Critical Condition of Sensorless Induction Generator using Flux Weakening in Wind Turbine Application

International Journal of Power Electronics and Drive System (IJPEDS)
Vol. 7, No. 1, March 2016, pp. 202~216
ISSN: 2088-8694  202
Journal homepage: http://iaesjournal.com/online/index.php/IJPEDS
Critical Condition of Sensorless Induction Generator using Flux
Weakening in Wind Turbine Application
Nanda Avianto Wicaksono, Abdul Halim, Aries Subiantoro, Feri Yusivar
Department of Electrical Engineering, University of Indonesia (UI), Indonesia
Article Info ABSTRACT
Article history:
Received Oct 10, 2015
Revised Dec 21, 2015
Accepted Jan 11, 2016
This paper was intended to examine thoroughly a critical condition of the
sensorless induction generator using flux weakening in wind turbine
application. The critical condition would happen when the rotor speed
reached the critical rotor speed reference. The critical rotor speed reference
was the highest of the rotor speed reference that still caused the stable
response. It was obtained by increasing the rotor speed reference until the
system response became unstable. In the low speed range of wind showed
that there was no unstable condition whatever a rotor speed reference was
set. On the other hand, there was a critical rotor speed reference in the
medium and high speed range of wind. The unstable condition was caused by
the induction generator that received a power higher than its capacity, so its
rotor speed couldn't be maintained at reference value. The first solution was
suggested that the stable condition would be made by setting the rotor speed
reference at the minimum critical reference. The second solution was
suggested that the controlling rotor speed in triangle area between the critical
condition and the operation that used the minimum critical reference for the
rotor speed reference. In the triangle area, the rotor speed was controlled by
setting the tip speed ratio.
Keyword:
Generated power
Rotor speed
Sensorless induction generator
Stability
Wind speed
Copyright © 2016 Institute of Advanced Engineering and Science.
All rights reserved.
Corresponding Author:
Nanda Avianto Wicaksono,
Department of Electrical Engineering,
University of Indonesia,
Kampus UI Depok, Jakarta, 16424 Indonesia.
Email: nandaavianto@gmail.com
1. INTRODUCTION
There was a potency of wind energy in many locations which were far from cities, areas with the
small and low density of population, and were having small economic activities. The locations fitted to be
developed by using small wind turbines. The small wind turbines should have low cost, high reliability and
low maintenance.
For efficient cost and simple purpose application, the small wind turbine was designed by using (a)
a sensorless squirrel cage induction generator (SCIG), (b) fixed pitch angle of blades, and (c) fixed gear ratio.
This configuration of the wind turbine was chosen because of its simple construction, simple maintenance,
reliable operation, and also low price.
The fixed pitch angle of blades caused that the wind turbine couldn’t limit a received power from
wind. The received power that was higher than the rated power caused the generator became over-voltages
and over-current. To protect the generator damage caused of the over-voltages and the over-current, the wind
turbine operation must be stopped when a wind speed became higher than its rated speed. The stopping
caused the generated power couldn’t be produced.
To make the wind turbine able to generate power in a high speed of wind, the system was equipped
by a flux weakening control algorithm. The flux weakening control would keep the voltages and currents

 ISSN: 2088-8694
IJPEDS Vol. 7, No. 1, March 2016 : 202 – 216
203
generator ranged within the rated values. In [1], Yusivar et al. had simulated the wind turbine induction
generator using flux weakening. The simulation showed the unstable condition at the over-speed operation.
To define the variable which caused the unstable condition, the Yusivar et al.'s research should be continued
by analyzing a boundary between the stable and the unstable conditions. The boundary between the stable
and the unstable conditions was called the critical condition.
The main contribution of this paper was to examine thoroughly the critical condition of the
sensorless induction generator using flux weakening in wind turbine application and define the variable
which caused the unstable condition.
2. RESEARCH METHOD
This research consisted of four stages, i.e. (1) mathematical modelling, (2) implementation, (3)
simulation, and (4) analysis.
2.1. The Mathematical Modelling
The model of the system consisted of three parts, i.e. (1) the mechanical of wind turbine, (2) the
induction generator, and (3) the controller (see Figure 1).
First, the mechanical of wind turbine was used for calculating a load torque that referred to the rotor
speed of induction generator and the wind speed.
Second, the model of induction generator was used to produce the stator currents that refer to stator
voltages and a load torque. The stator voltages were the output of the controller, while the load torque was
received from the mechanical of wind turbine.
Third, the controller consisted of (1) a three phases to two phases transform and vice versa, (2) a
pulse width modulation (PWM) generator, (3) a rotor flux orientation control (RFOC), (4) a flux weakening,
(5) a speed controller, and also (6) an observer to estimate the rotor speed. The block diagram of the
controller was shown in Figure 2.
Figure 1. The parts of system
This research used the wind turbine type of horizontal axis. Its blades were connected to the low
speed shaft that wasn’t directly connected to the induction generator. There was a gearbox between the low
speed shaft and the induction generator. The gearbox was used to increase the speed of the low speed shaft.
The model of the mechanical wind turbine used six equations (1-6). The equations was used to
calculate the capacity factor of the wind turbine (Cp), the turbine speed (wt), the tip speed ratio (), the
turbine power (Pwt), and the load torque (TL) [2], [3].
, 0.22
116
0.4 5 exp
12.5 (1)
(2)

(3)

IJPEDS ISSN: 2088-8694 
Critical Condition of Sensorless Induction Generator using Flux Weakening …(Nanda Avianto Wicaksono)
204
Flux
Weakening
RFOC
Speed
Controller
3
2
PWM
Generator
3
2
Observer
IG Gearbox
Blades
Figure 2. Block diagram of the controller
1 1
0.08
0.035
1
(4)
0.5 (5)
(6)
The model of the induction generator used seven differential equations, i.e. the differential equation
of the stator current in d-axis (isd) (7), the differential equation of the stator current in q-axis (isq) (8), the
differential equation of the rotor current in d-axis (ird) (9), the differential equation of the rotor current in q-
axis (irq) (10), the differential equation of the angular speed of the stator voltage (e, the differential
equation of the rotor speed (r) (12), and the differential equation of the angle of rotor (13) [4], [5].
1 1 1 1
(7)
1 1 1 1
(8)
(9)
(10)
(11)
1
. (12)
(13)

 ISSN: 2088-8694
205
Before using the differential equations, the stator voltages received from the PWM generator were
converted by the three phases (abc) to two phases (dq-axis) transformation. And after using the differential
equations, the rotor currents produced were converted by the two phases (dq-axis) to three phases (abc)
transformation.
To convert the stator voltage from the three phases (abc-axis) to the two phases (dq-axis) was used
the Clarke transform (14) and then the Park transform (15). Viceversa, to convert the stator current from the
two phases (dq-axis) to the three phases (abc-axis) was used the Park inverse transform (16) and then the
Clarke inverse transform (17). The converting voltages or currents from the three phases (abc-axis) to the two
phases (dq-axis) and viceversa were also used in the controller [6], [7].
2
3
1 0.5 0.5
0 0.5√3 0.5√3
(14)
cos sin
sin cos
(15)
cos sin
sin cos
(16)
2
3
1 0
0.5 0.5√3
0.5 0.5√3
(17)
The PWM generator was comparing the absolute value of the stator voltage references with the
triangle carrier. If the absolute value was higher than the triangle carrier, the switch was on. Vice versa, if the
absolute value was lower than the triangle carrier, the switch was off.
While the switch was on and the stator voltage reference was positive, the stator voltage of the
induction generator was the same as the PWM amplitude of +Vdc/2. Vice versa, the stator voltage of the
induction generator was -Vdc/2 while the switch was on and the stator voltage reference was negative (see
Figure 3) [8].

Figure 3. Generating PWM signal
The stator voltage references (vsd and vsq) that were received by the PWM generator contained the
linier references (usdlin and usqlin) and the nonlinier decoupling references (usddec and usqdec). The linier
references was adjusted from two PI controllers. The adjusting PI controllers referred to the error between the

206
rotor current references and the actual values. On the other hand, the nonlinier decoupling references were
calculated by equations (22) and (23) [8].
(18)
(19)
(20)
(21)
1 (22)
1 (23)
The equations (20) and (21) showed that the calculating linier references used the stator current
references (isdref and isqref) from the flux weakening. The flux weakening limited the stator current references
of the induction generator by using equations (24-28). The flux weakening received the stator current
reference in q-axis (isqref) from the speed controller of PI controller. The adjusting PI controller referred to the
error between the rotor speed reference (rref) and the actual rotor speed (r) (see equation (29)) [9], [10].
(24)
(25)
(26)
(27)
(28)
(29)
The equations (22) and (23) showed that the calculating nonlinier decoupling references needed the
stator speed (e) and the rotor magnetizing current (imr) form RFOC. Besides the stator speed and the rotor
magnetizing current, RFOC also calculated the electric torque of the induction generator (Te) and the stator
angle (e [8].
1
(30)
1 (31)
(32)
1 1
(33)

 ISSN: 2088-8694
207
The equation (29) showed that the speed controller needed the actual rotor speed ( ) from the
induction motor. But the actual rotor speed couldn’t be obtained form the sensorless induction generator. The
feed back signals form the the sensorless induction generator were mainly the stator currents.
To subtitute the actual rotor speed ( ), the controller was equipped with an observer. The observer
was used to estimate the rotor speed by using equation (34). The estimated rotor speed was calculated by
using the estimated rotor flux ( rdest and rqest) and the error (eisd and eisq) between the estimated stator
currents and the actual values [11].
(34)
(35)
In this research, the estimation of the rotor magnetizing fluxes and the stator currents used the
Luenberger observer. The Luenberger observer was able to estimate the state variables (X) and the output
variables (Y) in the state space (36) and (37) [11]. The estimated state variables (Xest) and the estimated
output variables (Yest) could be calculated by using equations (37) and (38) [12], [13].
(36)
(37)
(38)
(39)
By using the equations (38) and (39), the Luenberger observer in the equation (41) was used to
estimate the state variables of the induction generator (isd, isq, rd, and rq) in the equation (40). The
estimated state variables were isdest, isqest, rdest, and rqest [11], [12], [13].

1
1
0
0
1
1
1
0
0
1
0 0
0 0
(40)
1 0
0 1
0 0
0 0
0 0
0 0
1 0
0 1

208

1
1
0
0
1
1
1
0
0
1
0 0
0 0
2

(41)
1 0
0 1
0 0
0 0
0 0
0 0
1 0
0 1
The equation (41) used the gains g1, g2, g3, and g4 that were written as below:
1
(42)
1
(43)
1
1
(44)
1
1
(45)
2.2. Implementation
The mathematics model of wind turbine system was written in the C MEX S-Functions and
simulated by using MATLAB/SIMULINK. The realization of wind turbine model was shown in Figure 4.
The induction generator was a squirrel cage induction machine with the rated power of 1 hp power, the rated
speed of 140 rad/s, and the maximum torque of 5 Nm. The parameters of the wind turbine and the induction
generator were listed in Table 1.

Figure 4. The realization of wind turbine model

 ISSN: 2088-8694
209
Table 1. Wind turbine and induction generator parameters
Symbols Description Value Units
R Blades Radius 0.95 m
KG Gearbox ratio 6.65 -
Np Number of Pole Pairs 2 -
Ls Stator Inductance 234.9 mH
Lr Rotor Inductance 234.9 mH
Lm Mutual Inductance 227.9 mH
Rs Stator Resistance 2.75 
Rr Rotor Resistance 2.9 
2.3. Test Scenario
The testing of the system was done by giving a constant wind speed and a constant rotor speed
reference. There were two series of wind speed that were tested. The first series, the wind speeds were
between 3 m/s and 20 m/s with interval of 1 m/s. The second series, the wind speeds were between 25 m/s
and 45 m/s with interval of 5 m/s. For each wind speed, the incremental of rotor speed reference was applied
until the system response became unstable. The highest of the rotor speed reference that caused the stable
response was named the critical rotor speed reference.
3. RESULTS AND ANALYSIS
3.1. The System Stability
The simulation showed that there were two operating condition responses, i.e. (a) the unstable
condition response and (b) the stable condition response. The example of the unstable response was at the
rotor speed reference of 240 rad/s and wind speed changed from 6 m/s to 12 m/s (see Figure 5). On the other
hand, the example of stable response was at the rotor speed reference of 270 rad/s and wind speed changed
from 22 m/s to 30 m/s (see Figure 6). Both condition showed that the system stability didn’t only depend on
either the rotor speed reference or the wind speed.
(a) Wind speed changed from 6 m/s to 12 m/s (b) Rotor speed: the reference (green),
the actual (red), and the estimated (blue)
Figure 5. The unstable response at the rotor speed reference of 240 rad/s and
the wind speed changed from 6 m/s to 12 m/s

210
(a) Wind speed changed from 22 m/s to 30 m/s (b) Rotor speed: the reference (green),
Figure 6. The stable response at the rotor speed reference of 270 rad/s and
the wind speed changed from 22 m/s to 30 m/s
3.2. Analysis
The test resulted that there were two operating ranges. First, the low speed range of wind was at the
wind speed lower than 8 m/s. Second, the medium and high speed ranges of wind were at the wind speed
equal or higher than 8 m/s.
In the low speed range of wind, there wasn't unstable condition whatever a rotor speed reference was
set. The highest generated power was reached by the highest capacity factor of 43.8% (see Figure 7). The
highest capacity factor was named the optimum operation condition. The highest capacity factor was reached
by the optimum tip speed ratio of 6.35 (see Figure 8). At the optimum condition, the generated power was
still lower than the power capacity of the induction generator.
Figure 7. The optimal condition in the low speed range of wind
In Figure 8, the optimum operation condition was shown by point M. At the constant wind speed,
the increasing rotor speed reference caused the increasing tip speed ratio, then it would cause decreasing
capacity factor (see arrow a in Figure 8). On the other hand, the decreasing rotor speed reference caused a
decreasing tip speed ratio, and then it would cause a decreasing capacity factor too (see arrow b in Figure 8).

 ISSN: 2088-8694
211
Both decreasing capacity factor made the lower generated power. Therefore, the rotor speed reference that
was higher or lower than the optimum reference would decrease the generated power, so the generated power
never exceeded the power capacity of the induction generator.
In the medium and high speed range of wind, there was a critical rotor speed reference. A rotor
speed reference that was higher than the critical reference would cause unstable condition. The critical
condition occurred while the rotor speed reference was between 230 and 406 rad/s, the load torque of the
induction generator was between -2.5 and -5.2 Nm., the generated power was between 725 and 880 W., and
the capacity factor was between 0.19 and 42.44 % (see Figure 9). These showed that the load torque and the
capacity factor of the critical condition were depending on the rotor speed reference. On the other hand, the
generated power of the critical condition was a relative constant around the rated power of 1 hp or 746 W
(see Figure 9c). It showed that unstable condition was caused by the induction generator that received a
power higher than its capacity, so its rotor speed couldn't be maintained at reference value.
In Figure 8, the critical condition was shown by points in C range. The increasing rotor speed
reference caused increasing tip speed ratio and then it would cause increasing capacity factor (see arrow d in
Figure 8). The increasing capacity factor caused the increasing generated power then it made the generated
power was higher than the rating power motor, so the system would be unstable.
Figure 8. Relation between capacity factor and tip speed ratio
The simulation shown in Figure 5 could be described by two points A and B. Point A represented
the condition at the rotor speed reference of 240 rad/s and the wind speed 6 m/s. Point B represented the
condition at the rotor speed reference of 240 rad/s and wind speed 12 m/s. The stable condition at point A
changed into the unstable condition at point B (see arrow AB in Figure 9a).
On the other hand, the simulation shown in Figure 6 could be described by two points C and D.
Point C represented the condition at the rotor speed reference of 270 rad/s and the wind speed 22 m/s. Point
D represented the condition at the rotor speed reference of 270 rad/s and wind speed 30 m/s. Both conditions
at point C and D were stable (see arrow CD in Figure 9a).
Figure 9 also showed the minimum critical rotor speed reference that was 230 rad/s at the wind
speed of 12 m/s. A solution was suggested that the stable condition would be made by setting the rotor speed
reference at the minimum critical reference of 230 rad/s. If the rotor speed reference was set constantly at 230
rad/s, the generated power never exceeded the power capacity of the induction generator (see Figure 10). The
simulation showed that the system response was stable at the constant rotor speed reference of 230 rad/s
although wind speed changed in range 6-45 m/s (see Figure 11).

212
(a) Rotor speed reference (b) Torque
(c) Generated power (d) Capacity factor
Figure 9. The critical condition in the medium and high speed ranges of wind.
Figure 10. The generated power and the capacity factor with the constant rotor speed reference at 230 rad/s

 ISSN: 2088-8694
213
(a) Wind speed changed in range 6-45 m/s (b) Rotor speed: the reference (green),
Figure 11. The stable response at the constant rotor speed reference of 230 rad/s and
the wind speed changed in range 6-45 m/s
Figure 12 showed that there was a narrow gap of the generated power between the optimal condition
and the operation using the minimum critical reference for the rotor speed reference in the low speed range of
wind (see zone A). But, in the medium and high speed ranges of wind, there was a wide gap of the generated
power between the critical condition and the operation using the minimum critical reference for the rotor
speed reference (see gap B in Figure 12). The wide gap showed that there was a challenge to generate power
higher than the generated power using the constant rotor speed reference in the medium and high speed
ranges of wind. The higher power would be generated by the controlling rotor speed in triangle area C in
Figure 13.
Figure 12. Generated power comparison
Another solution was suggested that the rotor speed in triangle area C in Figure 13 was controlled
by setting the tip speed ratio. There was a wide gap of the tip speed ratio between the critical condition and
the operation using the minimum critical reference for the rotor speed reference (see Figure 15). The gap of
the tip speed ratio was wider than the gap of the capacity factor (see Figure 14). It showed that the controlling
tip speed ratio had a bigger opportunity. To control the tip speed ratio, the wind turbine system had to be
equipped with the wind speed sensor of anemometer.

214
Figure 13. Rotor speed reference comparison
Figure 14. Capacity factor comparison
Figure 15. Tip speed ratio comparison
4. CONCLUSION
In the medium and high speed range of wind, there was a critical condition of the sensorless
induction generator using flux weakening in the wind turbine application. The critical condition would
happen when the rotor speed reached the critical rotor speed reference. The critical rotor speed reference was
the highest of the rotor speed reference that still caused the stable response. The unstable condition was
caused by the induction generator that received a power higher than its capacity, so its rotor speed couldn't be
maintained at reference value. The first solution was suggested that the stable condition would be made by

 ISSN: 2088-8694
215
setting the rotor speed reference at the minimum critical reference. The second solution was suggested that
the controlling rotor speed in triangle area between the critical condition and the operation by using the
minimum critical reference for the rotor speed reference. In the triangle area, the rotor speed was controlled
by setting the tip speed ratio.
ACKNOWLEDGEMENTS
This research was supported by Ministry of Research, Technology and Higher Education, Republic
of Indonesia and University of Indonesia (0563/UN2.R12/HKP.05.00/2015).
REFERENCES
[1] Feri Yusivar, Ridwan Gunawan, Harry Sudibyo S., Aries Subiantoro, Arnol Sinaga. Over-speed Operation of Wind
Turbine Induction Generator Using Field Weakening Control under Speed Sensorless Condition. The 2nd
International Conference on Nano Electronics Research and Education. Hamamatsu. 2014; 70-74.
[2] K. Viswanadha S. Murthy, M. Kirankumar, G.R.K. Murthy. A Performance Comparison of DFIG using Power
Transfer Matrix and Direct Power Control Techniques. International Journal of Power Electronics and Drive
System. 2014; 5(2): 176-184.
[3] A Kavitha, AV Suresh. A Novel Inter Connection of DFIG withGrid in Separate Excitation SMES System with
Fuzzy Logic Control. Bulletin of Electrical Engineering and Informatics. 2015; 4(1): 43-52.
[4] Muldi Yuhendri, Mochamad Ashari, Mauridhi Hery Purnomo. Maximum Output Power Tracking of Wind Turbine
using Intelligent Control. TELKOMNIKA. 2011; 9(2): 217-226.
[5] Ibrahim M. Alsofyani, Tole Sutikno, Yahya A. Alamri, Nik Rumzi Nik Idris, Norjulia Mohamad Nordin, Aree
Wangsupphaphol. Experimental Evaluation of Torque Performance of Voltage and Current Models using Measured
Torque for Induction Motor Drives. International Journal of Power Electronics and Drive System. 2015; 5(3): 433-
440.
[6] K. Ramesh, Ch. Ravi Kumar, P. Bala Murali. Modeling and Implementation of Vector Control for Induction Motor
Drive. International Journal of Engineering Research and General Science. 2015; 3(2): 80-91.
[7] Ridwan Gunawan, Feri Yusivar, Zuhal A. Kadir. Field Oriented Control of An Induction Motor Speed Sensorless
with Current Vector Controller, direct-quadrature Current Compensator and Full Order Observer In direct-
quadrature Axis. The 2nd Indonesian Japan Joint Scientific Symposium. Jakarta. 2006.
[8] F. Yusivar, S. Wakao. Minimum Requirements of Motor Vector Control Modeling and Simulation Utilazing C MEX
S-Function in Matlab/Simulink. The 4th
IEEE International Conference on Power Electronics and Drive Systems.
Bali. 2001; 1: 315-321.
[9] A. Ben Ali, A. Khedher, M.F. Mimouni, R. Dhifaoui. Torque Maximization and Sensorless Control of Induction
Motor in a Flux Weakening Region. International Journal of Sciences and Techniques of Automatic Control &
Computer Engineering IJ-STA. 2009; 3(1): 972-985.
[10] Yusivar F, Wakao S. Field-Weakening Scheme Combined with Saturated Voltage Control Strategy. Proceedings of
the IASTED International Conference on Modelling, Identification and Control. Grindelwald, Switzerland. 2004;
23: 412-146.
[11] T. Raghu, J. Srinivas Rao, S. Chandra Sekhar. Simulation of Sensorless Speed Control of Induction Motor Using
APFO Technique. International Journal of Computer and Electrical Engineering. 2012; 4(4).
[12] Feri Yusivar, Kenji Uchida, Hiroyuki Haratsu, Shinji Wakao, Takashi Onuki. Speed Adaptive Observer for
Sensorless IM Drive using Combined Reference Frames. Fifteenth Annual IEEE Applied Power Electronics
Conference and Exposition (APEC). 2000; 1: 127-132.
[13] F. Yusivar, H. Haratsu, T. Kihara, S. Wakao, T. Onuki. Performance comparison of the controller configurations
for the sensorless IM drive using the modified speed adaptive observer. 8th International Conference on Power
Electronics and Variable Speed Drives. 2000; 495: 194-200.
BIOGRAPHIES OF AUTHORS
Nanda Avianto Wicaksono was born in Jakarta, Indonesia. He received his bachelor and master
degrees in Electrical Engineering at Institute Technology of Bandung, Indonesia in 1998 and
2002. He is currently the student in Electrical Engineering at University of Indonesia and the
researcher in Research and Development Agency, Ministry of Energy and Mineral Resources,
Republic of Indonesia. His research interests are in control system, electrical drive, power
electronics, and renewable energy.

216
Abdul Halim was born in Jakarta, Indonesia He received his bachelor and master degree in
Electrical Engineering at Keio University, Japan in 1995 and 1997, and also completed his
Doctor degree in 2000 at Tokyo Institute of Technology, Japan. He is currently the lecturer in
Electrical Engineering at University of Indonesia. His research interests are in control system
engineering, power system engineering, computer simulation, intelligent engineering, and
applied mathematics.
Aries Subiantoro was born in Jakarta, Indonesia. He received his Bachelor degree in Electrical
Engineering at University of Indonesia in 1995, and completed his Doctor degree in 2013 at
University of Indonesia. His research interests are in model and simulation, intelligent control
system, and model predictive control system.
Feri Yusivar was born in Bandung, Indonesia. He received his bachelor degree in Electrical
Engineering at University of Indonesia in 1992, and completed his Doctor degree in 2003 at
Waseda University, Japan. He is currently the Head of Control Laboratory in Electrical
Engineering at University of Indonesia. His research interests are in control system, electrical
drive, power electronics, and renewable energy.

Critical Condition of Sensorless Induction Generator using Flux Weakening in Wind Turbine Application

More Related Content

What's hot (17)

Similar to Critical Condition of Sensorless Induction Generator using Flux Weakening in Wind Turbine Application (20)

More from IAES-IJPEDS (20)

Recently uploaded (20)

Critical Condition of Sensorless Induction Generator using Flux Weakening in Wind Turbine Application