Novel Voltage Controller for Stand-alone Induction
Generator using PWM-VSI
G. V. Jayaramaiah
Research Scholar,
Energy Systems Engineering.
Indian Institute of Technology Bombay
Powai , Mumbai - 400 076
INDIA
Abstract— This paper presents a DSP-based constant voltage
controller for stand alone wind energy conversion system using
an induction generator. The system uses a pulse-width modulated
voltage source inverter (PWM-VSI) with a start-up battery. The
limitation of having stand alone wind energy conversion system
with self-excited induction generator (SEIG) is poor voltage
regulation which occurs with change in speed and load condition.
To overcome this problem, a DSP-based voltage controller is
developed. It regulates the voltage when SEIG is subjected to
a sudden application/removal of load. It is now possible to
operate the induction generator (IG) at constant voltage from
no load to full load. The amplitude of the terminal voltage of the
IG is regulated by varying the modulation index of the PWM
inverter. The system has an inherent current limiting feature
and it requires only sensing of dc link voltage. To predict the
performance of the proposed system, a MATLAB/SIMULINK-
based simulation study is carried out. The control algorithm is
implemented on a TMS320F243 DSP platform at the assembly
language level for optimum performance of the voltage controller.
Viability of the compensation process is ascertained through
experimental results obtained from the laboratory prototype.
I. I NTRODUCTION
Among the different renewable energy systems, wind
energy appears as the most promising one, due to both
technical and economic factors. Important progress in wind
energy conversion technologies has been achieved and more
efficient and more powerful wind generators are now available.
The selection of the generator depends upon many factors such
as type of application, machine characteristics, maintenance,
cost etc. Currently induction machines are more popular
compared to other machines. However its major disadvantage
is the requirement of excitation power. This reactive power
can be supplied by a variety of methods [1]–[3], ranging
from using simple capacitors to that of a VSI inverter with
complex power conversion techniques. Another limitation of
the SEIG in stand-alone systems is its inability to control
the terminal voltage and frequency under change in load and
wind speed. To overcome this problem, several methods are
proposed in [2]–[4]. The capacitor excitation [3], [4] is suitable
only when there is a constant load at the IG terminal and is
driven at constant mechanical speed. However any change in
load and rotor speed may result in a loss of excitation. To
overcome this problem, discrete blocks of capacitors can be
204
1-4244-0365-0/06/$20.00 (c) 2006 IEEE
B. G. Fernandes
Department of Electrical Engineering,
Indian Institute of Technology Bombay,
Powai , Mumbai - 400 076 , INDIA
Phone : +91 - 022 - 25767428,
Email : bgf@ee.iitb.ac.in
added or removed at the IG terminals either using contactors
or power electronic switches in series with the capacitors,
depending upon change in speed and load. Another method
of providing excitation involves using a saturable reactor [5]
connected in each phase of IG. This inductor with stepped
air gap gradually saturates with stator current. As a result, its
inductance decreases with load. However the overall system
is bulky and expensive.
In order to improve the performance of the system, use of
modern control techniques such as vector control and sliding
mode control have been suggested [6 & 7]. Though the use
of vector control technique improves the performance of the
system, the overall system becomes complex. It should be
noted that one of the key issues in standalone system is
reliability and simplicity in control structure. The use of these
techniques defeats this purpose.
The excitation schemes proposed in [6]–[10] involve power
electronic converters to source the required reactive power
to excite the IG. Though the voltage build-up process has
been discussed and the results have been presented in these
papers, the dynamic behaviour of SEIG is not discussed.
Moreover, the overall control structure is complex in nature.
Therefore the main objective of this paper is to develop a
simple control strategy to overcome the limitations of the
existing schemes. The proposed controller is also capable of
handling reactive loads and does not require the mechanical
speed sensors, ac voltage or current sensors, thereby reducing
the overall cost and hardware complexity. This also improves
the overall reliability of the system. The controller maintains
a constant voltage at the IG terminal during the change in
load by adjusting the inverter frequency. The amplitude of
the terminal voltage of IG is controlled by the modulation
index of the VSI. Unlike in most of the systems reported in
literature, the proposed system does not require a dump load.
Detailed Matlab/Simulink-based simulation studies are carried
out to demonstrate the effectiveness of the scheme. Viability
of the scheme is confirmed through experimental results using
a scaled down laboratory prototype. The control algorithm is
implemented on a TMS320F243 DSP.
 II. M ATHEMATICAL M ODELLING OF 3-Φ S ELF - EXCITED The set point of Vdc must be greater than the peak value
I NDUCTION G ENERATOR of the machine line voltage in order to force the desired line
The d-q axes equivalent circuits of an induction genera- currents. Total d. c. current Idc can be expressed in terms of
tor (IG) in synchronously rotating reference frame are shown inverter switching function as
in Fig. 1. The complete dynamic equations of IG, taking Idc = Sa iea + Sb ieb + Sc iec

ω eλqs (ω e− ω r)λ qr (suffix e identifies compensator phase currents)

Rs Lls Llr Rr The three switching functions take the value of 1 if the upper
− + + − switch of the inverter leg is on and it is 0 if the lower switch
I ds I dr
in the same inverter leg is on.
Vds λds Lm λdr Vdr
B. Model of the Voltage Source Inverter (VSI)
Using the switching function SF 1a,b,c the Vao , Vbo and Vco
can be obtained as:
∞
Vdc 
(a)
Vdc
ω eλds Lls Llr (ω e − ω r)λ dr R Vao = SF1a = An sin(nωt)
Rs r 2 2 n=1
+ − + −
∞
I qs I qr
Vdc Vdc 
Vbo = SF1b = An sin(nωt − 120o )
Vqs λqs Lm λqr 2 2 n=1
Vqr
∞
Vdc Vdc 
Vco = SF1c = An sin(nωt + 120o )
2 2 n=1
(b)
The Line-to-Line voltages generated by the inverter can be
derived as:
Fig. 1. d-q model of IG (a) d - axis (b) q - axis
Vab = Vao − Vbo
saturation into account, in synchronously rotating reference
frame [11], [12] is represented in matrix form as follows : Vbc = Vbo − Vco






d λds V i 0 −1 λds
= ds − Rs ds − ω
dt λqs Vqs iqs 1 0 λqs Vca = Vco − Vao





 C. Mathematical model of the Load

d λdr Vdr ids 0 −1 λdr Equations dealing with generator feeding a (R-L) load
= − Rr − (ω − ωr )
λ
dt qr V qr iqs 1 0 λqr in d-q frame are
Equations for electromagnetic torque and mechanical speed of
the SEIG are expressed as follows: d
Vds = RL iLds + LL iLds − ωe LL iLqs
3P dt
Te = (iqr λdr − λqr idr )
4 d
Vqs = RL iLqs + LL iLqs + ωe LL iLqs
d P ∗ (Tshaf t − Te ) dt
ωr = where iLds and iLqs are the d-axis and q-axis components
dt 2J
of the load current
IV. P RINCIPLE OF OPERATION OF THE CIRCUIT
III. M ATHEMATICAL M ODELING OF 3 − Φ PWM - VSI The proposed overall system block diagram is shown in
The complete mathematical modeling of the PWM-VSI and Fig. 2. A 12 V battery on the dc side of inverter is provided
the load are explained in the following sections. for initial excitation. The reactive power required by the IG
and load is provided by the voltage source inverter. Therefore
A. Representation of the d. c. side of the inverter the rating this inverter is chosen based on the excitation power
The capacitor voltage equation is governed by: of IG and reactive power requirements of the load. During
startup, the controller sets the stator frequency lower than the
d Idc rotor frequency so that the power produced by IG is used
Vdc = −
dt C to charge the capacitor connected across the dc link to a set
where Vdc is the voltage across the capacitor and Idc is the reference value. In this study this voltage is maintained at
current flowing through it, as shown in Fig. 2. 150 V. The error between the reference and actual capacitor

205
 PM−Prime mover Idc Charging circuit

IG − Induction Generator S1 S3 S5 Diode

1
0
0
1 11
00
00
11 1
0
0
1 1
0
0
1

R C
PM IG Y Vdc Vs 12 V

B
R S4 S6 S2
1
0
0
1 11
00
00
11 1
0
0
1
Y
B

Control Signals
R Y B

R S1 S4 S3 S6 S5 S2 Vdc *
Vdc

L PI

ωmax ωmin
N Vtri
V ma
Load 1 Vryb ryb Sin
OSC
0 dq Cos ωs
1 kHz
DSP TMS320F243

Fig. 2. Block diagram of the Proposed System

voltages is processed by the PI controller. If the measured
capacitor voltage is higher than the reference value, the stator
frequency is increased by the controller, thereby decreasing
the torque and power supplied by IG, and, if the measured
capacitor voltage is lower than the reference value the stator
frequency is decreased. The output of the PI regulator is fed
to the harmonic oscillator to generate the sine and cosine
waveforms. These waveforms are multiplied by the modulation
index (ma ) to get Vr , Vy and Vb as shown in Fig. 2. These
sinusoidal waveforms are compared with 1 kHz triangular
carrier signal to generate the switching pulses to the IGBT
inverter.
Any variation in the output power of IG is directly
indicated by the variation in the terminal voltage of the
generator. A decrease in capacitor voltage below the reference
value indicates that the active power drawn by the load is
higher than the power generated by IG. This difference in
power is supplied by the VSI and hence the dc link voltage
falls. Due to step change in load, the input power to the
induction generator decreases as speed of the prime mover
decreases. To maintain a constant voltage at the IG terminal,
the controller decreases the inverter frequency.
An increase in capacitor voltage indicates that the active
power required by the load is reduced due to the removal of
load. Under this condition active power generated by the IG
is higher than the power required by the load. In order to
decrease the active power the controller increases the inverter
frequency.
V. S IMULATION R ESULTS
The developed models of the sub-system are inte-
grated and the resulting system is simulated using MAT-
LAB/SIMULINK. The machine rating and its equivalent cir-
cuit parameters used for the study are given in appendix.
In order to prove the viability of the control scheme a
TMS320F243 DSP based scaled down laboratory prototype
is designed and developed. The simulated results showing the
variation of terminal voltage of the generator and the voltage
across dc link capacitor during start up are shown in Fig. 3
while, Fig. 4 shows these results obtained from the prototype
developed in the laboratory.
When the IG is suddenly loaded the frequency and
terminal voltage of IG, and the capacitor voltage tend
to fall. The terminal voltage is restored to the reference
value by adjusting the inverter frequency. The simulated
and experimental results are shown in Fig. 5 and Fig. 6
respectively. Similar behaviour of SEIG is observed when

206
 100

(V)
IG Voltage V (V)

100 0 (a)

RY
ry

−100

V
0 (a) −200
1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8

160

Vdc (V)
−100
150 (b)
140
0.2 0.4 0.6 0.8 1 1.2 1.4 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8

(A)
4

Phase−A
2
150 0
−2
(c)

IG
−4
(V)

100 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8
(b) Time (Sec)
dc
V

50
Trace a : Line-to-line voltage of IG
0 Trace b : Capacitor voltage Vdc
0.2 0.4 0.6 0.8 1 1.2 1.4 Trace c : phase current of IG
Time (Sec)

Trace a : Line-to-line voltage of IG Fig. 5. Simulated results of SEIG during step increase in load
Trace b : Capacitor voltage Vdc

Fig. 3. Simulated results of SEIG on no-load

Trace a : Line-to-line voltage of IG 200V/div

Trace b : Capacitor voltage Vdc 100V/div
Trace c : Load current 1A/div
Time : x-axis 200ms/div

Fig. 6. Experimental results of SEIG during step increase in load

are shown in Fig. 9.

Trace 1 : Line-to-line voltage of IG 200V/div VI. C ONCLUSION

Trace 2 : Capacitor voltage 50V/div
Time : x-axis 250ms/div
Fig. 4. Experimental results of DSP controlled SEIG on no-load
the load on IG is reduced. The terminal voltage and stator
frequency tend to increase. The terminal voltage and capacitor
voltage increase due to the mismatch in active power produced
by the IG, which is more than the power consumed by the
load. The closed loop controller adjusts the inverter frequency
so that the the terminal voltage is maintained at the reference
value. The simulated and experimental results are shown in
the Fig. 7 and Fig. 8 respectively. The variation of the stator
frequency and capacitor voltage during step change in load
A simple DSP based constant voltage controller for stand
alone wind energy conversion system using induction genera-
tor is proposed. The controller is able to maintain a constant
voltage at the terminals of IG during step change in load.
The developed controller has a fast dynamic response, robust
and reliable. The controller does not require any mechani-
cal speed sensor thereby reducing the cost and complexity
of the hardware. The control algorithm is implemented on
TMS320F243 EVM DSP platform using assembly level lan-
guage and simulated using MATLAB/SIMULINK. Viability of
the compensation process is ascertained through experimental
results obtained from the laboratory prototype.

207
 200 155
150

Vdc (V)
100

V (V)
0 145
(a)

ry
(a) 140
−100
135
−200 1.4 1.5 1.6 1.7 1.8 1.9 2
3.4 3.5 3.6 3.7 3.8 3.9 4
55
165

fs (Hz)
50 (b)
155
V (V)

(b)
dc

145 45
1.4 1.5 1.6 1.7 1.8 1.9 2
135 160
3.4 3.5 3.6 3.7 3.8 3.9 4

Vdc(V)
0.5 155
(c)
150
I (A)

0 (c) 145
L

3.4 3.5 3.6 3.7 3.8 3.9 4

50
−0.5
(d)

fs(Hz)
3.4 3.5 3.6 3.7 3.8 3.9 4
Time(Sec) 45

40
Trace a : Line-to-line voltage of IG Vry 3.4 3.5 3.6 3.7
Time (Sec)
3.8 3.9 4

Trace b : Capacitor voltage Vdc

Trace c : Load current Trace a and c : Capacitor voltage Vdc
Trace b and d : Stator frequency fs
Fig. 7. Simulated results of SEIG during step decrease in load
Fig. 9. Simulated results of SEIG during step change in load.

Trace a : Line-to-line voltage of IG 200V/div
Trace b : Capacitor voltage Vdc 100V/div
Trace c : Load current 1A/div
Time : x-axis 200ms/div
Fig. 8. Experimental results of SEIG during step decrease in load
VII. A PPENDIX
Parameters of the Induction Machine at 50 Hz
Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Φ
Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ∆
Rated voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 V
Rated current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.425 A
Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1430 RPM
Stator resistance(rs ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.83 (Ω)
Rotor resistance(rr ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.55 (Ω)
Stator inductance(ls ) . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.4535 (H)
Rotor inductance (lr ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.475 (H)
Magnetizing inductance(unsaturated) . . . . . . . . . . . . . . . . . . 0.3H
Rated power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1kW
Number of Poles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Rotor inertia(J) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.664 (kg.m2 )
R EFERENCES
[1] E.D.Basset and F.M.Potter, “Capacitance excitation of Induction Gen-
erators,” Transaction on American Institute of Electrical Engineering,
vol. 54, no. 1, pp. 540–545, 1935.
[2] H. C. Rai and A. K. Tandon, “Voltage Regulation of SEIG Using Passive
Elements,” 6th European Conference on Electrical Machines and Drives,
vol. 1, no. 1, pp. 240–245, 1983.
[3] M.B.Brennan and A.Abbondanti, “Static Excitation for Induction Gen-
erator,” IEEE Transactions on Industry Applications, vol. 13, no. 5, pp.
422–428, Sept/Oct, 1977.
[4] N. H. Malik and A. H. Al-Bahrani, “Influence of the Terminal Capacitor
on the Performance Characteristics of SEIG,” IEE Proceedings of
Instituteion of Electrical Engineering,Part C, vol. 137, no. 2, pp. 168–
173, March, 1990.
[5] S. M. Alghuwainew, “Steady State Analysis of a Self-Excited Induction
Generator by Shunt Saturable Reactor,” IEEE Conference on Electrical
Machines and Drives, vol. 1, no. 1, pp. 101–103, 1997.
[6] D. W. Novotny and G. H. Studtmann, “Self-excitation in Inverter
Driven Induction machine,” IEEE Transactions on Power Apparatus and
Systems, vol. 96, no. 4, pp. 1117–1125, July/Aug, 1977.
[7] S. N. Bhadra, K. V. Ratnam, and A. Manjunath, “Study of Voltage
Build up in a Self-Excited, Variable Speed Induction Generator/Static
Inverter System With D.C side Capacitor,” IEEE Conference on Power
Electronics, Drives and Energy System, vol. 2, no. 1, pp. 964–970, 1996.
[8] T.L.Maguire and A.M.Gole, “Apparatus for supplying an isolated dc
load from a variable speed Self-Excited Induction Generator,” IEEE
Transactions on Energy Conversion, vol. 8, no. 3, pp. 468–475, 1993.
[9] G. V. Jayaramaiah and B. G. Fernandes, “Dynamic Simulation of
Three Phase Self-Excited Induction Generator With PWM-VSI,” Na-
tional Conference on Control,Communication and Informations Systems,
vol. 1, no. 1, pp. 132–137,Jan,GOA, 2004.
[10] M.S.Miranda, R.O.Lyra, and S.R.Silva, “An Alternative Isolated wind
electric pumping system using Induction machine,” IEEE Transactions
on Energy Conversion, vol. 14, no. 4, pp. 1611–1616, Dec, 1999.
[11] P.C.Krause, O.Wasynczuk, and S. Sudhoff, Analysis of Electrical Ma-
chinery. New York: IEEE Press, 1994.
[12] O. Ojo, “Minimum Airgap Flux linkage Requirement for Self-Excitation
in Stand Alone Induction Generators,” IEEE Transactions on Energy
Conversion, vol. 10, no. 3, pp. 484–492, 1995.

208