Control Design and Dynamic Performance Analysis

of a Wind Turbine-Induction Generator Unit

Ezzeldin S. Abdin and Wilson Xu, Senior Member, IEEE

Abstract—This paper presents the modeling and control design

for a wind energy conversion scheme using induction generators.
The scheme consists of a three-phase induction generator driven
by a horizontal axis wind turbine and interfaced to the utility
through a double overhead transmission line. A static VAR
compensator was connected at the induction generator terminals
to regulate its voltage. The mechanical power input was controlled
using the blade pitch-angle. Both state and output feedback
controllers are designed using MATLAB software to regulate the
generator output. From the simulation results, the response of Fig. 1. Wind-Generator System.
closed loop system exhibited a good damping and fast recovery
under different type of large disturbances.
II. WIND-GENERATOR SYSTEM CONFIGURATION
I. INTRODUCTION Fig. 1 shows the wind-generator system configuration. The
system consists of a horizontal axis wind turbine (HAWT)

T HE CONTROL of a system, that generates power from an

unsteady input as the wind, presents a formidable problem.
The wind speed varies from time to time due to gusts, and is
and an induction generator (IG) interfaced to the utility grid
through a double overhead transmission line. A fixed-capacitor
thyristor-controlled reactor (FC-TCR) static VAR compensator
further disturbed by the effect of supporting tower shadow. At- is connected at the generator terminals for voltage regulation.
tempts have been made to overcome this problem for wind-syn-
chronous generator systems by keeping the rotor speed as nearly
III. SYSTEM MODELING
constant as possible using the blade pitch angle control [1], [2].
A mathematical model of a stand-alone wind energy conversion A. Wind Turbine Model
system with battery energy storage was developed in [3]. In [3], The wind turbine is characterized by nondimentional curves
a permanent magnet brushless synchronous generator was used of the power coefficient Cp as a function of both tip speed ratio,
and the wind turbine was considered as the only source of power.  and the blade pitch angle, . The tip speed ratio  is the ratio
A study of the application of induction generators in power of linear speed at the tip of blades to the speed of the wind. It
system was given in [4] where the effect of induction gener- can be expressed as follows [1]:
ator parameters and voltage support requirements for stability
R
under disturbance conditions were studied. The advantages of = (1)
induction generator versus synchronous generators lie in lower Vw
cost and higher reliability. They have simpler control and auxil- where R is the WT rotor radius, is the mechanical angular
iary equipment. The recent progress in the development of con- velocity of the WT rotor and Vw is the wind velocity. For the
trolled static VAR sources makes it possible to consider the in- wind turbine used in the study, the following form approximates
duction generator coupled with controlled VAR sources as an Cp as a function of  and [1]:
 
alternative to the use of synchronous generators [5], [6]. ( 0 3)
Cp = (0:44 0 0:0167 ) sin 0 0:00184( 0 3) :
In this paper, the dynamic modeling and control design for
a horizontal axis wind turbine induction generator is presented.
15 0 0:3
(2)
The effect of gating angle change of FC-TCR static VAR com- The output mechanical torque of the wind turbine, Tm can be
pensator on the generator voltage is studied. For output regula- calculated from the following equation [2]:
tion, state and output PI controllers are designed based on the
linear model and tested under a severe wind gust and large elec- Tm = 12 ARCp Vw2 = (3)
trical disturbances.
where  is the air density and A is the swept area by the blades.

B. Induction Generator Model

The wind turbine drives an induction generator whose flux
Manuscript received May 13, 1998. This work was supported by the Egyptian
Government.
linkage model can be described in d- and q -axis synchronous
The authors are with the Electrical Engineering Department, University of reference frame as [7]:
Alberta, Edmonton, Canada, T6G 2G7.
Publisher Item Identifier S 0885-8969(00)02218-X. p'ds = !b (Vds + Rsids + 'qs ) (4)
and and
p'qs = !b(Vqs + Rs iqs 0 'ds ) X
(5) Vqs = Vqb + RTLiqTL + XTL idTL + TL piqTL : (16)
!b
p'dr = !b (Vdr 0 Rr idr ) + (!b 0 !m )'qr (6)
IV. SMALL SIGNAL LINEARIZED MODEL
p'qr = !b (Vqr 0 Rr iqr ) 0 (!b 0 !m )'dr : (7)
The design techniques used for the control systems are based
In the above equations, all the rotor variables are referred to the on the state space linear model. The nonlinear model of the
stator side. The expression for the electromagnetic torque in per wind generator was linearized around an operating point. The
unit can be written as: linearized model takes the following form

Te = 'ds iqs 0 'qs ids : (8) px = Ax + Bu + d (17)

where
The wind turbine and induction generator rotors are represented
as a lumped one mass. So, the dynamic equation of motion can x = [1'ds ; 1'qs ; 1'dr ; 1'qr ; 1!r ; 1Vds ;
be written as: 1Vqs ; 1idl ; 1iql ; 1idTL ; 1iqTL ]t
! u = [1 ; 1 ]t
p!m = b (Tm 0 Te ) (9)
2H
d = Vw :
where H is the equivalent inertia constant of both wind turbine
and induction generator rotors.
V. STATE AND OUTPUT CONTROLLER DESIGN
C. Static VAR Compensator Model Two paths are available for control of the wind turbine. Static
In this paper, the static fixed-capacitor thyristor-controlled re- VAR control at the terminal of the induction generator allows
actor (FC-TCR) compensator was used to maintain a constant for regulation the terminal voltage to a set point. The second
voltage at the generator terminals by continuous adjustment of path is control of the blade pitch angle, which effects the me-
the reactive power using the gating angle of the thyristor. For chanical input power to the generator. To limit the mechanical
the FC, the relation between current and voltage in d- and q -axis stress in the wind turbine torsional system, a rate limit 610 deg/s
synchronous reference frame can be given by [8] should be put on the blade pitch actuator, which controls the me-
1 chanical power. The electrical power and voltage regulations are
Vqs = pV 0 Xc idc
!b ds
(10) achieved through PI controllers. The controllers’ design is based
on the linearized model. Two integrators are applied on the error
1 signals (Pref 0 Pg ) and (Vref 0 Vs ). These two integrators are
Vds = 0 pVqs + !bXc iqc : (11)
!b augmented with the state space linearized model to give the fol-
For the TCR, the full conduction is obtained with a gating firing lowing augmented model:
angle of 90 . The effect of increasing the gating firing angle is pz = 8z + 0u + Ed (18)
to reduce the fundamental harmonic component of the inductor
current, reducing its reactive power as well as its current. Using where
 
z= x ;
Fourier analysis, the fundamental component of the inductor
current can be calculated by the following equation [5]:
Z Z t
2( 0 ) 0 sin [2( 0 )] V
iL = Vs = s (12) = (1Pref 0 1Pg ) dt; (1Vref 0 1Vs ) dt
XL XLeq
     
where XLeq is the equivalent reactance of the TCR and is A 0 B 
the gating firing angle. The differential equations for the d- and
8= 0C 0 ; 0 = 0 ; E = 0 :
q -axis components of iL in synchronous reference frame can be Based on the augmented model, a state feedback controller,
written as follows: which takes the following form:
pidL = !b(Vds + XLeq iqL )=XLeq (13) u = 0 Ks z (19)

piqL = !b (Vqs 0 XLeq idL )=XLeq : (14) can be designed using modern control design techniques. In this
paper, the pole-assignment technique was used. By selecting
the desired closed loop poles, Ks can be obtained using the
D. Transmission Line Model MATLAB software.
The d and q equations, which describe the dynamics intro- The output feedback controller gains can be derived from
duced by the transmission line, are given as: the state feedback gains using Pseudoinverse matrix form. The
output vector y can be written as follows:
X
Vds = Vdb + RTL idTL 0 XTL iqTL + TL pidTL (15) y = Cz:
!b (20)
Fig. 2. State and output feedback controllers. Fig. 4. Comparison between nonlinear model and full-order linearized model
for small pulse change in .

state feedback controller was designed. After many iterations,

the desired closed loop poles were selected as shown in Table I.
By using the MATLAB software, the state feedback control
matrix, Ks was derived and given by the first equation at the
bottom of the next page. The output vector was selected to be
 Z
y= 1Pg ; 1Vs; 1Is; 1!r ; (1Pref 0 1Pg) dt;
Fig. 3. Stator voltage versus thyristor firing angle. Z t
(1Vref 0 1Vs) dt : (24)
If C C is invertible, the state vector z can be obtained as a func-
t

tion of output vector y as follows: The output feedback gains were derived from Ks and given by
the second equation at the bottom of the next page. The closed
z = [C tC ]01 Cy = pinv(C )y: (21) loop poles with output controller are calculated and given in
Table I. The state and output controller were implemented in
By defining the output feedback control law as the nonlinear simulation model. A three-phase short circuit at
the generator terminals for 100 ms following by switching off
u = 0 Ko y (22) the faulted line was chosen to compare the performance of the
two controllers. The response of the closed loop system under
and using (19) and (21), Ko can be derived from Ks as follows: this disturbance is shown in Fig. 6. The terminal voltage and
Ko = Ks pinv(C ):
electrical power were returned to its reference values in very
(23) short time and the deviation in rotor speed was reduced to a very
small value (about 0.7 rad/s). The response of the closed loop
Fig. 2 shows the block diagram representation of state and
system with output controller is very close to that of state feed-
output controllers.
back. In the practical, the output feedback is preferred because
all the output signals are available for measurement and there is
VI. SIMULATION RESULTS
no need for a full order or a reduced order observer as in the state
Fig. 3 shows the steady state variation of the stator voltage as feedback. Fig. 7 shows the response of closed loop system with
a function of the thyristor gate-firing angle, for different value output controller to a severe wind gust. The initial wind speed
of slips. As shown from Fig. 3, the stator voltage can be kept is 28 mile/h, near the rated value (24 mile/h) and the peak of
constant at rated value over a wide range of operating conditions the gust at 42 mile/h is near the cut out wind speed (45 mile/h).
by regulating the gating firing angle. A comparison between the The gust duration time is 3 s. After the gust, the wind velocity
nonlinear and linearized full-order models is shown in Figs. 4 was held constant at a new value superimposed on it a sinusoidal
and 5 for small pulse changes in the input signals and . The component to represent the aggregate effects of wind shear and
pulse duration times are 100 s for and 50 ms for . tower shadow. From the figure, the closed loop response has a
As shown from the figures, a close agreement between the good damping and the maximum terminal voltage and electrical
two model responses is found. Based on the linear model, the power excursions are significantly smaller and limited to about
Fig. 5. Comparison between nonlinear model and full-order linearized model
for small pulse change in .

TABLE I

0.45% for voltage and 1.25% for power. The rotor speed devia-
tion is very small and is equal to 0.25 rad/s.

VII. CONCLUSIONS
The paper has been presented the modeling and controller
design for a wind turbine induction generator unit. The voltage
regulation of induction generator has been achieved using a
fixed-capacitor thyristor controlled reactor static VAR com- Fig. 6. Response of the closed loop system for 3- short circuit following by
pensator over a wide range of operating conditions. State and switching off the faulted TL (Tf = 100 ms).

 
s =
034 75 01 294
: : 76:89 11:69 022 06
: 1:024 00 488 011 85 013 45 06 473 01 623 191 5 0241 75
: : : : : : :

0 001 01 001 00 117 0 007 00 007 07 711 05 738

K
0 009
: 0 029 : 1:00 0:521 0:058 0:002 : : : : : : :

 
=
11:667 07 278 010 905 022 058
: : : 191:502 241:748
Ko
0:109 00 037 0 427 0 058
: : : 0 7:711 0 5:738
 APPENDIX

A. Induction Generator Parameters

4 poles, 6.6 Kv, 60 Hz.
Xm = 4:161 pu.;
X1s = 0:135 pu.;
X1r = 0:075 pu.;
Rs = 0:0059 pu.;
Rr = 0:0339 pu.;
Hg = 1:975 s:

B. Static VAR Compensator Parameters

XL = 4:0 pu.; XC = 3:8 pu.

C. Transmission Line Parameters

XTL = 0:15 pu.;

RTL = 0:015 pu.;
VB = 1:04 pu.

D. Wind Turbine Parameters

Horizontal axis wind turbine (6 MW)
HT = 16:72 s;
R = 200 ft;
GR = 103:6
Vw (rated) = 24:0 mile/h
Vw (cut in) = 12:0 mile/h
Vw (cut out) = 45:0 mile/h:

E. List of Principal Symbols

'ds ; 'qs Stator flux linkage components in d- and q -axis (PU)
'dr ; 'qr Rotor flux linkage components in d- and q -axis (PU)
ids ; iqs Stator current components in d- and q -axis (PU)
idr ; iqr Rotor current components in d- and q -axis (PU)
idc ; iqc Capacitor current components in d- and q -axis (PU)
idL ; iqL TCR current components in d- and q -axis (PU)
idTL ; iqTL TL current components in d- and q -axis (PU)
Is ; V s Stator current and voltage respectively, (PU)
Pref Induction generator rated power (PU)
Pg Induction generator power (PU)
Rs; Rr Stator and rotor resistances (PU)
Fig. 7. Response of closed system with output controller for a gust change in
Vw .
RTL; XTL Resistance and reactance of the TL (PU)
XL ; Xc Inductive and capacitive reactances of the SVC (PU)
Vds ; Vqs Stator voltage components in d- and q -axis (PU)
output PI controllers were designed and implemented in the Vdb ; Vqb 1 -bus voltage components in d- and q -axis (PU)
nonlinear simulation model. A good damping performance has Vref Stator rated voltage (PU)
been achieved for the closed loop system during severe wind !b Base speed (377 rad/s)
gusts and large electrical system disturbances. !r Induction generator rotor speed (rad/s).
 ACKNOWLEDGMENT [7] I. Boldea and S. A. Nasar, Electric Machine Dynamics. New York, NY:
Macmillan Publishing Company, 1986.
E. S. Abdin would like to thank the University of Alberta for [8] M. A. Rahman, A. M. Osheiba, T. S. Radwan, and E. S. Abdin, “Model-
the equipment facilities. ling and controller design of an isolated diesel engine permanent magnet
synchronous generator,” IEEE Trans. on EC, vol. 11, no. 2, pp. 324–330,
June 1996.
REFERENCES
[1] J. R. Winkelman and S. H. Javid, “Control design and performance anal-
ysis of a 6 MW wind turbine generator,” IEEE Trans. on PAS, vol. 102, Ezzeldin S. Abdin received the B.Sc., M.Sc., and Ph.D. degrees in Electrical
no. 5, pp. 1340–1347, May 1983. Engineering from Menoufiya University, Egypt, in 1979, 1983, and 1989 respec-
[2] O. Wasynczuk, D. T. Man, and J. P. Sullivan, “Dynamic behavior of tively. In 1979, he joined the Department of Electrical Engineering, Menoufiya
a class of wind turbine generators during random wind fluctuations,” University, as an instructor. He was promoted to Assistant Lecturer in June
IEEE Trans. on PAS, vol. 100, no. 6, pp. 2837–2845, June 1981. 1983, Assistant Professor in April 1989, and Associate Professor in January
[3] B. S. Borowy and Z. M. Salameh, “Dynamic response of a stand-alone 1994. From January 1987 to December 1988, he was a Visiting Researcher at
wind energy conversion system with battery energy storage to a wind the University of New Brunswick, Fredericton, Canada. His current interests are
gust,” IEEE Trans. on EC, vol. 12, no. 1, pp. 73–78, Mar. 1997. in modeling of power system, controller design and renewable energy.
[4] F. P. de Mello, J. W. Feltes, L. N. Hannett, and J. C. White, “Application
of induction generators in power system,” IEEE Trans. on PAS, vol. 101,
no. 9, pp. 3385–3393, 1982.
[5] T. J. E. Miller, Reactive Power Control in Electric System: A Wiley- Wilsun Xu (SM’93) obtained the Ph.D. from the University of British Columbia
Interscience Publication, John Wiley & Sons, 1982. in 1989. From 1989 to 1996, he was with B.C. Hydro as an Electrical Engineer.
[6] Y. H. Song, A. T. Johns, and R. K. Aggarwal, “Nonlinear thyristor-con- He joined the University of Alberta as an Associate Professor in September
trolled static VAR compensation,” The European Power Electronics As- 1996. Dr. Xu’s research interest is in power quality, power system stability, and
sociation, Brighton, Sept. 13–16, 1993. distribution automation.