 m/s FuzzyMIT wind speed 9.45 m/s FuzzyMIT wind speed 12.45 m/s AnfisMIT wind speed 6.45 m/s AnfisMIT wind speed 9.45 m/s AnfisMIT wind speed 12.45 m/s

10

Stator Reactive Power (p.u.)

262

-2

8 6 4 2 0 -2

14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

14.9

15.3

14.95

15

15.05

Fig. 26. Stator active power during voltage sag in DFIG with various wind speed condition.

15.2

15.25

15.3

0.04

0.02

FuzzyMIT wind speed 6.45 m/s FuzzyMIT wind speed 9.45 m/s FuzzyMIT wind speed 12.45 m/s AnfisMIT wind speed 6.45 m/s AnfisMIT wind speed 9.45 m/s AnfisMIT wind speed 12.45 m/s

3

2

Torque (p.u.)

Rotor Active power (p.u.)

0.06

15.15

Fig. 29. Stator reactive power during voltage sag in DFIG with various wind speed condition.

FuzzyMIT wind speed 6.45 m/s FuzzyMIT wind speed 9.45 m/s FuzzyMIT wind speed 12.45 m/s AnfisMIT wind speed 6.45 m/s AnfisMIT wind speed 9.45 m/s AnfisMIT wind speed 12.45 m/s

0.08

15.1

Time (s)

Time (s)

1

0

0 -1 -0.02 -2 14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3 14.9

Time (s)

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

Time (s)

FuzzyMIT wind speed 6.45 m/s FuzzyMIT wind speed 9.45 m/s FuzzyMIT wind speed 12.45 m/s AnfisMIT wind speed 6.45 m/s AnfisMIT wind speed 9.45 m/s AnfisMIT wind speed 12.45 m/s

Rotor Reactive Power (p.u.)

0.04

0.02

0

-0.02

Fig. 30. Torque during voltage sag in DFIG with various wind speed condition.

FuzzyMIT wind speed 6.45 m/s FuzzyMIT wind speed 9.45 m/s FuzzyMIT wind speed 12.45 m/s AnfisMIT wind speed 6.45 m/s AnfisMIT wind speed 9.45 m/s AnfisMIT wind speed 12.45 m/s

1.08

1.06

Rotor speed (p.u.)

Fig. 27. Rotor active power during voltage sag in DFIG with various wind speed condition.

1.04

1.02

1 -0.04 14.9

0.98 14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

Time (s)

14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

Time (s)

Fig. 28. Rotor reactive power during voltage sag in DFIG with various wind speed condition.

Fig. 31. Rotor speed during voltage sag in DFIG with various wind speed.

Fig. 26 shows the Stator Active Power during voltage sag in DFIG with various wind speed conditions. The wind speed increases the power fed to the grid as shown in figure with negative sign. Fig. 27 shows the Rotor Active Power during voltage sag in DFIG with various wind speed conditions. The rotor active power increases with wind speed increases. Fig. 28 shows the Rotor Reactive Power during voltage sag in DFIG with various wind speed conditions. Rotor reactive power generation increases with wind speed increases.

Fig. 29 shows the Stator Reactive Power during voltage sag in DFIG with various wind speed conditions. The stator reactive power increases with the wind speed increases. Fig. 30 shows the Torque during voltage sag in DFIG with various wind speed conditions. The rotor torque decreases as wind speed increases. Fig. 31 shows the rotor speed during voltage sag in DFIG with various wind speed. The rotor speed increases when the wind speed increases.

N. Amuthan et al. / Electrical Power and Energy Systems 47 (2013) 255–263 0.02

FuzzyMIT wind speed 6.45 m/s FuzzyMIT wind speed 9.45 m/s FuzzyMIT wind speed 12.45 m/s AnfisMIT wind speed 6.45 m/s AnfisMIT wind speed 9.45 m/s AnfisMIT wind speed 12.45 m/s

Rotor Voltage (p.u.)

0.015

0.01

0.005

0 14.9

14.95

15

15.05

15.1

15.15

15.2

15.25

15.3

263

power using variable gain adjustment mechanism is presented in Table 4. 6. Conclusion This paper presents a simulated model for small voltage sag ride in DFIG wind farm using direct model reference adaptive internal model controller. The advantages of using variable gain adjustment mechanism are presented. The influence of variable gain adjustment mechanism is simulated and it is proved that linearly varying gain adjustment mechanism (AnfisMIT) yields better performance in case of the ride through voltage sag. Acknowledgements

Time (s) Fig. 32. Rotor voltage during voltage sag in DFIG with various wind speed.

Table 4 Comparison of reactive power using variable gain adjustment mechanism. S. No

Rotor reactive power FUZZYMIT

ANFISMIT

1. Very low Initial transient in Rotor reactive power

0.17%

2. Very low Final transient in Rotor reactive power

4.1%

3. Stator reactive power

4%

Very low

4. 37.7% Initial transient in stator reactive power

80.6%

5. Above P.U.value Final transient in stator reactive power

Above P.U.value

6.

Above P.U.value

Above P.U.value

Table 3 Comparison of active power using variable gain adjustment mechanism. S. No.

Rotor active power FUZZYMIT

AnfisMIT

1. Very low Initial transient in rotor active power

0.8%

2. Very low Final transient in rotor active power

1.6%

3. Stator Active power

3.7%

Very low

4. 77% Initial transient in stator active power

77%

5. Above the p.u value Final transient in stator active power

Above the p.u value

6.

Above the p.u value

Above the p.u value

Fig. 32 shows the rotor voltage during voltage sag in DFIG with various wind speed. The rotor voltage variation is very small as the wind speed increases. Comparison of active power using variable gain adjustment mechanism are presented in Table 3 and comparison of reactive

The authors greatly acknowledge the support given by PET Engineering College, Valliyoor, Tamil Nadu, India, Nallathambi science and technology charitable trust, Nagercovil, Tamil Nadu, India, and National Engineering College, Kovilpatti, Tamil Nadu, India. References [1] Zinger Donald S, Muljadi Eduard. Annualized wind energy improvement using variable speeds. IEEE Trans Ind Appl 1997;33(6):1444–7. [2] Rahimi Mohsen, Parniani Mostafa. Dynamic behavior analysis of doubly-fed induction generator wind turbines – the influence of rotor and speed controller parameters. Int J Electr Power Energy Syst 2010;32(5):464–77. [3] Mullane Alan, Lightbody Gordon, Yacamini R. Wind-turbine fault ride-through enhancement. IEEE Trans Power Syst 2005;20(4):1929–37. [4] Boukhezzar B, Siguerdidjane H. Comparison between linear and nonlinear control strategies for variable speed wind turbines, Elsevier. Control Eng Pract 2010;18:1357–68. [5] Morren Johan, de Haan Sjoerd WH. Ride through of wind turbines with doublyfed induction generator during a voltage dip. IEEE Trans Energy Convers 2005;20(2):435–41. [6] Andreas Petersson. Analysis, modeling and control of doubly-fed induction generators for wind turbines, Ph.D THESIS, Chalmers University of Technology; 2005. [7] Ghedamsi K, Aouzellag D. Improvement of the performances for wind energy conversions systems. Int J Electr Power Energy Syst 2010;32(9):936–45. [8] Senthil Kumar N, Gokulakrishnan J. Impact of FACTS controllers on the stability of power systems connected with doubly fed induction generators. Int J Electr Power Energy Syst 2011;33(5):1172–84. [9] Ramesh L, Chakraborthy N, Chowdhury SP, Chowdhury S. Intelligent DE algorithm for measurement location and PSO for bus voltage estimation in power distribution system. Int J Electr Power Energy Syst 2012;39(1):1–8. [10] Amuthan N, Singh SN. Direct model reference adaptive internal model controller for DFIG wind farms. Int J Recent Trends Eng 2009;1(1):7–11. [11] Tzeng Ching-Yaw. An internal model control approach to the design of yawrate-control ship-steering autopilot. IEEE J Oceanic Eng 1999;24(4):507–13. [12] Melba Mary P, Marimuthu NS. Design of self – tuning fuzzy logic controller for the control of an unknown industrial process. IET J Control Theory Appl 2009;3(4):428–36. [13] Angelov Plamen P, Dimitar P. Filev, an approach to online identification of takagi-sugeno fuzzy models. IEEE Trans Syst, Man, Cybern – Part B: Cybern 2004;34(1):484–97. [14] Amuthan N, Melba Mary P, Subburaj P, Sharmeela C. Ride through and direct model reference adaptive internal model controller with rule based adjustment mechanisum for DFIG Wind farms. Int J Sustainable Energy 2012;31(4):229–50. [15] Amuthan N, Krishnakumar S. Design of linear parametric internal model controller-polynomial approach. J Electr Electron Eng June 2009;2(2):114–7. [16] Saludes S, Fuente MJ. Fault tolerant fuzzy IMC control in a PH process. IEE Proceedings of European Control; 2003. . [17] Attya AB, Hartkopf T. Penetration impact of wind farms equipped with frequency variations ride through algorithm on power system frequency response. Int J Electr Power Energy Syst 2012;40(1):94–103.