GA Based Performance Enhancement of SHE Methods For Nine-Level Inverter
GA Based Performance Enhancement of SHE Methods For Nine-Level Inverter
GA Based Performance Enhancement of SHE Methods For Nine-Level Inverter
+
=
(1)
As the output voltage of the MLI is quarter wave, the
Fourier series constants a
0
,a
n
become (a
0
,a
n
=0)zero and
only b
n
is to be determined. The value of b
n
is established
using the equations (2) and (3)
( ) ( )
2
1
sin
0
0
t n td t
b v n
=
}
(2)
3 2
sin ( ) ( )sin ( )
1 1 2
1 1 2
/2 5
( )sin ( ) ( )sin ( )
1 2 3 1 2 3 4
5 4
v n td t v v n td t
b
n
v v v n td t v v v v n td t
(
(
+ + +
} }
(
(
=
(
(
+ + + + + +
} }
(
(
(3)
By analysis the constants b
n
, and then substituting in
equation (1) The Fourier analysis expression for the
output of the Nine-Level Inverter voltage waveform is
attainable as,
4
( ) ( cos .... cos )sin
1 1
1,3
v t v n v n t
s s
n n
= + +
=
(4)
S =number of DC sources
1 1
1,3
4
( ) ( cos .... cos )sin
dc
s s
n
V
v t v n v n t
n
=
= + +
(5)
The all sources are of equal value
V
dc
=V4=V
3
=V
2
=V
1
3.2 Estimation of Switching Angles
Fourier series of the quarter-wave symmetric H bridge
MLI output waveform is reported as given in equation (6)
in which S are the optimal switching angles, which must
content the following condition.
International Journal of EmergingTrends & Technology in Computer Science(IJETTCS)
Web Site: www.ijettcs.org Email: [email protected], [email protected]
Volume 2, Issue 5, September October 2013 ISSN 2278-6856
Volume 2, Issue 5 September October 2013 Page 148
1 2
....
2
s
< < <
The technique to solve the optimized selective harmonic
switching pulse will be explained in this segment . From
equation (1), the harmonic element in the waveform can
be detail as follows:
1. The amplitude of DC =0.
2. The amplitude of all odd harmonic components
including fundamental harmonic voltage, is given
1
1,3
4
( ) cos
dc
k
k
V
h n v n
n
=
=
(6)
3. The amplitude of all even harmonics=0. So, Only
the odd harmonics in the one fourth wave symmetric
Multilevel waveform wanted to be minimized. The
switching pulse of the waveform will be modified
to achieve the lowest distortion in the output.
1 2 3 4
4
( ) [cos cos cos cos ]...(7)
dc
V
h n
n
= + + +
If needed to control the maximum value of the output
voltage to be V1 and eliminate the third, fifth and seventh
order harmonics, the MI is given by
M =V1 the resulting harmonic equations are
1 2 3 4 1
1 2 3 4
1 2 3 4
1 2 3 4
4
[cos cos cos cos ] .................(8)
[cos3 cos3 cos3 cos3 ] 0.................(9)
[cos5 cos5 cos5 cos5 ] 0.................(10)
[cos7 cos7 cos7 cos7 ] .............
dc
V
V
n
+ + + =
+ + + =
+ + + =
+ + + = ....(11)
Equation (8) is rewritten as
1 2 3 4
cos cos cos cos ] .................(12) M + + + =
This paper deals with a four H-bridge cascaded inverter
because it 9-level output voltage can be approximately
sinusoidal. In that case the conventional method can
eliminate the 3
nd
, 5
th
and 7
th
harmonics except for the
fundamental wave. In Nine Levels Inverter and four DC
sources are needed so that the DC voltage levels are
chosen so as not to generate the third, fifth and seventh
order harmonics while achieving the desired basic
voltage. This is a system of four simultaneous equations
with five unknowns
1
,
2
,
3
and
4
. These values are
found by solving the simultaneous equations (810).
4. CLASSIC METHOD FOR HARMONIC
DEVALUATION
4.1 Newton Raphson Method
The ordinary method has the advantage of eliminating the
needed harmonic component but it has some problems.
First it is arduous to solve simultaneous equations which
are a set of nonlinear equations. In this equation can be
determined by an iterative method such as NR. If the
number of equations increases so carry the time and the
volume of calculations to get the conduction angles.
Furthermore the method is an exact one count on the
iteration which edge with the addition of some errors.
Second conducting angles are determined through an off-
line operation. Accordinarly they have to be systematized
in the look-up table. It needs more data in system to
implement switch pulse with an accurate determination.
If the range of the modulation index is divided into
design, the data of the conducting angles increases.
Particularly, the date of the conducting angles depends on
the determination of the modulation index. So the
ordinary method has a limitation in its application for an
adjustable motor drive. The ordinary method does not
determine the set of nonlinear equations but calculates
more trigonometric functions. The objective is to select
the levels of the DC voltages so as to obtain the required
fundamental voltages V
1
and specific higher order
harmonics of V (t) equals to zero.
4.2 Simulation for Nine Level Inverter
The single phase Nine Level Inverter using Pulse Width
Modulation technique and switching angle fluctuation
technique is simulated with the help of MATLAB
R2012a.For Nine Level Inverter four H-bridges are
essential for simulation. Power electronics switches are
used as power switches. Figure 3 displayed the simulation
circuit for Nine Level Inverter using Pulse Width
Modulation technique.
4.3 Simulation Results
Figure 6 displayed the circuit that is simulated in
MATLAB. The input voltage in every H-bridge is 400 V
and Semiconductor is used as a power switch. The load is
treated as pure resistance 50. The switching angles are
found using the Genetic Algorithm approach in the off
line. Figure 3 displayed the gate pulse generated in
Simulink for the power circuit shown in Figure 2. Figure
4 displayed the output voltage of the Inverter and the
harmonic of the output voltage. At the power circuit is
simulated with switching pulse determined by the Newton
- Raphson method.
International Journal of EmergingTrends & Technology in Computer Science(IJETTCS)
Web Site: www.ijettcs.org Email: [email protected], [email protected]
Volume 2, Issue 5, September October 2013 ISSN 2278-6856
Volume 2, Issue 5 September October 2013 Page 149
Figure 3. Gate Pulse
Figure 4. FFT analysis for Nine Level Inverter using
NR method
The circuit shown in Figure. 2 are simulated and the
results are resented in the Figure. 3, 4.
5. GA TECHNIQUE FOR SWITCHING
PULSE GENERATION
5.1 Genetic Algorithm to determined optimum
switching Angles.
The restraint of the Newton Raphson method is removed
by using the GA based optimization method. The
switching angles are calculated using Genetic Algorithm.
The steps for developing a problem and applying a
Genetic Algorithm is as follows:
1. Choose binary or floating point.
2. Discovery the number of variables particular to the
problem; her number will be, the number of genetic
codes in a chromosome. In this function the number of
variables is the number of controlling switching pulse
which is the number of Bridges in an MLI. A Nine-
Level Inverter need Four H-bridges; thus, each genetic
code for this application will have four switching
pulse, ie, (
1
,
2
,
3
,
4
).
3. Set a population range and load the population. High
population efficacy increases the rate of convergence
but it still increases the operation time. Then choose an
optimum-sized population wanted some trained in GA.
The population size in this paper has 20 genes, each
including four switching angles. The population is
loaded with random angles between 0 and /2 taking
into attention the one fourth symmetry of the output
voltage of MLI.
4. The most necessary element for the GA to calculate the
fitness of every chromosome is the cost function. The
impartial of this study is to eliminate specified
harmonics; thus the cost function has to be similar to
these harmonics.
In this work the third, fifth and seventh order harmonics
at the output voltage of a nine-level inverter are to be
eliminated. Then the cost function (CF) can be chosen as
the sum of these three harmonics establish to the
fundamental,
3 5 7
1 2 3 4
1
( , , , ,) 400 ................(13)
V V V
f
V
+ +
=
For each gene a multilevel output voltage is design using
the switching angles in the genetic code and the wanted
harmonic magnitudes is calculated using FFT techniques.
The FV is estimated for each chromosome including. In
this case,
3 5 7
1 2 3 4
1
( , , , ,) 400 ............(14)
V V V
FV
V
+ +
=
The switching angle set creating the maximum Fitness
Value is the best answer of the 1
st
iteration.
Figure 5. Flow chart of GA
The GA is normally set to run for a specific number
Iterations to find an answer. Afterwards the first iteration,
Fitness Value is used to find new offspring. These go
done crossover and mutation working and A population is
a new design which goes through the same period starting
from Fitness Value evaluation.
Commonly, the Genetic Algorithm can converge to a
solution well earlier 100 iterations are achieved. To save
time, the iterations have been stopped when the exact
value of the cost function going below 1, in which model
International Journal of EmergingTrends & Technology in Computer Science(IJETTCS)
Web Site: www.ijettcs.org Email: [email protected], [email protected]
Volume 2, Issue 5, September October 2013 ISSN 2278-6856
Volume 2, Issue 5 September October 2013 Page 150
the sum of the 3nd, 5
th
and the 7th harmonics is removed
Compared to the fundamental. Note that value after these
iterations, the GA discovery one solution; so, it has to be
run as much time as the number of solutions necessary to
mask the whole modulation index. The algorithm for
discovering the optimum switching value is detail
through the flow chart shown in Figure. 5.
Table 1. Switching angles for the different modulation
index
Modulation
index
Switching Pulse
1
(rad)
2
(rad)
3
(rad)
4
(rad)
0.1 0.5253 0.8594 0.997 1.2743
0.2 0.21 0.54 0.994 1.5689
0.3 0.194 0.465 0.7932 1.1176
0.4 0.0943 0.6092 0.774 1.3673
0.5 0.091 0.605 0.7745 1.374
0.6 0.3128 0.6663 0.9543 1.1698
0.7 0.0820 0.3549 0.493 0.953
0.8 0.6287 0.892 1.133 1.482
0.9 0.091 0.5893 0.781 1.3823
1.0 0.2023 0.645 1.1072 1.5638
5.2 Simulation Results
The Nine Level Inverter, switching value which is
suppressing the third, fifth and seventh order harmonics
are shown in Table. 1. For various ranges of modulation
indices using GA. The power circuit shown in Figure.2 is
simulated using the switching pulse estimated from
genetic algorithm. Figure 6 displays the output voltage
and its harmonic profile. From the spectrum analysis it is
complete that the THD in GA based is 14.75% and that
for Newton-Raphson is 21.58%. By comparing the two
Figures. 4 and 6 it is clearly identified that the harmonics
are reduced effectively by computing the switching angles
by GA compared to Newton Raphson method. By
comparing the two Figures. 4 and 6 it is clearly identified
that the harmonics are reduced effectively by computing
the switching angles by GA compared to the Newton
Raphson method.
Figure 6. FFT analysis for Nine Level Inverter using
GA method
6. CONCLUSION
The main objective of this project is to establish programs
on different modern technique to solve the Selective
Harmonic Elimination problem with equal D.C sources in
H-bridge Nine level inverter. In this work GA
optimization technique is used to find the optimum
switching angles of the Nine Level Inverter for the
Eliminating the harmonics. The results achieved show
that third, fifth and seventh order harmonics are removed
effectively. Genetic Algorithm located solution of
switching angles gives minimum Total Harmonic
Distortion in the output voltage waveform compared with
the typical N R method. As with this way, GA can be
used for any type of optimization problems. Genetic
Algorithm compared reduces the harmonic content more
especially than any other typical technique such as a
Newton Raphson method. This work can be extended by
implementing Genetic Algorithm to reduce the harmonics
in inverters with any number of evaluations.
References
[1] Mohammed H. Rashid, Power Electronics,
Prentice Hall of India Pvt. Ltd, 2nd Edition,
1994.
International Journal of EmergingTrends & Technology in Computer Science(IJETTCS)
Web Site: www.ijettcs.org Email: [email protected], [email protected]
Volume 2, Issue 5, September October 2013 ISSN 2278-6856
Volume 2, Issue 5 September October 2013 Page 151
[2] R.Baker and L. H. Bannister, Electric power
converter, U.S. Patent 3867643, Feb. 1975.
[3] S. Khomfoi, L. M Tolbert, Chapter31, Multilevel
Power Converters The University of Tennessee.
Paper page .31-1 to 31-50.
[4] Rodriguez, J.Lai, and J.Peng, F. Z.: Multilevel
Inverters: a Survey of Topologies, Controls and
Applications, IEEE Transaction India Electron. 49
No. 4 (Aug 2002), 724738.
[5] K. El-Naggar and T. H. Abdelhamid, Selective
harmonic elimination of new family of multilevel
inverters using genetic algorithms, Energy
Conversion and Management, Volume 49, No. 1,
pp.89-95, Jan. 2008.
[6] Holmes, D. G., & Lipo, T. A, Pulse width
modulation for power converters: Principles and
practice, NJ: IEEE Press. John Wiley, 2003.
[7] H. S. Patel and R. G. Hoft, Generalized harmonic
elimination and voltage control in thyristor
inverters: Part IHarmonic elimination, IEEE
Transaction India Application, volume. IA-9,
number. 3, pp. 310 317, May/Jun. 1973.
[8] H. S. Patel and R. G. Hoft, Generalized harmonic
elimination and voltage control in thyristor
inverters: Part IIVoltage control technique,
IEEE Transaction India Application, volume IA-10,
number. 5, pp. 666 673, Sep. /Oct 1974.
[9] Enjeti, P. N.Lindsay, J. F.: Solving Nonlinear
Equation of Harmonic Elimination PWM in Power
Control, Electron. Letter 23 No. 12 (June 1987),
656657.
[10] Shi, K. L.Hui Li: Optimized PWM Strategy
Based on Genetic Algorithms, IEEE Trans.
Industrial Electronics 52 No. 5 (Oct 2005), 1458
1461.
[11] A. I. Maswood, Shen Wei and M. A. Rahman, A
Flexible Way to Generate PWM-SHE Switching
Patterns Using Genetic Algorithm, Conference
Proceedings of IEEE (APEC), 2001, pp. 1130-
1134.
[12] R. Salehi, N. Farokhia, M. Abedi, and S.H. Fathi,
Elimination of Low Order Harmonics in Multilevel
Inverters Using Genetic Algorithm, Journal of
Power Electronics, volume 11, no. 2, Mar. 2011,
pp.132-139.
[13] S.N.Sivanandam, & S.N.Deepa (2011). Principles of
Soft Computing, Second edition page(s):385-401.
[14] Ramkumar, Kamaraj, Thamizharasan, GA Based
Optimization and Critical Evaluation SHE Methods
for Three-level Inverter, 2011 1
st
International
Conference on Electrical Energy Systems.
[15] M. J. Schutten, D. A. Torrey, Genetic Algorithms
for control of Power Converters, Conference
Proceedings on IEEE PESC, 1955, pp. 1321-
1326.
BIOGRAPHIES
N.Veeramuthulingam was born on 16th Feb
1988, in, Pondicherry India. He received his
B.Tech degree in Electrical and Electronics
Engineering from Pondicherry University in
2009 and M.Tech Degree in Electric Drives and Control
from Pondicherry Engineering College in 2011. He
presently works as Assistant Professor in the Department
of Electrical Engineering at, Surya Group of Institutions,
Tamilnadu, India. His research area includes Harmonics
Analysis in power converter, efficient pulse width
modulation strategies.
S. Sivajanani @ Santhoshma received her
B.Tech in Electrical and Electronics
Engineering from Pondicherry University in
2010 and also she completed her M.Tech in
Electrical Drives & Control in Pondicherry Engineering
College in 2012. Presently, she is working as Assistant
Professor in the Department of Electrical and Electronics
Engineering, Manakula Vinayagar Institute of
Technology, Pondicherry, India. Her area of interest
includes power electronics, harmonic analysis and drives.
V. Thebinaa obtained her Bachelor degree in
Electrical and Electronics and Master Degree
in Power Systems Engineering from
Annamalai University, Chidambaram. She is
currently working as Assistant Professor in the
Department of Electrical and Electronics Engineering,
Manakula Vinayagar Institute of Technology,
Pondicherry. Her area of interest includes power system,
power quality issues.
N. Arunkumar has received his B.E from
Madras University in 2003 and M. Tech from
Vellore Institute of Technology in 2007. He
has six years of teaching experience. He is
currently working as Assistant Professor in the
Department of Electrical and Electronics Engineering,
Manakula Vinayagar Institute of Technology,
Pondicherry. His area of interest includes power
electronics and drives, machine